Patterning and Cell Type Specification in the Developing CNS and PNS: Comprehensive Developmental Neuroscience [2 ed.] 012814405X, 9780128144053

Table of contents :
Cover
Patterning and Cell Type Specification in the Developing CNS and PNS
Copyright
Contributors
Part I: Induction and patterning of the CNS and PNS
1 - Morphogens, patterning centers, and their mechanisms of action
1.1 General principles of morphogen gradients
1.1.1 History of the morphogen and morphogenetic field
1.1.2 How morphogen gradients pattern tissues
1.1.3 How morphogens are distributed
1.1.4 How morphogen signaling is transduced and interpreted
1.1.5 How morphogen gradients are converted into sharp boundaries
1.1.6 Summary-general principles of morphogen gradients
1.2 Local signaling centers and probable morphogens in the telencephalon
1.2.1 Early forebrain patterning
1.2.2 The RPC
1.2.3 The telencephalic roof plate and cortical hem
1.2.4 The antihem
1.3 BMPs as morphogens in telencephalic patterning
1.3.1 Performance objectives for a BMP gradient in the dorsal telencephalon
1.3.2 Midline expression and homeogenetic expansion of BMP production
1.3.3 BMP signaling gradient in the dorsal telencephalon
1.3.4 BMPs as dorsal telencephalic morphogens
1.3.5 Linear conversion of BMP signaling by cortical cells
1.3.6 Nonlinear conversion of BMP signaling by DTM cells
1.3.7 Summary-the BMP signaling gradient
1.4 FGF8 as a morphogen in telencephalic patterning
1.5 Interactions among signaling centers in telencephalic patterning
1.5.1 FGF8, Shh, and BMP signaling
1.5.2 Cross-regulation of BMP, FGF, and WNT signaling
1.5.3 Interactions of Shh, FGFs, and Gli3
1.6 Morphogens in human brain disease
1.6.1 Holoprosencephaly and Kallmann syndrome
1.6.2 Gradients in holoprosencephaly neuropathology
1.6.3 Gradients in other human brain disorders
References
2 - Telencephalon patterning
2.1 Introduction
2.2 Telencephalon induction
2.2.1 The anterior neural ridge
2.2.2 FGF signaling
2.2.3 Wnt antagonism
2.2.4 Interactions of low Wnt with FGFs and BMPs
2.3 Overview of early telencephalic subdivisions
2.4 Establishing dorsal versus ventral domains
2.4.1 Shh and Gli3, two key players
2.4.2 Foxg1 and FGFs cooperatively promote ventral development
2.4.3 Establishing the dorsal telencephalic domain
2.4.4 Sharpening the dorsal-ventral border
2.4.5 The olfactory bulbs
2.5 Boundary structures as organizing centers and CR cell sources
2.5.1 Nomenclature of domains in the early telencephalic neuroepithelium
2.5.2 Specification of the hem and the antihem
2.5.2.1 Molecular mechanisms that act to position and specify the cortical hem
2.5.2.2 Molecular mechanisms that act to specify and position the antihem
2.5.3 Cajal-Retzius cells arise from four telencephalic boundary structures
2.5.4 Organizer functions of telencephalic boundary structures
2.5.4.1 Rostral signaling center/septum
2.5.4.1.1 Hem
2.5.4.2 Antihem
2.6 Subdividing ventral domains
2.6.1 The striatum and pallidum
2.6.2 The amygdala
2.6.3 An evolutionary perspective for how the neocortex arose
2.6.4 Lineage and fate mapping in the ventral telencephalon
2.7 Conclusions
Acknowledgments
References
3 - Area patterning of the mammalian neocortex
3.1 Introduction
3.1.1 Basic principles
3.1.2 Classic neocortical area patterning models
3.2 Indications that intrinsic mechanisms pattern the neocortical primordium
3.3 Morphogens impart position to the neocortical primordium
3.3.1 Morphogen signaling
3.3.2 Neocortical patterning by FGFs
3.3.3 Fgf8 regulates neocortical guidance of thalamic axons
3.3.4 Neocortical patterning by the cortical hem
3.4 Patterning genes downstream of morphogen signaling
3.4.1 Emx2 and Pax6
3.4.2 Dmrt5/Dmrta2
3.4.3 Couptf1/Nr2f1
3.4.4 Sp8
3.4.5 Pbx
3.5 Do neocortical areas arise from dedicated progenitor cell pools?
3.5.1 Transcription factors known to pattern the NP appear in gradients, not domains
3.5.2 Mapping the cortical primordium with forebrain enhancers
3.6 The influence of thalamic innervation
3.6.1 Guidance of thalamocortical axons and area formation
3.6.2 Thalamic innervation determines the function of a cortical area
3.6.3 Effects of thalamocortical afferents on area size and cortical progenitor cells
3.6.4 Thalamic dependence of an area-specific feature
3.6.5 Two mechanisms united
3.7 Spontaneous activity and neocortical patterning
3.8 Conservation of patterning mechanisms among different mammalian species
3.9 Conclusions
References
4 - Patterning of thalamus
4.1 Introduction
4.2 Insights into diencephalic patterning
4.2.1 Columnar and neuromeric models
4.2.2 Morphologic segmentation of the diencephalon in the prosomeric model
4.2.3 Molecular regionalization of the diencephalon
4.2.3.1 Prosomere 1
4.2.3.2 Prosomere 2: the epithalamic domain
4.2.3.3 Prosomere 3
4.3 Prosomere 2: the thalamic domain
4.3.1 Cell lineages in the p2 alar plate
4.3.2 Signaling molecules during the initial patterning phase
4.3.2.1 Shh
4.3.2.2 Wnt
4.3.2.3 Fibroblast growth factor
4.3.3 Transcription factor control for neuronal identity
List of acronyms and abbreviations
References
5 - Midbrain patterning: polarity formation of the tectum, midbrain regionalization, and isthmus organizer
5.1 Introduction: brief description about midbrain
5.2 Tectum laminar formation
5.3 Optic tectum as a visual center for the lower vertebrate
5.3.1 Retinotectal projection in a retinotopic manner
5.3.2 Polarity formation in the optic tectum
5.4 Development of midbrain from the mesencephalic brain vesicle
5.4.1 Transcription factors that determine the midbrain
5.4.2 Midbrain-hindbrain boundary formation
5.4.3 Diencephalon-mesencephalon boundary formation
5.4.4 Dorsoventral patterning in the midbrain
5.5 Isthmus organizer
5.5.1 Isthmus emanates organizing signal
5.5.2 Competence of the neural tube to Fgf8 signaling is determined by preexisting transcription factors
5.5.3 Intracellular signal transduction
5.5.4 How tectum and cerebellum are organized by isthmus organizing signal?
5.6 Concluding remarks
List of abbreviations of genes and molecules
List of abbreviations (general)
Glossary
References
6 - Cerebellar patterning
6.1 Introduction
6.2 Early formation of cerebellum
6.2.1 Morphogenetic aspect of first steps of cerebellar formation
6.2.2 Molecular mechanisms underlying initial formation of cerebellum
6.3 Three types of cerebellar patterning in adult mammals
6.3.1 Cerebellar anterior-posterior patterning
6.3.1.1 Lobes
6.3.1.2 Lobules (I-X)
6.3.1.3 Functional roles of lobes
6.3.2 Cerebellar medial-lateral patterning
6.3.2.1 Parasagittal zones
6.3.2.2 Parasagittal stripes
6.3.2.3 Correspondence between parasagittal zones and parasagittal stripes
6.3.2.4 Functional roles of parasagittal zones and stripes
6.3.3 Cerebellar outer-inner patterning
6.3.3.1 The molecular layer
6.3.3.2 The Purkinje cell layer
6.3.3.3 The granular layer
6.3.3.4 The white matter
6.3.3.5 The cerebellar nuclei
6.3.3.6 Roles of cerebellar outer-inner patterning
6.4 Formation of cerebellar patterning
6.4.1 Formation of cerebellar anterior-posterior patterning
6.4.1.1 Formation of lobes and lobules
6.4.1.2 Cellular mechanisms underlying the formation of lobes and lobules
6.4.2 Formation of cerebellar medial-lateral patterning
6.4.2.1 Formation of parasagittal zones
6.4.2.2 Cellular and molecular mechanisms underlying the formation of parasagittal zones
6.4.2.3 Formation of parasagittal stripes
6.4.2.4 Critical roles of Purkinje cell birth date in the formation of embryonic and adult parasagittal stripes and parasagittal zones
6.4.3 Formation of cerebellar outer-inner patterning
6.4.3.1 Formation of the molecular layer
6.4.3.2 Formation of the Purkinje cell layer
6.4.3.3 Formation of the granular layer
6.4.3.4 Formation of the white matter and the cerebellar nuclei
6.4.3.5 Mechanisms underlying the control of neuronal migration
6.4.3.6 The deficits of neuronal migration by exposure to toxic substances and natural environmental factors result in abnormal O-I ...
References
7 - Patterning and generation of neural diversity in the spinal cord
7.1 Introduction
7.2 Spatial signals and the generation of neuronal diversity
7.2.1 Dorsoventral patterning and the induction of progenitor domains
7.2.1.1 Induction of neural progenitor ventral fate: Shh signaling
7.2.1.2 Induction of dorsal progenitor fate: Bmp and Wnt signaling
7.2.2 Rostrocaudal patterning and regional identity
7.2.2.1 Rostrocaudal antiparallel signaling
7.2.2.2 Hox function in neuronal diversity
7.3 Transcription factor combinatorial codes
7.3.1 Transcriptional codes in spinal cord progenitor fate
7.3.2 Transcription factor combinatorial codes in the diversification of postmitotic motor neurons
7.3.3 Transcriptional signatures in spinal cord interneuron diversification
7.4 Local signals and cell-cell interactions
7.4.1 The role of notchdelta signaling in interneuron and motor neuron subtype specification
7.4.2 Retinoid signaling in motor neuron subtype specification
7.5 Temporal signals in the specification of spinal cord glia
7.5.1 Specification of oligodendrocytes
7.5.2 Astrogenesis in the spinal cord
7.6 Application of spinal cord developmental programs to advance therapies for human diseases
7.7 Conclusions
List of abbreviations
Glossary
References
8 - Formation and maturation of neuromuscular junctions
8.1 Introduction
8.2 The neuromuscular junction is comprised of three cell types
8.3 Origin and initial interaction among cells that form the neuromuscular junction
8.4 Formation of a differentiated postsynaptic membrane: the agrin-MuSK hypothesis
8.5 Interplay between agrin and ACh in sculpting the postsynaptic region
8.6 Molecules involved in nAChR prepatterning
8.7 Additional molecules important for clustering and stabilizing developing neuromuscular junctions
8.8 Synapse elimination at the neuromuscular junction
8.9 Synapse elimination: structural and functional changes at neuromuscular junctions
8.10 Synapse elimination: activity-dependent competition and molecular mechanisms
8.11 Synapse elimination: role of T/PSCs
8.12 Maturation and maintenance of neuromuscular junctions
8.13 Summary
List of abbreviations
References
9 - Neural induction of embryonic stem/induced pluripotent stem cells
9.1 Introduction
9.2 Introduction to embryonic stem cells and induced pluripotent stem cells
9.2.1 Reprogramming
9.2.2 Discovery of induced pluripotent stem cells
9.3 Neural induction
9.4 Patterning of neural progenitors
9.4.1 Neuronal progenitor specification along the D-V axis
9.4.2 Neuronal progenitor specification along the A-P axis
9.4.3 Patterning using multiple morphogens gradients
9.4.4 Temporal patterning
9.5 Differentiation to specific regional identities
9.5.1 Differentiation to forebrain cell types
9.5.1.1 Cerebral cortex
9.5.1.2 Hippocampus
9.5.1.3 Basal ganglia
9.5.2 Differentiation to midbrain cell types
9.5.3 Differentiation to hindbrain cell types
9.5.4 Differentiation to spinal cord cell types
9.6 Differentiation to neural crest stem cells
9.7 Differentiation to astrocytes and oligodendrocytes
9.7.1 Astrocytes
9.7.2 Oligodendrocytes
9.8 Direct conversion of fibroblasts to induced neurons
9.9 Conclusion
Acknowledgment
References
10 - Brain organoids as a model system for human neurodevelopment in health and disease
10.1 Recapitulation of in vivo neurodevelopment
10.1.1 Stage I: Neural induction and patterning
10.1.2 Stage II: Lumen formation and apical-basal polarity
10.1.3 Stage III: Proliferation of neural progenitors, interkinetic nuclear motion, and cortical expansion
10.1.4 Stage IV: Neurogenesis, cortical layers formation, and neuronal migration
10.1.5 Stage V: Neuronal maturation and network activity
10.1.6 Evolutionary neurodevelopmental biology in organoids
10.2 Organoids for neurodevelopmental disease modeling
10.2.1 Modeling diseases associated with brain structure
10.2.1.1 Microcephaly (small brains)-genetic mutations
10.2.1.2 Microcephaly-ZIKA virus, mechanisms, and potential therapies
10.2.1.3 Macrocephaly (large brains)
10.2.1.3.1 Lissencephaly (smooth brain)
10.2.2 Modeling of neuropsychiatric disorders
10.2.2.1 Autism spectrum disorders and schizophrenia
Acknowledgments
References
11 - Formation of gyri and sulci
11.1 Introduction
11.2 Timing of the formation of gyri and sulci
11.3 Cortical folding in evolution
11.4 Cellular mechanisms of cortical folding
11.4.1 Outer subventricular zone and basal progenitors
11.4.2 Gene expression profiles
11.4.3 Human- and primate-specific genes
11.4.4 Differential growth and proliferation
11.4.4.1 Cell cycle and the length of the neurogenic period
11.4.4.2 Growth patterns
11.4.4.3 Migration and cell adhesion
11.5 Mechanical mechanisms
11.6 Model systems in which to study cortical folding
11.6.1 Cerebral organoids
11.6.2 Ferret
11.6.3 Nonhuman primates
11.6.4 Human fetal tissue
11.7 Neurodevelopmental disorders
11.7.1 Lissencephaly
11.7.2 Polymicrogyria
11.7.3 Other folding disorders
11.8 Conclusions
Acknowledgments
References
Part II: Generation of neuronal diversity
12 - Cell biology of neuronal progenitor cells
12.1 Introduction
12.2 Location of neuronal progenitors
12.2.1 Multipotent progenitor cells in the ventricular zone generate CNS neurons
12.2.1.1 Neuroepithelial cells
12.2.1.2 Radial glia are neuronal progenitor cells
12.2.2 Neuronal progenitor cells in the subventricular zone
12.2.3 Other non-VZ/SVZ neuronal progenitor cells
12.2.3.1 The dentate gyrus
12.2.3.2 The external granule layer in the cerebellum
12.2.3.3 The retina
12.2.4 The peripheral nervous system
12.2.5 Adult neurogenesis
12.3 Creating different types of neuronal progenitor cells
12.3.1 Neuronal progenitor diversification begins with a regional address
12.3.2 Neuronal progenitor cells are specified temporally
12.3.2.1 Temporal order of neuron generation in the cerebral cortex
12.3.3 Molecular heterogeneity in neuronal progenitor cells
12.4 Cell lineage analysis reveals the fate of individual neuronal progenitor cells
12.4.1 Leading the way: cell lineage analysis in the invertebrate nervous system
12.4.2 Cell lineage analysis in the mammalian nervous system
12.4.3 Lineage analysis, the movie
12.5 Structure and dynamism of neuronal progenitor cells
12.5.1 Interkinetic nuclear migration
12.5.2 Nuclear movement of non-APCs progenitor cells
12.5.3 The structure of radial glia cells
12.5.3.1 Apical-basal processes
12.5.3.2 Adherens junctions
12.5.3.3 Gap junctions
12.5.4 Morphological transitions of neural progenitor cells
12.6 Asymmetric cell division for neuronal diversity
12.6.1 Establishing cell polarity and mitotic spindle orientation
12.6.2 Spindle orientation and cell fate
12.6.3 Asymmetric segregation of the centrosome and the primary cilium membrane
12.6.4 Asymmetric inheritance of the midbody
12.6.5 Asymmetric localization of cell fate determinants
12.7 Progenitor microenvironment and regulating neuronal progenitor number
12.7.1 Fgfs regulate brain size
12.7.2 Shh and cerebellar granule neuron generation
12.7.3 β-Catenin and Wnt pathway
12.7.4 Apoptosis
12.8 Summary
Acknowledgments
References
13 - Notch and neural development
13.1 History of Notch signaling
13.2 Molecular mechanisms
13.2.1 Notch pathway components
13.2.2 Ligand activation of the Notch receptor
13.2.3 Notch and the balancing act
13.3 Signaling diversity and cis-inhibition
13.4 Timing and feedback are everything
13.5 Notch and the maintenance of neural stem cells during nervous system development
13.6 Notch and the generation of interneuron diversity
13.7 Postnatal neurogenesis and gliogenesis
13.8 Notch, glial cell fate, and maturation
13.9 Notch and neuronal migration
13.10 Notch and dendrite morphogenesis
13.11 Synaptic plasticity and Notch signaling
13.12 Embryonic stem cells and clinical perspectives
13.13 Conclusion
References
14 - bHLH factors in neurogenesis and neuronal subtype specification
14.1 Overview of review content
14.2 Identification of neural bHLH transcription factors: History and evolutionary conservation between fly and mammal
14.2.1 The proneural bHLH factors
14.2.2 The E-proteins: heterodimeric partners for proneural bHLH factors
14.2.3 HES, HEY, and ID bHLH factors: inhibitors of neural differentiation
14.3 bHLH factor function in neuronal differentiation
14.3.1 Interplay between notch and proneural bHLH proteins
14.3.2 Refinements in the models for transition from progenitor to differentiated neuron
14.4 Functions of bHLH transcription factors in neuronal subtype specification
14.5 Molecular characteristics of bHLH transcription factors
14.5.1 Crystal structure of bHLH proteins: DNA recognition and dimer selectivity
14.5.2 Structure function analysis of proneural bHLH proteins
14.6 Protein-Protein interactions modulating cell type-specific functions of neural bHLH factors
14.7 Transcriptional targets of proneural bHLH factors
14.8 Transcriptional regulation of bHLH gene expression
14.9 Posttranslational control of neural bHLH transcription factor function
14.10 Reprogramming activities of proneural bHLH factors
14.11 Perspective
References
15 - The specification and generation of neurons in the ventral spinal cord
15.1 Introduction and general organization
15.2 Induction of spinal cord tissue and initiation of regional pattern
15.2.1 The emergence and organization of cell subtypes in the ventral spinal cord
15.2.2 Shh signaling and ventral cell fate specification
15.2.3 Transcriptional control of progenitor gene expression
15.2.4 Additional signaling influences over progenitor gene expression patterns
15.3 Spinal cord neurogenesis
15.3.1 Control of cell cycle progression and exit in neuronal progenitors
15.3.2 Coordination of cell fate and neurogenesis
15.4 The generation of differentiated neuronal cell subtypes
15.4.1 Motor neuron axial subclass specification: rostral-caudal patterning of the spinal cord influences cell fate within a dorsa ...
15.4.2 Genetic programs in postmitotic cells
15.4.3 Motor neuron subclass diversification
15.4.4 Correlation between cell fate and locomotor circuits
References
16 - Neurogenesis in the cerebellum
16.1 Introduction to the cerebellum
16.2 Overview of cerebellar development
16.3 Establishing the cerebellar territory
16.3.1 Establishing the cerebellar territory along the anterior-posterior axis: the isthmic organizer
16.3.2 Establishing the cerebellar territory along the dorsal-ventral axis
16.4 The cerebellar ventricular zone and its derivatives
16.4.1 Ventricular zone development and neurogenesis in ventricular zone
16.4.2 Molecular mechanisms that regulate the differentiation and migration of Purkinje cells and GABAergic neurons of CN
16.4.3 Molecular mechanisms that regulate development of PWM and GABAergic interneurons
16.5 The cerebellar rhombic lip and its derivatives
16.5.1 Rhombic lip induction and neurogenesis within the rhombic lip
16.5.2 Regulation of granule cell development
16.5.2.1 Regulation of tangential migration of granule neuron precursors from the rhombic lip
16.5.2.2 Regulation of proliferation and differentiation of GNPs in the EGL
16.5.2.3 Regulation of radial migration of granule cells from the EGL to the IGL
16.5.3 Regulation of differentiation and migration of glutamatergic neurons of CN and UBCs
16.6 Cerebellar stem cells and regeneration of the cerebellum
16.7 Conclusions and future perspectives
References
17 - The generation of midbrain dopaminergic neurons
17.1 Introduction
17.1.1 Dopamine
17.1.2 Dopamine system in the brain
17.1.2.1 Midbrain dopamine neurons-anatomically defined groups
17.1.2.2 Midbrain dopamine neurons-groups defined by molecular profiles
17.2 The development of midbrain dopaminergic neurons-general overview
17.3 Generation of midbrain dopaminergic progenitors: patterning, specification, and proliferation
17.3.1 Patterning
17.3.2 Specification and proliferation
17.3.2.1 The role of signaling centers and secreted factors
17.3.2.2 The role of transcription factors
17.3.2.3 Diversity in midbrain dopaminergic progenitors
17.4 Generation of immature and mature midbrain dopaminergic neurons
17.4.1 Regulation of maturation
17.4.2 Migration of midbrain dopaminergic neurons
17.4.3 Axonal pathfinding of midbrain dopaminergic neurons
17.5 The terminal differentiation of the mature dopaminergic neuron
17.6 Maintenance of midbrain dopaminergic neurons
17.7 Perspectives
References
18 - Neurogenesis in the basal ganglia
18.1 Introduction
18.2 Organization of embryonic subdivisions and their relationship to mature structures and cell types
18.2.1 Subdivisions of the mature and embryonic basal ganglia
18.2.2 Cellular organization of the developing basal ganglia
18.2.3 Fate analysis of the GEs and their subdivisions
18.3 Regional specification of subdivisions of the embryonic basal ganglia
18.3.1 Morphogen and growth/differentiation factor signaling in the developing basal ganglia
18.3.1.1 Shh signaling
18.3.1.2 Receptor tyrosine kinase signaling
18.3.1.3 Wnt signaling
18.3.1.4 Tgf-β signaling
18.3.1.5 Retinoid signaling
18.3.1.6 Notch signaling
18.3.2 Basal ganglia specification
18.3.3 LGE and CGE specification
18.3.4 MGE and POA specification
18.3.5 Septum specification
18.4 Generation of neuronal subtypes
18.4.1 LGE and CGE neuronal derivatives
18.4.1.1 Medium-sized striatal projection neurons
18.4.1.2 Olfactory bulb interneurons
18.4.1.3 Cortical and amygdalar interneurons
18.4.2 MGE and POA neuronal derivatives
18.4.2.1 Globus pallidus projection neurons
18.4.2.2 Striatal interneurons
18.4.2.3 Cortical interneurons
18.4.3 Cis-regulatory elements and epigenetics of basal ganglia development
18.4.4 Engineering basal ganglia neurons in vitro
18.5 Summary
References
19 - Specification of cortical projection neurons: transcriptional mechanisms
19.1 Introduction
19.2 Neocortical progenitors
19.3 Neocortical progenitor cell-fate acquisition and plasticity
19.4 Molecular controls over neocortical projection neuron subtype specification, development, and diversity
19.4.1 Subtype specification of corticofugal projection neurons
19.4.2 Subtype specification of callosal projection neurons
19.4.3 Areal controls over diversity of neocortical projection neuron subtypes
19.5 Progressive restriction and refinement of cortical projection neuron subtypes
19.6 Generation of cortical projection neuron subtypes in vitro from human pluripotent stem cells
19.7 Subtype-specific circuit wiring by growth cones
19.8 Conclusions
References
20 - The generation of cortical interneurons
20.1 Diversity of mature cortical interneurons
20.1.1 Parvalbumin interneurons
20.1.2 Somatostatin interneurons
20.1.3 Vasoactive intestinal peptide interneurons
20.1.4 Lamp5 interneurons
20.1.5 Gamma-synuclein and Serpinf1 interneurons
20.2 Developmental origin of cortical interneurons
20.2.1 The ventral origin of cortical neurons
20.2.2 Genetic determinants involved in the specification of the MGE and CGE
20.2.3 Place and time of origins of cortical interneurons
20.2.4 Fate mapping strategies to assess the origin of cortical interneurons
20.2.5 Genetic programs underlying the developmental emergence of interneurons
20.3 Migration of cortical interneurons
20.3.1 The influence of non-cell-autonomous signals on interneurons development
20.4 Postnatal cortical interneuron development
20.4.1 GABA is depolarizing during development
20.4.2 Early patterns of network activity
20.4.3 Role of activity in interneuron development
20.4.4 Interneuron development and neurological disorders
Acknowledgments
References
21 - Specification of retinal cell types
21.1 Introduction
21.2 Retinal progenitor cell competence
21.2.1 Establishment of retinal neuron and Müller glia birth order
21.2.2 Clonal analyses in the developing retina
21.2.3 Intrinsic versus extrinsic control of neurogenesis in the mammalian retina
21.3 Intrinsic regulation of retinal development
21.3.1 Early eye formation
21.3.2 Retinal neurogenesis
21.3.3 Intrinsic factor regulation of RGC development
21.3.4 Intrinsic factors regulating photoreceptor development
21.3.5 Epigenetic control of retinogenesis
21.3.6 MicroRNA-mediated regulation of retinal genes
21.4 Extrinsic regulation of retinogenesis
21.4.1 Bmp/Tgfβ superfamily signaling
21.4.2 Fgf signaling
21.4.3 Notch signaling
21.4.4 Retinoic acid signaling
21.4.5 Hh signaling
21.4.6 Wnt/β-catenin signaling
21.5 Regenerative capacity of the retina
21.6 Perspective
Glossary
References
22 - Neurogenesis in the postnatal V-SVZ and the origin of interneuron diversity
22.1 Newborn neurons are generated in the V-SVZ of the adult brain
22.2 Identification and origin of adult neural stem cells
22.3 OB interneurons are heterogeneous
22.4 Spatial specification of OB interneuron identity
22.5 Temporal regulation of OB interneuron production
22.6 Conclusion
Acknowledgments
References
23 - Neurogenesis in the damaged mammalian brain
23.1 Introduction
23.2 Persistent versus injury-induced neurogenesis in the adult brain
23.2.1 Neurogenesis in the intact brain
23.2.1.1 Active neurogenic regions
23.2.1.2 Common and distinct features of adult neurogenic niches
23.2.1.3 Cryptic or less active neurogenic regions
23.3 Neurogenesis in the injured brain
23.3.1 Stimulation of ongoing neurogenesis after damage
23.3.2 Ectopic production of new neurons and glia in damaged brains
23.3.2.1 Acute central nervous system injury
23.3.2.1.1 Neocortex
23.3.2.1.2 Striatum
23.3.2.1.3 Hippocampus
23.3.2.1.4 Substantia nigra
23.3.2.1.5 Spinal cord
23.3.2.1.6 Retina
23.3.2.1.7 Other regions of the central nervous system
23.3.2.2 Neurogenesis in chronic neurodegenerative conditions
23.3.2.2.1 Alzheimer's disease
23.3.2.2.2 Huntington disease
23.3.2.2.3 Other neurodegenerative disorders
23.4 Identity, integration, and extent of regeneration of new neurons
23.4.1 Neocortex and hippocampus
23.4.2 Striatum
23.4.3 Other regions of the central nervous system
23.5 Contribution of injury-induced neurogenesis to functional recovery
23.5.1 Attenuation of neurogenesis
23.5.2 Enhancement of neurogenesis
23.5.3 Just a correlation or the cause?
23.6 How widespread is injury-induced neurogenesis?: technical issues
23.7 Cellular origins of injury-induced neurogenesis
23.7.1 Contribution of NPCs in known neurogenic niches
23.7.2 Identity of cells that generate new neurons
23.7.3 Possible cellular sources outside neurogenic niches
23.8 Gliogenesis after injury
23.8.1 Oligodendrogenesis
23.8.2 Astrogenesis
23.9 Mechanisms underlying injury-induced neurogenesis
23.9.1 Cell-intrinsic limitation of NPCs
23.9.1.1 Limited number and expansion of NPCs
23.9.1.2 Limited plasticity of NPCs
23.9.1.3 Intrinsic fate determinants of NPCs
23.9.1.3.1 Maintenance and proliferation of NSCs
23.9.1.3.2 Differentiation of NSCs
23.9.1.3.3 Neuronal subtype specification
23.9.2 Environmental restrictions
23.9.2.1 Growth factors
23.9.2.2 Differentiation factors
23.9.2.3 Migratory cues
23.9.2.4 Survival and maturation signals
23.9.2.5 Inflammatory and immune signals
23.9.2.6 Neurotransmitter signals
23.9.2.6.1 Glutamate and GABA
23.9.2.6.2 Dopamine
23.9.2.6.3 Serotonin
23.9.2.6.4 Neuropeptides and other neurotransmitters
23.9.2.6.5 Specific neuronal populations
23.9.2.7 Hormones
23.9.2.8 Other signals
23.9.2.8.1 Nitric oxide
23.9.2.8.2 Lipid mediators
23.9.2.8.3 Cell grafts
23.10 Neuronal cell reprogramming
23.11 Link between neurodegeneration and neurogenesis
23.12 Neurovascular niche
23.13 Nonneurogenic roles of adult NPCs in brain repair
23.14 Future perspectives
Acknowledgments
References
24 - Neuronal identity specification in the nematode Caenorhabditis elegans
24.1 Introduction
24.2 Neuron classification
24.3 Neuronal cell lineages
24.4 Genes controlling lineage decisions
24.4.1 Neuronal versus nonneuronal lineage transformations
24.4.2 Neuron lineage alterations and losses
24.5 Terminal selectors control neuron class specification
24.6 Genes controlling neuron subclass diversification
24.6.1 Diversifying motor neuron classes
24.6.2 Neuronal identity diversification across the left/right axis
24.7 Other regulatory routines operating during neuronal differentiation
24.8 Linking neuronal class specification to lineage
24.9 Concluding remarks
Acknowledgments
References
25 - Development of the Drosophila melanogaster embryonic CNS: from neuroectoderm to unique neurons and glia
25.1 Introduction
25.2 Patterning of the neuroectoderm: breaking the homogeneity
25.2.1 Patterning the ventral neuroectoderm
25.2.2 Patterning the brain neuroectoderm
25.3 Homologous neuromeres: same but different
25.4 The chosen one: lateral inhibition
25.4.1 Delamination of VNC neuroblasts
25.4.2 Delamination of brain neuroblasts
25.5 Unequal legacy: asymmetric cell division
25.6 One thing at a time: the temporal cascade
25.7 Regulation of neuroblast and daughter cell proliferation
25.7.1 NB cell cycle exit and daughter cell proliferation switches: the role of cell cycle genes
25.7.2 NB cell cycle exit and daughter cell proliferation switches: the role of late temporal and Hox genes
25.7.3 NB exit and daughter cell proliferation switches: the role of the Notch pathway
25.7.4 NB exit and daughter cell proliferation switches: the role of early temporal and pan-neural genes
25.7.5 Brain-specific NB behavior: type II NBs
25.7.6 Brain-specific NB behavior: mushroom body and IPC NBs
25.8 The role of programmed cell death in the Drosophila embryonic VNC
25.9 Finishing the picture: specification of unique cell types
25.9.1 Specifying brain cells
25.9.2 Specifying VNC neuropeptide cells
25.9.3 Specifying motor neurons
25.9.4 Specifying midline neurons
25.9.5 Specifying glia cells
25.9.5.1 Specifying lateral glia cells
25.9.5.2 Specifying midline glia cells
25.10 Conclusions
25.11 Outstanding issues
Acknowledgments
References
26 - Neurogenesis in zebrafish
26.1 Neural plate induction and patterning
26.1.1 Formation of the neural tube
26.1.2 Neural plate induction
26.1.3 Neural plate patterning along the anteroposterior axis
26.2 Establishment of the primary neuronal scaffold
26.2.1 Organization of the primary neuronal scaffold
26.2.2 Formation of the primary neuronal scaffold
26.2.2.1 Identification of competent proneural domains within the neural plate
26.2.2.2 Neurogenesis control within the proneural clusters
26.2.2.2.1 Lateral inhibition in Drosophila
26.2.2.2.2 Lateral inhibition in vertebrates
26.2.2.2.3 Regulation of notch signaling
26.2.2.3 Determination of primary neuronal identities
26.2.2.3.1 Morphogens
26.2.2.3.2 Notch signaling
26.3 Secondary neurogenesis
26.3.1 Functional anatomy of secondary neurogenesis
26.3.1.1 Motor and sensory systems
26.3.1.2 Neuromodulatory, neurohormone, and neuropeptide systems
26.3.2 Molecular and cellular mechanisms of secondary neurogenesis
26.3.2.1 Secondary neurogenesis: balance between proliferation and differentiation
26.3.2.2 Neuroblast migration
26.3.2.2.1 Facial branchiomotor neurons migration
26.3.2.2.2 Migration of precursor cells in the cerebellum
26.3.2.3 Neuronal subtype specification
26.3.2.3.1 Specification of subtypes in the spinal cord
26.3.2.3.2 Neuromodulatory systems
26.3.2.3.2.1 DA neurons
26.3.2.3.2.2 NA neurons
26.3.2.3.2.3 5-HT and HA neurons
26.3.2.3.2.4 Diencephalic/hypothalamic neurohormones and neuropeptides
26.4 Adult neurogenesis and plasticity
26.4.1 Anatomy of adult neurogenesis
26.4.1.1 Neurogenesis domains
26.4.1.2 Influence of physiological parameters on neurogenic activity
26.4.2 Molecular and cellular mechanisms of adult neurogenesis
26.4.2.1 Localization, identity, and properties of adult progenitor cells
26.4.2.1.1 NSCs in the adult telencephalon: markers and lineages
26.4.2.1.1.1 Continuous lineages from embryo to adult contribute to generate an ``ordered'' pallial structure
26.4.2.1.1.2 Changes in neurogenesis with aging
26.4.2.1.2 NSCs at the adult MHB: markers and lineages
26.4.2.1.3 NSCs in the adult cerebellum: markers, lineage
26.4.2.2 Molecular pathways of adult neural progenitor maintenance and recruitment
26.4.2.2.1 Notch
26.4.2.2.2 microRNA-9
26.4.2.2.3 Fezf2
26.4.2.2.4 Fgf
26.4.2.2.5 Steroids
26.4.2.2.6 BDNF
26.4.2.2.7 Id (inhibitor of DNA binding)
26.4.2.3 Adult neurogenesis and plasticity upon brain or spinal injury
26.4.2.3.1 Neurogenesis and regeneration in the telencephalon
26.4.2.3.2 Neurogenesis and regeneration in the diencephalon (DA neurons)
26.4.2.3.3 Neurogenesis and regeneration in the optic tectum
26.4.2.3.4 Neurogenesis and regeneration in the cerebellum
26.4.2.3.5 Neurogenesis and regeneration in the spinal cord
References
27 - Gene regulatory networks controlling neuronal development: enhancers, epigenetics, and functional RNA
27.1 Introduction-genomic control of cell identity in the brain
27.2 Overview of gene regulation and the control of neuronal diversity
27.3 Interactions between transcription factors, regulatory DNA, and epigenetics
27.4 Enhancers
27.4.1 Mapping and functional prediction of enhancers in the brain
27.4.2 Enhancer activity in brain development
27.4.3 Combinatorial enhancer binding of transcription factors activates or represses
27.4.4 Comparative genomics-evolutionary conservation and novelty of brain enhancers
27.4.5 Example: ARX expression is regulated by coordinated activity of distal enhancers
27.4.6 Role of enhancer variation in neurodevelopmental and psychiatric disorders
27.4.7 Current questions regarding enhancer function
27.5 Epigenetics
27.5.1 How chromatin state contributes to gene regulation
27.5.2 Functional genome annotation
27.5.2.1 DNA methylation
27.5.2.2 Histone modification
27.5.2.3 Chromatin accessibility
27.5.3 Lineage specification and chromatin in the brain
27.5.4 Interaction between transcription factors and chromatin
27.5.5 Role of chromatin remodelers in neurodevelopmental disorders
27.5.6 Current questions regarding epigenetics
27.6 Regulatory RNA in brain development
27.6.1 Functional RNA: miRNA, lncRNA, eRNA
27.6.2 miRNA: a brief overview
27.6.3 lncRNA-evidence for function
27.6.4 eRNA-transcriptional artifacts or functional molecules?
27.6.5 Current questions regarding functional RNA
27.7 Putting it all together-gene regulatory networks
27.7.1 Example: Nkx2-1 in the basal ganglia
27.8 Conclusion
References
28 - Posttranscriptional and translational control of neurogenesis: roles for RNA-binding proteins
28.1 Introduction
28.1.1 Neurogenesis
28.1.2 Posttranscriptional regulation
28.2 Alternative splicing
28.2.1 Global and dynamic splicing patterns
28.2.2 Trans-regulators of splicing
28.2.3 Summary I
28.3 From nucleus to cytoplasm
28.3.1 The exon junction complex
28.3.2 Nonsense-mediated decay
28.3.3 Summary II
28.4 Translational control
28.4.1 Core translational machinery
28.4.2 The elavl family members
28.4.3 RNA localization, transport, and translation
28.4.4 Summary III
28.5 The epitranscriptome
28.5.1 Readers and writers
28.5.2 Summary IV
28.6 Perspectives
References
29 - Human neurogenesis: single-cell sequencing and in vitro modeling
29.1 Introduction
29.2 Single-cell sequencing modalities
29.2.1 Whole-cell RNA-sequencing to identify molecular signatures of known and novel cell types
29.2.2 Nuclei sequencing to discover novel human cell types
29.2.3 Multimodal integration of transcriptomic, morphologic, and physiologic features highlights functional significance of cellu ...
29.2.4 ATAC-seq, methylation, and other measures of chromatin state
29.2.5 Other modalities
29.2.6 In situ sequencing and other imaging strategies
29.3 Overview of analytical approaches and strategies
29.3.1 Clustering and basic analysis strategies
29.3.2 Approaches to lineage reconstruction
29.3.2.1 In vitro modeling of human neurogenesis
29.4 Cell culture strategies
29.4.1 Stem cells and reprogramming
29.4.2 Adherent culture systems
29.4.3 Brain organoid models
29.5 Modeling development in organoids
29.5.1 Regionalization
29.5.2 Timing of maturation compared to normal development
29.5.3 Developmental trajectories and neuronal differentiation
29.5.4 Cellular diversity
29.5.5 Architectonics
29.5.6 Cellular dynamics and migration
29.5.7 Reproducibility
29.6 Regional interactions
29.6.1 Whole brain organoids
29.6.2 Organoid fusing
29.7 Functional activity
29.7.1 Modeling circuits
29.7.2 Single-cell analysis of in vitro cerebral organoid models
29.7.3 Organoid models to study human evolution
29.8 Disease phenotypes
29.9 Engineering organoids
29.10 Conclusion
References
Part III: Development of glia, blood vessels, choroid plexus, immune cells in the nervous system
30 - A golden age for glial biology
30.1 Overview
30.2 Brief summary of section chapters
30.2.1 Chapters 31-33: neural stem cells and astrocytes
30.2.2 Chapters 34-40: myelinating cells
30.2.3 Chapters 41-43: microglia, ependyma, perivascular cells, and meninges
30.3 Conclusion
31 - Neural stem cells among glia
31.1 Introduction
31.2 NSCs among glia in the developing brain
31.2.1 Neuroepithelial cells
31.2.2 Radial glia
31.2.3 Intermediate (basal) progenitor cells
31.2.4 Outer radial glia
31.3 Molecular regulation of progenitor proliferation, cell fate, and polarity
31.3.1 Mapping progenitor cell fates
31.3.2 Role of apical-basal polarity in progenitors
31.3.2.1 Regulation at the apical surface
31.3.2.2 Role of the basal process
31.3.3 New models of molecular regulation in progenitors
31.4 NSCs among glia in the postnatal brain
31.4.1 RG persist after birth and function as NSCs in some vertebrates
31.4.2 NSCs (Type B1 cells) in the adult mammalian V-SVZ
31.4.3 NSCs (radial astrocytes) in the adult hippocampus
31.4.4 Regulation of adult NSCs
31.5 Link between embryonic and adult glial cells that function as NSCs
31.6 Origin of oligodendrocytes from RG and adult V-SVZ astrocytes
31.7 Evolutionary perspective
31.8 Perspective for brain repair
31.9 Conclusion
Acknowledgments
References
32 - Mechanisms of astrocyte development
32.1 Introduction
32.1.1 Overview of astrocyte function in the central nervous system
32.1.2 Why is the study of astrocytes uniquely challenging?
32.1.2.1 Interspecies differences in astrocyte developmental lineages
32.1.2.2 The absence of a clear developmental endpoint
32.1.2.3 The lack of molecular tools
32.1.3 Overview of the chapter
32.2 The origins of astrocytes
32.2.1 Use of in vitro culture methods to generate astrocytes
32.2.2 Use of induced pluripotent stem cell technology to generate astrocytes in vitro
32.2.3 Molecular mechanisms of astrocyte specification and initiation
32.2.3.1 1996-99: Role of signaling molecules
32.2.3.2 1996-99: Suppression of astrocyte fate and epigenetic states
32.2.3.3 2000-04: Discovery of the role of Notch signaling to promote astrocytes
32.2.3.4 2005: Feedback mechanisms controlling astrocyte fate
32.2.3.5 2006: Discovery of NFIA, which controls the neuron-glia switch
32.2.3.5.1 2009: NFIA also promotes differentiation of astrocytes, after the neuron-glia switch
32.2.3.5.2 2012: Relationship of NFIA with transcription factor Sox9
32.2.3.5.3 2014: Relationship of NFIA with transcription factors Sox10 and Olig2 to control oligodendrocyte fate
32.2.3.6 2006-present: discoveries of other pathways, transcription factors, and mechanisms of astrocyte fate determination
32.2.3.6.1 Receptors and signaling pathways: ErbB4 and MEK/ERK pathway
32.2.3.6.2 Transcription factors: Coup-TFI, Lhx2, and Zbtb20
32.2.3.6.3 Epigenetic controls: Hdac3 in the astrocyte-oligodendrocyte fate decision and the role of chromatin loops
32.2.4 Patterning of the neural tube and astrocytes
32.2.4.1 Are astrocytes patterned?
32.2.4.2 Patterning as a mechanism to generate astrocyte diversity
32.3 Mechanisms of astrocyte differentiation
32.3.1 The search for stage-specific and subtype-specific pan-astrocytic markers
32.3.1.1 Classical markers of astrocytes
32.3.1.2 Newly identified transcription factors as astrocyte markers
32.3.1.3 Functional proteins as mature astrocyte markers
32.3.1.4 Emerging astrocyte markers based on transcriptional profiling
32.3.2 Defining the intermediate phases of astrocyte lineage trajectory
32.3.2.1 Directionality of astrocyte migration from the subventricular zone
32.3.2.2 Location of astrocyte precursor proliferation
32.3.2.3 Molecular regulation of the intermediate phases of astrocyte development
32.4 Morphologic and functional maturation of astrocytes
32.4.1 Morphologic maturation of astrocytes
32.4.2 Functional maturation of astrocytes
32.4.2.1 Lessons from the fly about neuron-glia interactions
32.4.2.2 Neuronal activity sculpts astrocyte maturation
32.5 The development of astrocyte diversity
32.5.1 Morphological diversity across the adult central nervous system
32.5.2 Regional and functional diversity across the adult central nervous system
32.5.3 Does regional diversity control function of spatially separated astrocytes?
32.5.4 Local diversity at specific regions and their contribution to astrocyte function
32.5.5 Other aspects of astrocyte diversity
32.6 Conclusions and future directions
References
33 - Astrocyte-neuron interactions in synaptic development
33.1 Developmental stages of synapse formation and maturation
33.2 Role of astrocytes in synaptic development
33.2.1 Contact-mediated astrocyte synaptogenic signals
33.2.1.1 Integrin-protein kinase C
33.2.1.2 Neurexin
33.2.1.3 Gamma protocadherins
33.2.1.4 Neuroligins
33.2.1.5 Eph/ephrin
33.2.2 Astrocyte-secreted synapse-regulating signals
33.2.2.1 Synapse number
33.2.2.1.1 Thrombospondin
33.2.2.1.2 Sparcl1
33.2.2.1.3 Transforming growth factor beta
33.2.2.2 Presynaptic function
33.2.2.2.1 Cholesterol and lipid metabolism
33.2.2.3 Postsynaptic function
33.2.2.3.1 Glypicans
33.2.2.4 Tumor necrosis factor alpha
33.2.2.4.1 Chordin-like 1
33.2.2.4.2 Chondroitin sulfate proteoglycans
33.2.2.5 Negative synaptic regulators
33.2.2.5.1 SPARC
33.2.2.6 Additional astrocyte-derived signals
33.2.2.7 Inhibitory synapses
33.2.3 Astrocyte elimination of synapses
33.3 Region, temporal, and neuronal regulation of astrocyte synaptogenic cues
33.3.1 Regional heterogeneity of astrocyte synaptogenic gene expression
33.3.2 Temporal changes in astrocyte synaptogenic gene expression
33.3.3 Neuronal regulation of synaptogenic cue expression in astrocytes
33.4 Conclusion
References
34 - Specification of oligodendrocytes
34.1 Introduction
34.2 Determinants of oligodendroglial fate
34.3 Determinants of oligodendroglial identity
34.4 Determinants of progenitor state maintenance
34.5 Determinants of progression from the progenitor state
34.6 Determinants of terminal differentiation and the fully differentiated state
34.7 Concluding remarks perspectives
References
35 - Signaling pathways that regulate glial development and early migration-oligodendrocytes
35.1 Introduction
35.2 Signaling pathways regulating the initial appearance of oligodendrocyte precursors
35.2.1 Timing and localization of appearance of OPCs
35.2.2 Molecular control of early OPC appearance
35.2.2.1 Sonic hedgehog
35.2.2.2 Bone morphogenetic proteins
35.2.2.3 Wnts
35.2.2.4 Neuregulin
35.2.2.5 FGF
35.3 Regulation of OPC migration
35.3.1 Mechanisms of OPC dispersal: engagement of the vasculature
35.3.2 Molecular guidance of OPC dispersal
35.3.2.1 Netrins
35.3.2.2 Semaphorins
35.3.3 Molecular control of OPC motility
35.3.3.1 Growth factors
35.3.3.2 Neurotransmitters and channels
35.3.3.3 Chemokines
35.3.4 Signals regulating the final localization of oligodendrocytes
35.3.4.1 CXCL1
35.3.4.2 Tenascin C
35.4 Regulation of OPC differentiation
35.4.1 Cell extrinsic regulation of oligodendrocyte differentiation
35.4.1.1 LINGO-1
35.4.1.2 PSA-NCAM
35.4.1.3 Notch/delta
35.4.2 Cell-intrinsic regulators of oligodendrocyte differentiation
35.4.3 Transcriptional regulators of OPC terminal differentiation
35.4.3.1 Negative transcriptional regulators of OPC terminal differentiation
35.4.3.2 Positive regulators of OPC terminal differentiation
35.4.3.3 Intrinsic transcriptional regulation of oligodendrocyte maturation and myelination
35.5 Epigenetic regulation of oligodendrocyte development
35.5.1 ATP-dependent chromatin remodelers
35.5.2 Histone-modifying enzymes
35.5.3 miRNAs in oligodendrocyte development
35.5.4 lncRNAs in oligodendrocyte development
35.6 Conclusions
References
36 - Neuron-glial interactions and neurotransmitter signaling to cells of the oligodendrocyte lineage
36.1 Introduction
36.2 Distinguishing characteristics of OPCs, premyelinating oligodendrocytes, and mature oligodendrocytes
36.2.1 OPC distribution, morphology, and proliferation
36.2.2 Distribution and morphology of premyelinating oligodendrocytes and oligodendrocytes
36.2.3 Physiological properties of oligodendrocyte lineage cells
36.2.4 Transcriptional expression profiles across the oligodendrocyte lineage
36.3 Neurotransmitter signaling within the oligodendrocyte lineage: glutamate
36.3.1 AMPA receptor signaling within oligodendrocyte lineage cells
36.3.2 NMDA receptor signaling within oligodendrocyte lineage cells
36.3.3 Metabotropic glutamate receptors within oligodendrocyte lineage cells
36.3.4 Glutamate receptor expression during progenitor differentiation
36.4 Neurotransmitter signaling within the oligodendrocyte lineage: GABA, acetylcholine, and ATP
36.5 Synaptic signaling between neurons and OPCs
36.5.1 A surprising discovery: evidence for the existence of neuron-OPC synapses
36.5.2 Do neuron-OPC synapses regulate oligodendrogenesis?
36.5.3 Activity-dependent myelination
36.5.4 Additional features of neuron-OPC synapses: signaling functions beyond oligodendrogenesis?
36.6 Oligodendrocyte lineage cells in the context of disease and injury
36.6.1 OPC reactivity and vulnerability of oligodendrocyte lineage cells to pathology
36.6.2 Perinatal hypoxia and ischemia
36.6.3 OPCs and hypomyelination/demyelination
36.6.4 Tumorigenesis and gliomas
36.7 Conclusions/future directions
References
37 - Nonmammalian model systems of zebrafish
37.1 History and attributes of the zebrafish model system
37.1.1 Establishment of a new animal model
37.1.2 The zebrafish toolbox
37.2 Zebrafish glial classification
37.3 Zebrafish oligodendrocyte development
37.3.1 Oligodendrocyte specification
37.3.2 Oligodendrocyte lineage cell migration, proliferation, and differentiation
37.4 Zebrafish peripheral glia
37.4.1 Schwann cells and the zebrafish lateral line system
37.4.2 Genetic control of peripheral glial development
37.4.3 Motor root perineurial cells originate as CNS glia
37.4.4 Glial cell interactions at the CNS-PNS interface
37.5 Zebrafish radial glia
37.6 Zebrafish microglia
37.7 Conclusion
References
38 - Specification of macroglia by transcription factors: Schwann cells
38.1 Introduction
38.2 Specification of Schwann cells from neural crest
38.2.1 Alternate developmental fates of Schwann cell precursors
38.3 Immature Schwann cells: radial sorting and transition to myelination
38.4 Signaling pathways regulating the myelin program
38.4.1 Neuregulin
38.4.2 G protein-coupled receptor 126 signaling
38.4.3 Mitogen-activated protein kinase signaling. ERK1/2
38.4.4 PI-3 kinase and mTOR signaling
38.4.5 Calcium and prostaglandin signaling converging on nuclear factor of activated T-cell (NFAT) transcription factors in Schwan ...
38.4.6 Negative regulators of myelination
38.5 Integration of signaling pathways at myelin genes
38.6 Epigenetic regulation of Schwann cell differentiation
38.7 Reprogramming Schwann cell behavior in pathology
38.8 Conclusion
List of acronyms and abbreviations
References
39 - Signaling pathways that regulate glial development and early migration-Schwann cells
39.1 Introduction
39.1 Overview of Schwann cell development
39.1.1 Schwann cell precursors, the glial cells of early embryonic nerves
39.1.2 Immature Schwann cells
39.1.3 Axonal signals
39.1.4 Boundary cap cells
39.2 Developmental potential and Schwann cell plasticity
39.3 Major differences among migrating neural crest cells, SCP, and iSch
39.4 Gliogenesis from crest cells: the appearance of SCP
39.4.1 HDAC1/2
39.4.2 Sox10
39.4.3 NRG1
39.4.4 Notch
39.5 NRG1 and Notch signaling IN SCP
39.5.1 Survival
39.5.2 Migration
39.5.3 NRG1 on developing axons
39.5.4 NRG1 and Notch interact to promote SCP survival and iSch generation
39.6 Schwann cell generation and the architectural reorganization of peripheral nerves
39.7 SCP and early Schwann cells control neuronal survival, nerve fasciculation, and synapse formation
39.7.1 Neuronal survival
39.7.2 Fasciculation and synapse formation
39.8 Schwann cells in late embryonic and perinatal nerves
39.9 Signals that drive Schwann cell proliferation in vivo
39.9.1 Notch
39.9.2 TGFβ
39.9.3 YAP/TAZ pathway
39.9.4 NRG1
39.9.5 Laminin and GPR126
39.10 Signals that promote Schwann cell death and survival in vivo
39.11 Radial sorting
39.11.1 Laminin and integrins
39.11.2 NRG1
39.11.3 Lgi4
39.11.4 GPR126
39.11.5 Sox10
39.11.6 HDAC1/2
39.11.7 Zeb2
39.11.8 The HIPPO pathway
39.11.9 Jab 1
39.11.10 Wnt/beta-catenin signaling
39.12 The onset of myelination
39.12.1 Positive regulators
39.12.2 The onset of myelination: negative regulators
39.13 Conclusions
Acknowledgments
References
40 - Structure and function of myelinated axons
40.1 Introduction
40.2 Evolution of the myelinated axon
40.2.1 Ion channel clustering in the axon
40.2.2 Myelin-enabling ``wrap-id'' advances in cognition
40.3 Myelinating glial cells and axoglial interactions
40.4 Nodes of Ranvier: structure, composition, and function
40.4.1 Nodes of Ranvier
40.4.2 Paranodal junctions
40.4.3 Juxtaparanodes
40.5 Assembly of nodes of Ranvier
40.5.1 Clustering of Na+ channels at nodes of Ranvier in the PNS
40.5.2 Clustering of Na+ channels at nodes of Ranvier in the CNS
40.6 Long-term maintenance of nodes in the PNS and CNS
40.7 Function of nodes in AP propagation and initiation
40.7.1 Developmental maturation of Na+ channel complexes at nodes of Ranvier
40.7.2 Nodal spacing contributes to neuronal computations
40.7.3 Proximal nodes of Ranvier in determining neuronal firing patterns
40.8 Nodes of Ranvier in nervous system disease and injury
40.8.1 Autoimmune disorders
40.8.2 Developmental neuropsychiatric disorders
40.9 Conclusions and outlook
References
41 - Microglia
41.1 Introduction
41.2 Origin and maintenance of microglia
41.2.1 Developmental origins of microglia
41.2.2 Microglia in different species
41.2.3 Microglia turnover in the adult brain
41.3 Microglia as dynamic cells in the CNS
41.3.1 Challenging the term ``resting'' microglia in the healthy CNS
41.3.2 Microglial responses to localized trauma in vivo
41.4 Microglial activation
41.5 Microglial interactions with other cell types
41.6 Microglia and disease
41.6.1 Microglia in multiple sclerosis
41.6.2 Microglia in stroke
41.6.3 Microglia in Alzheimer's disease
41.6.4 Microglia in neuropathic pain
41.6.5 Single-cell approaches to understand microglia heterogeneity
41.7 Concluding remarks
List of abbreviations
References
42 - Ependyma
42.1 Introduction
42.2 Structure of cells in contact with the ventricles
42.2.1 Structure of multiciliated ependymal cells
42.2.1.1 Structure of tanycytes
42.2.1.2 Structure of other cells in contact with ventricles
42.2.2 Origin and developmental mechanisms
42.2.2.1 Ependymal cell specification
42.2.2.2 Ependymal cell differentiation
42.2.2.3 Ependymal cell maturation
42.2.3 Functions in the brain
42.2.3.1 Ependymal epithelium: interface between brain and CSF
42.2.3.1.1 The ependymal junctions
42.2.3.1.2 A filter for brain-CSF exchange
42.2.3.1.3 A regulator of osmotic pressure
42.2.3.1.4 A barrier against harmful substances
42.2.3.1.5 A regulator of peptide concentrations
42.2.3.2 Trophic and metabolic support by ependymal cells
42.2.3.3 Can ependymal cells function as neural stem cells?
42.2.4 Associated pathologies
42.2.4.1 Ependymoma
42.2.4.2 Hydrocephalus
42.3 Summary
References
43 - Meninges and vasculature
43.1 Meninges in development
43.1.1 Meninges assembly to adult structure: histology and molecular signaling
43.1.1.1 Emergence and maturation of the meningeal fibroblast layers
43.1.1.2 Developmental timeline and function of nonfibroblast cells of the meninges
43.1.2 Meninges-brain interface: signals from the meninges regulate development of the CNS
43.1.2.1 Meningeal Cxcl12 in fore- and hindbrain development
43.1.2.2 Meningeal retinoic acid in forebrain and hindbrain development
43.1.2.3 Meningeal bone morphogenic proteins in forebrain development
43.1.2.4 Meningeal deposition and maintenance of the pial BM
43.1.3 Perspectives on the meninges as an interface between the immune system and the brain
43.2 Development of the CNS vasculature
43.2.1 Timing and molecular mechanisms of CNS angiogenesis
43.2.1.1 Developmental timing of CNS vascularization
43.2.1.2 VEGF ligands regulate CNS vascular growth and patterning
43.2.1.3 Endothelial Wnt-β-catenin signaling is CNS vascular development
43.2.1.4 Integrin αvβ8 in CNS vascular development
43.2.1.5 Retinoic acid in cerebrovascular development
43.2.2 Establishment of the BBB
43.2.2.1 Developmental timing of BBB emergence
43.2.2.2 Molecular control of BBB development
43.2.2.3 Mural cells in regulation of vascular development and BBB maturation
43.2.3 Vascular contribution to neurodevelopmental events
43.2.3.1 Vascular regulation of neuro- and oligodendrogenesis
43.2.3.2 The embryonic vasculature as a migratory scaffold in the forebrain
43.2.3.3 The brain vasculature shapes axonal architecture
43.2.4 hiPSC-based BBB culture models: lessons from CNS vascular development
43.2.5 Summary and conclusions
References
Index
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
X
Y
Z
Back Cover

Citation preview

Patterning and Cell Type Specification in the Developing CNS and PNS Comprehensive Developmental Neuroscience Second Edition Senior Editors-in-Chief

John Rubenstein

Department of Psychiatry & Weill Institute for Neurosciences University of California, San Francisco, San Francisco, CA, United States

Pasko Rakic

Department of Neuroscience & Kavli Institute for Neuroscience Yale School of Medicine, New Haven, CT, United States

Editors-in-Chief

Bin Chen

Department of Molecular, Cell & Developmental Biology University of California, Santa Cruz, Santa Cruz, CA, United States

Kenneth Y. Kwan

Michigan Neuroscience Institute & Department of Human Genetics University of Michigan, Ann Arbor, MI, United States

Section Editors:

Elizabeth A. Grove

Department of Neurobiology & Grossman Institute for Neuroscience University of Chicago, Chicago, IL, United States

Shubha Tole Department of Biological Sciences Tata Institute of Fundamental Research, Mumbai, India

Francois Guillemot The Francis Crick Institute, London, United Kingdom

Kenneth Campbell Division of Developmental Biology, Cincinnati Children’s Hospital Medical Center University of Cincinnati College of Medicine, Cincinnati, OH, United States

Arturo Alvarez-Buylla Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research University of California, San Francisco, San Francisco, CA, United States

David Rowitch Department of Paediatrics, University of Cambridge and Wellcome-MRC Cambridge Stem Cell Institute, Cambridge, United Kingdom Adjunct Professor of Pediatrics, UCSF

Academic Press is an imprint of Elsevier 125 London Wall, London EC2Y 5AS, United Kingdom 525 B Street, Suite 1650, San Diego, CA 92101, United States 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom Copyright © 2020 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www. elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-12-814405-3 For information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Nikki Levy Acquisitions Editor: Natalie Farra Editorial Project Manager: Andrae Akeh Production Project Manager: Surya Narayanan Jayachandran Cover Designer: David Tastad Typeset by TNQ Technologies

Contributors Katerina Akassoglou, Gladstone Institute of Neurological Disease and Department of Neurology, University of California, San Francisco, CA, United States Nicola J. Allen, Molecular Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA, United States

Aparna Bhaduri, University of California, San Francisco, CA, United States S. Blaess, Institute of Reconstructive Neurobiology, LIFE & BRAIN Center, University of Bonn, Medical Faculty and University Hospital Bonn, Bonn, Germany

Fernando C. Alsina, Department of Molecular Genetics and Microbiology

Stephanie Bonney, Department of Pediatrics, Section of Developmental Biology, University of Colorado, Anschutz Medical Campus, Aurora, CA, United States

Alessandro Alunni, Zebrafish Neurogenetics Unit, Developmental & Stem Cell Biology Department, Institut Pasteur, UMR3738, CNRS, Paris, France

Bernadett Bosze, Department of Cell Biology and Human Anatomy, University of California Davis, Davis, CA, United States

A. Alvarez-Buylla, University of California, San Francisco, CA, United States

Joshua J. Breunig, Board of Governors Regenerative Medicine Institute, Los Angeles, CA, United States; Department of Biomedical Sciences, Los Angeles, CA, United States; Samuel Oschin Comprehensive Cancer Institute, Los Angeles, CA, United States; Department of Medicine, David Geffen School of Medicine, UCLA, Los Angeles, CA, United States

Madeline G. Andrews, University of California, San Francisco, CA, United States S.-L. Ang, Francis Crick Institute, London, United Kingdom B. Appel, University of Colorado School of Medicine, Aurora, CO, United States Badrul Arefin, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden Shahrzad Bahrampour, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden Q.-R. Bai, Tongji University, Shanghai, China Laure Bally-Cuif, Zebrafish Neurogenetics Unit, Developmental & Stem Cell Biology Department, Institut Pasteur, UMR3738, CNRS, Paris, France Renata Batista-Brito, Dominick P. Purpura Department of Neuroscience, Albert Einstein College of Medicine, Bronx, NY, United States Magnus Baumgardt, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden Jonathan Benito-Sipos, Departamento de Biología, Universidad Autónoma de Madrid, Madrid, Spain

Nadean L. Brown, Department of Cell Biology and Human Anatomy, University of California Davis, Davis, CA, United States S.A. Buffington, Baylor College of Medicine, Houston, TX, United States C.L. Call, Johns Hopkins School of Medicine, Baltimore, MD, United States K. Campbell, Cincinnati Children’s Hospital Medical Center, University of Cincinnati College of Medicine, Cincinnati, OH, United States Astrid E. Cardona, UTSA Brain Health Consortium and South Texas Center for Emerging Infectious Diseases, Department of Biology, The University of Texas at San Antonio, San Antonio, TX, United States Catarina Catela, Department of Neurobiology, University of Chicago, Chicago, IL, United States A. Cebrián-Silla, University of California, San Francisco, CA, United States; Univeristat de València, CIBERNED, Valencia, Spain

D.E. Bergles, Johns Hopkins School of Medicine, Baltimore, MD, United States

xxi

xxii Contributors

Yi-Ting Cheng, Center for Cell and Gene Therapy, Baylor College of Medicine, One Baylor Plaza, Houston, TX, United States

Elizabeth A. Grove, Department of Neurobiology, The Grossman Institute for Neuroscience, University of Chicago, Chicago, IL, United States

Victor V. Chizhikov, University of Tennessee Health Science Center, Department of Anatomy and Neurobiology, Memphis, TN, United States

J.L. Haigh, University of California, Davis, CA, United States

Marion Coolen, Zebrafish Neurogenetics Unit, Developmental & Stem Cell Biology Department, Institut Pasteur, UMR3738, CNRS, Paris, France Jesús Rodriguez Curt, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden Dimitrios Davalos, Neuroinflammation Research Center, Department of Neurosciences, Lerner Research Institute, Cleveland Clinic, Cleveland, OH, United States L.M. De Biase, Johns Hopkins School of Medicine, Baltimore, MD, United States Benjamin Deneen, Center for Cell and Gene Therapy, Department of Neuroscience, Baylor College of Medicine, One Baylor Plaza, Houston, TX, United States

Jean Hébert, Neuroscience, Genetics, Stem Cells, Albert Einstein College of Medicine, Bronx, NY, United States Oliver Hobert, Department of Biological Sciences, Howard Hughes Medical Institute, Columbia University, New York, NY, United States Robert B. Hufnagel, Medical Genetics and Ophthalmic Genomics Unit, National Eye Institute, Bethesda, MD, United States Wieland B. Huttner, Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany Yasuhiro Itoh, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States K.R. Jessen, University College London, London, United Kingdom

Omer Durak, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States

Jane E. Johnson, Department of Neuroscience, University of Texas Southwestern Medical Center, Dallas, TX, United States

Ryann M. Fame, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States; Department of Pathology, Boston Children’s Hospital, Boston, MA, United States

Eyal Karzbrun, Kavli Institute of Theoretical Physics and Department of Physics, University of California, Santa Barbara, CA, United States

Stephen P.J. Fancy, Neurology and Pediatrics, University of California, San Francisco, San Francisco, CA, United States Gord Fishell, Department of Neurobiology, Blavatnik Institute, Harvard Medical School, Boston, MA, United States; Stanley Center for Psychiatric Research, Broad Institute, Cambridge, MA, United States Isabelle Foucher, Zebrafish Neurogenetics Unit, Developmental & Stem Cell Biology Department, Institut Pasteur, UMR3738, CNRS, Paris, France L. Fuentealba, University of California, San Francisco, CA, United States Fred H. Gage, Salk Institute for Biological Studies, La Jolla, CA, United States Ludovic Galas, Normandie University, UNIROUEN, INSERM, PRIMACEN, Mont-Saint-Aignan, France Andrew W. Grande, Department of Neurosurgery, University of Minnesota, Minneapolis, MN, United States

Yutaro Komuro, Department of Neurology, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA, United States Hitoshi Komuro, Department of Neuroscience, Yale University School of Medicine, New Haven, CT, United States Arnold R. Kriegstein, University of California, San Francisco, CA, United States J.T. Lambert, University of California, Davis, CA, United States Katherine R. Long, Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany Guillermina López-Bendito, Instituto de Neurociencias de Alicante, Universidad Miguel Hernández-Consejo Superior de Investigaciones Científicas (UMH-CSIC), Sant Joan d’Alacant, Spain Jessica L. MacDonald, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States; Department of Biology, Syracuse University, Syracuse, NY, United States

Contributors

Jeffrey D. Macklis, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States; Bauer Laboratory, Cambridge, MA, United States Maria Carolina Marchetto, Salk Institute for Biological Studies, La Jolla, CA, United States Francisco J. Martini, Instituto de Neurociencias de Alicante, Universidad Miguel Hernández-Consejo Superior de Investigaciones Científicas (UMH-CSIC), Sant Joan d’Alacant, Spain Michael P. Matise, Department of Neuroscience & Cell Biology, Rutgers-Robert Wood Johnson Medical School, Piscataway, NJ, United States F.T. Merkle, Metabolic Research Laboratories and Medical Research Council Metabolic Diseases Unit, Wellcome Trust-Medical Research Council Institute of Metabolic Science, and the Wellcome Trust-Medical Research Council Cambridge Stem Cell Institute, University of Cambridge, Cambridge, United Kingdom A. Meunier, Institut National de la Santé et de la Recherche Médicale, Paris, France; Centre National de la Recherche Scientifique, Paris, France; Institut de Biologie de l’Ecole Normale Supérieure (IBENS), Paris, France Kathleen J. Millen, Seattle Children’s Hospital Research Institute Center for Integrative Brain Research, Seattle, WA, United States Robert H. Miller, Anatomy and Cell Biology, School of Medicine and Health Sciences, George Washington University, Washington, DC, United States R. Mirsky, University College London, London, United Kingdom Swati Mishra, Department of Pediatrics, Section of Developmental Biology, University of Colorado, Anschutz Medical Campus, Aurora, CA, United States; Department of Pathology, Institute for Stem Cell & Regenerative Medicine, University of Washington, Seattle, WA, United States Anna Victoria Molofsky, Laboratory of Molecular Neurobiology, Centre of New Technologies, University of Warsaw, Warsaw, Poland

xxiii

Masato Nakafuku, Division of Developmental Biology, Cincinnati Children’s Hospital Medical Center, Departments of Pediatrics and Neurosurgery, University of Cincinnati College of Medicine, Cincinnati, OH, United States Harukazu Nakamura, Laboratory of Organ Morphogenesis, Graduate School of Life Sciences, Tohoku University, Aoba-ku, Sendai, Japan Branden R. Nelson, Center for Integrative Brain Research, Seattle Children’s Research Institute, Seattle, WA, United States A.S. Nord, University of California, Davis, CA, United States K. Obernier, University of California, San Francisco, CA, United States Nobuhiko Ohno, Department of Anatomy, Division of Histology and Cell Biology, Jichi Medical University, Shimotsuke-Shi, Tochigi, Japan; Division of Ultrastructural Research, National Institute for Physiological Sciences, Okazaki, Aichi, Japan Abdulkadir Ozkan, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States David B. Parkinson, Medicine and Dentistry, Plymouth University, Plymouth, Devon, United Kingdom Manuel Peter, Department of Stem Cell and Regenerative Biology, and Center for Brain Science, Harvard University, Cambridge, MA, United States Samuel J. Pleasure, Department of Neurology, Programs in Neuroscience and Developmental Biology, Institute for Regenerative Medicine, University of California, San Francisco, CA, United States M.N. Rasband, Baylor College of Medicine, Houston, TX, United States Orly Reiner, Department of Molecular Genetics, The Weizmann Institute of Science, Rehovot, Israel D.H. Rowitch, University of California, San Francisco, CA, United States J.L.R. Rubenstein, University of California at San Francisco, San Francisco, CA, United States

Ignacio Monedero Cobeta, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden

Debosmita Sardar, Center for Cell and Gene Therapy, Baylor College of Medicine, One Baylor Plaza, Houston, TX, United States

K. Monk, Vollum Institute, Oregon Health Science Center, Portland, OR, United States

Anindita Sarkar, Salk Institute for Biological Studies, La Jolla, CA, United States

Edwin S. Monuki, Pathology & Laboratory Medicine, Developmental & Cell Biology, University of California Irvine, Irvine, CA, United States

K. Sawamoto, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan; National Institute for Physiological Sciences, Okazaki, Japan

xxiv

Contributors

Kamal Sharma, Department of Anatomy & Cell Biology, University of Illinois at Chicago, Chicago, IL, United States Q. Shen, Tongji University, Shanghai, China Julie A. Siegenthaler, Department of Pediatrics, Section of Developmental Biology, University of Colorado, Anschutz Medical Campus, Aurora, CA, United States Debra L. Silver, Department of Molecular Genetics and Microbiology; Department of Cell Biology; Department of Neurobiology; Duke Institute for Brain Sciences, Duke University Medical Center, Durham, NC, United States N. Spassky, Institut National de la Santé et de la Recherche Médicale, Paris, France; Centre National de la Recherche Scientifique, Paris, France; Institut de Biologie de l’Ecole Normale Supérieure (IBENS), Paris, France

S. Temple, Neural Stem Cell Institute, Rensselaer, NY, United States Stefan Thor, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden; School of Biomedical Sciences, University of Queensland, St Lucia, QLD, Australia Shubha Tole, Department of Biological Sciences, Tata Institute of Fundamental Research, Mumbai, Maharashtra, India Gregorio Valdez, Brown University, Providence, RI, United States David Vaudry, Normandie University, UNIROUEN, INSERM, PRIMACEN, Mont-Saint-Aignan, France; Normandie University, UNIROUEN, INSERM, U1239, DC2N, Mont-Saint-Aignan, France

S.R.W. Stott, The Cure Parkinson’s Trust, London, United Kingdom

Claire Ward, Dominick P. Purpura Department of Neuroscience, Albert Einstein College of Medicine, Bronx, NY, United States

Johannes Stratmann, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden

Michael Wegner, Institut für Biochemie, Emil-FischerZentrum, Universität Erlangen-Nürnberg, Erlangen, Germany

L. Subramanian, University of California, San Francisco, CA, United States

Behzad Yaghmaeian Salmani, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden

John Svaren, Department of Comparative Biosciences and Waisman Center, University of Wisconsin, Madison, WI, United States Lukasz Mateusz Szewczyk, Department of Psychiatry and Weill Institute for Neurosciences, University of California, San Francisco, San Francisco, CA, United States; Laboratory of Molecular Neurobiology, Centre of New Technologies, University of Warsaw, Warsaw, Poland

Chapter 1

Morphogens, patterning centers, and their mechanisms of action Elizabeth A. Grove1 and Edwin S. Monuki2 Department of Neurobiology, University of Chicago, Chicago, IL, United States; 2Pathology & Laboratory Medicine, Developmental & Cell

1

Biology, University of California Irvine, Irvine, CA, United States

Chapter outline 1.1. General principles of morphogen gradients 1.1.1. History of the morphogen and morphogenetic field 1.1.2. How morphogen gradients pattern tissues 1.1.3. How morphogens are distributed 1.1.4. How morphogen signaling is transduced and interpreted 1.1.5. How morphogen gradients are converted into sharp boundaries 1.1.6. Summarydgeneral principles of morphogen gradients 1.2. Local signaling centers and probable morphogens in the telencephalon 1.2.1. Early forebrain patterning 1.2.2. The RPC 1.2.3. The telencephalic roof plate and cortical hem 1.2.4. The antihem 1.3. BMPs as morphogens in telencephalic patterning 1.3.1. Performance objectives for a BMP gradient in the dorsal telencephalon

3 3 4 5 6 6 7 7 8 8 8 9 9 9

1.3.2. Midline expression and homeogenetic expansion of BMP production 1.3.3. BMP signaling gradient in the dorsal telencephalon 1.3.4. BMPs as dorsal telencephalic morphogens 1.3.5. Linear conversion of BMP signaling by cortical cells 1.3.6. Nonlinear conversion of BMP signaling by DTM cells 1.3.7. Summarydthe BMP signaling gradient 1.4. FGF8 as a morphogen in telencephalic patterning 1.5. Interactions among signaling centers in telencephalic patterning 1.5.1. FGF8, Shh, and BMP signaling 1.5.2. Cross-regulation of BMP, FGF, and WNT signaling 1.5.3. Interactions of Shh, FGFs, and Gli3 1.6. Morphogens in human brain disease 1.6.1. Holoprosencephaly and Kallmann syndrome 1.6.2. Gradients in holoprosencephaly neuropathology 1.6.3. Gradients in other human brain disorders References

10 11 11 12 12 13 13 14 15 15 15 15 15 17 17 18

1.1 General principles of morphogen gradients 1.1.1 History of the morphogen and morphogenetic field The concept of a morphogen can be traced to the turn of the 20th century, when Morgan postulated the presence of “formative substances” as the basis for different regeneration rates in worms (Morgan, 1901). Very soon thereafter, Boveri entertained this idea for normal development (Boveri, 1901). A seminal event for this field was the discovery of a localized source for morphogens known as the Spemann organizer (Spemann and Mangold, 1924). The term “morphogen” was coined by Turing, who described how uniformly distributed signals made by cells can spread, self-organize, and generate pattern (Turing, 1952). Turing patterns remain highly relevant in development, but for this chapter and the developing forebrain, the more relevant concept is that of nonuniform graded distributions of morphogens, an idea formalized in the famous “French flag” model of Wolpert (Fig. 1.1) (Wolpert, 1969).

Patterning and Cell Type Specification in the Developing CNS and PNS. https://doi.org/10.1016/B978-0-12-814405-3.00001-1 Copyright © 2020 Elsevier Inc. All rights reserved.

3

Morphogen

4 PART | I Induction and patterning of the CNS and PNS

2

1

Distance from source FIGURE 1.1 The French flag model. Schematic of how a diffusible morphogen can assign positional values and instruct cells fates. Morphogen (green) secreted from a source cell forms a concentration gradient within a tissue. At intermediate concentrations above threshold 1, responding cells adopt “white” fate. At high concentrations above threshold 2, cells adopt “blue” fate. Based on Kicheva, A., Gonzalez-Gaitan, M., 2008. The Decapentaplegic morphogen gradient: a precise definition. Curr. Opin. Cell Biol. 20, 137e143; Rogers, K.W., Schier, A.F., 2011. Morphogen gradients: from generation to interpretation. Annu. Rev. Cell Dev. Biol. 27, 377e407.

In this model, Wolpert described smoothly declining gradients of morphogen concentration within a “morphogenetic field” of cells. These gradients were imagined to arise via diffusion from a localized source toward a sink, thus giving cells within the morphogenetic field different positional values based on morphogen concentration. The positional values then determined the fates adopted by cells in the field (Fig. 1.1). It was not until the 1980s that the molecular identity of a morphogen was defined (bicoid) (Driever and Nusslein-Volhard, 1988a, 1988b). The first secreted morphogen was identified soon thereafter (decapentaplegic or dpp) (Ferguson and Anderson, 1992). Since then, many more morphogens have been discovered. Most, but not all, are secreted proteins; examples of other molecular classes include transcription factors (bicoid and dorsal) and a vitamin derivative (retinoic acid).

1.1.2 How morphogen gradients pattern tissues The concept of positional identity is important for understanding how morphogen gradients work because it is erroneous to consider morphogens as the sole determinants of cell fate. As it turns out, the same limited repertoire of morphogens is used over and over again across ontogeny and phylogeny to generate the dizzying array of cell types found in the animal kingdom. Thus, morphogens could not possibly be instructive of cell fate on their own. Rather, morphogens act upon tissues with different prepatterns and competencies, and these competencies in combination with the positional information provided by morphogens determine cell fate. For example, in this chapter, we discuss in detail how bone morphogenetic proteins (BMPs, the orthologues of dpp) and fibroblast growth factors (FGFs) provide positional information to dorsal telencephalic cells with restricted neural potential. The defining property of a morphogen is the ability to specify two or more cell fates in a concentration-dependent manner. For some, at least three fates are necessary to ensure that a morphogen is truly instructive (Freeman and Gurdon, 2002). Morphogens often specify between three to seven fates within a tissue (Ashe and Briscoe, 2006), which are separated by sharp, discrete boundaries. The acquisition of mature cell fates and boundaries is preceded by cell-intrinsic differences in the expression of “selector” genes (most often transcription factors) that specify cell fates in particular ways (Garcia-Bellido, 1975). Understanding how graded morphogenic information is converted into sharp (switch-like or ultrasensitive) changes in downstream gene expression remains a central problem for developmental biologists, although, as we will see below, several mechanisms underlying such “switches” have been defined. One important objective for many tissues patterned by morphogens is the establishment of secondary organizers or signaling centers (Meinhardt, 2009). These secondary sites of morphogen production expand the ranges over which morphogens can act, provide for finer subdivisions of pattern, or both, and are located at boundaries established by the primary morphogen gradient and selector genes. Other consequences of primary patterning include apoptosis, cell sorting to further refine borders, and other forms of cell-to-cell signaling (Lander et al., 2009a). These events can be regulated by morphogens or may become largely cell-autonomous and immune to extrinsic control. An interesting potential use of

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

5

morphogen gradients is the control of proliferation and growth, which is often relatively uniform within tissues, but the jury remains out on this issue (Schwank and Basler, 2010; Dekanty and Milan, 2011). (Note: Interestingly, generating uniform growth from graded information would be exactly the opposite problem of making sharp borders!) Over the last decade or so, the application of mathematical modeling and computer simulations has provided deep insights not only into the “what’s” and “how’s” of morphogen gradients but also into the many interesting “why’s”. These approaches have provided insights into systems-level features such as robustness (insensitivity to perturbations), adaptability or resilience (the ability to adapt to perturbations), precision, noise buffering, and scaling of pattern to tissue size, which would be impossible or impractical to address via wet lab experimentation alone. (Consider the many advantages of a computer, rather than a bench scientist, testing 1000 points in parameter space.) Several general principles have emerged from this work, which have been reviewed by several others (Barkai and Shilo, 2009; Lander et al., 2009a; Wartlick et al., 2009; Briscoe and Small, 2015). Among its many lessons, morphogen systems biology has taught us (1) that defining individual operations within a morphogen system (e.g., whether gene x is necessary and sufficient for function y) provides little understanding of the system itself; (2) that the performance objectives of morphogen systems differ as a result of the unique factors and forces that impinge upon them; (3) that morphogen systems can use very different mechanisms to meet similar objectives; (4) that every mechanism, no matter how advantageous, has trade-offs; and (5) that understanding how morphogen systems work and balance conflicting priorities is impossible without analyzing systems as a whole. Does it make sense that morphogen systems have evolved so many different mechanisms to achieve similar goals? From engineering and evolutionary perspectives, the answer is yes. Engineers are very familiar with the “no free lunch” principledi.e., every mechanism has trade-offs, and mechanisms that confer robustness in one setting might increase fragility in another. Thus, despite attempts to make systems resistant to most everything, highly engineered systems are inescapably fragile (Carlson and Doyle, 2002). Furthermore, fragility is not all baddin fact, fragility is necessary for adaptabilitydso the “robust, yet fragile” trade-off is a common feature of complex systems. Given that evolution acts upon preexisting biological templates rather than clean slates, mechanistic multiplicity and redundancy would also be logical, necessary, and unavoidable.

1.1.3 How morphogens are distributed How are morphogens distributed to generate concentration gradients? As originally envisioned (Wolpert, 1969; Crick, 1970), extracellular diffusion is a predominant mechanism (Lander et al., 2002; Rogers and Schier, 2011; Zhou et al., 2012). The advantages of diffusion are many. It is simple, fast, and occurs via random walk rather than being ballistic, which negates impediments such as tortuosity of the extracellular space (Lander, 2007). (Random walk is the reason why diffusion fills a maze almost as fast as it does an open space.) The speed of diffusion can also explain how long-range gradients can form in the presence of high-affinity receptors because ligand-receptor binding rates for morphogens are often much slower than diffusion (e.g., for activin, it takes 30 min to load 0.5% of available receptors) (Freeman and Gurdon, 2002). Accordingly, diffusivity is an important point of regulation and can occur by modifying the morphogens themselves (e.g., via lipid modification) or their binding to cofactors and extracellular matrix (Rogers and Schier, 2011). Diffusion can be used in interesting ways to generate signaling gradientsde.g., by giving a morphogen and its inhibitor different diffusion vectors for facilitated transport (Holley et al., 1996; Shimmi et al., 2005) or different diffusivities, which generates different ranges of activity (Meinhardt, 2009). Clearly, however, mechanisms other than diffusion are also used to generate morphogen gradients. For retinoic acid, spatially regulated intracellular degradation leads to its gradient (White et al., 2007). Cellular rather than molecular mechanisms have also been invoked, such as the progressive dilution of intracellular morphogens resulting from cell division, and the use of cellular filopodia-like extensions or “cytonemes” (Rogers and Schier, 2011). Cell-to-cell “transcytosis” has been proposed as an alternative to diffusion, although transcytosis is difficult to cleanly dissociate from diffusion, and it has been argued that the data supporting transcytosis can be entirely explained by diffusion (Lander et al., 2002; Kicheva and Gonzalez-Gaitan, 2008; Zhou et al., 2012). Nevertheless, nondiffusion mechanisms certainly exist and are likely to rectify whatever deficiencies diffusion has in specific systems (Lander, 2007). What do morphogen gradients look like? In the French flag model, Wolpert illustrated a declining exponential function (Fig. 1.1), and this turns out to accurately describe many morphogen systems. For diffusive systems, three parameters describe gradient profile: morphogen production rate or flux, diffusion, and clearance. When clearance occurs via a sink, as postulated by Crick (Crick, 1970), gradient profile is linear. In contrast, uniform clearance within a tissue leads to a declining exponential gradient. Nonuniform diffusion or clearance can lead to other distributions, such as power law functions (Kicheva and Gonzalez-Gaitan, 2008).

6 PART | I Induction and patterning of the CNS and PNS

The size of morphogenetic fields (i.e., the amount of tissue patterned by morphogen gradients) can also be estimated. The length scale, or decay length, for an exponential gradient is the distance over which morphogen concentration falls by e 1 (w37%). Length scale is determined by diffusivity and clearance rate, but it is independent of synthesis rate, and the morphogenetic fields for bicoid and Gurken/EGFR in Drosophila are w3e5 times the length scale (Goentoro et al., 2006). Using reasonable parameters, others have suggested a few hundred microns as a theoretical maximum, which matches in vivo situations reasonably well (Lander et al., 2009a).

1.1.4 How morphogen signaling is transduced and interpreted How is extracellular morphogen concentration measured by a cell? In most cases, signaling intensity is determined by the absolute number of occupied receptors (Dyson and Gurdon, 1998), although in the case of hedgehog (Hh), the ratio of occupied:unoccupied receptors has been implicated as the key determinant (Rogers and Schier, 2011). Intuitively, the use of absolute rather than relative numbers of occupied receptors would be useful at low morphogen concentrations and could allow for larger morphogenetic fields. Signaling intensity can be modulated in many ways, including via changes to the extracellular matrix or receptor numbers. When absolute receptor numbers are used, morphogen receptor signaling is directly proportional to extracellular morphogen concentration. Importantly, for most morphogens, this proportionality is maintained all the way down to their transcriptional effectors in the nucleus (Ashe and Briscoe, 2006; Rogers and Schier, 2011). A likely explanation for this proportionality is the linear rather than branched construction of most morphogen signaling pathways. Indeed, for many morphogen pathways, the signal transducer doubles as the transcriptional effector (e.g., Smad for Nodal and BMP, Gli for Sonic hedgehog/Shh) (Ashe and Briscoe, 2006). The absence of significant branching or cascading reduces the possibilities for nonlinear signal amplification. In this way, the positional values imparted by extracellular morphogen concentrations are directly and proportionally transmitted into responding cells. Responding cells are also quite sensitive to morphogen concentration. For example, cells can sense large concentration differences in Shh or activin (25-50X), and relatively small changes in concentration or activated receptor number (2-3X) are sufficient for fate transformations or major shifts in boundary position (Ferguson and Anderson, 1992; Dyson and Gurdon, 1998; Ashe and Briscoe, 2006). High sensitivity also expands the range over which morphogens can act, up to the point where signal and robustness mechanisms are overcome by noise. For many gradients, binding noise due to low receptor occupancy is a likely limiting factor (Lander et al., 2009a). Time is another crucial factor for interpreting morphogens. For practical reasons, spatial gradients are often considered at their steady states, but these states represent oversimplifying assumptions in many cases. Indeed, patterning in vivo can be quite fast (even less than a few hours), and some gradients are decoded during rising, preesteady state conditions (Barkai and Shilo, 2009). Temporal integration and dynamic interpretations of signaling can also be critical (Sagner and Briscoe, 2017). In some cases, fate transformations caused by increased concentration can be mimicked by increased duration; this is particularly well established for Shh. In other systems, however, cell fates reflect the highest concentration ever seen by a cell rather than a temporal integral (Ashe and Briscoe, 2006; Rogers and Schier, 2011).

1.1.5 How morphogen gradients are converted into sharp boundaries The need to generate discrete and well-separated cell types is nearly universal, and morphogen systems have evolved many different ways to convert graded extracellular (and intracellular) information into nonlinear fate decisions and boundaries. Some mechanisms generate nonlinearity in extracellular morphogen distribution; these include facilitated transport, alterations in morphogen degradation or clearance, and regulation of receptor availability. However, most nonlinear conversion mechanisms are intracellular, and many are transcriptional. This makes some sense because most morphogen transduction pathways retain linearity all the way down to the nucleus. Nonlinear transcriptional mechanisms include cooperativity, differential binding site affinity, autoregulatory positive feedback, feedforward loops, sign switching (e.g., from transcriptional activator to repressor), and cross-repression (Ashe and Briscoe, 2006). Many positive feedback mechanisms also generate bistability, the property of having two potential stable states at some stimulus concentrations. Bistability represents a form of cell memory and provides robustness to “on” states, which enables cells to maintain their fates after the inducing morphogen is no longer available. In addition to these transcriptional sharpening mechanisms, morphogen systems also employ cellular-level mechanisms to sharpen boundaries, including sorting, death, and respecification of mislocalized cells near boundaries (Ashe and Briscoe, 2006).

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

7

1.1.6 Summarydgeneral principles of morphogen gradients Morphogen gradient systems are complex but share many common features, which can be summarized as follows (Rogers and Schier, 2011): 1. Morphogens are released from dynamic localized sources, assemble with other molecules, and move via diffusion through the extracellular space. 2. Gradient shape is determined by flux from the source, diffusivity, and clearance from tissues. 3. Morphogen concentration and duration are transmitted linearly to intracellular molecules, ultimately resulting in the graded and proportional activity of transcriptional effectors. 4. Transcriptional effectors participate in complex regulatory networks that involve preexisting intrinsic factors, which ultimately determine target gene responses. 5. Feedback mechanisms act to buffer fluctuations in morphogen production, regulate signaling interpretation, and confer scalability and robustness to morphogen-mediated patterning.

1.2 Local signaling centers and probable morphogens in the telencephalon Developmental neurobiologists were relatively slow to adopt the concept that morphogens secreted from signaling centers can pattern complex structures of the embryonic brain. This was particularly true of the telencephalon, given that this part of the embryonic brain gives rise to the cerebral cortex, long believed to be so functionally complex that unique mechanisms would be needed to pattern it. The recognition of several putative morphogen sources adjacent to the early embryonic forebrain, however, led to the proposal that these signaling centers and morphogens were critical to the structural organization of the telencephalon, just as they are for other parts of the embryo (Furuta et al., 1997; Grove et al., 1998; Crossley et al., 2001; Ragsdale and Grove, 2001; Ohkubo et al., 2002). Thus, even the most functionally complex part of the body is patterned by mechanisms that are commonly used elsewhere in the embryo. Half a dozen putative signaling centers have now been identified for the telencephalon, chiefly for the cerebral cortex, and substantial evidence supports a patterning role for several of these. The candidate signaling centers comprise (1) a source of Shh from the prechordal mesoderm that underlies the medial prosencephalic neural plate (Rubenstein and Beachy, 1998); (2) the anterior neural ridge (ANR), which appears before closure of the anterior neuropore at the edge of the neural plate and expresses Fgf8 (Crossley and Martin, 1995); (3) the rostral telencephalic patterning center (RPC, also known as the anterior cerebral pole) formed as the telencephalic vesicle generates two cerebral hemispheres, also

(A)

(C) Hem

acp t

Fgf8

d

Wnt3a

(D)

(B)

Hem Hem Antihem

Wnt3a

Wnt3a/sFrp2

FIGURE 1.2 Three telencephalic signaling centers. (A,B) Dorsal views of two E10.5 forebrains processed with in situ hybridization to show the genes indicated, anterior to the top. (C,D) E13.5 hemispheres viewed from the medial (C) or lateral (D) face, anterior to the left. Fgf8 is expressed at the rostral patterning center (also known as the anterior cerebral pole or acp) (A) and Wnt3a at the cortical hem (BeD). sFrp2 expression marks the antihem, which forms a pincer shape with the Wnt3a-expressing hem (D).

8 PART | I Induction and patterning of the CNS and PNS

expressing Fgf genes of the Fgf8 subfamily (Bachler and Neubuser, 2001); (4) the telencephalic roof plate, a source of BMP signals (Furuta et al., 1997; Monuki et al., 2001; Cheng et al., 2006); (5) lineally related successors to the roof plate at the dorsomedial edge of each cerebral cortical hemispheredthe choroid plexus epithelium (CPE) and cortical hem, expressing Wnt and BMP proteins (Furuta et al., 1997; Grove et al., 1998; Hebert et al., 2002; Currle et al., 2005); and (6) the “antihem,” at the junction of the dorsal and ventral telencephalon, secreting the Wnt inhibitor sFrp2; the EGF family members Tgf-a, neuregulins 1 and 3; and FGF7 (Assimacopoulos et al., 2003) (Fig. 1.2). We describe here the putative signaling sources and their constituent signaling molecules that control patterning of broad divisions of the telencephalon. Another chapter in this volume discusses, more specifically, the role of two of the signaling sources in patterning the neocortex into a map of distinct areas.

1.2.1 Early forebrain patterning The prechordal mesoderm, producing Shh, and the ANR influence the earliest stages of forebrain patterning. FGF8 from the ANR upregulates gene expression of the transcription factor, Foxg1, whose expression is the first marker of telencephalic tissue (Shimamura and Rubenstein, 1997; Rubenstein and Beachy, 1998). Shh from mesoderm underlying the medial prosencephalic neural plate divides the diencephalic eye field into two (Chiang et al., 1996). Furthermore, similar to the action of Shh in the spinal cord, early activity of Shh contributes to the specification of ventral cell fates in the ventral telencephalon (Sussel et al., 1999; Gulacsi and Anderson, 2006). Dissimilar from the caudal central nervous system (CNS), ventralizing the telencephalon appears also to require FGF signaling (Gutin et al., 2006; Danjo et al., 2011).

1.2.2 The RPC The signaling center we term the RPC expresses Fgf genes of the Fgf8 subfamily, similarly to the isthmic organizer (ISO) at the midbrain/hindbrain junction. At both sites, FGF8 and FGF17 have partly separate and partly complementary patterning roles. Fgf8 is expressed as the RPC forms, and FGF8 induces expression of Fgf18 and Fgf17 (Cholfin and Rubenstein, 2008), the latter of which is expressed more broadly than Fgf8. Fgf18 has a more limited expression domain and its telencephalic role has not yet been studied. Although the RPC may arise from cells of the ANR, the two signaling sources are distinguishable temporally and by morphology. That is, the ANR is evident when the neural tube is open, but the RPC is identified when the anterior neuropore has closed at about embryonic day (E) 9 in the mouse. The ANR is critical to the initial patterning of the forebrain, the RPC in later patterning of the telencephalon. The RPC has also been referred to as the “commissural plate,” a structure that also forms anteromedially, but later in development, as a channel for the major commissures of the hemispheres. Indeed, FGF8 is needed to position the commissural plate (Moldrich et al., 2010). The RPC does not appear to be a specific progenitor of the commissural plate. Fate mapping indicates the RPC gives rise to neurons that populate the prefrontal cortex and parts of the septum, as well as the likely commissural plate (Toyoda et al., 2010; Hoch et al., 2015). Evidence detailed in a subsequent chapter indicates that FGF8 and FGF17, dispersing from the RPC as morphogens, pattern the neocortical area map (Fukuchi-Shimogori and Grove, 2001; Garel et al., 2003; Cholfin and Rubenstein, 2007, 2008; Toyoda et al., 2010).

1.2.3 The telencephalic roof plate and cortical hem The telencephalic roof plate can be defined as the midline of the telencephalic vesicle before it has divided into two hemispheres. Once the two medial hemispheric walls are distinct, bilateral CPE and cortical hems become evident at the dorsomedial edge of each cerebral cortex. The roof plate produces several members of the BMP family of signaling molecules, and genetic ablation of either the type I BMP receptor BMPRIa or of the entire roof plate causes a loss of telencephalic CPE (Hebert et al., 2002). Ablation of the roof plate further causes a cortical phenotype that resembles middle interhemispheric (MIH) holoprosencephaly (Cheng et al., 2006; Monuki, 2007). The cortical hem secretes Wnts and BMPs from the dorsomedial edge of the cortical primordium (Furuta et al., 1997; Grove et al., 1998) and is both necessary and sufficient for specifying the hippocampus. Without a cortical hem, or if hem Wnt signaling is sufficiently depleted, the hippocampus fails to develop (Galceran et al., 1999; Lee et al., 2000; Yoshida et al., 2006). Suggesting that canonical Wnt signaling induces the different hippocampal fields, constitutively active b-catenin induces cortical cells to express genes normally characteristic of hippocampus (Machon et al., 2007). Most striking, an ectopic hem induces a small secondary, ectopic hippocampus (Mangale et al., 2008). BMP signaling also contributes to the formation of the hippocampus. In mice deficient for two of the type I BMP receptors, BMP signaling is reduced but not abrogated, given a third type I receptor is also present in the telencephalon (Caronia et al., 2010). The double-mutant mouse has a greatly reduced hippocampal dentate gyrus (DG) compared with

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

9

control mice and a proportionally smaller population of adult DG neural stem cells. Moderate reductions in Wnt signaling in the hem also cause a diminished or absent DG (Li and Pleasure, 2005). Beyond the hippocampus, the cortical hem also regulates the size and patterning of neocortex (Caronia-Brown et al., 2014). How Wnt and BMP signaling work together in early hippocampal and neocortical development still needs clarification.

1.2.4 The antihem The antihem lies at the opposite edge of the cerebral cortex to the hem, forming a narrow band surrounding the boundary between the dorsal and ventral telencephalon. Interestingly, in mice deficient in the transcription factor Lhx2, which promotes cortical identity, both the hem and the antihem expand into the vacant territory (Mangale et al., 2008), suggesting hem and antihem are in some sense equivalent structures. The antihem expresses sFrp2, encoding a soluble Wnt inhibitor (Kim et al., 2001), Fgf7, and the EGF genes, Tgfa, Nrg1, and Nrg3 (Kim et al., 2001; Assimacopoulos et al., 2003). The latter are orthologs of Drosophila Spitz and Vein, encoding EGF ligands that control neuronal specification in the Drosophila ventral nerve cord (Skeath, 1998; von Ohlen and Doe, 2000). The antihem is evident by gene expression a day or two after the hem, suggesting it is present too late for a role in early corticogenesis. Nonetheless, complete loss of the antihem in the small eye (Pax6-deficient) mutant suggests possible involvement in the cortical patterning and cell migration defects that occur in small eye and Pax6 null mice. Further implying that the EGF family regulates cortical regionalization, EGF induces a molecular marker of limbic cortical areas, LAMP, in explants of nonlimbic cortex (Ferri and Levitt, 1995). Greater understanding of the specific functions of this putative signaling center awaits conditional genetic manipulations specific to the antihem.

1.3 BMPs as morphogens in telencephalic patterning Do the general principles of morphogen gradients apply to the mammalian telencephalon? In the next two sections, we focus on BMP and FGF signaling in dorsal telencephalic patterning, for which there is substantial evidence that the answer is “yes.”

1.3.1 Performance objectives for a BMP gradient in the dorsal telencephalon Following neural induction and neural tube closure at E9 in mice, four distinct cell fates differentiate along the dorsoventral (DV) axis of the dorsal telencephalon by E12 (Fig. 1.3). Three of these form domains at or near the dorsal telencephalic midline (DTM)dfrom medial to lateral, these are the choroid plaque, CPE, and cortical hem. CPE produces the cerebrospinal fluid, and the cortical hem acts as an organizer for the hippocampus (Mangale et al., 2008) and neocortical patterning center (Caronia-Brown et al., 2014). (The choroid plaque is not known to have a specific function.) Lateral to these is the cortex or cortical primordium, which is much larger and constitutes most of the dorsal telencephalon. In mice, the critical period for specifying these dorsal telencephalic fates precedes the onset of cortical neurogenesis at E11. Excessive CPE and hem form when the transcription factor Lhx2 is inactivated by E8.5, but not after E10.5, and this same E8.5eE10.5 period defines the critical period for specifying cortical identity (Mangale et al., 2008). Forebrain competency for CPE fate also coincides with this period, based on culture studies of E8.5 and E9.5 forebrain cells (Thomas and Dziadek, 1993) or E9.5 and E10.5 dorsal forebrain explants (Srinivasan et al., 2014). Peak CPE competency in embryonic stem (ES) cellederived systems also correlates with preneurogenic neuroepithelial cells rather than neurogenic radial glia (Watanabe et al., 2012). Thus, if an instructive BMP gradient exists, it must exist in the preneurogenic dorsal telencephalon. In addition to specifying cell types, what other performance objectives might a BMP gradient in the dorsal telencephalon have? The specification of CPE and cortical hem fits with the performance objective of specifying secondary organizers because both tissues have specific signaling functions (Mangale et al., 2008; Lehtinen et al., 2011). Another objective could be graded patterning of the cortex. The neurogenic cortical primordium is well known for its transcriptional gradients rather than thresholds (Sansom et al., 2005), and the gradients impact cortical arealization in mature animals (Bishop et al., 2000; Mallamaci et al., 2000). One notable performance objective that the cortex lacks is regeneration. With the exception of hippocampus and olfactory bulb, significant neuronal regeneration does not occur in the dorsal telencephalon, which may reflect a positive selection during evolution for long-term memory storage (Spalding et al., 2005; Bhardwaj et al., 2006). (The idea is that neuronal regeneration and replacement would cause losses of memory/information stored within existing neuronal circuits.) The lack of regeneration also implies that the neocortex must have good negative feedback and maintenance systems to generate and maintain the right number of cells from the beginning (Lander et al., 2009b).

10

PART | I Induction and patterning of the CNS and PNS

(A)

(B)

BMP

(C) 150 pSmad (grayscale)

Control

2 1

le

la

qu

e

(E)

(D)

[Stimulus]

xu

C

or

0

te

x

0.2

0.4

0.6

0.8

1.0

Distance from dorsal midline (mm)

s

Msx1 nlacZ

nH = 3.3

8

nH = 3.8

4 0 Mutant

0

Relative levels (log2)

% Response

Ultrasensitive

Mutant

(F)

12

100 Graded

H em

.p

.p

50

0

ch

ch

100

Control

BMP

3

−4 1

10

100

1000

BMP4 (ng ml−1)

Msx1

FIGURE 1.3 A bone morphogenetic protein (BMP) signaling gradient and switch in the dorsal telencephalon. (A) Coronal schematics of the dorsal telencephalon. BMPs (green) produced at the dorsal midline diffuse over a naïve dorsal telencephalic neuroepithelium. Within 2e3 days in mice, four fates are specifieddchoroid plaque (green), choroid plexus epithelium (blue), cortical hem (orange), and cortex (red). (B) Modified French flag model. The BMP gradient generates three thresholds separating the four cell domains, which include cells that continue to produce BMPs, and also provides graded positional information to the cortex. (C) The BMP signaling gradient in E10.5 dorsal telencephalon (from Cheng et al., 2006). The pSmad gradient is a simple declining exponential (decay length w290 um), which becomes reduced and flattened following roof plate ablation (“mutant”). (D) Schematics of a graded response (gray) and an ultrasensitive “switch” (red). (E) Ultrasensitivity of E12.5 cortical progenitors to BMP4 (Msx1 and Msx1-nlacZ RT-qPCR from Hu et al., 2008). (F) Msx1 ultrasensitivity in vivo at E10.5, as evidenced by its sharp border in normal embryos (arrows) and as isolated highly expressing cells following roof plate ablation (arrowheads) (Msx1 ISH from Hu et al., 2008).

Collectively, these observations suggest the following potential performance objectives for a BMP gradient in the preneurogenic dorsal telencephalon (Fig. 1.3). The BMP gradient might specify up to four discrete fates separated by three thresholds, which includes the positioning of two secondary organizers (CPE and cortical hem). Nonlinear conversions of graded BMP information would be needed for these thresholds. Coincidentally, the BMP orthologue dpp is thought to be responsible for three thresholds in both the Drosophila embryo and wing imaginal disc (Ashe et al., 2000; Affolter and Basler, 2007). During the neurogenic period, BMP gradients might also regulate transcriptional gradients within the cortex, which would require linear or sublinear interpretations of the BMP gradient.

1.3.2 Midline expression and homeogenetic expansion of BMP production Is BMP production sufficiently localized to generate a gradient in the dorsal telencephalon? Before and after neural tube closure in mice, at least six BMPs (BMP 2, 4, 5, 6, 7, and 12, which is also known as GDF7) are transcribed by roof plate or DTM cells (Furuta et al., 1997). The expression epicenter for all of these BMPs is the midline. Localized BMP production at the midline would be predicted to form a preneurogenic BMP gradient that is highest at the midline and lower more laterally (Fig. 1.3). Two additional features of BMP expression in the dorsal telencephalon increase the spatial and temporal ranges over which BMP gradients might act. First, some BMPs are expressed beyond the midline in the cortical primordium, and this

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

11

expression is also graded (Furuta et al., 1997). Second, early BMP-producing cells of the roof plate induce later BMPproducing CPE and hem cells (Currle et al., 2005), a form of “homeogenetic” induction (i.e., like inducing like) akin to those described in the midbrain, spinal cord, and Drosophila (Liem et al., 1995; Alexandre and Wassef, 2003; Bier and De Robertis, 2015). On the other hand, genetic lineage tracing suggests that lineage-based mechanisms do not play a major role in expanding BMP production, at least within progenitor domains (Currle et al., 2005). Intuitively, these mechanisms for expanding BMP production would be useful accompaniments to the evolutionary enlargement of the telencephalon.

1.3.3 BMP signaling gradient in the dorsal telencephalon Is there a BMP gradient in the dorsal telencephalon? As of now, there remains no direct evidence for BMPs themselves having a graded distribution. However, there is evidence for a gradient of BMP signaling based on the distribution of phosphorylated Smad1/5/8 (pSmad), the direct readout of BMP signaling. (Note: In the dorsal telencephalon, Smad1 and Smad5, but not Smad8, are probably the relevant Smads) (Arnold et al., 2006). At E10.5, the pSmad gradient peaks at the midline, where BMP production is highest, and exhibits a simple exponential decline away from the midline (Fig. 1.3C) (Cheng et al., 2006) with a length scale of 290 um (Srinivasan et al., 2014). The high dorsomedial-to-low ventrolateral (“DV”) orientation of this gradient is consistent with the midline-centered BMP production mentioned earlier. Perturbations to the system have confirmed the existence of this signaling gradient. First, genetic roof plate ablations, which reduce midline BMP production, lead to a corresponding reduction and flattening of the exponential pSmad gradient (Fig. 1.3C) (Cheng et al., 2006). Second, BMP4-soaked beads in dorsal telencephalic explants induce concentration-, position-, and orientation-dependent responses that imply an underlying BMP signaling gradient within the explanted tissue (Hu et al., 2008). Lastly, differences in the EC50 values in vitro and border positions in vivo for BMP target genes Msx1 and Msx2 lead to a BMP length scale calculation of 270 um, which agrees well with the 290 um value derived from the pSmad signaling gradient (Srinivasan et al., 2014). These observations on the BMP signaling gradient in the dorsal telencephalon have several implications. The exponential rather than power-law shape of the gradient suggests uniform clearance of BMPs from dorsal telencephalic neuroepithelium. Using the rough estimator of 3e5 times the calculated length scale (Srinivasan et al., 2014), the BMP signaling gradient might then pattern 800e1400 um, which would suffice for the entire dorsal telencephalon at preneurogenic stages. The shape of the gradient also makes good sense because multiple cell fate thresholds are needed toward the DTM, and it is easier to make thresholds where gradients are steepest. The exponential pSmad gradient further suggests that dorsal telencephalic cells interpret extracellular BMP concentration in a linearly proportional fashion. In the Drosophila embryo and wing imaginal disc, GFP-Dpp and pMad distributions are both best fit by simple declining exponentials, implying a linear graded relationship (Kicheva and Gonzalez-Gaitan, 2008). Although BMP distributions are unknown in the dorsal telencephalon, nuclear pSmad levels in cultured E12.5 cortical progenitors correlate in a graded and positive fashion with extracellular BMP4 concentration (Hu et al., 2008). The aforementioned explant studies (Hu et al., 2008) also imply that BMP4-soaked beads have a graded and additive effect on BMP signaling.

1.3.4 BMPs as dorsal telencephalic morphogens Are BMPs and/or BMP signaling necessary and sufficient for dorsal telencephalic fates? While thinking about BMP signaling in these “operational” terms provides limited insight into how the system actually works, the model falls apart in the absence of such evidence. Fortunately, there is substantial operational evidence to support the model. Cellular ablations established a requirement for BMP-producing roof plate cells in the specification of all three DTM fates (i.e., roof plate, CPE, and cortical hem) and in normal cortical patterning (Currle et al., 2005; Cheng et al., 2006). Importantly, telencephalon-specific inactivations of BMP receptors (BMPRIa and BMPRIb) demonstrated similar requirements for BMP signaling (Hebert et al., 2002; Fernandes et al., 2007; Caronia et al., 2010). While DTM fate specification does not occur following roof plate or BMP receptor ablation, cortex is specified, albeit abnormally patterned. Thus, DTM fates are more dependent on BMP signaling than cortexda graded requirement for BMP signaling that correlates well with its normal gradient in the dorsal telencephalon. In addition to the requirements for BMP signaling, sufficiency is the sine qua non for an instructive morphogen. Initial evidence for sufficiency came from explant studies, which demonstrated the ability of BMP4 and other BMPs to impart DTM-associated properties upon the cortical primordium (apoptosis, proliferation, and gene expression) (Furuta et al., 1997). BMP4 alone is also sufficient to rescue CPE fate in roof plateeablated explants (Cheng et al., 2006), to induce CPE ectopically in wild-type explants (Srinivasan et al., 2014), and to induce CPE fate in mouse and human ES cellederived

12

PART | I Induction and patterning of the CNS and PNS

systems (Watanabe et al., 2012). More recently, BMP4 has been shown to suffice for concentration-dependent and temporally-appropriate induction of CPE and cortical hem fates in a reduced culture system (Watanabe et al., 2016). Additional evidence for sufficiency comes from nestin-driven expression of constitutively active BMP type I receptors, which results in excessive simple epithelium resembling choroid plaque and CPE (although perhaps more choroid plaquelike because little to no papillary histology was observed) (Panchision et al., 2001). When applied to E12.5 cortical progenitors, BMP4 can induce concentration-dependent changes in gene expression that characterize CPE and choroid plaque (Hu et al., 2008) and even different EC50 values for DTM genes (Msx1 and Msx2) that reflect their border positions in vivo (Srinivasan et al., 2014). These findings suggest BMP4 sufficiency to specify multiple DTM fates and border positions in a concentration-dependent fashion, as expected for a morphogen.

1.3.5 Linear conversion of BMP signaling by cortical cells While the DTM is not required for cortical fate, it is required for normal DV patterning of the early cortex. As described earlier, the neurogenic cortical primordium is characterized by myriad transcription factor (TF) gradients rather than thresholds, and the orientations of many of these TF gradients align with that of the BMP signaling gradient. TF gradients are not apparent until after E10.5 and therefore after the specification of DTM and cortical fates and the onset of cortical neurogenesis. Although linearity of BMP-to-cortical TF regulation has not been rigorously established, it is difficult to imagine otherwise. The simple fact that cortical TF expression patterns are graded implies linear interpretations of BMP signaling, if BMP signaling indeed regulates cortical TF expression, and there is evidence for this being the case. The gradient of Emx2, an essential cortical patterning gene, matches that of the BMP signaling gradient (i.e., it also has a DV orientation). Correspondingly, the Emx2 gene has a Smad-binding enhancer that upregulates its expression (Theil et al., 2002). The cortical selector gene Lhx2 also has a DV gradient, and Lhx2 can be upregulated in explants at a distance from BMP4-soaked beads (Monuki et al., 2001). Following roof plate ablation, the Emx2 and Lhx2 gradients are reduced and flattened in a fashion that correlates strongly with the abnormal BMP signaling gradient (Cheng et al., 2006). Interestingly, the gradients of three other TFs (Pax6, Foxg1, and Ngn2) with opposing VD polarities are not affected by roof plate ablation, and the same selectivity for DV graded genes was observed at intermediate BMP4 concentrations in a reduced culture system (Watanabe et al., 2016). Together, these findings suggest that BMPs selectively upregulate DV genes and do so in a linear graded fashion.

1.3.6 Nonlinear conversion of BMP signaling by DTM cells In contrast to cortical progenitors, DTM progenitors must somehow convert graded BMP signaling into threshold TF and cell fate responses, and a nonlinear conversion mechanismdcell-intrinsic ultrasensitivity to BMP4dhas been described (Fig. 1.3DeF) (Hu et al., 2008). Ultrasensitivity describes any process that displays nonlinear switch-like behavior in response to a graded stimulus. In the dorsal telencephalon, ultrasensitivity to BMP4 occurs at the level of the well-known BMP target and DTM gene, Msx1, in dissociated cortical progenitors, in explants treated with BMP4-soaked beads, and in normal and roof plateeablated contexts in vivo (Hu et al., 2008). Other features of this ultrasensitivity phenomenondincluding its magnitude and immediate-early like kineticsdsuggest its utility for making crude initial borders in the DTM, which are then refined and sharpened. The mechanism underlying this form of ultrasensitivity turns out to involve mutually inhibitory interactions between BMPs and other signaling pathwaysdspecifically, the FGF and EGF signaling pathwaysdwhich occurs intracellularly above the level of individual BMP target genes. In the absence of FGF/EGF signaling, Msx1 (and Msx2) dose responses to BMP4 become entirely graded (Fig. 1.3E) (Srinivasan et al., 2014). Is there evidence for later refinement of borders in the dorsal telencephalon? The answer here is also “yes.” A common mechanism used to refine borders is cell sorting between neighboring cells, which results from differential cell affinities specified by selector genes. In the dorsal telencephalon, Lhx2 acts as a classic selector gene for the cortex (Mangale et al., 2008). Among its selector activities, Lhx2 specifies a cortical affinity state that differs from Lhx2negative cortical hem cells. These differential affinities result in sharp segregations of Lhx2-on (cortical) from Lhx2-off (hem) cells in induced mosaic contexts in vivo and in vitro, which suggests an Lhx2-dependent cell sorting process at the normal cortex-hem boundary. Similar roles have been ascribed to Pax6 at a lateral cortical border (pallialesubpallial boundary) (Stoykova et al., 1997; Gotz et al., 1998).

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

13

1.3.7 Summarydthe BMP signaling gradient The described findings lead to the following model of how the early dorsal telencephalon is patterned by BMP signaling: 1. Several BMPs are produced by cells of the midline roof plate, with BMP production then extending to secondary signaling centers and into the cortical primordium via homeogenetic induction. 2. The various BMP sources give rise to a declining exponential gradient of BMP signaling, suggesting linear interpretation of extracellular BMP concentration by dorsal telencephalic cells and uniform BMP clearance from the preneurogenic neuroepithelium. 3. DTM cells convert the graded BMP signaling information into three nonlinear, ultrasensitive thresholds that separate choroid plaque, CPE, cortical hem, and hippocampus. 4. DTM cells use a cell-intrinsic form of ultrasensitivity to BMP4, which requires mutually inhibitory interactions with FGF/EGF signaling as an initial nonlinear conversion mechanism. 5. Dorsal telencephalic borders are then refined by complementary mechanisms such as cell sorting, which is governed by selector genes such as Lhx2. 6. After cell fates and borders are specified, BMP signaling is interpreted in a linear graded fashion to regulate DV patterning of the cortical primordium. Many features of the system remain to be discovered and understood. In addition to elucidating homeogenetic induction and cell-intrinsic ultrasensitivity mechanisms further, it will be important to determine whether links exist between the BMP signaling gradient and other graded phenomena in the cortex, such as cell cycle length, self-renewal probability, and neurogenesis. Understanding better the specific BMPs involved (BMP homodimers and heterodimers) and their interactions with BMP inhibitors (e.g., noggin), transporters (chordin) and other morphogens should also be illuminating, as will studies on BMP diffusivity, clearance, and distribution. Fascinating questions about time, scaling of pattern to tissue size, and telencephalic growth regulation also remain to be answered. Systems-level approaches will be required to understand how the system works and why the system evolved the way it did. As the best-studied morphogen systems tell us, attaining a full quantitative understanding of dorsal telencephalic patterning will not be easy. Nonetheless, systemslevel understanding will undoubtedly provide insights into many important questions, including how morphogenetic patterning minimizes fragilities while maximizing robustness and the adaptability needed to increase neocortical size in humans and other species.

1.4 FGF8 as a morphogen in telencephalic patterning As noted previously, there is now substantial evidence that FGF8 and FGF17, from the RPC, pattern the neocortical area map (Fukuchi-Shimogori and Grove, 2001; Garel et al., 2003; Cholfin and Rubenstein, 2007, 2008; Toyoda et al., 2010), and this topic is the subject of a subsequent chapter. Most relevant here is evidence that FGF8 acts as a classic morphogen to pattern the neocortex and possibly the earlier-stage telencephalon. Increasing evidence supports an early role for FGF8 in the larger division of the telencephalic vesicle into a dorsal part that will generate the cerebral cortex and a ventral part that will generate subcortical nuclei. For example, in zebrafish, knocking down either Fgf8 or Fgf8 and Fgf3 expression together causes reduction or loss of ventral telencephalic cells (Shanmugalingam et al., 2000; Walshe and Mason, 2003). In mouse, Fgf3 appears less important (Theil et al., 2008), but reduced Fgf8 causes a striking reduction of the septum and lateral and medial ganglionic eminences (LGE, MGE) (Storm et al., 2006). Furthermore, in mice deficient in Fgfr1 and Fgfr2, the ventral telencephalon is greatly reduced, reflecting a shift from a ventral to a dorsal telencephalic cell fate (Gutin et al., 2006). Finally, characteristically ventral gene expression can be induced by FGF8 in dorsal telencephalic explants (Kuschel et al., 2003) and in dorsal telencephalon in vivo (Assimacopoulos, Taylor, and Grove, unpublished results). These findings may be surprising in the context of the well-established role for Shh in specifying ventral cell types in the spinal cord (Fuccillo et al., 2006). The relative contribution of Shh, FGF8, and other factors has been investigated recently by recapitulating ventral telencephalic development in mouse ES cell culture (Danjo et al., 2011). Cortical and subcortical cell identities were judged by patterns of expression of genes characteristic of cells in vivo. Application of moderate levels of Shh beginning early in the ES cell culture and continuing for several days induced LGE progenitors. Stronger Shh signaling specified MGE and caudal ganglionic eminence (CGE) progenitors; however, MGE specification also required FGF8, and the

14

PART | I Induction and patterning of the CNS and PNS

CGE required FGF15. The extent to which FGF8, FGF15, and Shh each control telencephalic ventralization in vivo needs additional study and is likely to involve an increased understanding of interactions between the Shh and FGF signaling pathways, as discussed further below. Once the neocortical primordium (NP) is established, FGF8 disperses from the RPC through the entire NP, creating a protein gradient with an exponential decline from anterior to posterior (Fig. 1.4) (Toyoda et al., 2010). If we assume that FGF8 disperses by diffusion from source cells, the exponential decline implies uniform clearance of FGF8 throughout the NP (see above). This has not yet been shown for the mouse NP, but FGF8 dispersion from a source followed by ubiquitous endocytosis and degradation of the protein has been demonstrated in zebrafish (Scholpp and Brand, 2004; Yu et al., 2009). The RPC becomes evident just after neural tube closure at about E9 in the mouse. At E9.5, when the FGF8 gradient was quantified from FGF8 immunofluorescence, the NP is about 300 microns long from anterior to posterior (Fig. 1.4D), or 70e80 cell widths (Toyoda et al., 2010), consistent with the size of a classic morphogenetic field (see above). The measured half-decay of FGF8 in the E9.5 NP occurred over about 10 cell widths (Toyoda et al., 2010). Given that as little as a twofold difference in morphogen concentration can induce cells to adopt different fates (Green and Smith, 1990), the half-decline of FGF8 compared with the total length of the NP at E9.5 suggests that several different area fates could be obtained from the estimated FGF8 gradient. Ectopic sources of myc-tagged FGF8 were introduced by electroporation, and FGF8-myc immunofluorescence showed graded dispersion from the electroporation site. FGF8-myc must have diffused from the new source or utilized another mechanism for dispersion that results in a gradient. Ectopic FGF8-myc sources induced expression of different FGF8 target genes at different distances indicating that upregulation of the expression of these target genes required different levels of FGF8 (Toyoda et al., 2010). A key feature of a morphogen is that it acts directly at a distance, rather than by a relay of several signaling molecules. To test this, a dominant negative FGF receptor with high in vitro affinity for FGF8 (Ornitz et al., 1996; Chellaiah et al., 1999; Zhang et al., 2006) was electroporated into the central region of the NP (Toyoda et al., 2010). The dominant negative receptor, dnFGFR3c, captured FGF8 at a distance from the FGF8 source, and the reduction of FGF8 induced cells to adopt a more posterior area identity, demonstrating direct, long-range patterning by FGF8 (Toyoda et al., 2010). These observations support FGF8 as a classic diffusible morphogen in neocortex.

1.5 Interactions among signaling centers in telencephalic patterning Evidence to date indicates that interactions among signaling centers in the telencephalon (1) regulate the expression of genes encoding specific signaling molecules in the other signaling center and (2) modify the activity of the other signaling center by regulation of a downstream signaling component.

(A)

E9.5

(E) ncxp 160

Intensity (AU)

140 iso

(B)

(C)

50 a

100

150

200

Distance in µm

250

R2 = 0.99

120 100 2

80 60

255 191

0

1

3 4

40

Ventricle

300 128 64

p

0

(D)

50

100

150

200

250

300

40 µm

Distance in µm

FIGURE 1.4 Characterization of the FGF8 gradient in the neocortical primordium at E9.5. (A) Sagittal midline section through one of nine mouse brains used in quantification. FGF8 IFl was measured in a standardized segment of the neocortical primordium (ncxp) (yellow band; white lines mark anterior and posterior ncxp boundaries). (B,C) FGF8 IFl (grayscale) in a digitally straightened segment shows an A/P intensity gradient (B), as does FGF8 IFl (false-colored) averaged from nine segments (C). (D) Plot of mean FGF8 IFl in arbitrary units (AU) against distance from the anterior FGF8 source. An IFl intensity plateau in the most anterior portion (C) was not plotted in (D). A declining exponential curve (red) was fitted to the remaining data (blue). The x and y axes of the plot in D do not start at zero to allow easier identification of half-decline points of the gradient. Maximum FGF8 IFl intensity is (1), sequential half-decline points are labeled (2), (3), and (4). The A/P distance of the half-decline is about 45 microns (see broken green lines in D), roughly the width of 10 DAPI-stained nuclei (white arrows, E). Scale bars: 0.1 mm in (A); 0.04 mm in (E).

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

15

1.5.1 FGF8, Shh, and BMP signaling Two companion papers described the first direct study of interactions among juxtaposed telencephalic signaling centers in chick and mouse (Crossley et al., 2001; Ohkubo et al., 2002). These studies proposed that FGF8, BMP4, and MGE-derived Shh regulate each other’s gene expression, and thereby the growth of the telencephalic vesicle, in a manner similar to interactions of FGF8, BMPs, and Shh in the limb bud (Ohkubo et al., 2002). BMP4 was found to repress Fgf8 and Shh expression, whereas Shh repressed Bmp expression and maintained expression of Fgf8. Both experimentally increased and decreased BMP signaling correlated with decreased growth of the telencephalic vesicles, suggesting that normal growth requires a balance of BMP, FGF8, and Shh signaling (Ohkubo et al., 2002).

1.5.2 Cross-regulation of BMP, FGF, and WNT signaling One function of interactions between the RPC and the cortical hem is to regulate the other signaling center’s boundaries. Increased FGF8 signaling decreases Wnt2b, 3a, and 5a expression in the hem, with a consequent reduction in the size of the hippocampus. Conversely, BMP signaling from the hem region in both chick and mouse decreases Fgf8 expression (Ohkubo et al., 2002; Shimogori et al., 2004). Thus, Wnt signaling in the hem is protected by BMP activity that holds back Fgf8 expression. Surprisingly, the BMP antagonist, noggin, can cause widespread upregulation of Fgf8 expression, suggesting that a large part of the cortical primordium is competent to express Fgf8 and would do so in the absence of active repression (Shimogori et al., 2004). Given that the cortical selector gene Lhx2 suppresses cortical hem and antihem tissue fates (Mangale et al., 2008), the NP might default into a conglomeration of hem, antihem, and RPC, were it not blocked from doing so by BMP signaling and Lhx2. The transcription factor, EMX2 (an ortholog of Drosophila empty spiracles), seems to be at the hub of interactions between FGF and Wnt signaling in the dorsal telencephalon. FGF8 downregulates expression of Emx2, whereas Wnt3a upregulates its expression (Fukuchi-Shimogori and Grove, 2001; Caronia-Brown et al., 2014). Emx2, in turn, regulates the two signaling centers (Fukuchi-Shimogori and Grove, 2003). Mice deficient in Emx2 have a combination of signaling center defects. Fgf8 and Fgf17 expression increases, expanding past normal limits of the RPC, and expression of Wnt2b, 3a, 5a, and 8b is reduced in the dorsomedial telencephalon, including the hem, as does the other downstream components of canonical Wnt signaling such as Lef1 (Fukuchi-Shimogori and Grove, 2003). Cell cycle control, likely to be influenced by Wnt signaling, is aberrant in the Emx2 mutant telencephalon, and the hippocampus is shrunken (Tole et al., 2000b; Muzio et al., 2002, 2005; Fukuchi-Shimogori and Grove, 2003; Shimogori et al., 2004; Cholfin and Rubenstein, 2008). Reducing excess FGF8 partially rescues both Wnt gene expression and the hippocampal fields (Shimogori et al., 2004), and reactivating Wnt signaling with lithium treatment partially restores normal cell cycle dynamics in the Emx2 mutant (Muzio et al., 2005). Whether reduced Wnt signaling in the Emx2 mutant mouse is caused directly by loss of Emx2 (Muzio et al., 2005), indirectly by increased FGF8 signaling (Shimogori et al., 2004), or both (Cholfin and Rubenstein, 2008) remains to be fully clarified. What seems clear is that the hem and RPC have opposite effects on the expression of Emx2 and that this is likely to shape the caudomedial high to rostral low gradient of Emx2 expression in the NP. As will be described in a subsequent chapter, the NP Emx2 gradient has a major role in neocortical area patterning.

1.5.3 Interactions of Shh, FGFs, and Gli3 Shh signaling is mediated by the Gli genes, and in the presence of Shh, all three Gli transcription factors work as transactivators. In the absence of Shh, however, Gli3 remains in a repressor state (Gli3R) (Fuccillo et al., 2006). Observations of the extra-toes mutant mouse, which lacks Gli3 altogether, indicate that Gli3R represses expression of Fgf8, Fgf17, and Fgf15 (Aoto et al., 2002; Rash and Grove, 2007). That is, in extra-toes, the expression domain of each of these Fgf genes greatly expands. Notably, in the extra-toes mouse, the cerebral cortex has a highly disrupted morphology; moreover gene expression typical of the ventral telencephalon extends into the rostrodorsal telencephalon (Tole et al., 2000a; Kuschel et al., 2003). Thus, under normal circumstances, the dorsal telencephalon is protected by Gli3R-mediated suppression from the ventralizing effects of FGF signaling (Kuschel et al., 2003).

1.6 Morphogens in human brain disease 1.6.1 Holoprosencephaly and Kallmann syndrome A major human disorder associated with defective morphogen-mediated patterning is holoprosencephaly (HPE, OMIM 236100), the most common congenital birth defect of the human forebrain (1 in 16,000 live births, 1 in 250 conceptions)

16

PART | I Induction and patterning of the CNS and PNS

(Monuki, 2007). HPE has long fascinated clinicians and scientists because of its striking phenotype, which can include a single forebrain “holosphere” rather than hemispheres (Fig. 1.5) and midline craniofacial defects such as a single midline eye (cyclopia). These phenotypes reflect the primary failures in midline induction that define HPE (Golden, 1998). Human HPE is divided into two categories with distinct phenotypes: (1) classic and (2) middle interhemispheric (MIH, also known as syntelencephaly). The more-common classic HPE is divided into alobar, semilobar, and lobar subtypes based on severity, with alobar being the most severe. A new HPE subtype, referred to as septopreoptic HPE (Hahn et al., 2010), probably represents a milder “form fruste” of classic HPE. Other HPE spectrum disorders include arrhinencephaly (absence of olfactory bulbs), septooptic dysplasia (OMIM 182230), and Kallmann syndrome (OMIM 308700), which combines anosmia with hypogonadotropic hypogonadism due to defects in olfactory and hypothalamic development. Human and animal genetics directly implicate signaling by four families of morphogensdNODAL, SHH, FGF, and BMPdin the induction and pathogenesis of HPE (Monuki, 2007). Although chromosomal abnormalities (most often trisomies) are more often the genetic culprits, over a dozen single genes have now been directly associated with human HPE (Table 1.1) (Solomon et al., 2011; Dubourg et al., 2016). These genes include several factors that affect NODAL and TABLE 1.1 Holoprosencephaly genes implicated in morphogen signaling. NODAL

SHH

FGF

BMP

NODAL

SHH

FGF8

Chordin

TDGF1/CRIPTO

PTCH1

FGFR1

Noggin

FOXH1/FAST1

DHCR7

Twsg1

TGIF1

GLI2

BmpRIa/Ib

Cyclops

DISP1

Squint

GAS1

One-eyed pinhead

CDON SUFU Smo

(ZIC2)

(SIX3)

(SIX3)

(ZIC2) (TGIF1)

(Megalin)

(Megalin) Genes or well-established modifiers from humans and experimental organisms linked to holoprosencephaly and/or cyclopic phenotypes, grouped by morphogen signaling pathway. Human genes are shown in bold upper case; genes with tentative placements are parenthesized and listed towards the bottom.

(A)

(B)

(C)

Bmp4

Fgf8

Shh

Nodal FIGURE 1.5 Morphogen interactions in holoprosencephaly. (A,B) Anterior views of the central nervous system from a normal 13-week fetus (A) and an 18-week gestation human fetus with alobar holoprosencephaly (B). The fetus with holoprosencephaly has a single forebrain vesicle (“holosphere”) and a cortex that is continuous across the ventrobasal, rostral, and dorsal midlines. (C) Morphogen interaction network that can account for classic holoprosencephaly (HPE) subtypes with variable extension to the dorsal midline, as well as MIH HPE, which selectively involves the dorsal midline region. Modified from Monuki, E.S., Walsh, C.A., 2001. Mechanisms of cerebral cortical patterning in mice and humans. Nat. Neurosci. 4 Suppl., 1199e1206; and Monuki, E.S., 2007. The morphogen signaling network in forebrain development and holoprosencephaly. J. Neuropathol. Exp. Neurol. 66, 566e575.

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

17

SHH signaling and more recently FGF8 and FGFR1 (Arauz et al., 2010; Rosenfeld et al., 2010; McCabe et al., 2011; Dubourg et al., 2016). The association of FGF8 mutations with both mild HPE and Kallmann syndrome support the concept that these malformations lie on a disease continuum. At this point, animal studies alone implicate BMP signaling in classic and MIH forms (Cheng et al., 2006; Fernandes et al., 2007). Major human HPE genes that do not encode for morphogens themselves (TGIF, ZIC2, and SIX3) regulate morphogens or morphogen signaling. For example, Tgif mutations disrupt Shh and Nodal signaling (Taniguchi et al., 2012), while Zic2 mutations result in defective node and prechordal plate development, thereby impacting Shh and Nodal signaling (Warr et al., 2008). Six3, on the other hand, is required for Shh transcriptional activation (Geng et al., 2008). The fundamental defect in HPE is failed midline induction (Golden, 1998), and all four of the morphogen families implicated are expressed at highest levels in midline tissues (Monuki, 2007). As discussed above, NODAL and SHH are both mainly produced by ventral midline tissues (prechordal, preoptic, and hypothalamic regions), FGF8 is expressed in the RPC, and BMPs are produced both in ventral midline (BMP7 in the prechordal plate) and DTM (roof plate, CPE, and cortical hem). Importantly, interactions between the four morphogen families can account for human HPE phenotypes and, in particular, for the long-standing conundrum of how primary defects in ventral signaling can lead to DTM failures (Fig. 1.5C) (Monuki, 2007). Briefly, and as detailed above, NODAL is upstream of SHH, which then engages in positive feedback with rostral FGF8. On the other hand, dorsal BMPs negatively regulate the SHH-FGF8 circuit, while FGF8 can have both positive and negative effects on dorsal BMPs. This morphogen signaling network can then explain how NODAL, SHH, and FGF8 signaling defects can extend to include the DTM region, but why this extension is quite variable (perhaps due to the more complex FGFedorsal BMP interactions), even within families harboring identical mutations. This network can also explain why rostral and ventral regions are not affected in MIH HPE (i.e., when dorsal BMP signaling is the suspected culprit).

1.6.2 Gradients in holoprosencephaly neuropathology With regard to the morphogen gradient hypothesis, it is important to consider aspects of HPE neuropathology that are themselves graded. The graded neuropathology that first comes to mind is the graded involvement of ventral versus dorsal forebrain regions in classic HPE, in which ventrobasal and rostral regions are more severely affected than dorsal and posterior ones. This spatial gradient is simplest to visualize in semilobar HPE, but it is evident in most classic HPE cases (perhaps with the exception of the most complete forms of alobar HPE). However, this gradient of neuropathology cannot be explained by the spatiotemporal gradient of any single morphogen. Instead, network interactions between midline morphogens must be invoked (Fig. 1.5C) (Monuki, 2007). Another set of gradients observed in HPEdwhich are more directly relevant to the morphogen gradient hypothesisdare the neuropathological gradients relative to the midline. Because all of the morphogens implicated in classic HPE are expressed at highest levels in midline tissues, it makes sense that HPE neuropathology is greatest toward the midline (Monuki, 2007). Equally important, neuropathologic severity diminishes away from the midline (e.g., neocortex is always present in severe classic HPE cases), and in milder forms, midline tissues such as olfactory bulbs or septopreoptic tissues appear to be the only structures involved. The graded neuropathology in MIH HPE is also consistent with the morphogen gradient hypothesis. MIH HPE patients uniformly have total to near-total absence of CPE in the lateral ventricles, while their hippocampi are often, but not always, small and dysplastic (Cheng et al., 2006). Thus, as in classic HPE, MIH HPE neuropathology is more severe medially (CPE) than laterally (hippocampus).

1.6.3 Gradients in other human brain disorders Although less well appreciated than in HPE, many developmental diseases exhibit characteristic gradients in neuropathology. One example is adrenoleukodystrophy (OMIM 300100), an X-linked peroxisomal disorder that results in widespread dysmyelination/demyelination of CNS white matter. Within the cerebrum, adrenoleukodystrophy has a characteristic neuropathological gradient, with posterior (occipital) white matter more severely affected than anterior (frontal). Another example is lissencephaly (OMIM 607423), a severe cortical malformation. The two genes most commonly associated with lissencephaly are LIS1 and DCX (also known as LISX1), and the neuropathologies due to mutations in these two genes exhibit opposing gradientsdLIS1-associated lissencephaly is posterior predominant, whereas DCX mutations cause anterior predominant lissencephaly (Guerrini and Marini, 2006). Patterning disorders of the brainstem also display AP gradients in neuropathology (Barkovich et al., 2009).

18

PART | I Induction and patterning of the CNS and PNS

These and other examples reflect an aspect of developmental brain disorders that remains poorly understood, but experimentally tractable in animal modelsdthe role of morphogenetic gradients in determining the graded patterns of disease. Whether primary morphogen gradients are responsible for the graded neuropathologies in these and other disorders is unknown, and if responsible, it remains to be seen whether the responsibilities are direct or indirect (e.g., morphogen gradients in the dorsal telencephalon could indirectly determine the graded lissencephaly phenotypes by regulating gradients within the cortical plate). Regardless, careful examinations of the neuropathologies in humans and animal models should provide clues to the potential gradients impinging upon disease processes and whether linear or nonlinear conversions of graded information are involved. In this regard, understanding how graded or switch-like signaling is achieved could also impact other human brain diseases linked to morphogens, but not to gradients of these morphogens per se (e.g., the role of SHH and WNT signaling in medulloblastoma).

References Affolter, M., Basler, K., 2007. The Decapentaplegic morphogen gradient: from pattern formation to growth regulation. Nat. Rev. 8, 663e674. Alexandre, P., Wassef, M., 2003. The isthmic organizer links anteroposterior and dorsoventral patterning in the mid/hindbrain by generating roof plate structures. Development 130, 5331e5338. Aoto, K., Nishimura, T., Eto, K., Motoyama, J., 2002. Mouse GLI3 regulates Fgf8 expression and apoptosis in the developing neural tube, face, and limb bud. Dev. Biol. 251, 320e332. Arauz, R.F., Solomon, B.D., Pineda-Alvarez, D.E., Gropman, A.L., Parsons, J.A., Roessler, E., Muenke, M., 2010. A hypomorphic allele in the FGF8 gene contributes to holoprosencephaly and is allelic to gonadotropin-releasing hormone deficiency in humans. Mol. Syndromol. 1, 59e66. Arnold, S.J., Maretto, S., Islam, A., Bikoff, E.K., Robertson, E.J., 2006. Dose-dependent Smad1, Smad5 and Smad8 signaling in the early mouse embryo. Dev. Biol. 296, 104e118. Ashe, H.L., Briscoe, J., 2006. The interpretation of morphogen gradients. Development 133, 385e394. Ashe, H.L., Mannervik, M., Levine, M., 2000. Dpp signaling thresholds in the dorsal ectoderm of the Drosophila embryo. Development 127, 3305e3312. Assimacopoulos, S., Grove, E.A., Ragsdale, C.W., 2003. Identification of a Pax6-dependent epidermal growth factor family signaling source at the lateral edge of the embryonic cerebral cortex. J. Neurosci. 23, 6399e6403. Bachler, M., Neubuser, A., 2001. Expression of members of the Fgf family and their receptors during midfacial development. Mech. Dev. 100, 313e316. Barkai, N., Shilo, B.Z., 2009. Robust generation and decoding of morphogen gradients. Cold Spring Harb. Perspect. Biol. 1, a001990. Barkovich, A.J., Millen, K.J., Dobyns, W.B., 2009. A developmental and genetic classification for midbrain-hindbrain malformations. Brain 132, 3199e3230. Bhardwaj, R.D., Curtis, M.A., Spalding, K.L., Buchholz, B.A., Fink, D., Bjork-Eriksson, T., Nordborg, C., Gage, F.H., Druid, H., Eriksson, P.S., Frisen, J., 2006. Neocortical neurogenesis in humans is restricted to development. Proc. Natl. Acad. Sci. U.S.A. 103, 12564e12568. Bier, E., De Robertis, E.M., 2015. EMBRYO DEVELOPMENT. BMP gradients: a paradigm for morphogen- mediated developmental patterning. Science 348. https://doi.org/10.1126/science.aaa5838. Bishop, K.M., Goudreau, G., O’Leary, D.D., 2000. Regulation of area identity in the mammalian neocortex by Emx2 and Pax6. Science 288, 344e349. Boveri, T., 1901. Die polarität von ovocyte, ei, und larve des strongylocentrus lividus. Zool. Jahrb. - Abt. Anat. Ontog. Tiere 384. Briscoe, J., Small, S., 2015. Morphogen rules: design principles of gradient-mediated embryo patterning. Development 142, 3996e4009. Carlson, J.M., Doyle, J., 2002. Complexity and robustness. Proc. Natl. Acad. Sci. U.S.A. 99 (Suppl. 1), 2538e2545. Caronia, G., Wilcoxon, J., Feldman, P., Grove, E.A., 2010. Bone morphogenetic protein signaling in the developing telencephalon controls formation of the hippocampal dentate gyrus and modifies fear-related behavior. J. Neurosci. 30, 6291e6301. Caronia-Brown, G., Yoshida, M., Gulden, F., Assimacopoulos, S., Grove, E.A., 2014. The cortical hem regulates the size and patterning of neocortex. Development 141, 2855e2865. Chellaiah, A., Yuan, W., Chellaiah, M., Ornitz, D.M., 1999. Mapping ligand binding domains in chimeric fibroblast growth factor receptor molecules. Multiple regions determine ligand binding specificity. J. Biol. Chem. 274, 34785e34794. Cheng, X., Hsu, C.M., Currle, D.S., Hu, J.S., Barkovich, A.J., Monuki, E.S., 2006. Central roles of the roof plate in telencephalic development and holoprosencephaly. J. Neurosci. 26, 7640e7649. Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H., Beachy, P.A., 1996. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407e413. Cholfin, J.A., Rubenstein, J.L., 2007. Patterning of frontal cortex subdivisions by Fgf17. Proc. Natl. Acad. Sci. U.S.A. 104, 7652e7657. Cholfin, J.A., Rubenstein, J.L., 2008. Frontal cortex subdivision patterning is coordinately regulated by Fgf8, Fgf17, and Emx2. J. Comp. Neurol. 509, 144e155. Crick, F., 1970. Diffusion in embryogenesis. Nature 225, 420e422. Crossley, P.H., Martin, G.R., 1995. The mouse Fgf 8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439e451. Crossley, P.H., Martinez, S., Ohkubo, Y., Rubenstein, J.L., 2001. Coordinate expression of Fgf8, Otx2, Bmp4, and Shh in the rostral prosencephalon during development of the telencephalic and optic vesicles. Neuroscience 108, 183e206.

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

19

Currle, D.S., Cheng, X., Hsu, C.M., Monuki, E.S., 2005. Direct and indirect roles of CNS dorsal midline cells in choroid plexus epithelia formation. Development 132, 3549e3559. Danjo, T., Eiraku, M., Muguruma, K., Watanabe, K., Kawada, M., Yanagawa, Y., Rubenstein, J.L., Sasai, Y., 2011. Subregional specification of embryonic stem cell-derived ventral telencephalic tissues by timed and combinatory treatment with extrinsic signals. J. Neurosci. 31, 1919e1933. Dekanty, A., Milan, M., 2011. The interplay between morphogens and tissue growth. EMBO Rep. 12, 1003e1010. Driever, W., Nusslein-Volhard, C., 1988a. The bicoid protein determines position in the Drosophila embryo in a concentration-dependent manner. Cell 54, 95e104. Driever, W., Nusslein-Volhard, C., 1988b. A gradient of bicoid protein in Drosophila embryos. Cell 54, 83e93. Dubourg, C., Carre, W., Hamdi-Roze, H., Mouden, C., Roume, J., Abdelmajid, B., Amram, D., Baumann, C., Chassaing, N., Coubes, C., FaivreOlivier, L., Ginglinger, E., Gonzales, M., Levy-Mozziconacci, A., Lynch, S.A., Naudion, S., Pasquier, L., Poidvin, A., Prieur, F., Sarda, P., Toutain, A., Dupe, V., Akloul, L., Odent, S., de Tayrac, M., David, V., 2016. Mutational sepctrum in holoprosencephaly shows that FGF is a new major signaling pathway. Hum. Mutat. 37, 1329e1339. Dyson, S., Gurdon, J.B., 1998. The interpretation of position in a morphogen gradient as revealed by occupancy of activin receptors. Cell 93, 557e568. Ferguson, E.L., Anderson, K.V., 1992. Decapentaplegic acts as a morphogen to organize dorsal-ventral pattern in the Drosophila embryo. Cell 71, 451e461. Fernandes, M., Gutin, G., Alcorn, H., McConnell, S.K., Hebert, J.M., 2007. Mutations in the BMP pathway in mice support the existence of two molecular classes of holoprosencephaly. Development 134, 3789e3794. Ferri, R.T., Levitt, P., 1995. Regulation of regional differences in the differentiation of cerebral cortical neurons by EGF family-matrix interactions. Development 121, 1151e1160. Freeman, M., Gurdon, J.B., 2002. Regulatory principles of developmental signaling. Annu. Rev. Cell Dev. Biol. 18, 515e539. Fuccillo, M., Joyner, A.L., Fishell, G., 2006. Morphogen to mitogen: the multiple roles of hedgehog signalling in vertebrate neural development. Nat. Rev. Neurosci. 7, 772e783. Fukuchi-Shimogori, T., Grove, E.A., 2001. Neocortex patterning by the secreted signaling molecule FGF8. Science 294, 1071e1074. Fukuchi-Shimogori, T., Grove, E.A., 2003. Emx2 patterns the neocortex by regulating FGF positional signaling. Nat. Neurosci. 6, 825e831. Furuta, Y., Piston, D.W., Hogan, B.L., 1997. Bone morphogenetic proteins (BMPs) as regulators of dorsal forebrain development. Development 124, 2203e2212. Galceran, J., Farinas, I., Depew, M.J., Clevers, H., Grosschedl, R., 1999. Wnt3a-/–like phenotype and limb deficiency in Lef1(-/-)Tcf1(-/-) mice. Genes Dev. 13, 709e717. Garcia-Bellido, A., 1975. Genetic control of wing disc development in Drosophila. Ciba Found. Symp. 0, 161e182. Garel, S., Huffman, K.J., Rubenstein, J.L.R., 2003. Molecular regionalization of the neocortex is disrupted in Fgf8 hypomorphic mutants. Development 130, 1903e1914. Geng, X., Speirs, C., Lagutin, O., Inbal, A., Liu, W., Solnica-Krezel, L., Jeong, Y., Epstein, D.J., Oliver, G., 2008. Haploinsufficiency of Six3 fails to activate Sonic hedgehog expression in the ventral forebrain and causes holoprosencephaly. Dev. Cell 15, 236e247. Goentoro, L.A., Reeves, G.T., Kowal, C.P., Martinelli, L., Schupbach, T., Shvartsman, S.Y., 2006. Quantifying the Gurken morphogen gradient in Drosophila oogenesis. Dev. Cell 11, 263e272. Golden, J.A., 1998. Holoprosencephaly: a defect in brain patterning. J. Neuropathol. Exp. Neurol. 57, 991e999. Gotz, M., Stoykova, A., Gruss, P., 1998. Pax6 controls radial glia differentiation in the cerebral cortex. Neuron 21, 1031e1044. Grove, E.A., Tole, S., Limon, J., Yip, L., Ragsdale, C.W., 1998. The hem of the embryonic cerebral cortex is defined by the expression of multiple Wnt genes and is compromised in Gli3-deficient mice. Development 125, 2315e2325. Green, J.B., Smith, J.C., 1990. Graded changes in dose of a Xenopus activin A homologue elicit stepwise transitions in embryonic cell fate. Nature 347 (6291), 391e394. Guerrini, R., Marini, C., 2006. Genetic malformations of cortical development. Exp. Brain Res. 173, 322e333. Experimentelle Hirnforschung. Gulacsi, A., Anderson, S.A., 2006. Shh maintains Nkx2.1 in the MGE by a gli3-independent mechanism. Cerebr. Cortex 16 (Suppl. 1), i89e95. Gutin, G., Fernandes, M., Palazzolo, L., Paek, H., Yu, K., Ornitz, D.M., McConnell, S.K., Hebert, J.M., 2006. FGF Signalling Generates Ventral Telencephalic Cells Independently of SHH. Development. Hahn, J.S., Barnes, P.D., Clegg, N.J., Stashinko, E.E., 2010. Septopreoptic holoprosencephaly: a mild subtype associated with midline craniofacial anomalies. AJNR Am. J. Neuroradiol. 31, 1596e1601. Hebert, J.M., Mishina, Y., McConnell, S.K., 2002. BMP signaling is required locally to pattern the dorsal telencephalic midline. Neuron 35, 1029e1041. Hoch, R.V., Clarke, J.A., Rubenstein, J.L., 2015. Fgf signaling controls the telencephalic distribution of Fgf-expressing progenitors generated in the rostral patterning center. Neural Dev. 10, 8. Holley, S.A., Neul, J.L., Attisano, L., Wrana, J.L., Sasai, Y., O’Connor, M.B., De Robertis, E.M., Ferguson, E.L., 1996. The Xenopus dorsalizing factor noggin ventralizes Drosophila embryos by preventing DPP from activating its receptor. Cell 86, 607e617. Hu, J.S., Doan, L.T., Currle, D.S., Paff, M., Rheem, J.Y., Schreyer, R., Robert, B., Monuki, E.S., 2008. Border formation in a Bmp gradient reduced to single dissociated cells. Proc. Natl. Acad. Sci. U.S.A. 105, 3398e3403. Kicheva, A., Gonzalez-Gaitan, M., 2008. The Decapentaplegic morphogen gradient: a precise definition. Curr. Opin. Cell Biol. 20, 137e143. Kim, A.S., Anderson, S.A., Rubenstein, J.L., Lowenstein, D.H., Pleasure, S.J., 2001. Pax-6 regulates expression of SFRP-2 and Wnt-7b in the developing CNS. J. Neurosci. 21, RC132.

20

PART | I Induction and patterning of the CNS and PNS

Kuschel, S., Ruther, U., Theil, T., 2003. A disrupted balance between Bmp/Wnt and Fgf signaling underlies the ventralization of the Gli3 mutant telencephalon. Dev. Biol. 260, 484e495. Lander, A.D., 2007. Morpheus unbound: reimagining the morphogen gradient. Cell 128, 245e256. Lander, A.D., Nie, Q., Wan, F.Y., 2002. Do morphogen gradients arise by diffusion? Dev. Cell 2, 785e796. Lander, A.D., Lo, W.C., Nie, Q., Wan, F.Y., 2009a. The measure of success: constraints, objectives, and tradeoffs in morphogen-mediated patterning. Cold Spring Harb. Perspect. Biol. 1, a002022. Lander, A.D., Gokoffski, K.K., Wan, F.Y., Nie, Q., Calof, A.L., 2009b. Cell lineages and the logic of proliferative control. PLoS Biol. 7, e15. Lee, S.M., Tole, S., Grove, E., McMahon, A.P., 2000. A local Wnt-3a signal is required for development of the mammalian hippocampus. Development 127, 457e467. Lehtinen, M.K., Zappaterra, M.W., Chen, X., Yang, Y.J., Hill, A.D., Lun, M., Maynard, T., Gonzalez, D., Kim, S., Ye, P., D’Ercole, A.J., Wong, E.T., LaMantia, A.S., Walsh, C.A., 2011. The cerebrospinal fluid provides a proliferative niche for neural progenitor cells. Neuron 69, 893e905. Li, G., Pleasure, S.J., 2005. Morphogenesis of the dentate gyrus: what we are learning from mouse mutants. Dev. Neurosci. 27, 93e99. Liem Jr., K.F., Tremml, G., Roelink, H., Jessell, T.M., 1995. Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 82, 969e979. Machon, O., Backman, M., Machonova, O., Kozmik, Z., Vacik, T., Andersen, L., Krauss, S., 2007. A dynamic gradient of Wnt signaling controls initiation of neurogenesis in the mammalian cortex and cellular specification in the hippocampus. Dev. Biol. 311, 223e237. Mallamaci, A., Muzio, L., Chan, C.H., Parnavelas, J., Boncinelli, E., 2000. Area identity shifts in the early cerebral cortex of Emx2-/- mutant mice. Nat. Neurosci. 3, 679e686. Mangale, V.S., Hirokawa, K.E., Satyaki, P.R., Gokulchandran, N., Chikbire, S., Subramanian, L., Shetty, A.S., Martynoga, B., Paul, J., Mai, M.V., Li, Y., Flanagan, L.A., Tole, S., Monuki, E.S., 2008. Lhx2 selector activity specifies cortical identity and suppresses hippocampal organizer fate. Science 319, 304e309. McCabe, M.J., Gaston-Massuet, C., Tziaferi, V., Gregory, L.C., Alatzoglou, K.S., Signore, M., Puelles, E., Gerrelli, D., Farooqi, I.S., Raza, J., Walker, J., Kavanaugh, S.I., Tsai, P.S., Pitteloud, N., Martinez-Barbera, J.P., Dattani, M.T., 2011. Novel FGF8 mutations associated with recessive holoprosencephaly, craniofacial defects, and hypothalamo-pituitary dysfunction. J. Clin. Endocrinol. Metab. 96, E1709eE1718. Meinhardt, H., 2009. Models for the generation and interpretation of gradients. Cold Spring Harb. Perspect. Biol. 1, a001362. Moldrich, R.X., Gobius, I., Pollak, T., Zhang, J., Ren, T., Brown, L., Mori, S., De Juan Romero, C., Britanova, O., Tarabykin, V., Richards, L.J., 2010. Molecular regulation of the developing commissural plate. J. Comp. Neurol. 518, 3645e3661. Monuki, E.S., 2007. The morphogen signaling network in forebrain development and holoprosencephaly. J. Neuropathol. Exp. Neurol. 66, 566e575. Monuki, E.S., Porter, F.D., Walsh, C.A., 2001. Patterning of the dorsal telencephalon and cerebral cortex by a roof plate-lhx2 pathway. Neuron 32, 591e604. Morgan, T.H., 1901. Regeneration and liability to injury. Science 14, 235e248. Muzio, L., Soria, J.M., Pannese, M., Piccolo, S., Mallamaci, A., 2005. A mutually stimulating loop involving Emx2 and canonical Wnt signalling specifically promotes expansion of occipital cortex and hippocampus. Cerebr. Cortex 15, 2021e2028. Muzio, L., DiBenedetto, B., Stoykova, A., Boncinelli, E., Gruss, P., Mallamaci, A., 2002. Emx2 and Pax6 control regionalization of the pre-neuronogenic cortical primordium. Cerebr. Cortex 12, 129e139. Ohkubo, Y., Chiang, C., Rubenstein, J.L., 2002. Coordinate regulation and synergistic actions of BMP4, SHH and FGF8 in the rostral prosencephalon regulate morphogenesis of the telencephalic and optic vesicles. Neuroscience 111, 1e17. Ornitz, D.M., Xu, J., Colvin, J.S., McEwen, D.G., MacArthur, C.A., Coulier, F., Gao, G., Goldfarb, M., 1996. Receptor specificity of the fibroblast growth factor family. J. Biol. Chem. 271, 15292e15297. Panchision, D.M., Pickel, J.M., Studer, L., Lee, S.H., Turner, P.A., Hazel, T.G., McKay, R.D., 2001. Sequential actions of BMP receptors control neural precursor cell production and fate. Genes Dev. 15, 2094e2110. Ragsdale, C.W., Grove, E.A., 2001. Patterning the mammalian cerebral cortex. Curr. Opin. Neurobiol. 11, 50e58. Rash, B.G., Grove, E.A., 2007. Patterning the dorsal telencephalon: a role for sonic hedgehog? J. Neurosci. 27, 11595e11603. Rogers, K.W., Schier, A.F., 2011. Morphogen gradients: from generation to interpretation. Annu. Rev. Cell Dev. Biol. 27, 377e407. Rosenfeld, J.A., Ballif, B.C., Martin, D.M., Aylsworth, A.S., Bejjani, B.A., Torchia, B.S., Shaffer, L.G., 2010. Clinical characterization of individuals with deletions of genes in holoprosencephaly pathways by aCGH refines the phenotypic spectrum of HPE. Hum. Genet. 127, 421e440. Rubenstein, J.L., Beachy, P.A., 1998. Patterning of the embryonic forebrain. Curr. Opin. Neurobiol. 8, 18e26. Sagner, A., Briscoe, J., 2017. Morphogen interpretation: concentration, time, competence, and signaling dynamics. Wiley Interdiscip. Rev. Dev. Biol. 6 https://doi.org/10.1002/wdev.271. Sansom, S.N., Hebert, J.M., Thammongkol, U., Smith, J., Nisbet, G., Surani, M.A., McConnell, S.K., Livesey, F.J., 2005. Genomic characterisation of a Fgf-regulated gradient-based neocortical protomap. Development 132, 3947e3961. Scholpp, S., Brand, M., 2004. Endocytosis controls spreading and effective signaling range of Fgf8 protein. Curr. Biol. 14, 1834e1841. Schwank, G., Basler, K., 2010. Regulation of organ growth by morphogen gradients. Cold Spring Harb. Perspect. Biol. 2, a001669. Shanmugalingam, S., Houart, C., Picker, A., Reifers, F., Macdonald, R., Barth, A., Griffin, K., Brand, M., Wilson, S.W., 2000. Ace/Fgf8 is required for forebrain commissure formation and patterning of the telencephalon. Development 127, 2549e2561. Shimamura, K., Rubenstein, J.L.R., 1997. Inductive interactions direct early regionalization of the mouse forebrain. Development 124, 2709e2718. Shimmi, O., Umulis, D., Othmer, H., O’Connor, M.B., 2005. Facilitated transport of a Dpp/Scw heterodimer by Sog/Tsg leads to robust patterning of the Drosophila blastoderm embryo. Cell 120, 873e886.

Morphogens, patterning centers, and their mechanisms of action Chapter | 1

21

Shimogori, T., Banuchi, V., Ng, H.Y., Strauss, J.B., Grove, E.A., 2004. Embryonic signaling centers expressing BMP, WNT and FGF proteins interact to pattern the cerebral cortex. Development 131, 5639e5647. Skeath, J.B., 1998. The Drosophila EGF receptor controls the formation and specification of neuroblasts along the dorsal-ventral axis of the Drosophila embryo. Development 125, 3301e3312. Solomon, B.D., Gropman, A., Muenke, M., 2011. Holoprosencephaly Overview. Spalding, K.L., Bhardwaj, R.D., Buchholz, B.A., Druid, H., Frisen, J., 2005. Retrospective birth dating of cells in humans. Cell 122, 133e143. Spemann, H., Mangold, H., 1924. Über Induktion von Embryonanlagen durch Implantation artfremder Organisatoren. Roux’ Arch. F. Entw. Mech. 100, 599e638. Srinivasan, S., Hu, J.S., Currle, D.S., Fung, E.S., Hayes, W.B., Lander, A.S., Monuki, E.S., 2014. A BMP-FGF morphogen toggle switch drives the ultrasensitive expression of multiple genes in the developing forebrain. PLoS Comput. Biol. 10, e1003463. Storm, E.E., Garel, S., Borello, U., Hebert, J.M., Martinez, S., McConnell, S.K., Martin, G.R., Rubenstein, J.L., 2006. Dose-dependent functions of Fgf8 in regulating telencephalic patterning centers. Development 133, 1831e1844. Stoykova, A., Gotz, M., Gruss, P., Price, J., 1997. Pax6-dependent regulation of adhesive patterning, R-cadherin expression and boundary formation in developing forebrain. Development 124, 3765e3777. Sussel, L., Marin, O., Kimura, S., Rubenstein, J.L., 1999. Loss of Nkx2.1 homeobox gene function results in a ventral to dorsal molecular respecification within the basal telencephalon: evidence for a transformation of the pallidum into the striatum. Development 126, 3359e3370. Taniguchi, K., Anderson, A.E., Sutherland, A.E., Wotton, D., 2012. Loss of Tgif function causes holoprosencephaly by disrupting the SHH signaling pathway. PLoS Genet. 8, e1002524. Theil, T., Dominguez-Frutos, E., Schimmang, T., 2008. Differential requirements for Fgf3 and Fgf8 during mouse forebrain development. Dev. Dynam. 237, 3417e3423. Theil, T., Aydin, S., Koch, S., Grotewold, L., Ruther, U., 2002. Wnt and Bmp signalling cooperatively regulate graded Emx2 expression in the dorsal telencephalon. Development 129, 3045e3054. Thomas, T., Dziadek, M., 1993. Capacity to form choroid plexus-like cells in vitro is restricted to specific regions of the mouse neural ectoderm. Development 117, 253e262. Tole, S., Ragsdale, C.W., Grove, E.A., 2000a. Dorsoventral patterning of the telencephalon is disrupted in the mouse mutant extra-toesJ. Dev. Biol. 217, 254e265. Tole, S., Goudreau, G., Assimacopoulos, S., Grove, E.A., 2000b. Emx2 is required for growth of the hippocampus but not for hippocampal field specification. J. Neurosci. 20, 2618e2625. Toyoda, R., Assimacopoulos, S., Wilcoxon, J., Taylor, A., Feldman, P., Suzuki-Hirano, A., Shimogori, T., Grove, E.A., 2010. FGF8 acts as a classic diffusible morphogen to pattern the neocortex. Development 137, 3439e3448. Turing, A.M., 1952. The chemical basis of morphogenesis. Philos. Trans. R. Soc. Lond. Ser. B Biol. Sci. 237, 37e72. von Ohlen, T., Doe, C.Q., 2000. Convergence of dorsal, dpp, and egfr signaling pathways subdivides the drosophila neuroectoderm into three dorsalventral columns. Dev. Biol. 224, 362e372. Walshe, J., Mason, I., 2003. Unique and combinatorial functions of Fgf3 and Fgf8 during zebrafish forebrain development. Development 130, 4337e4349. Warr, N., Powles-Glover, N., Chappell, A., Robson, J., Norris, D., Arkell, R.M., 2008. Zic2-associated holoprosencephaly is caused by a transient defect in the organizer region during gastrulation. Hum. Mol. Genet. 17, 2986e2996. Wartlick, O., Kicheva, A., Gonzalez-Gaitan, M., 2009. Morphogen gradient formation. Cold Spring Harb. Perspect. Biol. 1, a001255. Watanabe, M., Kang, Y.-J., Davies, L.M., Meghpara, S., Lau, K., Chung, C.-Y., Kathirya, J., Hadjantonakis, A.-K., Monuki, E.S., 2012. BMP4 sufficiency to induce choroid plexus epithelial fate from embryonic stem cell-derived neuroepithelial progenitors. J. Neurosci. 32, 15934e15945. Watanabe, M., Fung, E.S., Chan, F.B., Wong, J.S., Coutts, M., Monuki, E.S., 2016. BMP4 acts as a dorsal telencephalic morphogen in a mouse embryonic culture system. Biol. Open 5, 1834e1843. White, R.J., Nie, Q., Lander, A.D., Schilling, T.F., 2007. Complex regulation of cyp26a1 creates a robust retinoic acid gradient in the zebrafish embryo. PLoS Biol. 5, e304. Wolpert, L., 1969. Positional information and the spatial pattern of cellular differentiation. J. Theor. Biol. 25, 1e47. Yoshida, M., Assimacopoulos, S., Jones, K.R., Grove, E.A., 2006. Massive loss of Cajal-Retzius cells does not disrupt neocortical layer order. Development 133, 537e545. Yu, S.R., Burkhardt, M., Nowak, M., Ries, J., Petrasek, Z., Scholpp, S., Schwille, P., Brand, M., 2009. Fgf8 morphogen gradient forms by a source-sink mechanism with freely diffusing molecules. Nature 461, 533e536. Zhang, X., Ibrahimi, O.A., Olsen, S.K., Umemori, H., Mohammadi, M., Ornitz, D.M., 2006. Receptor specificity of the fibroblast growth factor family. The complete mammalian FGF family. J. Biol. Chem. 281, 15694e15700. Zhou, S., Lo, W.C., Suhalim, J.L., Digman, M.A., Gratton, E., Nie, Q., Lander, A.D., 2012. Free extracellular diffusion creates the dpp morphogen gradient of the Drosophila wing disc. Curr. Biol. 22, 668e675.

Chapter 2

Telencephalon patterning Shubha Tole1 and Jean He´bert2 1

Department of Biological Sciences, Tata Institute of Fundamental Research, Mumbai, Maharashtra, India; 2Neuroscience, Genetics, Stem Cells,

Albert Einstein College of Medicine, Bronx, NY, United States

Chapter outline 2.1. Introduction 2.2. Telencephalon induction 2.2.1. The anterior neural ridge 2.2.2. FGF signaling 2.2.3. Wnt antagonism 2.2.4. Interactions of low Wnt with FGFs and BMPs 2.3. Overview of early telencephalic subdivisions 2.4. Establishing dorsal versus ventral domains 2.4.1. Shh and Gli3, two key players 2.4.2. Foxg1 and FGFs cooperatively promote ventral development 2.4.3. Establishing the dorsal telencephalic domain 2.4.4. Sharpening the dorsaleventral border 2.4.5. The olfactory bulbs 2.5. Boundary structures as organizing centers and CR cell sources 2.5.1. Nomenclature of domains in the early telencephalic neuroepithelium 2.5.2. Specification of the hem and the antihem

23 24 24 24 26 26 27 27 27 30 30 31 32 32 32 33

2.5.2.1. Molecular mechanisms that act to position and specify the cortical hem 2.5.2.2. Molecular mechanisms that act to specify and position the antihem 2.5.3. CajaleRetzius cells arise from four telencephalic boundary structures 2.5.4. Organizer functions of telencephalic boundary structures 2.5.4.1. Rostral signaling center/septum 2.5.4.2. Antihem 2.6. Subdividing ventral domains 2.6.1. The striatum and pallidum 2.6.2. The amygdala 2.6.3. An evolutionary perspective for how the neocortex arose 2.6.4. Lineage and fate mapping in the ventral telencephalon 2.7. Conclusions Acknowledgments References

33 34 34 35 36 37 37 37 38 39 40 41 41 41

2.1 Introduction The neural networks of the adult cerebral hemispheres, which are one of the most complex structures known to us, underlie the vast range of human behaviors. Despite this complexity, the cerebral hemispheres start off during embryonic development as a simple sheet of neuroepithelial cells. This neuroepithelium constitutes the nascent telencephalon, located toward the anterior end of the neural plate. As development proceeds, the telencephalic neuroepithelium becomes patterned into distinct progenitor regions, which later give rise to specific neuronal subtypes, a process that is essential for the proper wiring of the cerebrum. Defects in these early patterning processes, even subtle ones, can result in serious intellectual and behavioral deficits. Here, how the neuroepithelium at the anterior end of the neural plate is specified to become the telencephalon is discussed. Also, the mechanisms that pattern the telencephalic neuroepithelium into the discrete progenitor domains destined to generate specific neuron subtypes are reviewed. Finally, what is known about the migration of neuroblasts from several progenitor domains and how this results in distinct combinations of neurons in each functionally different telencephalic area are examined.

Patterning and Cell Type Specification in the Developing CNS and PNS. https://doi.org/10.1016/B978-0-12-814405-3.00002-3 Copyright © 2020 Elsevier Inc. All rights reserved.

23

24

PART | I Induction and patterning of the CNS and PNS

A molecular framework is emerging that explains how the fates of precursor cells located in different areas within the telencephalic neuroepithelium and at different developmental stages are regulated. An interplay between cell-extrinsic factors secreted from signaling centers and cell-intrinsic factors in the neuroepithelium is central to the regulation of the rates of cell proliferation, differentiation, and apoptosis of telencephalic precursors and the types of neurons that they generate. Not surprisingly, as in most developing tissues throughout the body, the cell-extrinsic factors include members of the fibroblast growth factor (FGF), bone morphogenetic protein (BMP), and Wnt families as well as Sonic hedgehog (SHH). However, the cell-intrinsic factors, which are thus far mostly transcription factors, are more specific to the developing anterior nervous system and include factors encoded by genes such as Foxg1, Gli3, Pax6, Lhx2, Gsx2 (Gsh2), Nkx2.1, and Emx2. This chapter attempts to describe the critical genetic interactions that link extrinsic and intrinsic factors in telencephalon patterning. As well as being a fascinating undertaking, understanding the molecular mechanisms that regulate telencephalon development could be key in designing effective regenerative therapies for a range of forebrain disorders from developmental to degenerative ones. In particular, knowing what factors control the fates of neural precursors during development provides a framework for how specific types of telencephalic neurons could potentially be obtained using culture paradigms for regenerative purposes.

2.2 Telencephalon induction 2.2.1 The anterior neural ridge The initial formation of the telencephalon shares features with the induction of other tissues, such as the limbs, branchial arches, and midbrainehindbrain. Most notably, in each case, there is a discrete group of adjacent cells that acts as an organizer to induce the formation of these tissues. For the telencephalon, the inducing cells are those of the anterior neural ridge (ANR; found in mice) or anterior neural border (found in zebrafish; for simplicity, ANR is used henceforth). The ANR is located at the anterior end of the embryo and comprises the edge between the neuroectoderm and the underlying ectoderm. Cells at the rostrolateral end of the neural plate are fated to become the telencephalon (Cobos et al., 2001; Eagleson et al., 1995; Inoue et al., 2000; Hoch et al., 2015a). These cells turn on expression of Foxg1, a transcription factor gene of the forkhead family. RNA in situ hybridization analysis and lineage tracing using Foxg1Cre mice have shown that Foxg1 expression in the anterior neuroepithelium specifically marks telencephalic precursor cells and delineates most of the embryonic telencephalon (Hébert and McConnell, 2000; Shimamura and Rubenstein, 1997; Shimamura et al., 1995; Tao and Lai, 1992). In cultured explants of mice, if the ANR is dissected away from the anterior neuroepithelium, expression of Foxg1 fails to be induced (Shimamura and Rubenstein, 1997). Similarly, removal of ANR cells in zebrafish results in a failure to induce the normal expression of other telencephalic markers, emx1 and dlx2, two transcription factor genes expressed in presumptive dorsal and ventral telencephalic domains, respectively (Houart et al., 1998). Conversely, if the ANR is transplanted to more caudal regions of the neural plate, it induces ectopic expression of emx1, dlx2, and foxg1 (Houart et al., 1998, 2002). Taken together, these studies suggest that the ANR is necessary and sufficient to induce telencephalic character in the adjacent anterior neural plate. In addition to the ANR, cell ablation studies in mice and chicks and RNAi knockdown of Smad1 in chicks indicate that the nonneural ectoderm next to the ANR as well as the facial neural crest also play roles in inducing telencephalic tissue (Cajal et al., 2014; Aguiar et al., 2014).

2.2.2 FGF signaling Telencephalon induction also shares molecular mechanisms with other tissues. For several parts of the embryo, including the limbs and midbrainehindbrain, the induction of the tissue by an organizer seems to be mediated in part by FGFs. FGFsoaked beads placed ectopically in the presumptive diencephalic region or flank of chick embryos induce ectopic midbrains and limbs, respectively, whereas deletion of Fgf8 in the inducing cells leads to loss of the midbrain and limbs (Chi et al., 2003; Cohn et al., 1995; Crossley et al., 1996a,b; Lewandoski et al., 2000). FGF genes, including Fgf8, also are expressed in the ANR, where they likely mediate organizer activity for the telencephalon (Fig. 2.1). A bead soaked in FGF8 and placed on a cultured explant of anterior neural plate can induce Foxg1 expression (Shimamura and Rubenstein, 1997). Conversely, abolishing FGF signaling in the anterior neural plate by knocking out three FGF receptor genes, Fgfr1, Fgfr2, and Fgfr3, leads to loss of most or all Foxg1-expressing cells and a failure to form the telencephalon (Paek et al., 2009). This phenotype is not observed when any single FGF receptor gene or a pair of them is deleted, indicating that at the earliest stages of telencephalon development, Fgfr1, Fgfr2, and Fgfr3, can compensate for each other functionally

Telencephalon patterning Chapter | 2

(A)

(B) Head

Anterior

25

FGF Telencephalon

Tail Post.

Ant.

Posterior

High Wnt

FIGURE 2.1 Illustration of the central role of FGFs in telencephalon formation. FGFs induce cells to adopt a telencephalic fate at the neural plate stage, but the effects of reducing FGF signaling in the anterior neural plate are better depicted later, as in the midgestation (E12.5) mouse telencephalon (A). With decreasing FGF signaling, there is a progressive loss of telencephalic tissue starting with anterioreventralemedial areas progressing to posterioredorsalelateral ones (B, coronal views; C, sagittal views). Only the most dorsomedial area (choroid plexus and cortical hem) is spared. See text for details.

(Gutin et al., 2006). The same is true for the ligands (Fig. 2.2). Deletion of Fgf8 alone does not lead to loss of the telencephalon (Storm et al., 2006), indicating that other FGF ligand genes must be compensating for its loss (e.g., Theil et al., 2008). Consistent with this possibility, four other FGF ligand genes, Fgf3, Fgf15, Fgf17, and Fgf18, are expressed at the anterior end of the developing neural tube (Crossley and Martin, 1995; Maruoka et al., 1998; McWhirter et al., 1997; Shinya et al., 2001). Although it is clear that FGFs play a central role in the earliest steps of telencephalon formation, it has not been strictly demonstrated that they mediate classic organizer activity at this stage. Spemann and Mangold (1924) defined an organizer as a group of cells that has the potential to ectopically induce neighboring cells to form a normally structured tissue, usually with mirror-image symmetry to the normal tissue, as was the case in their experiments with tadpole body axis duplication and as is the case with midbrain duplication due to a bead of FGF8 placed in the presumptive diencephalic region (Crossley et al., 1996a). For telencephalon induction, this experiment would be technically difficult to execute in the mouse, although perhaps possible in the chick. Nevertheless, evidence discussed in the “Rostral signaling center/septum” subsection supports a role for FGFs as mediators of organizer activity in the telencephalon. What remains almost entirely unresolved is how FGF activity itself translates into the complex unfolding of telencephalon growth and patterning. The regulation of cell survival is likely to play some role. If either the ANR is ablated or FGF signaling is abolished in the anterior neural plate, all or most telencephalic precursor cells undergo apoptosis (Houart et al., 1998; Paek et al., 2009), similar to what occurs in limb and midbrainehindbrain precursors, for example, when Fgf8 is deleted. Why the cells die in the absence of FGF signaling is not understood. The answer may come from elucidating how the expression of genes that control cell survival and proliferation is regulated. For example, in glioblastoma cells and perhaps forebrain neuroepithelial cells, the cytostatic/proapoptotic gene Cdkn1a (p21Cip1) is directly regulated in its promoter region by intracellular mediators of both cytostatic and mitogenic factors (Seoane et al., 2004). The Cdkn1a promoter contains an

(A)

(B) Control B

Fgfr1/ 

Fgfr1 / ;2 / or Fgf8 /

Fgfr1 / ;2 / ;3 /

C

(C) Choroid plexus Cortical hem Dorsal Ventral FIGURE 2.2 An early somiteestage mouse embryo is used to illustrate two key factors at work in inducing the telencephalon. (A) Whole embryo showing the plane of section of the schematic in (B). Low Wnt signaling (yellow) along with FGFs (green) induce cells in the anteriorelateral neural plate to adopt a telencephalic fate (blue).

26

PART | I Induction and patterning of the CNS and PNS

activation element to which binds a FoxO/SMAD complex and a repressor element to which binds a cMYC/MIZ complex. In proliferating cells, Cdkn1a is repressed by cMyc/Miz (presumably in response to mitogenic factors). In response to TGFb signals, however, SMADs inhibit cMyc and form a complex with FoxO to activate Cdkn1a expression (Seoane et al., 2004). Remarkably, FoxG1 competes with the FoxO/SMAD complex for binding to its cis element in the Cdkn1a promoter. Whereas FoxO/SMAD binding promotes Cdkn1a expression, FoxG1 represses it (Seoane et al., 2004). FGFs and FoxG1 positively regulate each other’s expression (Martynoga et al., 2005; Paek et al., 2009; Shimamura and Rubenstein, 1997; Storm et al., 2003). Thus, when FGF signaling is abolished and Foxg1 is turned off, the expression of cytostatic/proapoptotic genes such as Cdkn1a may be derepressed in the anterior neural plate, leading to cell death. It remains to be seen if Cdkn1a itself and/or other genes with related functions are regulated by FGFs and other extracellular signals to control the survival and proliferation of anterior neural plate cells as they are induced to become telencephalic.

2.2.3 Wnt antagonism At least in zebrafish, FGFs are not the only factors released from the ANR that induce anterior neural plate cells to adopt a telencephalic fate. A secreted frizzled-related Wnt antagonist, Tlc, that is expressed in the ANR is necessary and sufficient to induce the telencephalon (Houart et al., 2002). Antisense morpholinos against tlc lead to a loss of the telencephalon, and conversely tlc-expressing cells can rescue the loss of telencephalon and cell death in ANR-ablated embryos (Houart et al., 2002). Moreover, tlc-expressing cells can induce ectopic expression of emx1 and foxg1 in more posterior neural tissue. Tlc likely acts by inhibiting Wnts as expected since transplanting Wnt-expressing cells into the ANR inhibits expression of telencephalic genes (Houart et al., 2002). Consistent with the idea that inhibiting Wnts is necessary to induce the telencephalon, zebrafish embryos mutant for the masterblind gene, which encodes Axin, a negative regulator of Wnt signaling, lack a telencephalon (Heisenberg et al., 2001; Masai et al., 1997; van de Water et al., 2001). Mouse embryos that are mutant for Six3, a direct repressor of certain Wnt genes, also lack a telencephalon, suggesting that Wnt antagonism is a conserved mechanism of telencephalon induction (Lagutin et al., 2003). However, Six3 is also required to promote Shh expression and ventral development of not only the telencephalon but also the rest of the forebrain (Geng et al., 2008). It also remains unclear if secreted frizzledrelated proteins emanate from the ANR to specify the telencephalon in species other than zebrafish. The source of the Wnts that are antagonized in telencephalon induction is unclear. The requirement for low Wnt signaling in forming the telencephalon may reflect a continuation of the earlier low-anterior to high-posterior Wnt gradient involved in anterioreposterior patterning of the neural plate, which has been well characterized across several species (Wilson and Houart, 2004) and which is discussed in the accompanying chapters on early neural patterning. Alternatively, a Wnt source internal to the anterior neural plate may play a role, perhaps the cortical hem primordium, which may be specified very early prior to neural tube closure.

2.2.4 Interactions of low Wnt with FGFs and BMPs Low levels of Wnt signaling appear necessary to promote appropriate levels of FGF gene expression in the anterior tip of the embryo. Tlc is both necessary and sufficient to promote fgf8 expression in anterior neural tissue of the zebrafish embryo (Houart et al., 2002). In embryos in which the ANR is ablated, tlc-expressing cells rescue not only fgf8 expression but also cell survival (Houart et al., 2002). Hence, the evidence to date indicates that low Wnt signaling acts upstream of FGFs in telencephalon induction, although it remains possible that FGFs also restrict Wnt signaling in a negative feedback loop. Other factors are also likely to be critical to telencephalon induction. For example, extracellular matrix components are probably required to set up gradients for Wnts and FGFs and potentiate their activities, as well as intracellular mediators of these signaling pathways. Only the surface has been scratched. How the ANR itself is formed is also a poorly understood process. BMP signaling is likely to play an important role. In bmp2b mutant zebrafish embryos, despite an expansion of the neural plate at the expense of ectoderm, the telencephalon fails to form (Barth et al., 1999). Moreover, expression of tlc at the anterior tip of the embryo is lost when an exogenous BMP inhibitor, Noggin, is added (Houart et al., 2002). These results together suggest that tlc expression and perhaps the ANR itself require for their development at least a threshold of BMP from the surrounding ectoderm. BMP signaling drives this effect by inhibiting the transcription factor Rx3, which induces the eye rather than telencephalon fate, and restricts cxcr4a expression, which acts to separate presumptive eye and telencephalon cells (Bielen and Houart, 2012).

Telencephalon patterning Chapter | 2

27

2.3 Overview of early telencephalic subdivisions Once the anterior neural plate acquires a telencephalic fate and expresses Foxg1, it becomes further subdivided into domains distinguishable by the expression of molecular markers. These include genes encoding transcription factors that are expressed in specific telencephalic subdomains, such as Nkx2.1, Gsx2, Pax6, and Emx2, as well as extracellular factors that are expressed in signaling centers at the edges of these subdomains, such as Shh, Fgfs, Wnts, and Bmps. In most or all cases, not only do these genes serve as useful markers but their functions are also essential in patterning the telencephalic neuroepithelium, as discussed in this chapter. Prior to neural tube closure, two broad ventral and dorsal domains can already be distinguished molecularly. These roughly prefigure the dorsal neuroepithelium that primarily generates glutamatergic neurons and the ventral one that primarily generates GABAergic neurons (Fig. 2.3). Shortly after neural tube closure, these two domains become further subdivided. The dorsal telencephalic domain (pallium) gets split into several regions. The largest of these gives rise to the neocortex, which then becomes patterned into different functional areas as described in a separate chapter. Specification of other cortical regions, such as the hippocampus located caudomedially, occurs as a result of the action of neighboring signaling centers. One of these signaling centers, the cortical hem, not only acts as an organizer for the hippocampus but also generates some of the earliest-born telencephalic neurons, the CajaleRetzius cells (CR; Meyer et al., 2002; Takiguchi-Hayashi et al., 2004; Yoshida et al., 2006). Progenitors at the border between the ventral and dorsal telencephalon and those at the rostromedial end also contribute CR cells to the telencephalon, as discussed below. The olfactory bulb (OB) located at the anterior tip of the telencephalon is also considered a dorsal derivative. The ventral telencephalic domain (subpallium) can be divided into two early regions: a medial part designated as the medial ganglionic eminence (MGE) and a posteriorelateral part that forms the lateral and caudal ganglionic eminences (LGE and CGE). Each ventral region generates specific GABAergic populations of neurons that either come to reside in the basal ganglia and associated limbic structures, including the amygdala and nucleus accumbens, or migrate long distances to populate all areas of the cortex. The MGE produces somatostatin, parvalbumin, and some neuropeptide Yeexpressing interneurons that populate the basal ganglia and cortex; the CGE produces a diversity of interneurons, including calretinin/vasoactive intestinal peptideeexpressing and Reelin-expressing ones; the LGE produces a large fraction of the OB interneurons, as well as GABAergic projection neurons that reside in several ventral areas including the striatum and limbic structures (Miyoshi et al., 2010; Nery et al., 2002; Wichterle et al., 2001; Wonders and Anderson, 2006). The partitioning of the telencephalic neuroepithelium into functionally different dorsal and ventral domains is due to the effect of secreted factors emanating from signaling centers, as discussed below.

2.4 Establishing dorsal versus ventral domains 2.4.1 Shh and Gli3, two key players Although the mechanisms that divide the early telencephalic neuroepithelium into presumptive dorsal and ventral domains remain superficially understood, key players have been identified (Fig. 2.4). These include both cell-intrinsic and mes di tel

Glutamatergic Cajal–Retzius GABAergic np ba h

FIGURE 2.3 Schematic of a midgestational mouse head depicting the domains of progenitor cells that generate broad subtypes of neurons. Dorsal progenitors generate glutamatergic neurons and ventral progenitors GABAergic ones. Progenitors in areas between the hemispheres and between the dorsaleventral areas give rise to the earliest-born neurons, the CajaleRetzius cells. ba, branchial arch; di, diencephalon; h, heart; mes, mesencephalon; np, nasal pit; tel, telencephalon.

28

PART | I Induction and patterning of the CNS and PNS

Ant.

Wnt3a

Dorsal

Gli3 Shh

Pax6 Nkx2.1

Posteriormedial

Pax6 Gsx2

Pax6

Neural plate Dorsal view

Ventral

Dorsal

Pax6

(anterior-lateral)

Emx

Wnt

Fgf Foxg1

Gsx2 Nkx2.1

MGE Pre-neurogenesis

Early neural tube

Gli3

LGE

Nkx2.1

Nkx2.1 Shh

Posterior

Shh

Emx2

Anteriorlateral

Gli3

Dorsal (posterior-medial)

LGE

MGE

FIGURE 2.4 Model for the establishment of dorsaleventral telencephalic domains. (A) Rough representation of the early expression patterns of key factors in the developing telencephalon, from the neural plate stage (dorsal view) to the early neural tube stage (cross section) to a stage that corresponds to midgestation (preneurogenesis) in the mouse. (B) Salient interactions between these factors, as well as FGFs and Foxg1, are presented (see text for details). Lhx2, a key early player in establishing the dorsal domain, is discussed in a separate section on the hem and antihem.

cell-extrinsic factors. Two of these factors are GLI3, a zinc finger transcription factor that acts to dorsalize the telencephalon, and SHH, a secreted signaling protein that acts to ventralize the telencephalon. The Gli3 gene, which was first identified as the classical mouse mutation Extra toes (Hui and Joyner, 1993), is initially expressed throughout the telencephalic neuroepithelium before becoming progressively restricted to the dorsal domain (Aoto et al., 2002; Corbin et al., 2003). In the Gli3 mouse mutant, the dorsal telencephalon fails to develop normally. The caudomedial areas, including the choroid plexus, cortical hem, and hippocampus, fail to form, and the neocortical area is progressively lost (Fig. 2.4; Grove et al., 1998; Kuschel et al., 2003; Theil et al., 1999; Tole et al., 2000). This phenotype is recapitulated by loss of Rpgrip1l a ciliopathy gene, which results in decreased GLI3R, underscoring the importance of cilia in distributing soluble factors such as Shh that pattern the telencephalon (Besse et al., 2011). The Shh gene is expressed in the midline of the neural plate and continues to be expressed in the ventral midline after neural tube closure (Echelard et al., 1993). In Shh / mouse embryos, all ventral telencephalic precursors are missing as assessed by loss of expression of ventral markers including Dlx2, Gsx2, and Nkx2.1 (Chiang et al., 1996; Fuccillo et al., 2004; Ohkubo et al., 2002; Rallu et al., 2002; Rash and Grove, 2007). Conversely, ectopic expression of Shh induces these ventral markers in the dorsal telencephalon of zebrafish and mice (Barth and Wilson, 1995; Ericson et al., 1995; Hauptmann and Gerster, 1996; Kohtz et al., 1998; Shimamura and Rubenstein, 1997). In addition to a lack of ventral cell types, the telencephalon of Shh / mutants is severely reduced in size because of its requirement in maintaining cell proliferation and survival (Ericson et al., 1995; Litingtung and Chiang, 2000; Ohkubo et al., 2002; Rowitch et al., 1999). Hence, it remains unclear what the relative contributions of cell death, reduced proliferation, and cell-fate transformation are to the loss of ventral precursor cells in the Shh mutant. Along with SHH itself, a range of factors required for SHH activity are also needed to generate ventral telencephalic precursors (e.g., factors involved in SHH protein processing, Huang et al., 2007; ciliogenesis, Ashique et al., 2009; or factors that genetically interact such as E2f4, Ruzhynsky et al., 2007). Mutations of Shh in both mice and humans lead to holoprosencephaly (Chiang et al., 1996; Roessler et al., 1996), a disorder in which medial areas of both forebrain and craniofacial tissues do not develop normally, leading to the incomplete separation of bilaterally symmetrical structures including the telencephalic hemispheres. The morphological defects within the telencephalon in Shh mutants in both mice and humans appear to extend beyond the ventral regions into the dorsal telencephalon in that the nascent hemispheres fail to separate dorsally (Chiang et al., 1996; Ohkubo et al., 2002; Solomon et al., 2010). However, this is not due to a failure of dorsomedial cell types to initially be generated, because they can still be identified in the Shh mutant, but perhaps due instead to a lack of overall growth and expansion of the hemispheres (Fernandes et al., 2007; Hayhurst et al., 2008; Rash and Grove, 2007). Moreover, humans carrying a SHH mutation rarely, if ever, show a variant of holoprosencephaly (midline interhemispheric holoprosencephaly) in which the dorsal areas are affected (Solomon et al., 2010).

Telencephalon patterning Chapter | 2

29

Remarkably, in double mutants lacking both Shh and Gli3, early ventral patterning is by and large rescued (Aoto et al., 2002; Rallu et al., 2002; Rash and Grove, 2007). Therefore, Shh controls the relative size of the dorsal and ventral domains of the telencephalon in large part by restricting the dorsalizing function of Gli3. In this way, rather than directly promoting ventral cell fates, Shh functions early on to promote ventral identity by preventing dorsalization. The rescue of early ventral development in the Shh mutant by elimination of Gli3 implies that other factors function independently or downstream of Shh to generate ventral precursor cells (Rallu et al., 2002) (Figs. 2.5 and 2.6).

DP MP

LGE MGE

LP VP

dLGE PSB FIGURE 2.5 Pallial domains: medial, dorsal, lateral, and ventral (MP, DP, LP, and VP). Subpallial domains: medial and lateral ganglionic eminences (MGE and LGE).

hem antihem

Wild type

Chimera with Lhx2 null patches

Lhx2 null

FIGURE 2.6 The hem (green) and antihem (red) are normally located at the medial and lateral ends of the cortical neuroepithelium. In chimeric brains, Lhx2 null cells scattered in the dorsal telencephalon take on hem fate if they are located medially, and antihem fate if they are located laterally. In brains that are completely Lhx2 null, the hem and the antihem occupy the entire extent of the dorsal telencephalon. From Mangale, V.S., Hirokawa, K.E., Satyaki, P.R., Gokulchandran, N., Chikbire, S., Subramanian, L., Shetty, A.S., Martynoga, B., Paul, J., Mai, M.V., Li, Y., Flanagan, L.A., Tole, S., Monuki, E.S., 2008 Lhx2 selector activity specifies cortical identity and suppresses hippocampal organizer fate. Science 319, 304e309.

FIGURE 2.7 Sources of CajaleRetzius cells: hem (green); antihem (red); septum (yellow); and thalamic eminence (blue).

30

PART | I Induction and patterning of the CNS and PNS

2.4.2 Foxg1 and FGFs cooperatively promote ventral development Foxg1 and Fgfs are both required for generating ventral precursors independent of Shh. The knockout of Foxg1 in mice or morpholino knockdown in zebrafish results in the loss of ventral precursor cells (Danesin et al., 2009; Dou et al., 1999; Martynoga et al., 2005; Xuan et al., 1995). Unlike the Shh / phenotype, however, in the Foxg1 / mutants, ventral cells are not rescued by removal of Gli3 (Hanashima et al., 2007), suggesting that Foxg1 acts genetically downstream of Shh and Gli3. Consistent with this interpretation, in zebrafish, shh misexpression cannot rescue ventral development in the foxg1 knockdowns, whereas foxg1 can partially rescue ventral cells in embryos in which SHH signaling is blocked (Danesin et al., 2009). Notably, the telencephalon is entirely lost in the Foxg1;Gli3 mouse double mutant, indicating that Gli3 and Foxg1 are essential for generating the dorsal and ventral subdivisions of the telencephalon, respectively (Hanashima et al., 2007). FOXG1 is likely to directly interact with other factors in generating ventral cell types, for example, TLE2 (Roth et al., 2010). Fgfs are also required for generating ventral telencephalic cells. In zebrafish, without fgf8 and fgf3 the ventral telencephalon does not form (Shanmugalingam et al., 2000; Shinya et al., 2001; Walshe and Mason, 2003). In mouse, ventral cells fail to be generated in Fgfr1 / ;Fgfr2 / double mutants and Fgf8 hypomorphic and null mutants (Gutin et al., 2006; Storm et al., 2006). FGFs induce all ventral regions independently of Shh, because even when Shh is expressed and active, no ventral structures develop if Fgfr1 and Fgfr2 are disrupted (Gutin et al., 2006). Moreover, FGF8-soaked beads ectopically induce expression of ventral markers in the dorsal telencephalon, even in the absence of SHH signaling (Kuschel et al., 2003), and electroporation of an Fgf8-expressing construct can rescue expression of ventromedial genes in Shh / mutants (Okada et al., 2008). As for the Foxg1 mutant and unlike for the Shh mutant, the loss of Gli3 does not rescue the loss of ventral cells in the Fgfr1;Fgfr2 mutant, placing FGFs genetically downstream of Gli3 (Gutin et al., 2006). In mice, Foxg1 and Fgfs promote each other’s expression in the nascent telencephalon, forming a positive feedback loop that promotes ventral development. In Foxg1 / mutants, Fgf8 expression is lost (Martynoga et al., 2005). Conversely, FGF signaling is both necessary and sufficient to promote Foxg1 expression: FGF8-soaked beads induce expression of Foxg1 in cultured telencephalic explants (Shimamura and Rubenstein, 1997), and Foxg1 expression is reduced in Fgf8 mutants and almost absent in the Fgfr1 / ;Fgfr2 / ;Fgfr3 / triple mutant (Paek et al., 2009; Storm et al., 2006). In addition, although Shh is required to maintain normal expression levels of both Foxg1 and Fgf genes, it does so indirectly. The reduction in Foxg1 expression observed in Shh / mouse mutants is due to the gain of GLI3-repressor activity because in Shh / ;Gli3 / double mutants Foxg1 expression recovers despite the absence of Shh (Rash and Grove, 2007). The maintenance of Fgf expression by Shh is also accomplished indirectly via repression of GLI3. Shh is required to promote and maintain the expression of Fgf3, Fgf8, Fgf15, Fgf17, and Fgf18 in the anterior medial telencephalon (Aoto et al., 2002; Ohkubo et al., 2002; Rash and Grove, 2007). In the Shh / ;Gli3 / mutant, however, Fgf expression is recovered because it is no longer repressed by GLI3. Moreover, the fact that Fgf expression is similarly expanded in both the Gli3 / and Shh / ;Gli3 / mutants confirms that Shh promotes Fgf expression only indirectly by repressing GLI3 function (Aoto et al., 2002; Kuschel et al., 2003; Rash and Grove, 2007; Theil et al., 1999). Overall, these studies suggest that ventral telencephalic cells are lost in the Shh / mutant because of the unchecked action of the GLI3-repressor, which results in loss of Fgf expression and reduced Foxg1 expression. Other genes also play roles in ventral development. Sox2, for example, is required for ventromedial telencephalon development by promoting Nkx2.1 and Shh expression (Ferri et al., 2013). Similarly, Otx2, which encodes a homeodomain transcription factor, regulates Fgf signaling in the early anterior ventral telencephalon (Hoch et al., 2015a,b). Finally, Six3 also promotes ventral telencephalon fates by promoting Foxg1 expression and repressing Wnt signaling (Carlin et al., 2012). The hierarchical organization and interplay between these factors remain largely unclear. And there are likely additional, yet to be identified, factors.

2.4.3 Establishing the dorsal telencephalic domain Although Foxg1 and Fgfs are absolutely required for generating the ventral telencephalon, they also participate in forming the dorsal telencephalon. As discussed above, when FGF signaling is completely abolished by knocking out three FGF receptor genes in the mouse, not only the ventral telencephalon but also the dorsal domain is missing (Paek et al., 2009). Similarly, in Foxg1 / mutants, in addition to the lack of ventral cells, the dorsal area is reduced in size (Xuan et al., 1995), although this is less severe than with the complete loss of FGF signaling. Loss of dorsal cells in the Foxg / mutant is due in part to the premature differentiation of precursor cells into early-born neurons and in part to the loss of anteriorelateral precursor cells, as the remaining dorsal region appears to acquire medialecaudal features (Hanashima et al., 2004, 2007; Muzio and Mallamaci, 2005; Xuan et al., 1995). Notably, foxg1 has also been shown in zebrafish embryos to restrict dorsal

Telencephalon patterning Chapter | 2

31

development by directly repressing wnt8b expression (Danesin et al., 2009). However, whereas in Foxg1 / mouse embryos Fgf8 expression is lost, in zebrafish embryos treated with foxg1 morpholinos fgf8 is upregulated (Danesin et al., 2009; Martynoga et al., 2005). This difference could be accounted for by either a difference in species or a dose-dependent effect of Foxg1, as a null mutation is used in the mouse, whereas in zebrafish, foxg1 may still be expressed at a low level. Nevertheless, Foxg1 appears to have dual roles in the dorsal telencephalon: (1) expanding its size in conjunction with FGFs (mainly anteriorelateral regions) and (2) restricting posterioremedial cell fates by restricting Wnt expression. Consistent with a role for Wnts in telencephalic dorsalization, Wnts appear necessary and sufficient, together with FGFs, to induce progressively more mature dorsal identities to telencephalic precursor cells in chick embryos (Gunhaga et al., 2003). In this process, Wnts appear to act through the canonical intracellular signaling pathway, because in mice, gain- and loss-of-function mutations in b-catenin, an intracellular effector of canonical Wnt signaling, lead to gain and loss of dorsal telencephalic cell identities, respectively (Backman et al., 2005). Wnts may not be the only factors emanating from the dorsomedial telencephalic area that promote dorsal identity. BMP4- and BMP5-soaked beads placed in the ventral forebrain of chick embryos disrupt the ventral identity of surrounding cells and loss of megalin in mice, which results in increased Bmp4 expression, leads to a loss of ventral precursors (Golden et al., 1999; Spoelgen et al., 2005). However, more direct evidence for a requirement for BMPs in specifying dorsal identity is still lacking. In fact, dorsal precursors appear to be specified normally in mouse mutants lacking both Bmpr1a and Bmpr1b, although functional compensation by a more distantly related Bmpr type I gene cannot be excluded (Fernandes et al., 2007). It is interesting to note that Wnts repress telencephalon induction early on (as discussed in the Telencephalic Induction section above), but later promote dorsal telencephalic identities. In both cases, however, Wnt signaling may be acting in a temporal continuum to keep Foxg1 expression in check: early on repressing it in caudal regions of the neural plate and later repressing it in the dorsomedial regions of the telencephalon itself. In fact, forced expression of Foxg1 in the dorsomedial cortical hem region results in a fate change from hem-derived CR neurons to hippocampal neurons (Liu et al., 2018). Foxg1 is, of course, not the only transcription factor gene required to establish and maintain a cortical fate in the dorsal telencephalon. Pax6, Emx2, Lhx2, and Gli3 are also key determinants in regulating cortical fate (Godbole et al., 2017; Hasenpusch-Theil et al., 2017; and see Section 3.4.4 below), and other factors may contribute to this network in ways that remain to be explored.

2.4.4 Sharpening the dorsaleventral border In inducing dorsal identity to telencephalic cells in chick embryos, Wnt3A maintains expression of Pax6 (Gunhaga et al., 2003). Pax6, a paired-box transcription factor gene, is the key to establishing dorsal identities. Moreover, Pax6 is essential in sharpening the border that sets the dorsal apart from the ventral telencephalon. Pax6 is first expressed throughout the neuroepithelium of the neural plate that is destined to form the telencephalon (Inoue et al., 2000). After neural tube closure, its expression becomes more restricted to the presumptive dorsal telencephalon at the same time that another transcription factor gene, Nkx2.1, is upregulated ventrally (Corbin et al., 2003). Hence, the border between Pax6 and Nkx2.1 expression initially defines presumptive dorsal and ventral domains. Shortly after, the regions of Pax6 and Nkx2.1 expression are split by a domain of expression of Gsx2, another homeobox transcription factor gene (Corbin et al., 2003). Pax6 and Gsx2 then maintain a mirrored pattern of expression, with Pax6 showing a low-dorsal to high-ventral gradient and Gsx2 a low-ventral to high-dorsal gradient. The dorsaleventral boundary at this point is defined by the limits of expression of Pax6 and Gsx2, which slightly overlap at their highest points of expression (Corbin et al., 2000; Stoykova et al., 2000; Toresson et al., 2000; Yun et al., 2001). Pax6 and Gsx2 act antagonistically to position the dorsaleventral border: in Pax6 / mouse embryos, the border area of the dorsal telencephalon is ventralized and in the Gsx2 / mouse embryos, the border area of the ventral telencephalon is dorsalized (Corbin et al., 2000; Stoykova et al., 2000; Toresson et al., 2000; Yun et al., 2001; Cocas et al., 2011). Moreover, in Pax6 / ;Gsx2 / double mutants, patterning of the dorsaleventral border is not as severely disrupted compared to either single mutant, illustrating the functional antagonism between these two genes (Toresson et al., 2000). The function of Pax6 at the pallialesubpallial boundary (PSB) may also help explain how the actions of Foxg1 and FGFs can so abruptly transition from generating ventral cells, including Gsx2-expressing ones, to dorsal cells. However, it still remains to be determined whether deletion of Pax6 can rescue the loss of ventral cells observed in Foxg1 and FGF mutants and where Pax6 stands in the regulatory network. Loss of Pax6 can, however, partially rescue the Shh / phenotype, although to a lesser extent than loss of Gli3. Nevertheless, this suggests that Gli3 and Pax6 interact to promote dorsal fates at the boundary (Fuccillo et al., 2006). In addition, although Gli3 is not required to initiate Pax6 expression, it is required in part to maintain it (Aoto et al., 2002; Kuschel et al., 2003; Theil et al., 1999). The dorsalizing functions of Pax6 and Gli3 also involve interactions with Emx2, an

32

PART | I Induction and patterning of the CNS and PNS

empty spiracles-related transcription factor gene required for dorsal patterning. In Gli3 mutants, Emx2 expression is lost, suggesting that Emx2 lies genetically downstream of Gli3 (Theil et al., 1999). The requirement for Gli3 in promoting Emx2 expression, however, is likely indirect through activation of BMPs and Wnts (Theil et al., 2002). Importantly, in embryos mutant for Pax6 and Emx2, dorsal precursor cells fail to adopt or maintain a cortical fate and instead assume at least in part a ventral fate (Muzio et al., 2002). One copy of either gene is sufficient to maintain a cortical identity. It is important to note that the size of the dorsal telencephalon in the double mutant is drastically smaller, suggesting that Emx2 and Pax6 are required to maintain not only the cortical fate of these cells but also their proliferative state. Nevertheless, the fact that loss of Pax6 only partially rescues the Gsx2 / phenotype suggests that other factors participate in forming or maintaining the dorsaleventral boundary. As mentioned, Emx2 also plays a role in enforcing the dorsaleventral boundary, particularly in conjunction with Dmrta2 and Dmrt3, genes that encode two zinc finger transcription factors and are important in boundary formation and cortical specification (Desmaris et al., 2018; Konno et al., 2012; Saulnier et al., 2013). In addition, mouse embryos that lack the orphan nuclear receptor gene, Tlx, have a slightly ventralized telencephalon with a loss of features of the dorsaleventral border, a phenotype that is worsened when combined with the Pax6 / mutant (Stenman et al., 2003). Retinoic acid may also be required for inducing characteristics of the dorsaleventral border, as shown in chick embryos treated with a retinoic acid signaling inhibitor (Marklund et al., 2004). However, genetic studies in the mouse suggest that retinoic acid is not crucial for early telencephalon development, as embryos in which no retinoic acid signaling activity can be detected in the forebrain area, Raldh2 / ;Raldh3 / mutants, exhibit normal patterning (Molotkova et al., 2007). Other factors that play roles in the dorsaleventral patterning of the telencephalon undoubtedly remain to be identified.

2.4.5 The olfactory bulbs The OB, a dorsal derivative like the cortex, is composed of dorsally derived, Tbr1- and Emx1-expressing projection neurons and ventrally derived interneurons. Consistent with being a dorsal derivative, the OB is missing in Gli3 mutants (Franz, 1994). Interestingly, if either dorsally derived projection neurons, which require Tbr1 to form, or ventrally derived interneurons, which require Dlx genes to form, are lacking, a bulb still manages to form, albeit smaller and disorganized in its cellular arrangements (Bulfone et al., 1998). However, mutations in genes such as Fgfr1, Pax6, or Lhx2 that do not directly affect the production of either neuronal type disrupt OB morphogenesis, offering clues to how the process of bulb evagination may be regulated (Hébert et al., 2003; Jimenez et al., 2000; Nomura and Osumi, 2004; Saha et al., 2007). It is also worth mentioning that the OBs are unique in their high rate of interneuron turnover throughout life. Newborn interneurons arise in one of the key sites of adult neurogenesis in the brain, the anterior subventricular zone, and migrate into the OB throughout adulthood using a well-characterized route, the rostral migratory stream. This system beautifully illustrates parallels between developmental processes and “maintenance” phenomena that play out in maturity.

2.5 Boundary structures as organizing centers and CR cell sources As described in an earlier section, Fgf signaling from the rostral region of the neural plate (the ANR) is responsible for early patterning events in the telencephalon. With neural tube closure, Fgf expression is now found in a rostral signaling center, a region that will become the septum (henceforth called “septum” with the understanding that the signaling is likely to be initiated before the septum actually forms). Fgf8 signaling from this center is responsible for patterning ventral areas and for specifying cortical area identities and demonstrates the properties of an “organizer” (Fukuchi-Shimogori and Grove, 2001). Fgf8 signaling also imposes a rostral limit on the extent of a medial signaling center, the cortical hem (Shimogori et al., 2004). The hem itself serves as an organizer for hippocampal induction. These properties are discussed in detail in a later section. First, we will discuss how the hem and antihem are formed. Then we will focus on a unique feature of these telencephalic signaling centers, that they each produce the earliest-born neurons of the cortex, CR cells.

2.5.1 Nomenclature of domains in the early telencephalic neuroepithelium By the time the neural tube is closed, patterning mechanisms that have begun prior to closure bring about its division into distinct pallial (dorsal) and subpallial (ventral) domains. The dorsal telencephalic neuroepithelium can be divided into four different types of pallial tissue based on gene expression patterns. The medial pallium (MP) contains the hem and the hippocampal primordium; the dorsal pallium (DP) corresponds to the neocortical primordium; the lateral pallium (LP) is thought to give rise to the piriform cortex; and the ventral pallium (VP) together with the LP contributes to specific components of the claustroamygdaloid complex (Puelles et al., 2000; Medina et al., 2004). The ventricular zone of the VP

Telencephalon patterning Chapter | 2

33

is also identified as the antihem (Assimacopoulos et al., 2003). An adjacent subpallial region, the dLGE (dorsal portion of the lateral ganglionic eminence), is also thought to contribute to the amygdaloid complex (Medina et al., 2004; Tole et al., 2005). The VP and the dLGE lie on either side of the PSB (Yun et al., 2001) and share an enriched expression of Pax6 in the ventricular zone. A prominent palisade of radial glial fibers delineates the PSB, originating in a region of the ventricular zone termed the “corticostriatal junction” (Molnár and Butler, 2002), and extending up to the pial surface in the region of the piriform cortex and amygdala. The hem and the antihem were proposed to be important embryonic signaling centers on the basis of their locations, flanking the cortical neuroepithelium, and their enriched expression of several different types of signaling molecules. The hem is identified by specific expression of Wnt family members 2b, 3a, and 5a (Grove et al., 1998). It also expresses several members of the Bmp gene family in broader domains that include the adjacent choroid plexus and/or hippocampal primordium (Furuta et al., 1997; Grove et al., 1998). The antihem expresses epidermal growth factor (EGF) family members, a fibroblast growth factor Fgf7, as well as a Wnt signaling inhibitor (Kim et al., 2001; Assimacopoulos et al., 2003). Of the several EGF family members, ligands Tgfa, Nrg1, and Nrg3 are concentrated at the antihem. Egf is itself expressed throughout the ventral neuroepithelium, but is not concentrated at the PSB (Assimacopoulos et al., 2003; Kornblum et al., 1997). The antihem neuroepithelium also expresses the secreted frizzled-related gene, Sfrp2 (Kim et al., 2001; Assimacopoulos et al., 2003). Members of this family bind directly to Wnts and since they are secreted, act as Wnt antagonists (Rattner et al., 1997; Kawano and Kypta, 2003; Ladher et al., 2000). Sfrp2 is intensely expressed in the antihem, and more weakly in the rest of the telencephalic neuroepithelium (Kim et al., 2001; Assimacopoulos et al., 2003).

2.5.2 Specification of the hem and the antihem Along the rostrocaudal axis, the antihem is more pronounced rostrally, appearing at levels anterior to the hem. In contrast, the hem is seen from midlevels and is most prominent caudally, persisting at levels where the antihem is no longer present (Grove et al., 1998; Assimacopoulos et al., 2003; Mangale et al., 2008). These positions parallel the graded expression of developmental control molecules in the telencephalon: Pax6 is expressed in a rostrolateral(high) to caudomedial(low) gradient, whereas Lhx2 and Emx2 show the opposite gradients (Bishop et al., 2000; Nakagawa and O’Leary, 2001). Pax6 is required for the specification of the antihem (Kim et al., 2001; Assimacopoulos et al., 2003). Lhx2 suppresses both hem and antihem fate, and both structures are expanded in the Lhx2 mutant (Bulchand et al., 2001; Monuki et al., 2001; Mangale et al., 2008). The hem and the antihem are noncortical in that they do not contribute to the hippocampus or the neocortex. Their derivatives are described in a later section. However, cells of the prospective cortical primordium take on either hem or antihem fate in the absence of Lhx2, revealing a fundamental commonality between these two fates. Studies using embryonic stem cell chimeras have demonstrated that Lhx2 null cells become hem if located medially, and antihem if located laterally (Mangale et al., 2008). It is unknown how this positional control of hem versus antihem fate choice is regulated. Attractive candidates are earlyexpressing transcription factors that are themselves graded in expression, such as Pax6 and Emx2 (Muzio et al., 2002), or Foxg1, which suppresses hem fate, and appears to be required for lateral fates including that of the antihem (Dou et al., 1999; Muzio and Mallamaci, 2005). In the dorsal telencephalon of the Foxg1 mutant, medial fates are expanded and lateral fates are missing (Muzio and Mallamaci, 2005). In mosaic embryos created by tamoxifen-induced gene disruption of Lhx2, medially located Lhx2 null patches do not express Pax6 or Foxg1, whereas laterally located Lhx2 null patches express both these genes (Mangale et al., 2008). While this is entirely consistent with the requirement of Pax6 and Foxg1 for antihem fate, it still does not explain how these differences between medial and lateral Lhx2 null patches is brought about in the first place. This remains a fundamental open question: how the early patterning of the telencephalon is brought about, such that distinct signaling centers flank a territory of “responding” tissue, the cortical neuroepithelium.

2.5.2.1 Molecular mechanisms that act to position and specify the cortical hem Several molecular models have been proposed to explain position of the cortical hem in the telencephalon. These studies seek to explain the mechanisms which define the rostrocaudal as well as the mediolateral boundaries of this signaling center. Fgfs expressed by the ANR are fundamental regulators of midline patterning, acting to establish the midline domain within the telencephalon prior to invagination (Okada et al., 2008). The interregulation between early patterning signals is not entirely well understood, however. On the one hand, it appears that Bmps from the roof plate can restrict the extent of the Fgf expression and are themselves repressed by the Fgfs (Ohkubo et al., 2002; Shimogori et al., 2004; Storm et al., 2006). On the other hand, however, in Bmpr1a;Bmpr1b double mutants, there is no significant increase in Fgf8 expression. Both the hem and the choroid plexus are lost, however, indicating that the hem requires Bmp signaling

34

PART | I Induction and patterning of the CNS and PNS

(Fernandes et al., 2007). Data from Fgf8 hypomorphs or nulls suggest that Fgf8 promotes Bmp4 expression at the dorsal midline at E9.5, whereas by E11.5 Fgfs inhibit Bmp4 expression (Storm et al., 2003, 2006). With loss of Fgf signaling, the telencephalon is progressively reduced in size as one or both copies of Fgfr1, 2, and 3 are removed such that in Fgfr1/2/3 triple mutants, there is only a small telencephalic remnant detectable. Strikingly, midline fates, in particular the Wnt3a-expressing domain and the adjacent choroid plexus, are spared, indicating that the hem and its choroid plexus derivative can arise independent of Fgf signaling (Paek et al., 2009). Yet, in utero electroporation experiments at early stages demonstrate that Fgf8 can repress Wnt genes whose expression defines the cortical hem (Shimogori et al., 2004). In summary, the cross regulation between two groups of secreted signals presents a complex picture that is not yet completely understood, but it is likely that such early positive and negative interactions serve to define the final caudomedial position of the cortical hem. This caudomedial domain is further refined by cross-regulatory interactions between transcription factors Emx2 and Pax6 (Kimura et al., 2005). In particular, Emx2 appears to specify a caudomedial domain in the telencephalon which contains the cortical hem as well as the hippocampus. Emx2 may act by restricting the anterior region of Fgf gene expression (Shimogori et al., 2004) though this action appears to have a critical period in that Emx2 overexpression using a later acting (E11.5 onward) Nestin-Cre driver does not affect Fgf8 expression (Hamasaki et al., 2004). Furthermore, Emx2 functions as an effector of the canonical Wnt signaling from the hem to regulate proliferation within the caudomedial region (Muzio et al., 2005). Emx2 may have partially redundant functions with Emx1 in specifying the broad medial telencephalic domain. The Emx1,2 double mutant lacks the entire hemehippocampus domain, though lateral, neocortical tissue is intact (Shinozaki et al., 2004). Thus, Emx2 appears to act at two stages: to establish the domain where hem induction will occur and, later, to mediate the effects of hem signaling during further development of this region. While these mechanisms serve to define the rostrocaudal extent of the hem, other mechanisms act to define the mediolateral boundaries of the hem with the choroid plexus and the cortex. An early acting regulatory mechanism involving transcription factors of the bHLH family regulates the hemechoroid plexus boundary (Imayoshi et al., 2008). The hem and choroid plexus are defined by the differential expression of Hes and Ngn genes. At the time of boundary formation, Hes genes are enriched in the putative choroid plexus region, possibly as a result of the direct activation of these genes by Bmps from the roof plate. At the same time, this region downregulates Ngn gene expression, which continues to be maintained in the adjacent cortical hem. This downregulation of Ngn expression is important in establishing the choroid plexus fate and therefore delineating the hemechoroid plexus boundary (Imayoshi et al., 2008). The hem is restricted to its medial position by the actions of Lhx2, Foxg1, and Pax6, each of which suppresses hem fate. Clones of dorsal telencephalic cells that are Foxg1;Lhx2 double nulls, generated by low-dose tamoxifen administration to E8.5-E10.5 embryos carrying both conditional mutant alleles and CreER, transform into hem (Godbole et al., 2018). Similarly, in Lhx2 and Pax6 double mutants, the lateral extent of the hem increases compared to Lhx2 single mutants (Godbole et al., 2017). Therefore, the positioning of the hem is a finely regulated process with multiple players ensuring it does not appear ectopically.

2.5.2.2 Molecular mechanisms that act to specify and position the antihem Three transcription factors, Pax6, Tlx, and Gsx2, are known to regulate the specification and positioning of the antihem. Its location at the PSB makes the antihem vulnerable to perturbations that disrupt dorsoventral patterning in the telencephalon. The PSB is severely affected in the Pax6 mutant. There is a ventralization of the pallial neuroepithelium of the Pax6 mutant telencephalon, such that the VP and LP now express subpallial markers Mash1, Gsx2, and Dlx2 (Stoykova et al., 1996, 2000; Stoykova, 1997; Toresson et al., 2000; Yun et al., 2001; Kim et al., 2001). Tlx mutants exhibit a similar, but less severe phenotype to Pax6 mutants (Stenman et al., 2003). Tlx is expressed throughout the neuroepithelium, high at the lateral sulcus and on both sides of the PSB. As in the case of the Pax6 mutant, the Tlx mutant too exhibits LGE characteristics at the expense of those of the VP (Stenman et al., 2003). In contrast, the Gsx2 mutant displays the opposite phenotype, one in which pallial gene expression signatures are seen in subpallial domains such as the dLGE (Toresson et al., 2000; Yun et al., 2001). A detailed analysis of the interactions of Pax6 and Gsx2 reveals a cross-repressive mechanism, wherein Pax6 is required to induce VP-specific markers, and Gsx2 is necessary to suppress the expression of these genes in the dLGE, thereby restricting them to the VP (Carney et al., 2009).

2.5.3 CajaleRetzius cells arise from four telencephalic boundary structures To date, four distinct origins for CR cells have been identified. Of these, the cortical hem (Meyer et al., 2002; TakiguchiHayashi et al., 2004; Yoshida et al., 2006), the antihem, and the septum (Meyer et al., 2002; Bielle et al., 2005) are

Telencephalon patterning Chapter | 2

35

well-characterized telencephalic structures. A newly proposed source, the thalamic eminence (TE), is positioned at the boundary between the telencephalon and the diencephalon (Cabrera-Socorro et al., 2007; Abellan et al., 2009) and appears to localize within the diencephalon as development proceeds. These four structures are somewhat different in their molecular features and located distant from each other, but they share key characteristics. First, they are positioned at boundaries between two distinct tissues or brain regions. The hem and the TE each have the choroid plexus on one side and neuroepithelium on the other, while the antihem and the septum separate the dorsal and VP. The septum is also connected with the telencephalic choroid plexus early in development. Second, each of these origins expresses a distinct complement of signaling molecules, and these structures are either proposed or known to be organizing centers that regulate the development of the forebrain (Grove et al., 1998; Assimacopoulos et al., 2003; Mangale et al., 2008; Abbott and Jacobowitz, 1999). Third, these structures produce a range of different cell types, including the choroid plexus (hem; Louvi et al., 2007), parts of the amygdala (antihem; Soma et al., 2009; Hirata et al., 2009), septal nuclei, and neurons of the TE. Yet they display a unifying feature in that each structure also produces CR cells. It is not at all clear why this earliest-born cell population has multiple origins. Reelin expression is a common feature to all these different types of CR cells, but other features allow the four sources to be grouped in different ways. CR cells from the hem, antihem, and TE, but not those from the septum, express calretinin (Bielle et al., 2005; Abbott and Jacobowitz, 1999). The hem, septum, and TE lineage CR cells express p73 (Meyer et al., 2002, 2004; Cabrera-Socorro et al., 2007), but those from the antihem and septum do not (Bielle et al., 2005). Dbx1 expression, in contrast, marks cells from the antihem and the septum, but not those from the hem or the TE (Bielle et al., 2005). Er81 appears to be uniquely expressed by septum-derived CR cells, a feature regulated by rostrally expressed Fgf8 (Zimmer et al., 2010). An open question is whether these differences confer unique functions to each group of CR cells. Indeed, an attractive hypothesis proposes that the diversity of CR cell progenitor zones may correlate with or regulate the development of cytoarchitectonic differences between the neocortex, olfactory cortex, and the hippocampus (Bielle et al., 2005). This hypothesis was examined by selectively ablating specific subpopulations of CR cells (Bielle et al., 2005). When antihem- and septum-derived CR cells were ablated by expressing DTA (diphtheria toxin) via the Dbx1 locus, a significant loss of reelin expression was seen in the septum and piriform cortex at E11.5, but this was compensated for by E14.5, presumably by CR cells from other sources. However, there was a gross reduction in the thickness of the lateral cortex. This defect was selective for the lateral region, since the cingulate cortex appeared normal, indicating an important role for the antihem-derived CR cells in regional cortical development (Bielle et al., 2005). A surprising, counterintuitive result came from experiments in which hem-derived CR cells were ablated by expressing DTA via Wnt3a locus (Yoshida et al., 2006). This caused a massive and near-complete depletion of CR cells overlying the neocortex, which was apparently not rescued by migration of CR cells from other sources. Despite this, neocortical lamination was unaffected (Yoshida et al., 2006). Similarly, in p73 mutants, which also display a loss of cortical CR cells, neocortical lamination was normal except for the absence of the hippocampal fissure, which may be due to the loss of p73 itself (Meyer et al., 2004). Thus, the precise role of hem-derived CR cells continues to be elusive. More broadly, the positions of the four CR cell sources at important boundaries in the forebrain structures together motivate speculations about their roles in the evolution of the telencephalon, as discussed below.

2.5.4 Organizer functions of telencephalic boundary structures The hem, antihem, septum, and TE are evolutionarily ancient, having been identified in several vertebrate phyla. The hem has been identified in birds (Garda et al., 2002) and also in reptiles (Cabrera-Socorro et al., 2007). The antihem appears earlier and was an important discovery as a ventral pallial territory in amphibians (Smith-Fernandez et al., 1998; Moreno et al., 2004; Brox et al., 2004). TE and septum have been identified in fish and amphibians (González et al., 2002; Wullimann and Mueller, 2004). These structures are therefore in a position to influence not only the development but also the evolution of the forebrain. For example, the appearance of the hem and the antihem on either side of the pallium preceded and may have influenced the expansion of this structure in vertebrates. At the caudal telencephalon of the mouse embryo, where these two signaling centers approach each other, the intervening tissue produces an unusual stream of migrating cells, the caudal amygdaloid stream, that forms part of the amygdaloid complex, but has the molecular signature of the neocortex (Remedios et al., 2007). We will now discuss the known organizer functions of the rostral signaling center and the hem, and also current understanding about the role of the antihem. The TE, being a more recent addition to this group, has yet to be examined for such a role, but its position at the telencephalice diencephalic boundary makes it a good candidate to have played a role in the morphogenesis of these major brain subdivisions.

36

PART | I Induction and patterning of the CNS and PNS

2.5.4.1 Rostral signaling center/septum The rostral signaling center was the first of these structures to be identified as an organizer, by elegant experiments using in utero micro electroporation by Fukuchi-Shimogori and Grove (2001). When endogenous Fgf8 signaling was attenuated by electroporation of a construct encoding a secreted Fgfr receptor moiety that sequesters Fgf8, cortical areas were shifted forward. When Fgf8 was overexpressed rostrally, cortical areas shifted backward. Most dramatically, when a second source of Fgf8 was created by caudally directed electroporation, a mirror-image duplication of the somatosensory barrel fields was created (Fukuchi-Shimogori and Grove, 2001). Thus, cortical areas are specified by Fgf8 from the rostral signaling center. Fgf8, present in a gradient in the forebrain (Chen et al., 2009), therefore functions as a morphogen, from the rostral signaling center which has emerged as an organizer for cortical area patterning. These studies reveal yet again how complex patterns are achieved, or at least initiated, using simple mechanisms involving few players. 2.5.4.1.1 Hem The cortical hem was considered to be analogous to the dorsal signaling center of the spinal cord, the roof plate, which also secretes Wnt and Bmp family molecules (Grove et al., 1998; Chizhikov and Millen, 2005). Bmp signaling from the roof plate is responsible for patterning adjacent neuronal fates (Liem et al., 1997), and ablation of the roof plate causes loss of specific neuronal populations (Lee et al., 2000a). A similar a role for the cortical hem was proposed (Grove et al., 1998). Supporting this hypothesis, the entire hippocampus is missing when the hem is deleted (Yoshida et al., 2006) or when a particular hem-specific signaling molecule, Wnt3a, is disrupted (Lee et al., 2000b). When components of the Wnt signaling cascade Lef1 (Galceran et al., 2000) or Lrp6 (Zhou et al., 2004) are disrupted, the dentate precursor pool is diminished and does not mature or migrate properly. But highly reduced cell populations of the dentate precursors were detected in each mutant (Galceran et al., 2000; Zhou et al., 2004). Therefore, these studies were not able to separate a role for Wnt signaling in the expansion of the precursor population from one in which they act to specify of hippocampal cell fates (Li and Pleasure, 2005). The role of the cortical hem has also been tested in explant culture experiments in which the hem was either removed or transplanted to ectopic locations of medial telencephalic preparations (Tole and Grove, 2001). However, the age of the tissue used was E12.5, apparently too late for either perturbation to have any effect on hippocampal specification. Indeed, the authors concluded that the fine details of hippocampal field specification must have occurred by E12.5, even though overt differentiation of hippocampal fields occurs much later, from E15.5. Definitive evidence of the role of the cortical hem came from chimeras in which Lhx2 null cells, surrounded by wild-type cortical neuroepithelium, differentiated into ectopic hem tissue (Mangale et al., 2008). An ectopic hippocampus formed adjacent to each patch of hem, with spatially correct induction and positioning of multiple hippocampal fields. This consolidated the cortical hem is an organizer for the hippocampus. Which signaling molecules from the hem are critical for ectopic hippocampal induction? The literature strongly supports a role of Wnt signaling for this role. The Wnt3a, Lef1, and Lrp6 mutant studies all indicate that Wnt signaling is necessary for hippocampal development (Galceran et al., 2000; Lee et al., 2000b; Zhou et al., 2004). Furthermore, ectopic activation of Lef1 upregulated some hippocampal field markers in lateral neuroepithelium, demonstrating that Wnt signaling, is sufficient for this process (Machon et al., 2007). In contrast, Bmp signaling has not been implicated in hippocampal development. The Bmpr1a mutant, which lacks the telencephalic choroid plexus, appears to form a hippocampus (Hébert et al., 2002). Although the Bmpr1a;1b double mutant does not form a hippocampus, it is more likely to be directly due to the absence of the hem in these mice (Fernandes et al., 2007), rather than a direct role of Bmps in hippocampal patterning. In terms of regulating telencephalic neuronal development, Bmp signaling appears to act at earlier stages, including specifying the extreme medial fate of the choroid plexus (Panchision et al., 2001; Hébert et al., 2002; Cheng et al., 2006), causing cell death (Furuta et al., 1997; Fernandes et al., 2007), and possibly regulating the graded expression of Lhx2 itself (Monuki et al., 2001). However, mice mutant in both Bmpr1a and 1b still do express Lhx2 in the cortical promordium (Fernandes et al., 2007), so there is not an absolute requirement of Bmp signaling for maintaining Lhx2 expression. Thus, early actions of Bmp signaling may set in motion events which permit the formation of the cortical hem, which in turn induces the hippocampus. How signals from the hem bring about the specification of distinct hippocampal field identities remains an important open question. An important issue is how the hem can direct not only the specification but also the structural organization of multiple hippocampal fields. A clue comes from examining the radial glial palisade associated with the dentate migration. The organization of this palisade is thought to guide the dentate cells from their origin at the ventricular zone adjacent to the

Telencephalon patterning Chapter | 2

37

hem to their final location where they form the blades of the dentate gyrus (Rickmann et al., 1987). Mangale et al. (2008) report additional radial glial palisades associated with ectopic hem tissue, which appear to guide distinct migratory streams terminating at each dentate gyrus. The organization of the radial glial scaffolding itself is dependent on Wnt signaling (Zhou et al., 2004). Thus, each patch of hem may be responsible for orienting the scaffolding adjacent to it, which would then guide the ectopically induced dentate cells to form an ectopic gyrus.

2.5.4.2 Antihem In contrast to the organizer function of the hem, such a role for the antihem has yet to be identified. We present some speculations based on available evidence, such as the loss of the antihem correlating with severe disruption of the radial glial palisade at the PSB. This raises a strong parallel with the role of the hem in organizing the hippocampal radial glia. Tlx mutants have fewer radial glial fibers at the PSB which do not appear to fasciculate to form a palisade (Stenman et al., 2003). Pax6 is itself required for the differentiation of radial glia in the forebrain (Götz et al., 1998). Not surprisingly, the radial glial palisade at the PSB is disrupted in the Pax6 mutant (Stoykova et al., 1997; Hirata et al., 2002; Chapouton et al., 1999). Although markers identified the radial glial progenitor cell population, the fascicle itself could not be distinguished. Furthermore, interneurons produced within the subpallium were detected in greater numbers in the Pax6 mutant cortex, suggesting that some feature of the normal PSB serves to restrict the tangential migration of interneurons (Chapouton et al., 1999). Finally, Pax6 mutant also displays profound defects in thalamocortical and corticofugal axon pathfinding. The underlying cause of this defect was suggested to be a combination of structural abnormalities and alterations in the expression of specific pathfinding molecules of the Semaphorin family at the Pax6 mutant PSB (Jones et al., 2002). What other signaling molecules might mediate some of these defects? Nrg1, which is concentrated in the antihem, has been shown to be essential for the formation and maintenance of radial glial cells (Anton et al., 1997; Schmid et al., 2003). Drawing parallels with the hem, Wnt signaling might also regulate the radial glial palisades organized by the antihem. Wnt7b is expressed adjacent to the antihem, in the dLGE (Kim et al., 2001). The expression of the Wnt antagonist Sfrp2 in the antihem may lead to a concentration of the Wnt signal to the subpallial side of the PSB (Kim et al., 2001), providing a positional signal to the radial glial palisade. Together, these studies support an integral role for the antihem in mediating axon guidance and cell migration, that of interneurons into the cortex as well as that of the derivatives of the lateral telencephalon, such as the olfactory cortex, the claustrum, and the amygdala. The latter role is likely to arise from the regulation of the radial glial palisade at the PSB.

2.6 Subdividing ventral domains The ventral telencephalon consists of subpallial structures arising from transient bulges of the neuroepithelium, called the MGE, LGE, and CGE, respectively. Their major contributions are the projection neurons and interneurons of the structures they give rise to, as well as the interneurons of the entire dorsal telencephalon. Pallial domains also contribute neurons to ventral telencephalic structures. To understand the specification of ventral structures, first, mutant studies that reveal transcriptional and signaling mechanisms that regulate the different structures are examined. Then lineage and fatemapping studies that reveal the complex migration patterns involved in assembling the final structure are examined.

2.6.1 The striatum and pallidum Two of the major subdivisions of the ventral telencephalon are the MGE and LGE whose precursor cells contribute largely to the striatum and the pallidum. The CGE is a more recently defined structure that is morphologically a caudal extension of the LGE and as yet not molecularly distinguishable from it. However, it contributes distinct subsets of interneurons to the dorsal telencephalon (Miyoshi et al., 2010; Nery et al., 2002; Yozu et al., 2005) and is discussed in a later section on lineages and fate mapping. Like the rest of the telencephalon, the specification of the LGE and MGE is also regulated by secreted factors from signaling centers and transcription factors. As described above, SHH and FGFs are required to generate the ventral telencephalon. SHH acts primarily by repressing the dorsalizing effect of the GLI3 repressor, but not exclusively since loss of Gli2 or Gli1 and Gli2 also leads to ventral phenotypes, albeit much less severe compared with the Shh / mutant (Yu et al., 2009). FGFs, which presumably function genetically downstream of SHH (see earlier), appear to act in a dose-dependent manner to pattern the ventral subdivisions. Hypomorphic and null alleles of Fgf8 lead to phenotypes with progressively more severe loss of ventral markers with diminishing levels of Fgf8 expression (Storm et al., 2006). Similarly, a knockout of the Fgfr1 gene only results in a failure to produce MGE cells that express Lhx6 or Lhx7, two LIM-domain transcription factors necessary for generating MGE-derived

38

PART | I Induction and patterning of the CNS and PNS

interneurons (Fragkouli et al., 2005; Gutin et al., 2006; Liodis et al., 2007), whereas disruption of both Fgfr1 and Fgfr2 together leads to a more severe phenotype in which all or most MGE and LGE precursor cells are lost (Gutin et al., 2006). Notably, the expression of Nkx2.1, Gsx2, and Ascl1 is lost in these mutants. Nkx2.1, which is expressed specifically in MGE precursor cells, is essential for specifying their fate: in the Nkx2.1 / mouse mutant, the MGE area expresses molecular features of the LGE and generates striatal rather than pallidal neurons (Butt et al., 2008; Sussel et al., 1999). Gsx2 together with Gsh1 is essential for specifying LGE precursors. In Gsh1 / ;Gsx2 / mouse mutants, LGE precursors are dorsalized and fail to express Ascl1, which acts as an effector of GSH function (Toresson and Campbell, 2001; Wang et al., 2009). Here as well, other factors will undoubtedly be found to pattern the MGE and LGE into distinct domains (Tucker et al., 2008). The MGE, LGE, and CGE produce distinct groups of interneurons that populate the entire dorsal and ventral telencephalon. It is no surprise that the ventricular zones of these structures are not uniform but in fact can be divided into up to 18 distinct domains based on molecular expression patterns as well as transplantation-based fate-mapping studies (Fig. 2.8; Flames et al., 2007). Neurons and glia arising from these domains would have distinct molecular identities, and these domains therefore provide an appropriate spectrum from which the diversity of structures such as the amygdaloid complex, described below, can arise (Fig. 2.10).

2.6.2 The amygdala An important component of the ventral telencephalon is the amygdaloid complex, a collection of several distinct nuclei. Together, these represent possibly the most diverse collection of brain structures in terms of morphology, position,

pLGE2 pLGE1 pLGE4

pLGE2 pLGE3 pLGE4 pMGE1 pLGE1

pLGE3 pMGE2

pLGE1

pLGE2 pLGE3 pMGE3

pMGE3

pMGE3

pPOH1 pMGE5

pMGE4

pPOA1

pMGE1 pMGE5

pPOA2

FIGURE 2.8 The subpallial neuroepithelium is divided into distinct domains by the combinatorial expression of transcription factors. From Flames, N., Pla, R., Gelman, D.M., Rubenstein, J.L., Puelles, L., Marin, O., 2007. Delineation of multiple subpallial progenitor domains by the combinatorial expression of transcriptional codes. J. Neurosci. 27, 9682e9695 (pending permission from J. Neurosci).

VZ

DP Tbr1 (high) Tbr1 (low) Emx1 LP

Migrating streams

Emx2 Lhx2 Lhx9

dLGE VP

Pax6

FIGURE 2.9 Neuroepithelial domains in the vicinity of the pallial-subpallial boundary are distinguished by the combinatorial expression of transcription factors and give rise to molecularly distinct streams of migrating cells. Modified from Tole, S., Remedios, R., Saha, B., Stoykova, A., 2005. Selective requirement of Pax6, but not Emx2, in the specification and development of several nuclei of the amygdaloid complex. J. Neurosci. 25, 2753e2760 (pending permission from J. Neurosci).

Telencephalon patterning Chapter | 2

39

1

1 2 3

2 AAD

3 AAV nLOT2 nLOT1 ACo MePD MePV Ce

La BLa BM PMCo

FIGURE 2.10 Anterior amygdaloid area, dorsal (AAD) and ventral (AAV); anterior cortical (ACo); basolateral (BLa), basomedial (BM), central (Ce), lateral (La) nuclei; medial nucleus, postero dorsal (MePD), and postero ventral (MePV); nucleus of the lateral olfactory tract, layers 1 and 2 (nLOT1, nLOT2); postero medial cortical nucleus (PMCo). Modified from Remedios, R., Subramanian, L., Tole, S., 2004. LIM genes parcellate the embryonic amygdala and regulate its development. J. Neurosci. 24, 6986e6990.

developmental origins, neurotransmitter systems, and association with different functional systems (reviewed in Swanson and Petrovich, 1998). Morphologically, components of the amygdaloid complex range from “cortical” or layered, and considered extensions of the trilaminar olfactory cortex, to “nuclear” or clustered. In terms of location, they range from superficial to deep, and also spanning much of the rostrocaudal extent of the telencephalon from the region of the septum to the caudal pole of the hippocampus. Their developmental origins include not only pallial as well as subpallial contributions but also populations originating within the diencephalon. The neurotransmitter systems used vary depending on their origins, in that the projections from pallial-derived structures are excitatory (glutamatergic) and subpallial-derived structures are inhibitory (GABAergic). Finally, the amygdaloid complex participates in a wide range of functional systems, including olfaction, reproductive and defensive behavior, autonomic motor functions, memory, learning, and emotion. Precursors of amygdaloid cells must undergo extensive migration from diverse sites of origin and assemble in a coordinated fashion to form the mature structure. A correspondingly complex set of regulatory mechanisms are likely to regulate the development of the amygdala, and understanding these is of relevance not only to appreciate the diversity of developmental control mechanisms that give rise to distinct brain structures but also because understanding the development of the amygdala may offer broader evolutionary insights.

2.6.3 An evolutionary perspective for how the neocortex arose The amygdaloid complex preceded the neocortex in evolution and is present in reptiles and amphibians, and fish. The pallial component of the amygdaloid complex arises in most part from the LP and VP, domains that have the unique position of being at the interface of the neocortex-producing DP and the ventral telencephalic SP. While all the pallial domains do share several common pallial molecular signatures and developmental control mechanisms (Puelles et al., 2000; Yun et al., 2001), the mechanisms regulating the specification of the pallial amygdala reveal fundamental developmental distinctions between the MP/DP on the one hand and the LP/VP on the other. Therefore, such comparisons have been the focus of intense exploration, since they have the potential to offer hypotheses for how the neocortex may have arisen from a DP domain that produces very different dorsal tissue in nonmammalian species. The pallial amygdala comprises nuclei of the basolateral complex (lateral, basolateral, and basomedial nuclei) as well as some of the cortical nuclei including the nucleus of the lateral olfactory tract (nLOT). These are thought to arise from a multicomponent “lateral migratory stream” arising from domains near the PSB (the VP and LP; Puelles et al., 2000; Medina et al., 2004; Tole et al., 2005). These domains also produce the piriform cortex, the claustrum and endopiriform nuclei, and are likely to contribute to the anterior amygdaloid areas and cortical nuclei as well (Puelles et al., 2000; Medina et al., 2004). An exception is layer 2 of the nLOT (nLOT2), which has the molecular signature of the

40

PART | I Induction and patterning of the CNS and PNS

DP, and which also shares specification (Lhx2; Pax6; Emx1,2; Tbr1) and migration (Reelin; Cdk5) mechanisms with the DP-derivative, the neocortex (Remedios et al., 2007). Significantly, many of these mechanisms (Lhx2; Tbr1; Reelin; Cdk5) do not appear to be utilized by the rest of the amygdaloid complex, including the other pallial derivatives. The nLOT2 primordium has therefore been speculated to be a possible precursor to the rest of the DP, thus endowing it with a dependence on transcription factors and migration mechanisms that the neocortex is known to require (Remedios et al., 2007). The central and medial nuclei (including the dorsal and ventral subdivisions of the posterior portion of the medial nucleusdMePD and MePV) are distinct from the other amygdaloid components in that they are inhibitory in terms of their output and were thought to arise entirely from the subpallium. However, it is now clear that only the central nucleus is likely to be purely subpallial, arising from the LGE (Garcia-Lopez et al., 2008). The medial nuclei, together with cell groups termed the “extended amygdala,” contains components from the subpallium (MGE), the pallium (VP), and also the diencephalon (Garcia-Lopez et al., 2008; Cocas et al., 2009; Soma et al., 2009). Transcriptional mechanisms required for the development of amygdaloid nuclei arising from these lateral migratory streams include transcription factors involved in regulating the dorsoventral boundary. Gsx2, required for dLGE specification, is also required for the subpallial component of the lateral migratory stream (Carney et al., 2006). Pax6 and Tlx are each required for VP specification and are thence necessary for the specification of the entire of the basolateral complex (Stenman et al., 2003; Tole et al., 2005). Lhx2 and Pax6 are each required for the specification of the nLOT2 (Remedios et al., 2004, 2007; Tole et al., 2005). Surprisingly, Emx2 is not required for any component of the amygdaloid complex thus far examined (Tole et al., 2005). A redundancy with Emx1 does not explain this finding because Emx1 is not expressed in the VP, from which much of the pallial amygdala arises (Puelles et al., 2000; Yun et al., 2001). Viewed from an evolutionary perspective, it may be that Emx2 was harnessed later in evolution, for specific aspects of pallial development, in which it plays some redundant roles with Emx1 such as the specification of the medial telencephalic primordium (Shinozaki et al., 2004) and some unique roles such as in the growth of the hippocampus and migration of dentate granule cells (Pellegrini et al., 1996; Tole et al., 2000).

2.6.4 Lineage and fate mapping in the ventral telencephalon As is the case with the dorsal telencephalon, the ventral telencephalon too receives migratory streams originating from outside its boundaries. Prominent among these are cells derived from the Emx1 lineage (Gorski et al., 2002; Cocas et al., 2009). Emx1 is expressed in the MP, DP, and LP, and consistent with this pallial origin, cells of this lineage are excitatory, and populate the basolateral complex. Surprisingly, a second group of Emx1 lineage cells was also discovered in the striatum. Further examination revealed migratory properties in these cells, and also the unexpected ability to upregulate ventral markers such as Dlx2 and Gsx2, once the cells reached the striatum (Gorski et al., 2002; Willaime-Morawek et al., 2006; Cocas et al., 2009). Consistent with their new location and molecular identity, these cells are inhibitory, which is an unusual feature for the Emx1 lineage. The VP is the source of migrating streams with complex molecular identities as described in an earlier section. Dbx1, expressed in the VP, was used to fate map this lineage (Hirata et al., 2009). Consistent with their pallial origin, the Dbx1 lineage cells were Tbr1 positive, migrated via the lateral migratory stream, and contributed excitatory neurons to the basolateral complex and the cortical nuclei of the amygdaloid complex. This study also revealed a novel Dbx1 lineage contribution to the medial amygdala from the preoptic area of the hypothalamus. Cells from this stream are inhibitory and populate the medial amygdala (Hirata et al., 2009). The MGE is the source of Nkx2.1 lineage inhibitory neurons that migrate long distances to dorsal and ventral telencephalic structures (Xu et al., 2004, 2008; Flandin et al., 2010). The Nkx2.1 lineage produces two distinct sublineages, one of which is cholinergic, destined for the striatum, and is not repelled by striatal semaphorins, whereas the other is GABAergic, is repelled by semaphorins, and proceeds to migrate into the cortex. The persistent expression of Nkx2.1 itself in postmitotic neurons of the striatal population but not the cortical population ensures repression of the neuropilin 2 receptor and brings about this differential response (Nóbrega-Pereira et al., 2008). A combinatorial code of LIM-HD transcription factors Lhx6, 7/8, and Isl1 directs common precursors to execute the cell fate choice between these sublineages (Fragkouli et al., 2009). Finally, there are recent reports of migrations from the diencephalon that contribute to the medial amygdala. The precise origins of this stream are not known, since it was identified by electroporation of GFP widely into the diencephalic neuroepithelium in elegant experiments that ensured no telencephalic electroporation (Soma et al., 2009). A similar stream was identified independently, as part of the Otp lineage from the hypothalamus, which appears to terminate in the medial amygdala. In addition to this stream, the Otp lineage has a wider contribution to ventral telencephalic structures, and these

Telencephalon patterning Chapter | 2

41

cells require Otp to cross the diencephalicetelencephalic boundary (Garcia-Moreno et al., in press). Furthermore, the Foxb1 lineage, the origin of which is strictly diencephalic, also appears to contribute to the amygdala though the destination has not been fully characterized (Zhao et al., 2008). Since the medial amygdala projects to the hypothalamus, it is interesting that the two structures appear to have a shared lineage at least for some of their components. In summary, the amygdala itself, and the ventral telencephalon in general, appears to be at least as complex as the dorsal telencephalon in terms of the specification and migration mechanisms required by its components to assemble into the final mature structure.

2.7 Conclusions The development of any system begins with a simple structure, in this case the neuroepithelium, upon which and within which regulatory mechanisms act to bring about complexity as the structure matures. The apparent simplicity of the neuroepithelium is, however, deceptive. Though uniform in morphology and cell composition, the early neuroepithelium already has within it specified domains along the rostrocaudal and mediolateral axis, and its cells are already polarized along the apicobasal (what will be the radial) axis. The forebrain neuroepithelium, even before the neural tube forms, is already flanked by specialized tissue expressing signaling molecules at its rostral and lateral edges and extending beneath its midline. These molecules bring about extrinsic effects that are anything but simple: their cross regulation has borne the focus of intense exploration for several years, and still offers new insights. Intrinsic transcription factors act via cross regulation, redundancy, and combinatorial codes, all of which serve to specify unique domains in the neuroepithelium. Despite all this complexity, there is a broad framework within which these mechanisms act: the dorsal telencephalon produces excitatory neurons, while the ventral telencephalon produces inhibitory neurons. Reciprocal contributions between these two broad telencephalic domains ensure that both types of neurons are available in each telencephalic structure. The lineage relationships, subpopulations, and mechanisms of migration of these neurons are the least well-understood aspect of early telencephalic development. Once the neurons are assembled in their respective structures, the next frontier they face is a problem perhaps even more challenging than their previous developmental history presented: connecting to the correct targets. Whereas migrating neurons navigate through younger, therefore smaller structures, axons typically have to pathfind through territories that have developed considerably and present them with several decision points. Yet some of the same mechanisms can be reutilized between the two pathfinding problemsdthe guidance of cells themselves, and later on, their axons. Each requires modulation of both signals and receptors in appropriate spatiotemporal sequence. After the circuit is assembled, once again phenomena that have parallels with development play out: the maintenance of the circuit and its modification and consolidation based on experience. The mechanistic complexity of early development therefore sets the stage for, and presages, multiple levels of intricate regulation to provide the organism with a functional nervous system. Understanding the molecular mechanisms that regulate telencephalon development is not only a fascinating undertaking but it is also key in designing effective regenerative therapies for a range of forebrain disorders from developmental to degenerative ones. Knowing what factors control the fates of neural precursors during development provides a framework for how specific types of telencephalic neurons could potentially be obtained using culture paradigms for therapeutic purposes.

Acknowledgments Portions of the section “Boundary structures as organizing centers and CR cell sources” were taken or adapted from Subramanian et al. (2009), as permitted by the copyright agreement with ST. We thank Achira Roy and Parul Chachra for preparing the schematics in Figs. 2.7 and 2.9 and Rajlakshmi Marpakwar for assistance with the references.

References Abbott, L.C., Jacobowitz, D.M., 1999. Developmental expression of calretinin-immunoreactivity in the thalamic eminence of the fetal mouse. Int. J. Dev. Neurosci. 17, 331e345. Abellan, A., Menuet, A., Dehay, C., Medina, L., Rétaux, S., 2009. Differential expression of LIM-homeodomain factors in Cajal-Retzius cells of primates, Rodents, and birds. Cerebr. Cortex. Aguiar, D.P., Sghari, S., Creuzet, S., 2014. The facial neural crest controls fore- and midbrain patterning by regulating Foxg1 expression through Smad1 activity. Development 141 (12), 2494e2505.

42

PART | I Induction and patterning of the CNS and PNS

Anton, E.S., Marchionni, M.A., Lee, K.F., Rakic, P., 1997. Role of GGF/neuregulin signaling in interactions between migrating neurons and radial glia in the developing cerebral cortex. Development 124, 3501e3510. Aoto, K., Nishimura, T., Eto, K., Motoyama, J., 2002. Mouse GLI3 regulates Fgf8 expression and apoptosis in the developing neural tube, face, and limb bud. Dev. Biol. 251, 320e332. Ashique, A.M., Choe, Y., Karlen, M., May, S.R., Phamluong, K., Solloway, M.J., Ericson, J., Peterson, A.S., 2009. The Rfx4 transcription factor modulates Shh signaling by regional control of ciliogenesis. Sci. Signal. 2 (95), ra70. Assimacopoulos, S., Grove, E.A., Ragsdale, C.W., 2003. Identification of a Pax6-dependent epidermal growth factor family signaling source at the lateral edge of the embryonic cerebral cortex. J. Neurosci. 23, 6399e6403. Backman, M., Machon, O., Mygland, L., van den Bout, C.J., Zhong, W., Taketo, M.M., Krauss, S., 2005. Effects of canonical Wnt signaling on dorsoventral specification of the mouse telencephalon. Dev. Biol. 279, 155e168. Barth, K.A., Wilson, S.W., 1995. Expression of zebrafish nk2.2 is influenced by sonic hedgehog/vertebrate hedgehog-1 and demarcates a zone of neuronal differentiation in the embryonic forebrain. Development 121, 1755e1768. Barth, K.A., Kishimoto, Y., Rohr, K.B., Seydler, C., Schulte-Merker, S., Wilson, S.W., 1999. Bmp activity establishes a gradient of positional information throughout the entire neural plate. Development 126, 4977e4987. Besse, L., Neti, M., Anselme, I., Gerhardt, C., Rüther, U., Laclef, C., Schneider-Maunoury, S., 2011. Primary cilia control telencephalic patterning and morphogenesis via Gli3 proteolytic processing. Development 138 (10), 2079e2088. Bielen, H., Houart, C., 2012. BMP signaling protects telencephalic fate by repressing eye identity and its Cxcr4-dependent morphogenesis. Dev. Cell 23 (4), 812e822. Bielle, F., Griveau, A., Narboux-Nême, N., Vigneau, S., Sigrist, M., Arber, S., Wassef, M., Pierani, A., 2005. Multiple origins of Cajal-Retzius cells at the borders of the developing pallium. Nat. Neurosci. 8, 1002e1012. Bishop, K.M., Goudreau, G., O’Leary, D.D., 2000. Regulation of area identity in the mammalian neocortex by Emx2 and Pax6. Science 288, 344e349. Brox, A., Puelles, L., Ferreiro, B., Medina, L., 2004. Expression of the genes Emx1, Tbr1, and Eomes (Tbr2) in the telencephalon of Xenopus laevis confirms the existence of a ventral pallial division in all tetrapods. J. Comp. Neurol. 474, 562e577. Bulchand, S., Grove, E.A., Porter, F.D., Tole, S., 2001. LIM-homeodomain gene Lhx2 regulates the formation of the cortical hem. Mech. Dev. 100, 165e175. Bulfone, A., Wang, F., Hevner, R., Anderson, S., Cutforth, T., Chen, S., Meneses, J., Pedersen, R., Axel, R., Rubenstein, J.L., 1998. An olfactory sensory map develops in the absence of normal projection neurons or GABAergic interneurons. Neuron 21, 1273e1282. Butt, S.J., Sousa, V.H., Fuccillo, M.V., Hjerling-Leffler, J., Miyoshi, G., Kimura, S., Fishell, G., 2008. The requirement of Nkx2-1 in the temporal specification of cortical interneuron subtypes. Neuron 59 (5), 722e732. Cabrera-Socorro, A., Hernandez-Acosta, N.C., Gonzalez-Gomez, M., Meyer, G., 2007. Comparative aspects of p73 and Reelin expression in Cajale Retzius cells and the cortical hem in lizard, mouse and human. Brain Res. 1132, 59e70. Cajal, M., Creuzet, S.E., Papanayotou, C., Sabéran-Djoneidi, D., Chuva de Sousa Lopes, S.M., Zwijsen, A., Collignon, J., Camus, A., 2014. A conserved role for non-neural ectoderm cells in early neural development. Development 141 (21), 4127e4138. Carlin, D., Sepich, D., Grover, V.K., Cooper, M.K., Solnica-Krezel, L., Inbal, A., 2012. Six3 cooperates with Hedgehog signaling to specify ventral telencephalon by promoting early expression of Foxg1a and repressing Wnt signaling. Development 139 (14), 2614e2624. Carney, R.S., Alfonso, T.B., Cohen, D., Dai, H., Nery, S., Stoica, B., Slotkin, J., Bregman, B.S., Fishell, G., Corbin, J.G., 2006. Cell migration along the lateral cortical stream to the developing basal telencephalic limbic system. J. Neurosci. 26, 11562e11574. Carney, R.S., Cocas, L.A., Hirata, T., Mansfield, K., Corbin, J.G., 2009. Differential regulation of telencephalic pallialesubpallial boundary patterning by Pax6 and Gsh2. Cerebr. Cortex. Chapouton, P., Gärtner, A., Götz, M., 1999. The role of Pax6 in restricting cell migration between developing cortex and basal ganglia. Development 126, 5569e5579. Cheng, X., Hsu, C.M., Currle, D.S., Hu, J.S., Barkovich, A.J., Monuki, E.S., 2006. Central roles of the roof plate in telencephalic development and holoprosencephaly. J. Neurosci. 26, 7640e7649. Chi, C.L., Martinez, S., Wurst, W., Martin, G.R., 2003. The isthmic organizer signal FGF8 is required for cell survival in the prospective midbrain and cerebellum. Development 130, 2633e2644. Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H., Beachy, P.A., 1996. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407e413. Chizhikov, V.V., Millen, K.J., 2005. Roof plate-dependent patterning of the vertebrate dorsal central nervous system. Dev. Biol. 277, 287e295. Cobos, I., Shimamura, K., Rubenstein, J.L., Martínez, S., Puelles, L., 2001. Fate map of the avian anterior forebrain at the four-somite stage, based on the analysis of quail-chick chimeras. Dev. Biol. 239 (1), 46e67. Cocas, L.A., Miyoshi, G., Carney, R.S., Sousa, V.H., Hirata, T., Jones, K.R., Fishell, G., Huntsman, M.M., Corbin, J.G., 2009. Emx1-lineage progenitors differentially contribute to neural diversity in the striatum and amygdala. J. Neurosci. 29, 15933e15946. Cocas, L.A., Georgala, P.A., Mangin, J.M., Clegg, J.M., Kessaris, N., Haydar, T.F., Gallo, V., Price, D.J., Corbin, J.G., 2011. Pax6 is required at the telencephalic pallial-subpallial boundary for the generation of neuronal diversity in the postnatal limbic system. J. Neurosci. 31 (14), 5313e5324. Cohn, M.J., Izpisúa-Belmonte, J.C., Abud, H., Heath, J.K., Tickle, C., 1995. Fibroblast growth factors induce additional limb development from the flank of chick embryos. Cell 80 (5), 739e746. Corbin, J.G., Gaiano, N., Machold, R.P., Langston, A., Fishell, G., 2000. The Gsh2 homeodomain gene controls multiple aspects of telencephalic development. Development 127, 5007e5020.

Telencephalon patterning Chapter | 2

43

Corbin, J.G., Rutlin, M., Gaiano, N., Fishell, G., 2003. Combinatorial function of the homeodomain proteins Nkx2.1 and Gsh2 in ventral telencephalic patterning. Development 130, 4895e4906. Crossley, P.H., Martin, G.R., 1995. The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439e451. Crossley, P.H., Martinez, S., Martin, G.R., 1996a. Midbrain development induced by FGF8 in the chick embryo. Nature 380, 66e68. Crossley, P.H., Minowada, G., MacArthur, C.A., Martin, G.R., 1996b. Roles for FGF8 in the induction, initiation, and maintenance of chick limb development. Cell 84, 127e136. Danesin, C., Peres, J.N., Johansson, M., Snowden, V., Cording, A., Papalopulu, N., Houart, C., 2009. Integration of telencephalic Wnt and hedgehog signaling center activities by Foxg1. Dev. Cell 16 (4), 576e587. Desmaris, E., Keruzore, M., Saulnier, A., Ratié, L., Assimacopoulos, S., De Clercq, S., Nan, X., Roychoudhury, K., Qin, S., Kricha, S., Chevalier, C., Lingner, T., Henningfeld, K.A., Zarkower, D., Mallamaci, A., Theil, T., Campbell, K., Pieler, T., Li, M., Grove, E.A., Bellefroid, E.J., 2018. DMRT5, DMRT3, and EMX2 cooperatively repress Gsx2 at the pallium-subpallium boundary to maintain cortical identity in dorsal telencephalic progenitors. J. Neurosci. 38, 9105e9121. Dou, C., Li, S., Lai, E., 1999. Dual role of brain factor-1 in regulating growth and patterning of the cerebral hemispheres. Cerebr. Cortex 9, 543e550. Eagleson, G., Ferreiro, B., Harris, W.A., 1995. Fate of the anterior neural ridge and the morphogenesis of the Xenopus forebrain. J. Neurobiol. 28 (2), 146e158. Echelard, Y., Epstein, D.J., St-Jacques, J.B., Shen, L., Mohler, J., McMahon, J.A., McMahon, A.P., 1993. Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75, 1417e1430. Ericson, J., Muhr, J., Placzek, M., Lints, T., Jessell, T.M., Edlund, T., 1995. Sonic hedgehog induces the differentiation of ventral forebrain neurons: a common signal for ventral patterning within the neural tube. Cell 81, 747e756. Fernandes, M., Gutin, G., Alcorn, H., McConnell, S.K., Hébert, J.M., 2007. Mutations in the BMP pathway in mice supports the existence of two molecular classes of holoprosencephaly. Development 134, 3789e3794. Ferri, A., Favaro, R., Beccari, L., Bertolini, J., Mercurio, S., Nieto-Lopez, F., Verzeroli, C., La Regina, F., De Pietri Tonelli, D., Ottolenghi, S., Bovolenta, P., Nicolis, S.K., 2013. Sox2 is required for embryonic development of the ventral telencephalon through the activation of the ventral determinants Nkx2.1 and Shh. Development 140 (6), 1250e1261. Flames, N., Pla, R., Gelman, D.M., Rubenstein, J.L., Puelles, L., Marin, O., 2007. Delineation of multiple subpallial progenitor domains by the combinatorial expression of transcriptional codes. J. Neurosci. 27, 9682e9695. Flandin, P., Kimura, S., Rubenstein, J.L., 2010. The progenitor zone of the ventral medial ganglionic eminence requires Nkx2-1 to generate most of the globus pallidus but few neocortical interneurons. J. Neurosci. 30, 2812e2823. Fragkouli, A., Hearn, C., Errington, M., Cooke, S., Grigoriou, M., Bliss, T., Stylianopoulou, F., Pachnis, V., 2005. Loss of forebrain cholinergic neurons and impairment in spatial learning and memory in LHX7-deficient mice. Eur. J. Neurosci. 21, 2923e2938. Fragkouli, A., van Wijk, N.V., Lopes, R., Kessaris, N., Pachnis, V., 2009. LIM homeodomain transcription factor-dependent specification of bipotential MGE progenitors into cholinergic and GABAergic striatal interneurons. Development 136, 3841e3851. Franz, T., 1994. Extra-toes (Xt) homozygous mutant mice demonstrate a role for the Gli-3 gene in the development of the forebrain. Acta Anat. 150, 38e44. Fuccillo, M., Rallu, M., McMahon, A.P., Fishell, G., 2004. Temporal requirement for hedgehog signaling in ventral telencephalic patterning. Development 131, 5031e5040. Fuccillo, M., Joyner, A.L., Fishell, G., 2006. Morphogen to mitogen: the multiple roles of hedgehog signalling in vertebrate neural development. Nat. Neurosci. Rev. 7, 772e783. Fukuchi-Shimogori, T., Grove, E.A., 2001. Neocortex patterning by the secreted signaling molecule FGF8. Science 294, 1071e1074. Furuta, Y., Piston, D.W., Hogan, B.L., 1997. Bone morphogenetic proteins (BMPs) as regulators of dorsal forebrain development. Development 2203e2212. Galceran, J., Miyashita-Lin, E.M., Devaney, E., Rubenstein, J.L., Grosschedl, R., 2000. Hippocampus development and generation of dentate gyrus granule cells is regulated by LEF1. Development 127, 469e482. Garcia-Lopez, M., Abellan, A., Legaz, I., Rubenstein, J.L., Puelles, L., Medina, L., 2008. Histogenetic compartments of the mouse centromedial and extended amygdala based on gene expression patterns during development. J. Comp. Neurol. 506, 46e74. Garda, A.L., Puelles, L., Rubenstein, J.L., Medina, L., 2002. Expression patterns of Wnt8b and Wnt7b in the chicken embryonic brain suggest a correlation with forebrain patterning centers and morphogenesis. Neuroscience 113, 689e698. Geng, X., Speirs, C., Lagutin, O., Inbal, A., Liu, W., Solnica-Krezel, L., Jeong, Y., Epstein, D.J., Oliver, G., 2008. Haploinsufficiency of Six3 fails to activate Sonic hedgehog expression in the ventral forebrain and causes holoprosencephaly. Dev. Cell 15 (2), 236e247. Godbole, G., Roy, A., Shetty, A.S., Tole, S., 2017. Novel functions of LHX2 and PAX6 in the developing telencephalon revealed upon combined loss of both genes. Neural Dev. 12 (1), 19. Godbole, G., Shetty, A.S., Roy, A., D’souza, L., Chen, B., Miyoshi, G., Fishell, G., Tole, S., 2018. Hierarchical genetic interactions between FOXG1 and LHX2 regulate the formation of the cortical hem in the developing telencephalon. Development 145. https://doi.org/10.1242/dev.154583. Golden, J.A., Bracilovic, A., McFadden, K.A., Beesley, J.S., Rubenstein, J.L., Grinspan, J.B., 1999. Ectopic bone morphogenetic proteins 5 and 4 in the chicken forebrain lead to cyclopia and holoprosencephaly. Proc. Natl. Acad. Sci. U.S.A. 96, 2439e2444. González, A., López, J.M., Sánchez-Camacho, C., Marín, O., 2002. Regional expression of the homeobox gene Nkx2-1 defines pallidal and interneuronal populations in the basal ganglia of amphibians. Neuroscience 114, 567e575.

44

PART | I Induction and patterning of the CNS and PNS

Gorski, J.A., Talley, T., Qiu, M., Puelles, L., Rubenstein, J.L., Jones, K.R., 2002. Cortical excitatory neurons and glia, but not GABAergic neurons, are produced in the Emx1-expressing lineage. J. Neurosci. 22, 6309e6314. Götz, M., Stoykova, A., Gruss, P., 1998. Pax6 controls radial glia differentiation in the cerebral cortex. Neuron 21, 1031e1044. Grove, E.A., Tole, S., Limon, J., Yip, L., Ragsdale, C.W., 1998. The hem of the embryonic cerebral cortex is defined by the expression of multiple Wnt genes and is compromised in Gli3-deficient mice. Development 125, 2315e2325. Gunhaga, L., Marklund, M., Sjodal, M., Hsieh, J.C., Jessell, T.M., Edlund, T., 2003. Specification of dorsal telencephalic character by sequential Wnt and FGF signaling. Nat. Neurosci. 6, 701e707. Gutin, G., Fernandes, M., Palazzolo, L., Paek, H., Kai, Y., Ornitz, D., McConnell, S.K., Hébert, J.M., 2006. FGF acts independently of SHH to generate ventral telencephalic cells. Development 133, 2937e2946. Hamasaki, T., Leingärtner, A., Ringstedt, T., O’Leary, D.D., 2004. EMX2 regulates sizes and positioning of the primary sensory and motor areas in neocortex by direct specification of cortical progenitors. Neuron 43 (3), 359e372. Hanashima, C., Li, S.C., Shen, L., Lai, E., Fishell, G., 2004. Foxg1 suppresses early cortical cell fate. Science 303, 56e59. Hanashima, C., Fernandes, M., Hébert, J.M., Fishell, G., 2007. The role of Foxg1 and dorsal midline signaling in the generation of Cajal-Retzius subtypes. J. Neurosci. 27, 11103e11111. Hasenpusch-Theil, K., Watson, J.A., Theil, T., 2017. Direct interactions between Gli3, Wnt8b, and Fgfs underlie patterning of the dorsal telencephalon. Cerebr. Cortex 27 (2), 1137e1148. Hauptmann, G., Gerster, T., 1996. Complex expression of the zp-50 pou gene in the embryonic zebrafish brain is altered by overexpression of sonic hedgehog. Development 122, 1769e1780. Hayhurst, M., Gore, B.B., Tessier-Lavigne, M., McConnell, S.K., 2008. Ongoing sonic hedgehog signaling is required for dorsal midline formation in the developing forebrain. Dev. Neurobiol. 68 (1), 83e100. Hébert, J.M., McConnell, S.K., 2000. Targeting of cre to the Foxg1 (BF-1) locus mediates loxP recombination in the telencephalon and other developing head structures. Dev. Biol. 222, 296e306. Hébert, J.M., Mishina, Y., McConnell, S.K., 2002. BMP signaling is required locally to pattern the dorsal telencephalic midline. Neuron 35, 1029e1041. Hébert, J.M., Lin, M., Partanen, J., Rossant, J., McConnell, S.K., 2003. FGF signaling through FGFR1 is required for olfactory bulb morphogenesis. Development 130, 1101e1111. Heisenberg, C.P., Houart, C., Take-Uchi, M., Rauch, G.J., Young, N., Coutinho, P., Masai, I., Caneparo, L., Concha, M.L., Geisler, R., Dale, T.C., Wilson, S.W., Stemple, D.L., 2001. A mutation in the Gsk3-binding domain of zebrafish Masterblind/Axin1 leads to a fate transformation of telencephalon and eyes to diencephalon. Genes Dev. 15 (11), 1427e1434. Hirata, T., Nomura, T., Takagi, Y., Sato, Y., Tomioka, N., Fujisawa, H., Osumi, N., 2002. Mosaic development of the olfactory cortex with Pax6dependent and -independent components. Brain Res. Dev. Brain Res. 136, 17e26. Hirata, T., Li, P., Lanuza, G.M., Cocas, L.A., Huntsman, M.M., Corbin, J.G., 2009. Identification of distinct telencephalic progenitor pools for neuronal diversity in the amygdala. Nat. Neurosci. 12, 141e149. Hoch, R.V., Clarke, J.A., Rubenstein, J.L., 2015a. Fgf signaling controls the telencephalic distribution of Fgf-expressing progenitors generated in the rostral patterning center. Neural Dev. 10, 8. Hoch, R.V., Lindtner, S., Price, J.D., Rubenstein, J.L., 2015b. OTX2 transcription factor controls regional patterning within the medial ganglionic eminence and regional identity of the septum. Cell Rep. 12 (3), 482e494. Houart, C., Westerfield, M., Wilson, S.W., 1998. A small population of anterior cells patterns the forebrain during zebrafish gastrulation. Nature 391, 788e792. Houart, C., Caneparo, L., Heisenberg, C., Barth, K., Take-Uchi, M., Wilson, S., 2002. Establishment of the telencephalon during gastrulation by local antagonism of Wnt signaling. Neuron 35, 255e265. Huang, X., Litingtung, Y., Chiang, C., 2007. Region-specific requirement for cholesterol modification of sonic hedgehog in patterning the telencephalon and spinal cord. Development 134 (11), 2095e2105. Hui, C.C., Joyner, A., 1993. A mouse model of Greig cephalopolysyndactyly syndrome: the extra-toesJ mutation contains an intragenic deletion of the Gli3 gene. Nat. Genet. 3, 241e246. Imayoshi, I., Shimogori, T., Ohtsuka, T., Kageyama, R., 2008. Hes genes and neurogenin regulate non-neural versus neural fate specification in the dorsal telencephalic midline. Development 135, 2531e2541. Inoue, T., Nakamura, S., Osumi, N., 2000. Fate mapping of the mouse prosencephalic neural plate. Dev. Biol. 219, 373e383. Jimenez, D., Garcia, C., de Castro, F., Chedotal, A., Sotelo, C., de Carlos, J.A., Valverde, F., Lopez-Mascaraque, L., 2000. Evidence for intrinsic development of olfactory structures in Pax-6 mutant mice. J. Comp. Neurol. 428, 511e526. Jones, L., López-Bendito, G., Gruss, P., Stoykova, A., Molnár, Z., 2002. Pax6 is required for the normal development of the forebrain axonal connections. Development 129, 5041e5052. Kawano, Y., Kypta, R., 2003. Secreted antagonists of the Wnt signalling pathway. J. Cell Sci. 116, 2627e2634. Liem Jr., K.F., Tremml, G., Jessell, T.M., 1997. A role for the roof plate and its resident TGFbeta-related proteins in neuronal patterning in the dorsal spinal cord. Cell 91, 127e138. Kim, A.S., Anderson, S.A., Rubenstein, J.L., Lowenstein, D.H., Pleasure, S.J., 2001. Pax-6 regulates expression of SFRP-2 and Wnt-7b in the developing CNS. J. Neurosci. 21, RC132. Kimura, J., Suda, Y., Kurokawa, D., Hossain, Z.M., Nakamura, M., Takahashi, M., Hara, A., Aizawa, S., 2005. Emx2 and Pax6 function in cooperation with Otx2 and Otx1 to develop caudal forebrain primordium that includes future archipallium. J. Neurosci. 25, 5097e5108.

Telencephalon patterning Chapter | 2

45

Kohtz, J.D., Baker, D.P., Corte, G., Fishell, G., 1998. Regionalization within the mammalian telencephalon is mediated by changes in responsiveness to Sonic Hedgehog. Development 125, 5079e5089. Konno, D., Iwashita, M., Satoh, Y., Momiyama, A., Abe, T., Kiyonari, H., Matsuzaki, F., 2012. The mammalian DM domain transcription factor Dmrta2 is required for early embryonic development of the cerebral cortex. PLoS One 7 (10), e46577. Kornblum, H.I., Hussain, R.J., Bronstein, J.M., Gall, C.M., Lee, D.C., Seroogy, K.B., 1997. Prenatal ontogeny of the epidermal growth factor receptor and its ligand, transforming growth factor alpha, in the rat brain. J. Comp. Neurol. 380, 243e261. Kuschel, S., Ruther, U., Theil, T., 2003. A disrupted balance between Bmp/Wnt and Fgf signaling underlies the ventralization of the Gli3 mutant telencephalon. Dev. Biol. 260, 484e495. Ladher, R.K., Church, V.L., Allen, S., Robson, L., Abdelfattah, A., Brown, N.A., Hattersley, G., Rosen, V., Luyten, F.P., Dale, L., Francis-West, P.H., 2000. Cloning and expression of the Wnt antagonists Sfrp-2 and Frzb during chick development. Dev. Biol. 218, 183e198. Lagutin, O.V., Zhu, C.C., Kobayashi, D., Topczewski, J., Shimamura, K., Puelles, L., Russell, H.R., McKinnon, P.J., Solnica-Krezel, L., Oliver, G., 2003. Six3 repression of Wnt signaling in the anterior neuroectoderm is essential for vertebrate forebrain development. Genes Dev. 17, 368e379. Lee, K.J., Dietrich, P., Jessell, T.M., 2000a. Genetic ablation reveals that the roof plate is essential for dorsal interneuron specification. Nature 403, 734e740. Lee, S.M., Tole, S., Grove, E., McMahon, A.P., 2000b. A local Wnt-3a signal is required for development of the mammalian hippocampus. Development 127, 457e467. Lewandoski, M., Sun, X., Martin, G.R., 2000. Fgf8 signalling from the AER is essential for normal limb development. Nat. Genet. 26, 460e463. Li, G., Pleasure, S.J., 2005. Morphogenesis of the dentate gyrus: what we are learning from mouse mutants. Dev. Neurosci. 27, 93e99. Liodis, P., Denaxa, M., Grigoriou, M., Akufo-Addo, C., Yanagawa, Y., Pachnis, V., 2007. Lhx6 activity is required for the normal migration and specification of cortical interneuron subtypes. J. Neurosci. 27, 3078e3089. Litingtung, Y., Chiang, C., 2000. Control of Shh activity and signaling in the neural tube. Dev. Dynam. 219 (2), 143e154. Liu, B., Xiao, H., Zhao, C., 2018. Forced expression of Foxg1 in the cortical hem leads to the transformation of Cajal-Retzius cells into dentate granule neurons. J. Dev. Biol. 6 (3). Louvi, A., Yoshida, M., Grove, E.A., 2007. The derivatives of the Wnt3a lineage in the central nervous system. J. Comp. Neurol. 504, 550e569. Machon, O., Backman, M., Machonova, O., Kozmik, Z., Vacik, T., Andersen, L., Krauss, S., 2007. A dynamic gradient of Wnt signaling controls initiation of neurogenesis in the mammalian cortex and cellular specification in the hippocampus. Dev. Biol. 311. Mangale, V.S., Hirokawa, K.E., Satyaki, P.R., Gokulchandran, N., Chikbire, S., Subramanian, L., Shetty, A.S., Martynoga, B., Paul, J., Mai, M.V., Li, Y., Flanagan, L.A., Tole, S., Monuki, E.S., 2008. Lhx2 selector activity specifies cortical identity and suppresses hippocampal organizer fate. Science 319, 304e309. Marklund, M., Sjodal, M., Beehler, B.C., Jessell, T.M., Edlund, T., Gunhaga, L., 2004. Retinoic acid signalling specifies intermediate character in the developing telencephalon. Development 131, 4323e4332. Martynoga, B., Morrison, H., Price, D.J., Mason, J.O., 2005. Foxg1 is required for specification of ventral telencephalon and region-specific regulation of dorsal telencephalic precursor proliferation and apoptosis. Dev. Biol. 283, 113e127. Maruoka, Y., Ohbayashi, N., Hoshikawa, M., Itoh, N., Hogan, B.M., Furuta, Y., 1998. Comparison of the expression of three highly related genes, Fgf8, Fgf17 and Fgf18, in the mouse embryo. Mech. Dev. 74, 175e177. Masai, I., Heisenberg, C.P., Barth, K.A., Macdonald, R., Adamek, S., Wilson, S.W., 1997. Floating head and masterblind regulate neuronal patterning in the roof of the forebrain. Neuron 18, 43e57. McWhirter, J.R., Goulding, M., Weiner, J., Chun, J., Murre, C., 1997. A novel fibroblast growth factor gene expressed in the developing nervous system is a downstream target of the chimeric homeodomain oncoprotein E2A-Pbx1. Development 124, 3221e3232. Medina, L., Legaz, I., González, G., De Castro, F., Rubenstein, J.L., Puelles, L., 2004. Expression of Dbx1, Neurogenin 2, Semaphorin 5A, Cadherin 8, and Emx1 distinguish ventral and lateral pallial histogenetic divisions in the developing mouse claustroamygdaloid complex. J. Comp. Neurol. 474, 504e523. Meyer, G., Perez-Garcia, C.G., Abraham, H., Caput, D., 2002. Expression of p73 and Reelin in the developing human cortex. J. Neurosci. 22 (12), 4973e4986. Meyer, G., Cabrera Socorro, A., Perez Garcia, C.G., Martinez Millan, L., Walker, N., Caput, D., 2004. Developmental roles of p73 in CajaleRetzius cells and cortical patterning. J. Neurosci. 24, 9878e9887. Miyoshi, G., Hjerling-Leffler, J., Karayannis, T., Sousa, V.H., Butt, S.J., Battiste, J., Johnson, J.E., Machold, R.P., Fishell, G., 2010. Genetic fate mapping reveals that the caudal ganglionic eminence produces a large and diverse population of superficial cortical interneurons. J. Neurosci. 30 (5), 1582e1594. Molnár, Z., Butler, A.B., 2002. The corticostriatal junction: a crucial region for forebrain development and evolution. Bioessays 24, 530e541. Molotkova, N., Molotkov, A., Duester, G., 2007. Role of retinoic acid during forebrain development begins late when Raldh3 generates retinoic acid in the ventral subventricular zone. Dev. Biol. 303 (2), 601e610. Monuki, E.S., Porter, F.D., Walsh, C.A., 2001. Patterning of the dorsal telencephalon and cerebral cortex by a roof plate-Lhx2 pathway. Neuron 32, 591e604. Moreno, N., Bachy, I., Rétaux, S., González, A., 2004. LIM-homeodomain genes as developmental and adult genetic markers of Xenopus forebrain functional subdivisions. J. Comp. Neurol. 472, 52e72. Muzio, L., Mallamaci, A.J., 2005. Foxg1 confines Cajal-Retzius neuronogenesis and hippocampal morphogenesis to the dorsomedial pallium. J. Neurosci. 25, 4435e4441.

46

PART | I Induction and patterning of the CNS and PNS

Muzio, L., DiBenedetto, B., Stoykova, A., Boncinelli, E., Gruss, P., Mallamaci, A., 2002. Conversion of cerebral cortex into basal ganglia in Emx2(-/-) Pax6(Sey/Sey) double-mutant mice. Nat. Neurosci. 5, 737e745. Nakagawa, Y., O’Leary, D.D., 2001. Combinatorial expression patterns of LIM homeodomain and other regulatory genes parcellate developing thalamus. J. Neurosci. 21, 2711e2725. Nery, S., Fishell, G., Corbin, J.G., 2002. The caudal ganglionic eminence is a source of distinct cortical and subcortical cell populations. Nat. Neurosci. 5, 1279e1287. Nóbrega-Pereira, S., Kessaris, N., Du, T., Kimura, S., Anderson, S.A., Marin, O., 2008. Postmitotic Nkx2-1 controls the migration of telencephalic interneurons by direct repression of guidance receptors. Neuron 59, 733e745. Nomura, T., Osumi, N., 2004. Misrouting of mitral cell progenitors in the Pax6/small eye rat telencephalon. Development 131, 787e796. Ohkubo, Y., Chiang, C., Rubenstein, J.L., 2002. Coordinate regulation and synergistic actions of BMP4, SHH and FGF8 in the rostral prosencephalon regulate morphogenesis of the telencephalic and optic vesicles. Neuroscience 111, 1e17. Okada, T., Okumura, Y., Motoyama, J., Ogawa, M., 2008. FGF8 signaling patterns the telencephalic midline by regulating putative key factors of midline development. Dev. Biol. 320, 92e101. Paek, H., Gutin, G., Hébert, J.M., 2009. FGF signaling is strictly required to maintain early telencephalic precursor cell survival. Development 136 (14), 2457e2465. Panchision, D.M., Pickel, J.M., Studer, L., Lee, S.H., Turner, P.A., Hazel, T.G., McKay, R.D., 2001. Sequential actions of BMP receptors control neural precursor cell production and fate. Genes Dev. 15, 2094e2110. Pellegrini, M., Mansouri, A., Simeone, A., Boncinelli, E., Gruss, P., 1996. Dentate gyrus formation requires Emx2. Development 122, 3893e3898. Puelles, L., Kuwana, E., Puelles, E., Bulfone, A., Shimamura, K., Keleher, J., Smiga, S., Rubenstein, J.L., 2000. Pallial and subpallial derivatives in the embryonic chick and mouse telencephalon, traced by the expression of the genes Dlx-2, Emx-1, Nkx-2.1, Pax-6, and Tbr-1. J. Comp. Neurol. 424, 409e438. Rallu, M., Machold, R., Gaiano, N., Corbin, J.G., McMahon, A.P., Fishell, G., 2002. Dorsoventral patterning is established in the telencephalon of mutants lacking both Gli3 and Hedgehog signaling. Development 129, 4963e4974. Rash, B.G., Grove, E.A., 2007. Patterning the dorsal telencephalon: a role for sonic hedgehog? J. Neurosci. 27, 11595e11603. Rattner, A., Hsieh, J.C., Smallwood, P.M., Gilbert, D.J., Copeland, N.G., Jenkins, N.A., Nathans, J., 1997. A family of secreted proteins contains homology to the cysteine-rich ligand-binding domain of frizzled receptors. Proc. Natl. Acad. Sci. U.S.A. 94, 2859e2863. Remedios, R., Subramanian, L., Tole, S., 2004. LIM genes parcellate the embryonic amygdala and regulate its development. J. Neurosci. 24, 6986e6990. Remedios, R., Huilgol, D., Saha, B., Hari, P., Bhatnagar, L., Kowalczyk, T., Hevner, R.F., Suda, Y., Aizawa, S., Ohshima, T., Stoykova, A., Tole, S., 2007. A stream of cells migrating from the caudal telencephalon reveals a link between the amygdala and neocortex. Nat. Neurosci. 10, 1141e1150. Rickmann, M., Amaral, D.G., Cowan, W.M., 1987. Organization of radial glial cells during the development of the rat dentate gyrus. J. Comp. Neurol. 264, 449e479. Roessler, E., Belloni, E., Gaudenz, K., Jay, P., Berta, P., Scherer, S.W., Tsui, L.C., Muenke, M., 1996. Mutations in the human Sonic Hedgehog gene cause holoprosencephaly. Nat. Genet. 14 (3), 357e360. Rowitch, D.H., S-Jacques, B., Lee, S.M., Flax, J.D., Snyder, E.Y., McMahon, A.P., 1999. Sonic hedgehog regulates proliferation and inhibits differentiation of CNS precursor cells. J. Neurosci. 19 (20), 8954e8965. Ruzhynsky, V.A., McClellan, K.A., Vanderluit, J.L., Jeong, Y., Furimsky, M., Park, D.S., Epstein, D.J., Wallace, V.A., Slack, R.S., 2007. Cell cycle regulator E2F4 is essential for the development of the ventral telencephalon. J. Neurosci. 27 (22), 5926e5935. Saha, B., Hari, P., Huilgol, D., Tole, S., 2007. Dual role for LIM-homeodomain gene Lhx2 in the formation of the lateral olfactory tract. J. Neurosci. 27, 2290e2297. Saulnier, A., Keruzore, M., De Clercq, S., Bar, I., Moers, V., Magnani, D., Walcher, T., Filippis, C., Kricha, S., Parlier, D., Viviani, L., Matson, C.K., Nakagawa, Y., Theil, T., Götz, M., Mallamaci, A., Marine, J.C., Zarkower, D., Bellefroid, E.J., 2013. The doublesex homolog Dmrt5 is required for the development of the caudomedial cerebral cortex in mammals. Cerebr. Cortex 23 (11), 2552e2567. Schmid, R.S., McGrath, B., Berechid, B.E., Boyles, B., Marchionni, M., Sestan, N., Anton, E.S., 2003. Neuregulin 1-erbB2 signaling is required for the establishment of radial glia and their transformation into astrocytes in cerebral cortex. Proc. Natl. Acad. Sci. U.S.A. 100, 4251e4256. Seoane, J., Le, H.V., Shen, L., Anderson, S.A., Massagué, J., 2004. Integration of Smad and forkhead pathways in the control of neuroepithelial and glioblastoma cell proliferation. Cell 117 (2), 211e223. Shanmugalingam, S., Houart, C., Picker, A., Reifers, F., Macdonald, R., Barth, A., Griffin, K., Brand, M., Wilson, S.W., 2000. Ace/Fgf8 is required for forebrain commissure formation and patterning of the telencephalon. Development 127, 2549e2561. Shimamura, K., Rubenstein, J.L.R., 1997. Inductive interactions direct early regionalization of the forebrain. Development 124, 2709e2718. Shimamura, K., Hartigan, D.J., Martinez, S., Puelles, L., Rubenstein, J.L., 1995. Longitudinal organization of the anterior neural plate and neural tube. Development 121, 3923e3933. Shimogori, T., Banuchi, V., Ng, H.Y., Strauss, J.B., Grove, E.A., 2004. Embryonic signaling centers expressing BMP, WNT and FGF proteins interact to pattern the cerebral cortex. Development 131, 5639e5647. Shinozaki, K., Yoshida, M., Nakamura, M., Aizawa, S., Suda, Y., 2004. Emx1 and Emx2 cooperate in initial phase of archipallium development. Mech. Dev. 121, 475e489. Shinya, M., Koshida, S., Sawada, A., Kuroiwa, A., Takeda, H., 2001. Fgf signalling through MAPK cascade is required for development of the subpallial telencephalon in zebrafish embryos. Development 128, 4153e4164.

Telencephalon patterning Chapter | 2

47

Smith-Fernandez, A.S., Pieau, C., Reperant, J., Boncinelli, E., Wassef, M., 1998. Expression of the Emx-1 and Dlx-1 homeobox genes define three molecularly distinct domains in the telencephalon of mouse, chick, turtle and frog embryos: implications for the evolution of telencephalic subdivisions in amniotes. Development 125, 2099e2111. Solomon, B.D., Mercier, S., Vélez, J.I., Pineda-Alvarez, D.E., Wyllie, A., Zhou, N., Dubourg, C., David, V., Odent, S., Roessler, E., Muenke, M., 2010. Analysis of genotype-phenotype correlations in human holoprosencephaly. Am. J. Med. Genet. C Semin. Med. Genet. 154C (1), 133e141. Soma, M., Aizawa, H., Ito, Y., Maekawa, M., Osumi, N., Nakahira, E., Okamoto, H., Tanaka, K., Yuasa, S., 2009. Development of the mouse amygdala as revealed by enhanced green fluorescent protein gene transfer by means of in utero electroporation. J. Comp. Neurol. 513, 113e128. Spemann, H., Mangold, H., 1924. Über induktion von Embryonalanlagen durch Implantation artfremder Organisatoren. Wilhelm Roux Arch. Entwicklungsmech. Org. 100, 599e638. Spoelgen, R., Hammes, A., Anzenberger, U., Zechner, D., Andersen, O.M., Jerchow, B., Willnow, T.E., 2005. LRP2/megalin is required for patterning of the ventral telencephalon. Development 132, 405e414. Stenman, J., Yu, R.T., Evans, R.M., Campbell, K., 2003. Tlx and Pax6 co-operate genetically to establish the pallio-subpallial boundary in the embryonic mouse telencephalon. Development 130, 1113e1122. Storm, E.E., Rubenstein, J.L., Martin, G.R., 2003. Dosage of Fgf8 determines whether cell survival is positively or negatively regulated in the developing forebrain. Proc. Natl. Acad. Sci. U.S.A. 100, 1757e1762. Storm, E.E., Garel, S., Borello, U., Hébert, J.M., Martinez, S., McConnell, S.K., Martin, G.R., Rubenstein, J.L.R., 2006. Dosage dependent functions of Fgf8 in regulating telencephalic patterning centers. Development 133, 1831e1844. Stoykova, A., Fritsch, R., Walther, C., Gruss, P., 1996. Forebrain patterning defects in small eye mutant mice. Development 122, 3453e3465. Stoykova, A., Götz, M., Gruss, P., Price, J., 1997. Pax6-dependent regulation of adhesive patterning, R-cadherin expression and boundary formation in developing forebrain. Development 124, 3765e3777. Stoykova, A., Treichel, D., Hallonet, M., Gruss, P., 2000. Pax6 modulates the dorsoventral patterning of the mammalian telencephalon. J. Neurosci. 20, 8042e8050. Sussel, L., Marin, O., Kimura, S., Rubenstein, J.L., 1999. Loss of Nkx2.1 homeobox gene function results in a ventral to dorsal molecular respecification within the basal telencephalon: evidence for a transformation of the pallidum into the striatum. Development 126, 3359e3370. Swanson, L.W., Petrovich, G.D., 1998. What is the amygdala? Trends Neurosci. 21, 323e331. Takiguchi-Hayashi, K., Sekiguchi, M., Ashigaki, S., Takamatsu, M., Hasegawa, H., Suzuki-Migishima, R., Yokoyama, M., Nakanishi, S., Tanabe, Y., 2004. Generation of reelin-positive marginal zone cells from the caudomedial wall of telencephalic vesicles. J. Neurosci. 24 (9), 2286e2295. Tao, W., Lai, E., 1992. Telencephalon-restricted expression of BF-1, a new member of the HNF-3/fork head gene family, in the developing rat brain. Neuron 8, 957e966. Theil, T., Alvarez-Bolado, G., Walter, A., Ruther, U., 1999. Gli3 is required for Emx gene expression during dorsal telencephalon development. Development 126, 3561e3571. Theil, T., Aydin, S., Koch, S., Grotewold, L., Ruther, U., 2002. Wnt and Bmp signalling cooperatively regulate graded Emx2 expression in the dorsal telencephalon. Development 129, 3045e3054. Theil, T., Dominguez-Frutos, E., Schimmang, T., 2008. Differential requirements for Fgf3 and Fgf8 during mouse forebrain development. Dev. Dynam. 237 (11), 3417e3423. Tole, S., Grove, E.A., 2001. Detailed field pattern is intrinsic to the embryonic mouse hippocampus early in neurogenesis. J. Neurosci. 21, 1580e1589. Tole, S., Ragsdale, C.W., Grove, E.A., 2000. Dorsoventral patterning of the telencephalon is disrupted in the mouse mutant extra-toes. Dev. Biol. 217, 254e265. Tole, S., Remedios, R., Saha, B., Stoykova, A., 2005. Selective requirement of Pax6, but not Emx2, in the specification and development of several nuclei of the amygdaloid complex. J. Neurosci. 25, 2753e2760. Toresson, H., Campbell, K., 2001. A role for Gsh1 in the developing striatum and olfactory bulb of Gsh2 mutant mice. Development 128, 4769e4780. Toresson, H., Potter, S., Campbell, K., 2000. Genetic control of dorsal-ventral identity in the telencephalon: opposing roles for Pax6 and Gsh2. Development 127, 4361e4371. Tucker, E.S., Segall, S., Gopalakrishna, D., Wu, Y., Vernon, M., Polleux, F., Lamantia, A.S., 2008. Molecular specification and patterning of progenitor cells in the lateral and medial ganglionic eminences. J. Neurosci. 28 (38), 9504e9518. Walshe, J., Mason, I., 2003. Unique and combinatorial functions of Fgf3 and Fgf8 during zebrafish forebrain development. Development 130, 4337e4349. Wang, B., Waclaw, R.R., Allen 2nd, Z.J., Guillemot, F., Campbell, K., 2009. Ascl1 is a required downstream effector of Gsx gene function in the embryonic mouse telencephalon. Neural Dev. 4, 5. van de Water, S., van de Wetering, M., Joore, J., Esseling, J., Bink, R., Clevers, H., Zivkovic, D., 2001. Ectopic Wnt signal determines the eyeless phenotype of zebrafish masterblind mutant. Development 128 (20), 3877e3888. Wichterle, H., Turnbull, D.H., Nery, S., Fishell, G., Alvarez-Buylla, A., 2001. In utero fate mapping reveals distinct migratory pathways and fates of neurons born in the mammalian basal forebrain. Development 128, 3759e3771. Willaime-Morawek, S., Seaberg, R.M., Batista, C., Labbe, E., Attisano, L., Gorski, J.A., Jones, K.R., Kam, A., Morshead, C.M., van der Kooy, D., 2006. Embryonic cortical neural stem cells migrate ventrally and persist as postnatal striatal stem cells. J. Cell Biol. 175, 159e168. Wilson, S.W., Houart, C., 2004. Early steps in the development of the forebrain. Dev. Cell 6, 167e181. Wonders, C.P., Anderson, S.A., 2006. The origin and specification of cortical interneurons. Nat. Rev. Neurosci. 7, 687e696. Wullimann, M.F., Mueller, T., 2004. Identification and morphogenesis of the eminentia thalami in the zebrafish. J. Comp. Neurol. 471 (1), 37e48.

48

PART | I Induction and patterning of the CNS and PNS

Xu, Q., Cobos, I., De La Cruz, E., Rubenstein, J.L., Anderson, S.A., 2004. Origins of cortical interneuron subtypes. J. Neurosci. 24, 2612e2622. Xu, Q., Tam, M., Anderson, S.A., 2008. Fate mapping Nkx2.1-lineage cells in the mouse telencephalon. J. Comp. Neurol. 506, 16e29. Xuan, S., Saptista, C.A., Balas, G., Tao, W., Soares, V.C., Lai, E., 1995. Winged helix transcription factor BF-1 is essential for the development of the cerebral hemispheres. Neuron 14, 1141e1152. Yoshida, M., Assimacopoulos, S., Jones, K.R., Grove, E.A., 2006. Massive loss of Cajal-Retzius cells does not disrupt neocortical layer order. Development 133, 537e545. Yozu, M., Tabata, H., Nakajima, K., 2005. The caudal migratory stream: a novel migratory stream of interneurons derived from the caudal ganglionic eminence in the developing mouse forebrain. J. Neurosci. 25 (31), 7268e7277. Yu, W., Wang, Y., McDonnell, K., Stephen, D., Bai, C.B., 2009. Patterning of ventral telencephalon requires positive function of Gli transcription factors. Dev. Biol. 334 (1), 264e275. Yun, K., Potter, S., Rubenstein, J.L., 2001. Gsh2 and Pax6 play complementary roles in dorsoventral patterning of the mammalian telencephalon. Development 128, 193e205. Zhao, T., Szabo, N., Ma, J., Luo, L., Zhou, X., Alvarez-Bolado, G., 2008. Genetic mapping of Foxb1-cell lineage shows migration from caudal diencephalon to telencephalon and lateral hypothalamus. Eur. J. Neurosci. 28, 1941e1955. Zhou, C.J., Zhao, C., Pleasure, S.J., 2004. Wnt signaling mutants have decreased dentate granule cell production and radial glial scaffolding abnormalities. J. Neurosci. 24, 121e126. Zimmer, C., Lee, J., Griveau, A., Arber, S., Pierani, A., Garel, S., Guillemot, F., 2010. Role of Fgf8 signalling in the specification of rostral Cajal-Retzius cells. Development 137, 293e302.

Chapter 3

Area patterning of the mammalian neocortex Elizabeth A. Grove Department of Neurobiology, The Grossman Institute for Neuroscience, University of Chicago, Chicago, IL, United States

Chapter outline 3.1. Introduction 3.1.1. Basic principles 3.1.2. Classic neocortical area patterning models 3.2. Indications that intrinsic mechanisms pattern the neocortical primordium 3.3. Morphogens impart position to the neocortical primordium 3.3.1. Morphogen signaling 3.3.2. Neocortical patterning by FGFs 3.3.3. Fgf8 regulates neocortical guidance of thalamic axons 3.3.4. Neocortical patterning by the cortical hem 3.4. Patterning genes downstream of morphogen signaling 3.4.1. Emx2 and Pax6 3.4.2. Dmrt5/Dmrta2 3.4.3. Couptf1/Nr2f1 3.4.4. Sp8 3.4.5. Pbx

49 49 50 51 52 52 53 54 55 55 56 57 57 58 58

3.5. Do neocortical areas arise from dedicated progenitor cell pools? 3.5.1. Transcription factors known to pattern the NP appear in gradients, not domains 3.5.2. Mapping the cortical primordium with forebrain enhancers 3.6. The influence of thalamic innervation 3.6.1. Guidance of thalamocortical axons and area formation 3.6.2. Thalamic innervation determines the function of a cortical area 3.6.3. Effects of thalamocortical afferents on area size and cortical progenitor cells 3.6.4. Thalamic dependence of an area-specific feature 3.6.5. Two mechanisms united 3.7. Spontaneous activity and neocortical patterning 3.8. Conservation of patterning mechanisms among different mammalian species 3.9. Conclusions References

59 59 59 60 60 61 62 62 62 63 63 64 64

3.1 Introduction 3.1.1 Basic principles A spectacular advance in biology has been to uncover principles and molecular mechanisms that underlie embryonic patterning of the vertebrate and invertebrate body plans (Wolpert, 1996). These principles have been applied to understanding morphogenesis of structures as diverse as the fly wing and the vertebrate spinal cord. In the embryonic vertebrate spinal cord, progenitor cells respond to different levels of a signaling molecule, or morphogen, by expressing different transcription factors (TFs) (Wolpert, 1996). Neighboring TFs regulate one another’s gene expression to establish distinct progenitor domains that generate different neuron subtypes (Briscoe and Ericson, 2001). The pattern discussed here is the “area map” of the neocortex, the brain structure that controls our higher cerebral functions, including memory, cognition, and decision-making. The functional organization of neocortex is its division into different areas that form a map over the cortical sheet. To date, most studies of how the area map is assembled have used

Patterning and Cell Type Specification in the Developing CNS and PNS. https://doi.org/10.1016/B978-0-12-814405-3.00003-5 Copyright © 2020 Elsevier Inc. All rights reserved.

49

50

PART | I Induction and patterning of the CNS and PNS

FIGURE 3.1 The mouse area map at P6 revealed by expression of selected genes and proteins. (A) Schematic of a P6 forebrain, dorsolateral view, showing the positions of Fr, S1, V1, and A1. Anterior (a), posterior (p), medial (m), and lateral (l) in the map indicated at lower left are roughly the same for (AeD). (B, C) Lateral view of P6 hemispheres. (D) Horizontal section through flattened P6 cortex. (BeD) Cux1 expression (B) and immunoreactivity for the serotonin transporter (SERT-IR) (D) delineate subfields of S1 and individual barrels of the larger barrel field (AeE). V1 is a posterior, positively stained triangular shape (B, D). A1 is a roughly oval field posterolateral to S1 (D). Cdh8 expression marks frontal cortex (Fr) (C), which is negative for SERT-IR (D). Cdh8 expression also marks V1, which stands out against darker staining in surrounding higher-order visual areas. Scale bar: D (for BeD), 1 mm. fl, forelimb; hl, hindlimb; Li, limbic cortex; lj, lower jaw; pmbsf, posteromedial barrel subfield. From Assimacopoulos, S., Kao, T., Issa, NP., Grove, EA., 2012. Fibroblast growth factor 8 organizes the neocortical area map and regulates sensory map topography. J. Neurosci. 32, 7191e7201.

mice as their model system and have therefore focused on areas that take up most of the mouse neocortex, namely primary motor and primary sensory areas (Fig. 3.1). Although carnivores and primates have many additional prominent areas, these primary areas have similar relative positions with respect to the anterior to posterior (A/P) and medial to lateral (M/L) neocortical axes (Nauta and Feirtag, 1986; Krubitzer and Stolzenberg, 2014). That is, across examined mammalian species, primary visual cortex (V1) is posterior in the cortical hemisphere, primary somatosensory area (S1) is anterior to V1, and primary motor cortex (M1) is anterior to S1. Primary auditory cortex (A1) is lateral to V1. These similarities suggest that basic organization of the area map and perhaps the mechanisms of area patterning are conserved across mammals (Nauta and Feirtag, 1986). Other structures are patterned in the embryo by signaling centers that lie at the boundaries of the tissue primordium. These centers release signaling proteins, termed morphogens, which disperse through the tissue establishing protein gradients that confer positional information (Wolpert, 1996). Morphogens regulate the patterned expression of genes encoding TFs and other proteins, which continue the patterning process. Findings that fit development of the neocortical area map to such a model are reviewed in this chapter. Development of the area map cannot be understood, however, by focusing on the neocortical primordium (NP) in isolation. Because embryonic tissues interact to pattern one another, other parts of the brain would be expected to influence the neocortex. The cerebral cortex is the apex of a hierarchy of brain structures. Information from the outside world or from other brain structures must access the cerebral cortex via the thalamus, virtually exclusively (Nauta and Feirtag, 1986). A key feature of the area map is that different areas receive distinct sets of projections from different thalamic nuclei, relaying different modalities of information. The specific thalamic input an area receives is therefore central to its functional identity, but thalamocortical innervation also contributes in development to formation of the area. This chapter also surveys how thalamic axons and their activity, both prenatal and postnatal, refine area boundaries and establish area-specific features such as the barrel fields of S1.

3.1.2 Classic neocortical area patterning models The Protomap Model: The protomap model was proposed before advances in identifying the molecular mechanisms of NP patterning, but it already suggested that the neocortex, similar to other parts of the embryo, is patterned by “molecular

Area patterning of the mammalian neocortex Chapter | 3

51

FIGURE 3.2 The protomap and protocortex models of area formation. (A) Schematics of cross sections through the NP when only progenitor cells are present. (B) Cross sections through the neonatal NP when thalamocortical axons are entering the cortical plate (CP). (Left) Protomap model: Intrinsic area differences are specified in the progenitor cells of the VZ by molecular determinants. As pyramidal neurons (white ovals) migrate radially out of the VZ, they transfer the protomap to the CP. Area-specific axon guidance cues set up by the protomap (not in the original model, but see text) help to guide thalamocortical axons into the correct area (indicated by color coding of axons [curving lines] and protoareas). (Right) Protocortex model: The cortical primordium is essentially homogeneous while it is generated and is patterned into areas later by cues from axons growing in from the thalamus. Thalamic axons thus impart positional information and area identity to the CP (indicated by color coding of axons and areas).

mechanisms” as NP progenitor cells are dividing. Intrinsic area differences, specified by these molecular determinants, are established in the ventricular zone (VZ) forming the “protomap.” Newborn neurons migrate out of the VZ in radial arrays to form the incipient gray matter of the neocortex, carrying the area protomap with them (Rakic, 1988). The extraordinarily distinct columns of migrating neurons seen in embryonic monkey brains, reaching from the VZ to the cortical plate, and termed “radial units,” fostered this model (Rakic, 1988). The Protocortex Model: The cytoarchitectonic features that classically define areas do not appear in mice until a week or more after birth, similar to the more recently identified gene expression patterns that demarcate areas in the map (Fig. 3.1). Moreover, early transplant experiments suggested a high level of plasticity in neocortical area identity (Schlaggar and O’Leary, 1991). In the protocortex model, therefore, the NP is essentially homogeneous as it is generated (a “tabula rasa”) and is finally patterned into areas by activity-dependent mechanisms arising when thalamic afferents grow into the developing neocortex (O’Leary, 1989). Thalamic afferents arrive after growth of the cortical primordium is well underway, whereas, in other embryonic systems, basic patterning is initiated before major growth (Edgar and Lehner, 1996), which would distinguish the NP oddly from other structures. Each model is presented (Fig. 3.2) in an extreme version to emphasize key points. Both models as they were actually proposed suggested that the neocortex is predominantly patterned by one mechanism, but, ultimately, by both (Rakic, 1988; O’Leary, 1989). As an example, when thalamic afferents to area 17 (V1) were reduced by bilateral enucleation in embryonic monkeys, for example, a new cytoarchitectonic area appeared along the border of area 17 (V1), termed area X (Rakic et al., 1991). The deafferented area neither shrank nor became part of area 18 (V2) as a result of invasion by thalamic axons targeted for area 18. A plausible interpretation was that area 17 had already been at least partially determined by mechanisms intrinsic to the cortex so that inappropriate innervation created a hybrid area (Rakic et al., 1991). The two models of neocortical patterning were influential in bringing considerable attention to the question of how the neocortical area map is set up. The ultimate conclusion of this chapter is, of course, that intrinsic and extrinsic signals cooperate to build the normal area map. Currently, however, this conclusion is based on many intervening years of research into the problem.

3.2 Indications that intrinsic mechanisms pattern the neocortical primordium Although there was evidence for early plasticity in the NP (O’Leary and Stanfield, 1989; O’Leary et al., 1994), a variety of studies in the 1990s indicated that area identity is specified by mechanisms intrinsic to the embryonic NP. First, in several experiments, part of a presumptive area in the embryonic rat NP was dissected out and grafted into a different neocortical region in a neonatal rat. Presumptive rat V1 transplanted into parietal cortex attracted axonal input from the visual relay

52

PART | I Induction and patterning of the CNS and PNS

nucleus of the thalamus, the dorsal lateral geniculate nucleus (dLGN), but little or no input from the thalamic nuclei that innervated the host parietal cortex (Gaillard and Roger, 2000). Transplants of embryonic tissue grafted at even greater distances from their normal position attracted graft-appropriate innervation (Frappe et al., 1999). When a piece of frontal cortex (FC) was positioned in occipital cortex, at the opposite end of the hemisphere, thalamocortical axons correctly innervated the native FC in the host animal and then turned posteriorly, traveling through the NP, to innervate the graft. These findings suggested that area identity was determined in the embryonic cortical graft when it was explanted, and that the graft contained guidance cues for appropriate thalamic input. Second, although the early NP appeared at first glance to be morphologically homogeneous, closer observation suggested heterogeneity. Explant experiments in rat and mouse, similar to those above, showed that presumptive limbic cortex and S1 had already been determined to express, respectively, the limbic systemeassociated membrane protein (LAMP) and the H-2Z1 transgene (Barbe and Levitt, 1991; Cohen-Tannoudji et al., 1994). H-2Z1 drives galactosidase expression in an enhancer trap transgenic mouse line, and X-gal labels, postnatally, neurons in layer 4 of S1 (Cohen-Tannoudji et al., 1994). Grafting experiments indicated that H-2Z1 activity was determined in presumptive S1 as early as embryonic day (E) 12, about a week before birth and several days before thalamic axons reach the cortex (Cohen-Tannoudji et al., 1994; Narboux-Neme et al., 2012). Third, mutant mice in which thalamocortical axons fail to reach the neocortex were examined for evidence of regional patterning in the NP. In mice in which the TF Gbx2 was constitutively deleted, thalamocortical projections were essentially absent. Yet, region-specific expression of several genes appeared in normal patterns in the neocortex (Miyashita-Lin et al., 1999). This included genes encoding TFs, a classic cadherin and a calcium-binding protein, all regionally expressed in wild-type NP before birth and in some cases afterward. Observations of Mash1-deficient mice with aberrant thalamocortical pathfinding also revealed normal regional gene expression (Nakagawa et al., 1999). These studies provided the first evidence of early regional gene expression in the NP without the thalamus, but neither mutant examined lived after the day of birth. A more recent study reported that mouse neocortex lacking both thalamocortical and corticothalamic axons, the “cortex isole,” showed grossly normal region-specific gene expression at postnatal day (P) 6, though not at P20 when the cortex had deteriorated (Zhou et al., 2010). One study of postnatal mice without thalamocortical connectivity found a lack of clear area boundaries, however, which should be noted (Vue et al., 2013). Nonetheless, the three other sets of studies together strongly suggest that signals intrinsic to the NP provide initial regional patterning of the area map.

3.3 Morphogens impart position to the neocortical primordium 3.3.1 Morphogen signaling Morphogens are chemical factors that direct morphogenesis or pattern formation in the embryo (Turing, 1952). In the embryonic vertebrate spinal cord, several morphogens determine the dorsal/ventral (D/V) axis. One of these, sonic hedgehog (SHH), generated by the ventral floor plate and notochord, disperses in a high ventral to low dorsal gradient in the spinal cord primordium. The gradient of SHH upregulates expression of different TFs at different D/V positions in the ventral half of the cord. Next, pairs of TFs repress one another’s expression, creating progenitor domains with distinct boundaries. These progenitor cell pools generate a range of ventral neuron types, for example, motor neurons (Briscoe et al., 2000; Briscoe and Ericson, 2001). In the dorsal half of the spinal cord, members of other signaling molecule families, bone morphogenetic proteins (BMPs) and Wingless-related integration site (WNT) proteins, regulate dorsal cell type and the fate of neural crest cells originating in the dorsal cord (Liem et al., 1995, 1997; Dorsky et al., 1998). Potentially more relevant to patterning the neocortex, patterning the dorsal telencephalon in chick is coordinately regulated by sources of BMP4, SHH, and FGF8 (Ohkubo et al., 2002). Why should mammalian neocortex, which derives from the dorsal telencephalon, be different? Patterning molecules have been identified in and near the mouse NP at characteristic sites for embryonic signaling centers, namely at the poles and boundaries of the tissue to be patterned. Multiple Bmp genes are expressed in the dorsomedial telencephalon in the very early embryo, and subsequently both Bmp and Wnt genes are expressed in a corresponding region at the medial edge of the cortical primordium, termed the cortical hem (Furuta et al., 1997; Grove et al., 1998). Before anterior neural tube closure, several fibroblast growth factor (FGF) genes, including Fgf8, Fgf17, and Fgf18, are expressed at the anterior neural ridge (Crossley and Martin, 1995; Bachler and Neubuser, 2001). Once the tube closes, these Fgf genes continue to be expressed at the anterior pole of each developing telencephalic hemisphere, forming in each hemisphere the “rostral patterning center” (RPC) (Crossley and Martin, 1995; Bachler and Neubuser, 2001; Cholfin and Rubenstein, 2007, 2008). The cortical hem and RPC appear in other mammals, too. The cortical hem is present in humans (Abu-Khalil et al., 2004), and both the RPC and hem in the ferret (Jones et al., 2019).

Area patterning of the mammalian neocortex Chapter | 3

53

3.3.2 Neocortical patterning by FGFs The prominent role of FGF8 at the isthmic organizer, a signaling center between the midbrain and hindbrain, in A/P patterning of the midbrain (Crossley et al., 1996; Martinez et al., 1999; Harada et al., 2016), suggests that FGF8, potentially in concert with other FGFs, confers A/P positional values to the NP. At E9.5e10.5, FGF8 disperses across the NP from A/P as a gradient with an exponential decline (Toyoda et al., 2010), thus resembling a classic morphogen (Wolpert, 1996). Mice constitutively deficient in Fgf8 die at gastrulation (Sun et al., 1999). Therefore, to examine the functions of FGF8 in the brain and other parts of the embryo, an Fgf8 mutant allelic series was generated, including Fgf8 hypomorphs and a conditional telencephalic deletion of Fgf8 (Meyers et al., 1998). Reduced levels of FGF8 in Fgf8 hypomorphic mice caused a posterior shift in the patterns of neocortical gene expression (Garel et al., 2003). That is, anterior NP regions were shrunken and posterior regions enlarged (Garel et al., 2003). Among the genes examined were those encoding the TFs, Emx2 and Nr2f1, which are directly involved in area patterning, as described below. Mice hypomorphic for Fgf8 die at birth, before neocortical areas are distinct. Regulating levels of gene expression in the mouse NP with in utero microelectroporation (IUME), however, result in mouse pups that are born normally and live for several weeks. To alter levels of FGF8 in the mouse embryo NP, or to introduce new FGF8 sources, plasmids containing Fgf8 or a dominant negative FGF8 receptor (dnFgfr3c) were introduced into the NP at E10.5 or E11.5 with IUME (Fukuchi-Shimogori and Grove, 2001; Assimacopoulos et al., 2012). Excess anterior FGF8 augments endogenous FGF8, and DN-FGFR3c sequesters FGF8, reducing available FGF8 levels (Toyoda et al., 2010). Because at these early embryonic ages, plasmids are quickly diluted out by cell division, NP cells were effectively exposed to excess, ectopic, or decreased FGF8 for only about 2 days, commensurate with the normal time of exposure to the endogenous source. At postnatal day 6 (P6), expression of several genes and proteins clearly marks out sensory and motor area boundaries (Fig. 3.3). Thus, electroporated brains were examined at P6. Excess anterior FGF8 enlarged anterior areas and shrank more posterior areas, thereby pushing area boundaries posteriorly. Anterior sites of DN-FGFR3c produced opposite changes. Excess FGF8 shifted the whisker barrel fields of S1, normally centrally located in the cortical hemisphere, into the posterior half of the hemisphere. Diminished FGF8 shifted the fields into the anterior half (Fukuchi-Shimogori and Grove, 2001). These results, combined with observations of the Fgf8 hypomorph (Garel et al., 2003), suggested that FGF8 establishes A/P position in the NP. To test this definitively, Fgf8 was electroporated into the posterior NP creating a second FGF8 source opposite the endogenous anterior source. The new FGF8 source induced a duplicate of the S1 barrel fields that was mirror-reversed along the A/P axis relative to the native S1. That is, posterior FGF8 locally reset the A/P axis of the NP (Fukuchi-Shimogori and Grove, 2001) (Fig. 3.4). FGF8, therefore, imparts positional values to the NP, in particular along the A/P axis, although the medial position of the anterior FGF source suggests a possible influence on the M/L axis as well. Positional values are fine-grained enough that a second, posterior FGF8 source induces secondary barrel fields with the correct number of individual barrels in each distinct barrel row of the PMBSF. Further, FGF8-induced barrel fields are similar in size to the endogenous fields.

FIGURE 3.3 Cortical signaling centers. (A, B) Whole mount E10.5 mouse brains processed with in situ hybridization for Fgf8 or Wnt3a, dorsal view, anterior is up. Fgf8 is strongly expressed at the anterior pole of the telencephalon (A) and Wnt3a along the medial edge of both hemispheres in the cortical hem (B). (C) A schematic indicating the dispersion of FGFs from the RPC and of BMPs and WNTs from the hem (thick arrows). Arrows with broken lines indicate the direction of influence of theoretical additional centers.

54

PART | I Induction and patterning of the CNS and PNS

FIGURE 3.4 Ectopic FGF8 induces duplicate somatosensory barrel fields. (A, B) Horizontal sections, anterior to the left, through flattened brains labeled for SERT immunohistochemistry (A) or cytochrome oxidase (B). (C) E10.5 mouse brain in dorsal view, anterior up. (D) Chart of IUME sites plotted on lateral face of a standard P6 hemisphere. (A) Primary S1 in a wild-type mouse. SERT-IR shows the subfields of S1 including the pmbsf, representing the main whiskers (rows lettered) and the second subfield representing the smaller whiskers on the snout. (B) Brain electroporated with Fgf8 at a posterior site (yellow oval in (C) shows typical target). The new, posterior source of FGF8 has induced a second S1 (SI2), mirror image to the endogenous S1 (SI1) (B). Compare positions of subfields hl, fl, lj, and pmbsf in (A) to those in the two S1s in (B). (D) Posterior positions of 20 Fgf8 electroporation sites that induced duplicate barrels plotted on a standard P6 hemisphere. From Assimacopoulos, S., Kao, T., Issa, NP., Grove, EA., 2012. Fibroblast growth factor 8 organizes the neocortical area map and regulates sensory map topography. J. Neurosci. 32, 7191e7201.

FGF17 belongs to the same subfamily of FGFs as FGF8, and Fgf17 is also expressed at the RPC. FGF17 is less active than FGF8 in establishing the A/P axis of the NP (Toyoda et al., 2010), but it is particularly important for patterning FC (Cholfin and Rubenstein, 2007, 2008). The FC contains areas responsible for motor function and social behaviors and in higher mammals for cognition and short-term memory. Mice deficient in Fgf17 show a selective reduction of the dorsomedial FC (Cholfin and Rubenstein, 2007), whereas Fgf8 mutants have an A/P patterning defect, in which dorsomedial FC is transformed into a cingulate cortex. Mice lacking one or two copies of Fgf17 are viable and normal in a range of behavioral tests but display striking decreases in a range of social interactions, presumably caused by defects in the FC (Scearce-Levie et al., 2007). Several FGFs appear at or near the RPC and although their functional interactions have been clarified, the significance on NP patterning requires further study. FGF15, for example, belongs to a different FGF subfamily from FGF8 and has opposite effects from FGF8 on the balance of proliferation versus differentiation. FGF15 represses proliferation and promotes neural differentiation, and additionally promotes expression of a patterning TF Nr2f1 (see below) that is repressed by FGF8 (Borello et al., 2008). FGF8 promotes expression of genes encoding FGF negative regulators, including Sprouty1-2 (Spry1, Spry2) (Faedo et al., 2010). Spry1/2 in turn inhibit the signaling pathways that FGF8 activates, including the Ras-Erk and PI3 kinase-Akt pathways. Mice lacking both Spry1 and Spry2 thus show a neocortical area phenotype similar to that of mice with excess anterior FGF8 (Faedo et al., 2010). Levels of FGF8, and other FGFs in the NP, clearly need to be precisely controlled. The complex pathways by which this is accomplished warrant further investigation.

3.3.3 Fgf8 regulates neocortical guidance of thalamic axons Axons from the somatosensory relay nucleus in the thalamus (ventroposterior medial nucleus, VPM) faithfully tracked changes in S1 area position and innervated both shifted and duplicated S1 (Fukuchi-Shimogori and Grove, 2001).

Area patterning of the mammalian neocortex Chapter | 3

55

Innervation was at least partially functional, given that ablating whisker follicles obliterated the corresponding barrels in both endogenous and duplicate barrel fields (Assimacopoulos et al., 2012). These observations indicated that shifted or duplicated areas could attract thalamic axons to ectopic targets as the axons grew in, raising the question of where in the NP ectopic guidance cues were positioned. The subplate (SP) is a layer of neurons beneath the developing canonical layers of the NP. A large body of work indicates that thalamic axons interact with the SP to innervate target areas (Friauf et al., 1990; Ghosh et al., 1990; Ghosh and Shatz, 1993; Allendoerfer and Shatz, 1994; Herrmann et al., 1994; Finney et al., 1998). Local damage of the SP prevents the entry of axons into their target cortical area (Ghosh et al., 1990), and in reeler mutants, thalamocortical axons grow to the displaced SP near the cortical surface, before diving down to innervate the layers of the neocortex (Caviness and Frost, 1983). Evidently, thalamocortical axons are required to contact the SP before synapsing with neurons in the true layers of the neocortex. Additionally, active synaptic interactions between thalamic axons and the SP are needed for neocortical target selection (Friauf et al., 1990; Herrmann et al., 1994; Finney et al., 1998). Blocking activity in SP neurons in cats, for example, greatly perturbs the density, pattern, and precision of dLGN axons innervating V1 (Catalano and Shatz, 1998). These authors left open, however, whether activity is instructive for growing thalamocortical axons or permissive, allowing the axons to read target-derived molecular cues (Catalano and Shatz, 1998). Altering levels of FGF8 in the NP using IUME provided evidence that FGF8 imparts A/P positional values to the progenitor cells of the SP as well as the NP (Shimogori and Grove, 2005). The SP and the true layers of the NP are generated in temporal order, with SP neurons born earliest. Thus, regional gene expression in the SP and putative regional identity are shifted by IUME of Fgf8 at E10.5, just as progenitor cells begin to give rise to SP neurons. A day later, however, at E11.5, electroporation of Fgf8 shifts regional gene expression and regional identity in the NP but no longer in the SP where most neurons are now generated (Shimogori and Grove, 2005). This serendipitous discovery allowed experiments that pulled the regional identities of the NP and SP in or out of register. Anterior Fgf8 electroporation at E10.5 caused both regions in the SP and areas in the NP to shift posteriorly in register. DiI staining from a deposit in the somatosensory VPM at P5 labeled axons that reached the cortex and traveled posteriorly in the SP to a position beneath the shifted S1. The axons then crossed the SP correctly and innervated the nascent S1 (Shimogori and Grove, 2005). These observations confirmed that FGF8 imparts A/P positional values to the SP, which appear likely to be represented by area-specific molecular guidance cues instructive to thalamocortical axons. Anterior electroporation of Fgf8 at E11.5 altered FGF8 levels after SP neurons are born, pulling the NP and SP out of register. In this case, VPM axons crossed the SP at the expected S1 location and then apparently recognized the shifted S1 and traveled posteriorly, within the neocortex, to innervate the shifted target area and form barrels in layer 4 (Shimogori and Grove, 2005). The latter behavior is reminiscent of that of heterochronic grafts described at the beginning of this chapter. Explants were taken from presumptive areas in embryonic rats and grafted into a host animal a few days old. In the host neocortex, thalamocortical axons had already passed through the SP and innervated the correct host area. These axons, or their axonal branches, were therefore ready to recognize and travel through the neocortex to innervate an appropriate ectopic graft. Thalamic axons therefore do not enter the NP and induce area identity on a blank slate. Instead, they depend on molecular and activity-based interactions with the SP, and further guidance cues from within the neocortex itself to find and innervate the correct area and layers.

3.3.4 Neocortical patterning by the cortical hem A second candidate signaling center, the cortical hem, lies along the medial edge of the cortex, next to the hippocampus, and expresses multiple Wnt and Bmp genes (Furuta et al., 1997; Grove et al., 1998). The hem is most important as an organizer for the hippocampus (Mangale et al., 2008), and without the hem, or b-catenin-mediated WNT signaling from the hem, the hippocampus fails to develop (Galceran et al., 2000; Lee et al., 2000). Moreover, ectopic hem structures can induce orderly hippocampal fields (Mangale et al., 2008). In neocortex, genetic deletion of the hem causes a reduction of dorsomedial neocortex and an expansion of ventrolateral cortex, largely composed of the olfactory cortex. The hem therefore influences the M/L axis of the cerebral cortex and to some extent the A/P axis. The greatest loss of neocortex in hem-ablated animals is in the posterior occipital region, where primary and secondary visual areas will develop.

3.4 Patterning genes downstream of morphogen signaling As signaling sources, both the hem and the RPC shrink as embryogenesis progresses and the brain grows larger. Local sources of FGF8 play other roles in cortical development, for example, in the generation of anterior CajaleRetzius cells

56

PART | I Induction and patterning of the CNS and PNS

(Zimmer et al., 2009). The germinal zone of the NP as a whole, however, becomes too large to allow widespread dispersion of FGF8 or WNT proteins. Positional information is carried on by TFs expressed in gradients across the NP that are initially shaped by FGF8 or WNT3a signaling or both.

3.4.1 Emx2 and Pax6 The first TF genes proposed to pattern the area map were Emx2, an ortholog of the Drosophila genes empty spiracles (ems), and Pax6 (small eye, orthologous to Drosophila eyeless) (Bishop et al., 2000). Both ems and eyeless have central roles in generating head structures in the fly. Emx2 has a posteromedial high to anterolateral low expression gradient in the NP, and accordingly, in mice lacking Emx2 function, posteromedial NP regions shrink, and anterior frontal NP expands. Pax6 is expressed in high anterolateral to low posteromedial gradient in the NP, precisely opposite to the Emx2 expression gradient, and mice lacking Pax6 function (sey/sey) show opposite shifts to those in the Emx2 mutant (Bishop et al., 2000) (Fig. 3.5). A detailed study of the two mutant mice assessed the regional expression of several more genes regionally expressed in the E18.5 NP. Posterior shifts of gene expression in the Emx2 mutant, opposed to anterior shifts in the Pax6 mutant, were particularly striking in the case of gene pairs that are expressed in complementary patterns along the A/P axis of the NP (Rorb and p75; Efna5 and EphA7) (Bishop et al., 2000, 2002). Mice constitutively deficient in Emx2 or Pax6 die at or before birth. A transgenic mouse line carrying excess Emx2 under the control of the nestin promoter, however, lives past birth. In this mouse, excess Emx2 is introduced throughout the NP but still shows a high-posterior to low-anterior gradient. Excess Emx2 has the opposite effect of Emx2 loss, enlarging V1 and shrinking more anterior areas such as M1 (Hamasaki et al., 2004). Substantial evidence therefore confirms that, in mouse, Emx2 regulates the size and position of areas in the neocortical map, in particular promoting the development of posterior areas (Hamasaki et al., 2004). Consistent with the opposing actions of Emx2 and Pax6 on area size and position, the gradients of Emx2 and Pax6 expression depend on an interaction between the Emx2 and Pax6 proteins (Muzio et al., 2002). Overexpression of Emx2 in the NP, for example, represses Pax6 expression (Hamasaki et al., 2004). Finally, both signaling sources described above

FIGURE 3.5 Effect of excess Emx2 on the neocortical area map. (A) P7 mouse brains in dorsal view, anterior is up. Brain on left is wild-type (wt), on right is a brain overexpressing Emx2 from a nestin-Emx2 (ne-Emx2) transgenic mouse. The transgenic mouse has higher levels of EMX2 protein throughout the NP. The triangular V1 is much larger in the transgenic than in the wild-type mouse, and motor cortex (M) is concomitantly smaller (B). Overall, the two brains are of similar size. Scale bar ¼ 1 mm. Slightly adapted from Hamasaki, T., Leingartner, A., Ringstedt, T., O’Leary, DD., 2004. EMX2 regulates sizes and positioning of the primary sensory and motor areas in neocortex by direct specification of cortical progenitors. Neuron 43, 359e372, with permission from Neuron.

Area patterning of the mammalian neocortex Chapter | 3

57

appear to set the Emx2 gradient. FGF8 inhibits and WNT3a increases expression of Emx2, consistent with the latter’s high posteromedial to low-anterior expression gradient (Fukuchi-Shimogori and Grove, 2003; Caronia-Brown et al., 2014). Pax6 is critical to several processes in neocortical development (Yun et al., 2001; Englund et al., 2005; Asami et al., 2011; Manuel et al., 2015; Ypsilanti and Rubenstein, 2016; Elsen et al., 2018). For example, Pax6 plays a major role in dorsalizing the telencephalon (Yun et al., 2001). Judged by characteristic gene expression, the ventral telencephalon is expanded in Pax6 null embryos, and the cortical primordium decreased (Yun et al., 2001). By contrast, further analysis is required to clarify the role of Pax6 in area patterning. Conditional deletion of Pax6 in the dorsal telencephalon utilizing the Emx1-IRES-Cre mouse (Gorski et al., 2002) allowed the assessment of the area map at postnatal ages (Pinon et al., 2008). Gene expression patterns demarcating area boundaries appeared somewhat shifted, but thalamocortical trajectories did not, suggesting the NP Pax6 gradient did not regulate area identity (Pinon et al., 2008). A second study of a highly similar mouse line with conditional deletion of Pax6 revealed a much smaller cortex overall, and a disproportional shrinkage of S1, in particular of the barrel fields (Zembrzycki et al., 2013). Again there was no obvious gross A/P shift of area boundaries. Finally, overexpression of Pax6, using a YAC transgene, had no apparent effect on the area map, the topography of thalamocortical projections, or the A/P position of the S1 barrel fields (Manuel et al., 2007). The original proposal that Emx2 and Pax6 interact to determine size and position of neocortical areas (Bishop et al., 2000, 2002) has been highly influential. Subsequent studies have made the role of Pax6 in these processes less clear. The stage of development at which Pax6 is deleted or augmented in the different experiments could account for the discrepancies, but further work is needed to reconcile the discrepant findings effectively.

3.4.2 Dmrt5/Dmrta2 The Dmrt (doublesex and mab-3-related transcription factor) genes encode evolutionarily conserved TFs much studied for their roles in sexual development, but, more recently, in developmental regulation and patterning of the cerebral cortex (Saulnier et al., 2012; Bellefroid et al., 2013). Consistent with the latter role, in humans, DMRTA2/DMRT5 has been associated with microcephaly and region-specific neocortical pachygyria. Both Dmrt3 and Dmrt5 are expressed in the mouse NP in high-posterior to low-anterior gradients, similar to Emx2 expression. Also similar to Emx2, FGF8 and WNT signaling regulate the gradients of Dmrt5 and Dmrt3 expression in the NP (Caronia-Brown et al., 2014). Conditional deletion of Dmrt5 in the NP using Nestin-Cre or Emx1-IRES-Cre mice (Tronche et al., 1999; Gorski et al., 2002) reduced the size of the cortical hemisphere with a disproportionate reduction of V1 (De Clercq et al., 2018). Overexpression of Dmrt5, using a transgenic approach, enlarged V1 at the expense of more anterior areas, including M1. These observations indicate a role for Dmrt5, similar to that of Emx2, in regulating area size and position along the A/P axis of the NP.

3.4.3 Couptf1/Nr2f1 CoupTFI, now Nr2f1 (nuclear receptor subfamily 2, group F, member 1), is a nuclear orphan receptor that functions as a transcriptional repressor or an activator in different contexts with different cofactors (Tsai and Tsai, 1997). After midgestation in the mouse, Nr2f1 is expressed in a steep high-posterior to low-anterior gradient in the NP, forming a relatively sharp boundary, about two-thirds of the way along the posterior to anterior length of the hemisphere (Jones et al., 2019). The primary sensory areas, V1, A1, and S1, arise from posterior and central progenitor cells that express Nr2f1 strongly, whereas M1 and other areas of FC are generated by anterior progenitors that show little or no Nr2f1 expression. In mice in which Nr2f1 was conditionally deleted in the cerebral cortex using the Emx1-IRES-Cre line, dramatic changes appeared in the area map (Armentano et al., 2007). In postnatal neocortex, sensory areas were tiny and pushed to the back of the cortical hemisphere. The diminutive primary sensory areas received correct thalamic input, and barrels defined a miniature S1. The rest of the neocortex adopted features of motor cortex with respect to gene expression, input from VA/VL motor thalamic nuclei, and projections characteristic of motor cortex. These findings strongly suggested that Nr2f1 balances the territory allotted to motor versus sensory areas and more specifically support a role for Nr2f1 in repressing frontal/motor area identity (Armentano et al., 2007). Further analysis of the mutant focused on the role of Nr2f1 in differentiation of corticospinal motor neurons (CSMNs) (Tomassy et al., 2010), which predominate in layer 5 of M1. CSMN gene expression and projections were compared between M1 and S1 in wild-type mice and, in the conditional Nr2f1 mutant, the regions that would have become M1 and S1. Strong expression of the TF genes Fezf2 and Ctip2 (COUP-TF1 interacting protein 2) is typical of CSMNs. In wildtype M1, expression of Fezf2, and Ctip2 marks out a broad, dense layer 5, contrasting with a much thinner layer 5 in S1. In conditional Nr2f1 null mutants, the NP at the normal coordinates of S1 displayed a massive increase in layer 5 of neurons

58

PART | I Induction and patterning of the CNS and PNS

expressing Fezf2 and Ctip2 and other CSMN gene expression markers. That is, presumptive S1 had been “motorized.” Consistent with this, retrograde axonal labeling revealed that neurons with atypical expression of Fezf2 and Ctip2 altered their axonal trajectories to project to subcerebral targets that included the spinal cord (Tomassy et al., 2010). Focusing next on layer 6, corticothalamic neurons revealed that loss of Nr2f1 subverted neuronal differentiation programs. In wild-type mice, corticothalamic neurons in layer 6 express the TF genes, Tbr1 and Foxp2. Thus, Foxp2/Tbr1 and Ctip2 are strictly expressed in different neuron populations, confined to layers 6 and 5, respectively. In conditional Nr2f1 mutants, the number of hybrid neurons expressing a mix of these genes greatly increased. Meanwhile, genuine CSMNs in layer 5 of the conditional mutant mice developed abnormally, failing to send projections to the spinal cord. In the absence of Nr2f1, therefore, neurons did not differentiate separately along one of the two distinct cell type pathways (Tomassy et al., 2010). Nr2f1 regulates the balance of early- and late-born neurons and the relative sizes of difference layer 5 cell populations (Faedo et al., 2008; Tomassy et al., 2010). These functions also come into play in mediating the apparent area shifts in the conditional Nr2f1 mice. Corticothalamic projection neurons in wild-type mice are born before CSMNs at about E12.5. Loss of Nr2f1 increased the number of E12.5 neurons that strongly express Ctip2 in layer 6, as well as permitting neurons with a hybrid identity, suggesting that Nr2f1 functions to repress CSMN fate during the generation of corticothalamic neurons (Tomassy et al., 2010). From a large-scale perspective, therefore, loss of Nr2f1 caused an enlarged M1 at the expense of sensory areas. At the finer scale of the neuron, abnormally large and widespread populations of neurons differentiated with CSMN or hybrid features. Conditional loss of Nr2f1 in early postnatal neurons also caused a large expansion of M1 in P0 neocortex, with primary sensory areas again pushed back in the hemisphere and shrunken (Alfano et al., 2014). Further, overexpression of Nr2f1 in constitutively Nr2f1 null mice after neurogenesis rescued at least some of the deficits in the mutant cortex, suggesting Nr2f1 is needed in postmitotic neurons to regulate sensory versus motor neuronal differentiation. This begs the question of the role of Nr2f1 expression in progenitor neurons. The answer awaits a method of gene deletion in progenitor cells but not their postmitotic progeny. One possible outcome is that proper repression of the CSMN identity requires postmitotic cell expression of Nr2f1, but that Nr2f1-expressing progenitor cells direct this expression in their daughter neurons. The primary finding is that Nr2f1 is necessary and sufficient to repress the formation of CSMNs, a conclusion that neatly links area, layer, and cell type development (Faedo et al., 2008).

3.4.4 Sp8 The zinc-finger TF Sp8 is a member of the Sp1 (specificity protein 1) family, orthologous to Drosophila buttonhead, and is involved in varied embryonic processes, including limb and craniofacial development (Kawakami et al., 2004; Kasberg et al., 2013). FGF8 and WNT3a upregulate Sp8 expression in mouse (Caronia-Brown et al., 2014; Dunty et al., 2014), and, consistent with this, Sp8 expression appears in the anteromedial NP. Mice with conditional deletion of Sp8 display defects in both the ventral and dorsal telencephalon and show several deficits in cortical development (Zembrzycki et al., 2007). With respect to the NP area map, expression gradients of the patterning genes Emx2 and Pax6, as well as the regional domains of other genes, indicate shrinkage of the anterior NP, with enlargement of the posterior NP (Zembrzycki et al., 2007; Borello et al., 2014). Linking Sp8 with other patterning TFs, Sp8 may directly interact with Emx2 in patterning the area map (Zembrzycki et al., 2007); and Sp8 and Nr2f1 may reciprocally regulate patterning (Borello et al., 2013) (see below). In other contexts, Sp8 induces Fgf8 expression (Kawakami et al., 2004), which would complicate the interpretation of Sp8 actions on A/P patterning of the NP. In the Sp8 null mutant mouse telencephalon, however, expression of Fgf8 appeared normal (Zembrzycki et al., 2007). Similarly, when Sp8 was expressed throughout the mouse telencephalon, Fgf8 expression was not abnormally widely induced (Borello et al., 2013). Nr2f1 expression, however, was repressed, and further assessment showed that Sp8 and Nr2f1 reciprocally inhibit one another. Finally, excess Sp8 enhanced the expression of genes downstream of FGF8, and the MAPK signaling cascade, which mediates FGF8 activity. Together these findings indicate that in the context of NP patterning, Sp8 is a downstream effector of FGF8, promoting anterior NP identity and mediating the high-posterior to low-anterior expression gradients of both Emx2 and Nr2f1 (Zembrzycki et al., 2007; Borello et al., 2013).

3.4.5 Pbx Pbx genes are atypical homeodomain-containing TFs. Of four Pbx genes in vertebrates, Pbx1 and Pbx2 are expressed in the mouse cortical primordium. To determine the function of Pbx genes in cortical development, two lines of conditional Pbx1

Area patterning of the mammalian neocortex Chapter | 3

59

mutants were generated in which Pbx1 function was lost either at the progenitor cell stage or later in postmitotic neurons. To render the phenotype more dramatic, mice also lacked one copy of Pbx2. In postnatal neocortex, the Pbx mutants showed posterior to anterior regional shifts and ventral expansion of the dorsomedial cingulate and retrosplenial areas (Golonzhka et al., 2015). A role for Pbx1 in promoting frontal cortical areas was seen when Pbx1 was deleted either in progenitor cells or postmitotic neurons, indicating that newborn neurons retain a choice of area fate. Dorsoventral patterning, by contrast, required Pbx activity exclusively in progenitor cells. Importantly, unlike the TFs described previously, the patterning effects of Pbx1/2 were at least partially indirect, caused by altered gradients of other TFs, including Emx2, which were further shown to be likely target genes of Pbx1/2 in Chip-Seq experiments (Golonzhka et al., 2015). These findings therefore place Pbx1 upstream of other NP patterning genes.

3.5 Do neocortical areas arise from dedicated progenitor cell pools? The mechanisms of embryonic patterning have been most thoroughly studied in the vertebrate ventral spinal cord. During patterning of the ventral cord, signaling molecules regulate expression of particular TF genes in different regions in the primordium of the ventral cord. At first, gene expression boundaries are relatively indistinct, but as development progresses, neighboring pairs of TFs repress one another’s expression, creating progenitor domains with increasingly sharp boundaries (Briscoe et al., 1999, 2000). Each progenitor domain, identified by its unique mix of TFs, gives rise to a specific neuronal cell type of the ventral spinal cord (Briscoe et al., 1999, 2000).

3.5.1 Transcription factors known to pattern the NP appear in gradients, not domains So far, the graded expression of patterning genes for the NP has been emphasized. If NP patterning resembled that in the ventral spinal cord, patterning genes with opposing gradients would interact to form sharply bounded domains, which would give rise to different areas. Such boundaries, however, have not been identified either among the progenitors cells in the VZ or in the subventricular zone (SVZ), where intermediate progenitor (IP) cells, generated from the VZ, are located. Gene expression in the SVZ is graded just as in the VZ (Elsen et al., 2013). A caveat here is that the overlap of expression of known patterning genes has not been as carefully examined as might be. At E9.5, for example, the expression domains of Sp8 and Nr2f1 along the A/P axis look distinctly complementary with minimal overlap (Borello et al., 2013). Directed exploration for genes with expression confined to distinct progenitor domains has not been helpful so far (Donoghue and Rakic, 1999; Sansom et al., 2005; Kudo et al., 2007). These searches have revealed more genes expressed in gradients in the germinal layers of the NP. Nonetheless, to date, relatively few TFs have been identified with a clear role in patterning the area map. Given that the discovery of Dmrt5 as an NP patterning gene was quite recent, however, additional patterning genes are highly likely to be discovered.

3.5.2 Mapping the cortical primordium with forebrain enhancers In contrast to the graded expression of patterning TFs, strikingly bounded domains of enhancer regulatory activity have been identified in the mouse NP in a series of investigations. Putative enhancers were identified in noncoding sequences of the human genome, based on the ultraconservation of these sequences across human, mouse, and rat (Pennacchio et al., 2006). Candidates were screened for spatially restricted enhancer activity by linking the human DNA fragments to a minimal promoter fused to a lacZ reporter gene and generating transgenic mice. Tissue-specific reporter gene expression was evaluated in mouse embryos at E11.5, when whole embryo staining and visualization is straightforward and when neurogenesis is just beginning in the NP. Enhancer activity was spatially restricted, with several putative enhancers active in the mouse telencephalon (Pennacchio et al., 2006). Subsequent work provided an “enhancer atlas” of the developing mouse telencephalon showing reporter mouse lacZ expression for many reproducible forebrain enhancers. The activity patterns of several enhancers marked out clear domains in the VZ of the medial, dorsal, and lateral cortical primordium at E11.5 and subsequently in the cortical plate (Visel et al., 2013; Pattabiraman et al., 2014) (Fig. 3.6). Fate-mapping enhancer activity domains demonstrated that precisely positioned domains, identified at E11.5, can give rise to similarly located subdivisions of the hippocampus and neocortex in late embryonic and postnatal mice. Further assays indicated that Nr2f1 and Pax6, two of the patterning TFs described above, bound to particular sets of the dorsal telencephalic enhancers. A proposal, based on these and further findings (Pattabiraman et al., 2014), is that the graded patterning TFs interact to regulate the activity of enhancer elements. Subsequently the enhancers regulate expression of further genes, presumably in cortical protodomains, which ultimately give rise to specific areas or regions of the cerebral cortex.

60

PART | I Induction and patterning of the CNS and PNS

FIGURE 3.6 Mapping the cortical primordium with forebrain enhancers. (A) Schematic showing structures visible in sections (BeR). DP, dorsal pallium; LGE, lateral ganglionic eminence; LP, lateral pallium; MP, medial pallium; Se, septum; VP, ventral pallium. The MP is roughly equivalent to the hippocampal primordium (HP); DP and LP to the neocortical primordium; and VP to the source of olfactory cortex and specific extracortical structures. (BeR) Coronal sections through the anterior forebrain of transgenic mouse embryos at E11.5, stained for lacZ. Blue staining reflects the activity of select forebrain enhancers that reproducibly label particular regions of the developing cortical primordium and are numbered in each panel. Enhancers are assembled in order from medialedorsalelateraleventral.

3.6 The influence of thalamic innervation 3.6.1 Guidance of thalamocortical axons and area formation In mice, connections between thalamus and neocortex develop between late embryogenesis and early postnatal life. A major step forward in understanding thalamocortical pathway development was the recognition of intermediate targets for thalamocortical axons and the guidance cues at work at each stage. Axons are guided along the pathway out of the thalamus, through the ventral telencephalon, and into and through the incipient internal capsule (IC). In parallel, signals were found that regulate the topography of thalamocortical axons en route. The cues along the trajectory included chemorepellent and chemoattractant axon guidance molecules and guidepost cells that form an essential permissive “corridor” through the otherwise nonpermissive basal telencephalon (Garel and Rubenstein, 2004; Vanderhaeghen and Polleux, 2004; Lopez-Bendito et al., 2006; Powell et al., 2008; Deck et al., 2013; Anton-Bolanos et al., 2018). Thalamocortical axons are ordered such that axons from different thalamic nuclei remain segregated and exit the IC at correct positions relative to the A/P and M/L axes of the NP. Indeed, the orderliness of the thalamic axons headed for different parts of the neocortex (Vanderhaeghen and Polleux, 2004; Garel and Lopez-Bendito, 2014; Anton-Bolanos et al., 2018) suggests renewed support for the “protocortex” model and gives an explanation of how it might work in practice. That is, presorted thalamocortical axons arrive beneath the neocortex, interact synaptically with the SP, and require perhaps one more permissive cue to enter their target. Once there, thalamocortical synaptic activity could stamp an identity on each area. Evidence already described in this chapter indicates, however, that the sequence of thalamocortical axon molecular guidance cues does not stop when the axons arrive beneath the cortex. Grafting experiments indicated the presence of areaspecific attractant signals (Frappe et al., 1999; Gaillard and Roger, 2000). The effects on thalamocortical trajectories of shifting the positional identity of the SP and the NP in or out of register by electroporating anterior Fgf8 at different embryonic ages further indicate the presence of instructive, region-specific molecular cues in both NP and the SP (Shimogori and Grove, 2005).

Area patterning of the mammalian neocortex Chapter | 3

61

A seeming superfluity of thalamocortical guidance cues is consistent with the complexity of the system. Thalamocortical axons do not generally exit the IC directly below the area to be innervated. Instead, the axons travel through the SP and only enter the NP when appropriate cues are available. Further, projections from thalamus to cortex do not come in a one-to-one ratio. Individual areas can receive input from several thalamic nuclei, and a single thalamic nucleus may innervate more than one area (Nauta and Feirtag, 1986). Additionally, areas can be innervated by thalamic nuclei that do not correspond in a simple way to the position of the area. As examples, (1) the mediodorsal nucleus, in the center of the thalamus, projects to prefrontal cortex, which is far anterior; (2) the limbic cingulate and retrosplenial areas are medial and stretch almost the whole A/P length of the mouse neocortex, yet receive input from anterior and lateral thalamic nuclei as well as the medial intralaminar nuclei and the mediodorsal nucleus (Nauta and Feirtag, 1986). A complete map of the 3D thalamus onto a (comparatively) 2D cerebral cortex has not been generated.

3.6.2 Thalamic innervation determines the function of a cortical area Thalamic innervation is central to area identity, given that the function of an area depends on the type of information it receives from the thalamus, as well as from other areas of neocortex. Manipulating the kind of input that reaches an area through the thalamus, moreover, can change not only its function but also induce anatomical features that are characteristic of a different area. In a classic study of the developing ferret, retinal inputs were forced to innervate the auditory nucleus of the thalamus, the MGN. To route visual projections to auditory cortex, the retina was deprived of its major targets, the superior colliculus and dLGN. Meanwhile, normal auditory input to the MGN was interrupted surgically. Retinal afferents populated the deafferented MGN so that the MGN now relayed visual information to A1. A systematic representation of visual space was created in rewired A1, the direction and orientation selectivity of visual stimuli was similar to that in V1 neurons, and rewired A1 contained simple and complex receptive fields in the same proportion as V1. Impressively, even intraarea circuitry characteristic of V1 developed in visual A1. The specific sensory modality of thalamic input therefore controlled not only the function of a developing neocortical area but also, perhaps more surprisingly, specific cytoarchitectural features (Fig. 3.7). Nonetheless, A1 with visual input did not make abnormal connections with visual areas and instead retained its connections with other auditory areas. Even visually rewired A1 maintained some of its original area

FIGURE 3.7 The type of thalamic information conveyed controls area identity. (A) Top brain schematic shows normal projections from the retina and auditory periphery to the thalamus and from thalamus to cortex. Lower schematic illustrates a brain in which auditory inferior colliculus projections to the MGN have been removed. The MGN now receives visual input from the retina and relays it to a “visually responsive” A1. (B) Orientation maps in normal V1 and rewired A1. Different preferred orientations are shown in different colors. The pinwheel, around which cells preferring different orientations are represented, is a characteristic feature of the orientation map. Notably both V1 and rewired A1 contain pinwheels (see dotted circles). (C) Horizontal connections in upper cortical layers are also typical of V1, but not A1, where local connections are more proximal. Horizontal connections link neurons with similar orientation selectivity. In rewired A1, horizontal connections resemble those in V1, not the local connections in A1. Scale bars ¼ 0.5 mm. Adapted slightly from Sur, M., Rubenstein, J.L., 2005. Patterning and plasticity of the cerebral cortex. Science 310, 805e810.

62

PART | I Induction and patterning of the CNS and PNS

identity. These observations emphasize the importance of the modality and pattern of thalamic input on the development of a neocortical area and show that these factors interact with the intrinsic developmental program (Sur et al., 1988, 1990; Roe et al., 1990, 1992; Pallas and Sur, 1993; Sur, 1993; Sur and Rubenstein, 2005). A more recent study underscores the importance of modality-specific inputs. S1 and S2 layer 4 cells normally receive mutually exclusive input from the VPM or the posterior nucleus of the thalamus (PO). The VPM relays tactile information from the whiskers, whereas PO carries nociceptive input. After genetic ablation of the VPM in neonatal mice, axons from the PO innervated layer 4 cells in S2 as usual and also invaded layer 4 of S1, normally the target of VPM. In response, S1 layer 4 neurons downregulated genes they normally express and upregulated expression of characteristic layer 4 S2 genes. Functionally, rewired S1 layer 4 neurons responded to tactile input, though more weakly than usual, and to nociceptive input. Innervation by PO thus caused a convergence of normally segregated sensory modalities and gave S1 layer 4 cells a functionally hybrid identity. These observations indicate a strong influence of input modality on neocortical neurons but also reveal an underlying, earlier specification, though not an unchanging commitment, of area identity.

3.6.3 Effects of thalamocortical afferents on area size and cortical progenitor cells Classic studies in monkeys reported that removing the eyes reduced the dLGN and caused V1 (area 17) to shrink. The size of V2 (area 18) increased as axons grew into the abnormally large available space (Dehay et al., 1991, 1996). Genetically increasing or decreasing the size of the dLGN causes a corresponding change in the size of V1 in mice (Vue et al., 2013). In both monkey and mouse, areas normally differ in progenitor cell cycle kinetics, a feature of early area differences (Dehay and Kennedy, 2007). Signals from thalamic axons are proposed to underlie these differences, regulating the output of neurons from both radial glial cell progenitors and the IP cells that amplify neuron generation (Dehay and Kennedy, 2007). In mouse organotypic culture, thalamocortical axons regulate progenitor cell cycle dynamics via an apparently diffusible signal (Dehay et al., 2001), now identified as signaling between ephrin-A5 and EphA4 (Gerstmann et al., 2015). Thalamic signals regulating progenitor cell behavior, however, may be more significant in monkeys, in which thalamocortical axons innervate the SP much earlier than in mice, implying an earlier proximity of progenitor cells to signals from axons in the SP (Alzu’bi et al., 2019). Positional identity conferred by morphogen signals in or close to the NP, however, is presumed to occur still earlier. A reasonable conclusion is that the position, orientation, and size of an area are specified in early embryogenesis, but area size and cell proliferation can also be regulated by later developmental events, both normal and abnormal.

3.6.4 Thalamic dependence of an area-specific feature The position and orientation of S1 within the neocortex is determined by early embryonic mechanisms described above. The S1 barrel fields, however, are an example of an area-specific feature whose development is virtually wholly controlled by thalamic afferents. Rodents display neocortical “barrels” in S1, each of which corresponds to one of the large facial whiskers or the smaller whiskers on the snout. The layout of the barrels matches precisely the pattern of peripheral whiskers, making the barrel fields the most visually prominent element of the rodent area map. A barrel “hollow” contains clustered thalamocortical afferents relaying, via the VPM, input from one whisker. Layer 4 S1 neurons collect around each hollow to form the barrel “walls” and direct their dendrites inward to receive thalamic input. Formation of the barrels is activity-dependent. When glutamatergic activity is silenced in the thalamocortical pathway, barrels do not appear, and layer 4 neurons fail to develop a normal morphology (Li et al., 2013). On a finer scale too, a molecular mechanism that directs neuronal dendrites into the barrel hollow is activity-dependent (Matsui et al., 2013).

3.6.5 Two mechanisms united If intrinsic NP mechanisms do not generate areas, they might at least create larger regions that are refined into separate areas by thalamic innervation. In a study that tested this alternative, the pathway from the dLGN to V1 was progressively eliminated using a genetic approach (Chou et al., 2013). In wild-type mice, dLGN axons arrive by E16.5 in the SP, where they remained for some days before invading the CP of presumptive V1 just after birth. In the mutant mouse, the dLGN steadily decreased in size and was virtually absent by P7. Moreover, although dLGN axons reached the SP before the thalamic nucleus disappeared, few LGN axons ever entered the neocortex. The result was not a shrunken V1 or the absence of the region in which V1 normally appears. Rather, the entire region containing the presumptive V1 and surrounding “higher-order” visual areas was altered. The whole territory adopted gene expression levels and patterns that normally define the secondary and tertiary visual areas that receive input from V1. Further, the changes in gene expression were

Area patterning of the mammalian neocortex Chapter | 3

63

consistent in each neocortical layer. These findings suggest that mechanisms intrinsic to the NP determined a broad occipital “visual field” that is more or less homogeneous. Only when dLGN axons enter the neocortex is the region carved into V1 and the surrounding higher-order fields (Chou et al., 2013). An open question, however, is how dLGN axons know where to enter.

3.7 Spontaneous activity and neocortical patterning Spontaneous retinal activity coordinates activity-based patterning in the visual system (Ackman et al., 2012). Retinal waves mediate the development of eye-specific layers in the dLGN and ocular dominance columns in V1. Retinal waves include the entire retinal visual field, require cholinergic neurotransmission, and exhibit spatiotemporal correlations between the two hemispheres (Ackman et al., 2012). In contrast to the visual system, spontaneous activity in the somatosensory system and its possible roles in development have barely been touched upon. Two-photon calcium imaging of barrel cortex in neonatal mice has disclosed layer 4 neurons in the same barrel firing synchronously and revealed a barrel map of spontaneous activity (Mizuno et al., 2018). The map reflects activity specifically in thalamocortical axon terminals and is not coordinated with whisking behavior or dependent on input from the trigeminal axon terminals in the whisker pad. Whether the spontaneous wave pattern originates from the trigeminal nucleus or the somatosensory thalamus is unknown (Mizuno et al., 2018). A second study reports that, in the mouse, calcium waves appear to move across the thalamus from one nucleus to another, very different from the retinal waves that cover and propagate retinotopic position (Moreno-Juan et al., 2017). A new study, as of this writing, suggests that spontaneous calcium wave activity in VPM axons travels to nascent S1 and is responsible for forming columnar territories in S1 and ultimately the S1 barrel fields (Anton-Bolanos et al., 2018). When calcium waves were disrupted by extreme hyperpolarization of thalamic neurons, the barrel fields failed to develop, even when the mice were allowed to grow up and use their whiskers (Anton-Bolanos et al., 2019). These observations raise the possibility that spontaneous activity in the embryo prepatterns the barrel fields, although barrels are not clearly visible for a few days after birth. Nonetheless, blocking glutamatergic transmission also prevents barrel formation (Li et al., 2013), so either both types of activity are needed sequentially for barrel field development or experience-driven activity is required to maintain the barrel pattern. Spontaneous activity in the developing somatosensory system clearly deserves closer study.

3.8 Conservation of patterning mechanisms among different mammalian species The protomap model was based on observations of primate neocortex (Rakic, 1988), a reminder that the ultimate goal is to discover mechanisms that pattern cerebral cortex in all mammals, including humans. A question is whether the processes that pattern the small lissencephalic mouse neocortex also shape the larger gyrencephalic brains of carnivores and primates. One reason for doubt is that patterning molecules, or morphogens, such as FGF8 and WNT3a, have limited dispersion ranges of a few hundred microns (Scholpp and Brand, 2004; Farin et al., 2016; Parchure et al., 2018). Larger, more complex cortices presumably derive from larger cortical primordia whose size could prevent appropriate morphogen dispersion. On the other hand, different basic mechanisms for neocortical patterning in different species fail to fit with a central finding in developmental biology, namely the conservation of patterning mechanisms across vertebrates (Alberts et al., 2015). In addition, as noted, consistent with conserved patterning processes, primary sensory and motor areas adopt similar relative positions in the neocortical maps of a wide range of mammals (Krubitzer and Stolzenberg, 2014; Nauta and Feirtag, 1986). A way around the problem of patterning a larger brain would be to hold back growth at an early stage so that morphogen gradients can be effective in specifying brain components. This possibility has been tested by examining growth of the NP in the gyrencephalic carnivore, the ferret, Mustela putorius furo, whose mature brain is several times the volume of the mouse brain. The focus was the stage of embryogenesis at which FGF8 initiates NP patterning in the mouse, about E9.5 (Toyoda et al., 2010). Although the NP of the ferret is ultimately much larger than that of the mouse, at the critical stage of telencephalic development, the A/P length of the NP is the same in mouse and ferret (Jones et al., 2019). Thus, FGF8 could, in principle, disperse from rostral to caudal in the ferret NP and determine the A/P axis of the area map. The distance Wnt3a disperses along the M/L axis in the mouse CP is still unknown, obviating a quantitative approach to whether Wnt signaling regulates this axis in ferret as in mouse. However, a conserved function of Wnt3a in hippocampal development has been supported by the juxtaposition of the cortical hem to the presumptive hippocampal primordium in the ferret, as well as in humans (Abu-Khalil et al., 2004; Jones et al., 2019).

64

PART | I Induction and patterning of the CNS and PNS

Human embryo brains have not yet been shown to have an RPC, but this seems likely, given conservation of the RPC signaling center in mice, chick, and zebrafish (Shimamura and Rubenstein, 1997; Shanmugalingam et al., 2000; Crossley et al., 2001; Ohkubo et al., 2002). Consistent with an FGF8 gradient establishing the A/P axis in human neocortex, increasing or reducing FGF8 in neocortical cells derived from human pluripotent stem cells induced gene expression appropriate, respectively, to the anterior or posterior NP (Imaizumi et al., 2018). Further suggesting FGF8 plays a role in neocortical area patterning, abnormalities in human cortex are linked to dysregulation of FGF8 or FGF receptor 3 signaling (Frank et al., 2002; Hevner, 2005; Dubourg et al., 2016). Human embryo specimens in the Carnegie and Kyoto collections (Human Developmental Anatomy Center, Washington, DC) were assigned to Carnegie stages, measured, and imaged. Images, with scale bars, of first trimester embryos are available online and in e-book form. Initial comparisons of images of E9.5 mouse and CS13 human brains suggest the telencephalon and NP are roughly the same size (Jones et al., 2019). A more systematic study is warranted. Yet, mammals with mature cortices of widely varying size may be patterned by common mechanisms if the early cortical primordium is kept, transiently, at a common size.

3.9 Conclusions Studies in the 1990s indicated that neocortex is divided into regions, or even specific areas, by mechanisms intrinsic to the cortical primordium. These divisions precede or do not require thalamocortical innervation. Several intrinsic molecular signals have now been identified, including morphogens, such as FGF8 and Wnt3a, and a panel of downstream TFs, including Emx2, Pax6, Nr2f1, Dmrt5, and Sp8. FGF8 provides fine-grained positional values to the nascent neocortical area map. Downstream TFs regulate the size and position of areas or regulate cell type differentiation across areas. Meanwhile, great progress has been made in understanding how thalamocortical axons reach the NP and how axons are organized en route to exit the IC at the correct A/P and M/L coordinates. Evidence from both old and new studies, however, indicate that these axons are provided with final molecular guidance cues from the SP and from neurons within the perinatal neocortex. Thus, in this case, the cortex guides development of thalamic innervation rather than vice versa. Thalamic afferents are nonetheless required for the differentiation of area-specific features, such as the S1 barrel fields, and more critically, thalamic innervation is key to defining the function of an area. It remains unclear what the intrinsic map of the neocortex looks like in the embryonic NP. To date, no genes expressed in clear domains in the VZ or SVZ have been identified, marking “protoareas.” Nonetheless, the presence of axon guidance cues in the early NP implies that presumptive areas are at least roughly drawn out in the NP. Boundaries could be refined later by thalamic innervation. A more defined map is suggested by the activity of select forebrain enhancers in sharply bounded domains in the early NP. The genes these enhancers regulate may also be expressed in distinct domains. The debate continues regarding the degree to which the neocortical area map is generated by intrinsic mechanisms versus extrinsic influences from the thalamus. Many unknowns remain, one of which has only recently been explored, namely the nature and role of spontaneous activity in patterning the neocortex. This should draw great interest in the near future. Another imperative question is the meaning of the bounded enhancer activity in the NP. Might protoareas exist after all, and what would be the best way to reveal them? Whatever further discoveries await, for example, the function of genes downstream of the TFs already identified or the molecular mechanisms by which the thalamus contributes to refining the area map, the final conclusion is likely to be the same as that of today, namely that the neocortical area map is generated by an interplay of intrinsic and extrinsic mechanisms. The advance is that we will know how it is done.

References Abu-Khalil, A., Fu, L., Grove, E.A., Zecevic, N., Geschwind, D.H., 2004. Wnt genes define distinct boundaries in the developing human brain: implications for human forebrain patterning. J. Comp. Neurol. 474, 276e288. Ackman, J.B., Burbridge, T.J., Crair, M.C., 2012. Retinal waves coordinate patterned activity throughout the developing visual system. Nature 490, 219e225. Alberts, B., Johnson, A., Lewis, J., Morgan, D., Raff, M., Roberts, K., Walter, P., 2015. Molecular Biology of the Cell, sixth ed. Garland Science. Alfano, C., Magrinelli, E., Harb, K., Hevner, R.F., Studer, M., 2014. Postmitotic control of sensory area specification during neocortical development. Nat. Commun. 5, 5632. Allendoerfer, K.L., Shatz, C.J., 1994. The subplate, a transient neocortical structure: its role in the development of connections between thalamus and cortex. Annu. Rev. Neurosci. 17, 185e218. Alzu’bi, A., Homman-Ludiye, J., Bourne, J.A., Clowry, G.J., 2019. Thalamocortical afferents innervate the cortical subplate much earlier in development in primate than in rodent. Cerebr. Cortex 29, 1706e1718. Anton-Bolanos, N., Espinosa, A., Lopez-Bendito, G., 2018. Developmental interactions between thalamus and cortex: a true love reciprocal story. Curr. Opin. Neurobiol. 52, 33e41.

Area patterning of the mammalian neocortex Chapter | 3

65

Anton-Bolanos, N., Sempere-Ferrandez, A., Guillamon-Vivancos, T., Martini, F.J., Perez-Saiz, L., Gezelius, H., Filipchuk, A., Valdeolmillos, M., LopezBendito, G., 2019. Prenatal activity from thalamic neurons governs the emergence of functional cortical maps in mice. Science 364, 987e990. Armentano, M., Chou, S.J., Tomassy, G.S., Leingartner, A., O’Leary, D.D., Studer, M., 2007. COUP-TFI regulates the balance of cortical patterning between frontal/motor and sensory areas. Nat. Neurosci. 10, 1277e1286. Asami, M., Pilz, G.A., Ninkovic, J., Godinho, L., Schroeder, T., Huttner, W.B., Gotz, M., 2011. The role of Pax6 in regulating the orientation and mode of cell division of progenitors in the mouse cerebral cortex. Development 138, 5067e5078. Assimacopoulos, S., Kao, T., Issa, N.P., Grove, E.A., 2012. Fibroblast growth factor 8 organizes the neocortical area map and regulates sensory map topography. J. Neurosci. 32, 7191e7201. Bachler, M., Neubuser, A., 2001. Expression of members of the Fgf family and their receptors during midfacial development. Mech. Dev. 100, 313e316. Barbe, M.F., Levitt, P., 1991. The early commitment of fetal neurons to the limbic cortex. J. Neurosci. 11, 519e533. Bellefroid, E.J., Leclere, L., Saulnier, A., Keruzore, M., Sirakov, M., Vervoort, M., De Clercq, S., 2013. Expanding roles for the evolutionarily conserved Dmrt sex transcriptional regulators during embryogenesis. Cell. Mol. Life Sci. Bishop, K.M., Goudreau, G., O’Leary, D.D., 2000. Regulation of area identity in the mammalian neocortex by Emx2 and Pax6. Science 288, 344e349. Bishop, K.M., Rubenstein, J.L., O’Leary, D.D., 2002. Distinct actions of Emx1, Emx2, and Pax6 in regulating the specification of areas in the developing neocortex. J. Neurosci. 22, 7627e7638. Borello, U., Cobos, I., Long, J.E., McWhirter, J.R., Murre, C., Rubenstein, J.L., 2008. FGF15 promotes neurogenesis and opposes FGF8 function during neocortical development. Neural Dev. 3, 17. Borello, U., Madhavan, M., Vilinsky, I., Faedo, A., Pierani, A., Rubenstein, J., Campbell, K., 2014. Sp8 and COUP-TF1 reciprocally regulate patterning and Fgf signaling in cortical progenitors. Cerebr. Cortex 24, 1409e1421. Briscoe, J., Ericson, J., 2001. Specification of neuronal fates in the ventral neural tube. Curr. Opin. Neurobiol. 11, 43e49. Briscoe, J., Pierani, A., Jessell, T.M., Ericson, J., 2000. A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101, 435e445. Briscoe, J., Sussel, L., Serup, P., Hartigan-O’Connor, D., Jessell, T.M., Rubenstein, J.L., Ericson, J., 1999. Homeobox gene Nkx2.2 and specification of neuronal identity by graded Sonic hedgehog signalling. Nature 398, 622e627. Caronia-Brown, G., Yoshida, M., Gulden, F., Assimacopoulos, S., Grove, E.A., 2014. The cortical hem regulates the size and patterning of neocortex. Development 141, 2855e2865. Catalano, S.M., Shatz, C.J., 1998. Activity-dependent cortical target selection by thalamic axons. Science 281, 559e562. Caviness Jr., V.S., Frost, D.O., 1983. Thalamocortical projections in the reeler mutant mouse. J. Comp. Neurol. 219, 182e202. Cholfin, J.A., Rubenstein, J.L., 2007. Patterning of frontal cortex subdivisions by Fgf17. Proc. Natl. Acad. Sci. U.S.A. 104, 7652e7657. Cholfin, J.A., Rubenstein, J.L., 2008. Frontal cortex subdivision patterning is coordinately regulated by Fgf8, Fgf17, and Emx2. J. Comp. Neurol. 509, 144e155. Chou, S.J., Babot, Z., Leingartner, A., Studer, M., Nakagawa, Y., O’Leary, D.D., 2013. Geniculocortical input drives genetic distinctions between primary and higher-order visual areas. Science 340, 1239e1242. Cohen-Tannoudji, M., Babinet, C., Wassef, M., 1994. Early determination of a mouse somatosensory cortex marker. Nature 368, 460e463. Crossley, P.H., Martin, G.R., 1995. The mouse Fgf 8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439e451. Crossley, P.H., Martinez, S., Martin, G.R., 1996. Midbrain development induced by FGF8 in the chick embryo. Nature 380, 66e68. Crossley, P.H., Martinez, S., Ohkubo, Y., Rubenstein, J.L., 2001. Coordinate expression of Fgf8, Otx2, Bmp4, and Shh in the rostral prosencephalon during development of the telencephalic and optic vesicles. Neuroscience 108, 183e206. De Clercq, S., Keruzore, M., Desmaris, E., Pollart, C., Assimacopoulos, S., Preillon, J., Ascenzo, S., Matson, C.K., Lee, M., Nan, X., Li, M., Nakagawa, Y., Hochepied, T., Zarkower, D., Grove, E.A., Bellefroid, E.J., 2018. DMRT5 together with DMRT3 directly controls hippocampus development and neocortical area map formation. Cerebr. Cortex 28, 493e509. Deck, M., Lokmane, L., Chauvet, S., Mailhes, C., Keita, M., Niquille, M., Yoshida, M., Yoshida, Y., Lebrand, C., Mann, F., Grove, E.A., Garel, S., 2013. Pathfinding of corticothalamic axons relies on a rendezvous with thalamic projections. Neuron 77, 472e484. Dehay, C., Kennedy, H., 2007. Cell-cycle control and cortical development. Nat. Rev. Neurosci. 8, 438e450. Dehay, C., Savatier, P., Cortay, V., Kennedy, H., 2001. Cell-cycle kinetics of neocortical precursors are influenced by embryonic thalamic axons. J. Neurosci. 21, 201e214. Dehay, C., Horsburgh, G., Berland, M., Killackey, H., Kennedy, H., 1991. The effects of bilateral enucleation in the primate fetus on the parcellation of visual cortex. Dev. Brain Res. 62, 137e141. Dehay, C., Giroud, P., Berland, M., Killackey, H., Kennedy, H., 1996. Contribution of thalamic input to the specification of cytoarchitectonic cortical fields in the primate: effects of bilateral enucleation in the fetal monkey on the boundaries, dimensions, and gyrification of striate and extrastriate cortex. J. Comp. Neurol. 367, 70e89. Donoghue, M.J., Rakic, P., 1999. Molecular evidence for the early specification of presumptive functional domains in the embryonic primate cerebral cortex. J. Neurosci. 19, 5967e5979. Dorsky, R.I., Moon, R.T., Raible, D.W., 1998. Control of neural crest cell fate by the Wnt signalling pathway. Nature 396, 370e373. Dubourg, C., et al., 2016. Mutational spectrum in holoprosencephaly shows that FGF is a new major signaling pathway. Hum. Mutat. 37, 1329e1339. Dunty Jr., W.C., Kennedy, M.W., Chalamalasetty, R.B., Campbell, K., Yamaguchi, T.P., 2014. Transcriptional profiling of Wnt3a mutants identifies Sp transcription factors as essential effectors of the Wnt/beta-catenin pathway in neuromesodermal stem cells. PLoS One 9, e87018.

66

PART | I Induction and patterning of the CNS and PNS

Edgar, B.A., Lehner, C.F., 1996. Developmental control of cell cycle regulators: a fly’s perspective. Science 274, 1646e1652. Elsen, G.E., Hodge, R.D., Bedogni, F., Daza, R.A., Nelson, B.R., Shiba, N., Reiner, S.L., Hevner, R.F., 2013. The protomap is propagated to cortical plate neurons through an Eomes-dependent intermediate map. Proc. Natl. Acad. Sci. U.S.A. 110, 4081e4086. Elsen, G.E., Bedogni, F., Hodge, R.D., Bammler, T.K., MacDonald, J.W., Lindtner, S., Rubenstein, J.L.R., Hevner, R.F., 2018. The epigenetic factor landscape of developing neocortex is regulated by transcription factors Pax6/ Tbr2/ Tbr1. Front. Neurosci. 12, 571. Englund, C., Fink, A., Lau, C., Pham, D., Daza, R.A., Bulfone, A., Kowalczyk, T., Hevner, R.F., 2005. Pax6, Tbr2, and Tbr1 are expressed sequentially by radial glia, intermediate progenitor cells, and postmitotic neurons in developing neocortex. J. Neurosci. 25, 247e251. Faedo, A., Borello, U., Rubenstein, J.L., 2010. Repression of Fgf signaling by sprouty1e2 regulates cortical patterning in two distinct regions and times. J. Neurosci. 30, 4015e4023. Faedo, A., Tomassy, G.S., Ruan, Y., Teichmann, H., Krauss, S., Pleasure, S.J., Tsai, S.Y., Tsai, M.J., Studer, M., Rubenstein, J.L., 2008. COUP-TFI coordinates cortical patterning, neurogenesis, and laminar fate and modulates MAPK/ERK, AKT, and beta-catenin signaling. Cerebr. Cortex 18, 2117e2131. Farin, H.F., Jordens, I., Mosa, M.H., Basak, O., Korving, J., Tauriello, D.V., de Punder, K., Angers, S., Peters, P.J., Maurice, M.M., Clevers, H., 2016. Visualization of a short-range Wnt gradient in the intestinal stem-cell niche. Nature 530, 340e343. Finney, E.M., Stone, J.R., Shatz, C.J., 1998. Major glutamatergic projection from subplate into visual cortex during development. J. Comp. Neurol. 398, 105e118. Frank, D.U., Fotheringham, L.K., Brewer, J.A., Muglia, L.J., Tristani-Firouzi, M., Capecchi, M.R., Moon, A.M., 2002. An Fgf8 mouse mutant phenocopies human 22q11 deletion syndrome. Development 129, 4591e4603. Frappe, I., Roger, M., Gaillard, A., 1999. Transplants of fetal frontal cortex grafted into the occipital cortex of newborn rats receive a substantial thalamic input from nuclei normally projecting to the frontal cortex. Neuroscience 89, 409e421. Friauf, E., McConnell, S.K., Shatz, C.J., 1990. Functional synaptic circuits in the subplate during fetal and early postnatal development of cat visual cortex. J. Neurosci. 10, 2601e2613. Fukuchi-Shimogori, T., Grove, E.A., 2001. Neocortex patterning by the secreted signaling molecule FGF8. Science 294, 1071e1074. Fukuchi-Shimogori, T., Grove, E.A., 2003. Emx2 patterns the neocortex by regulating FGF positional signaling. Nat. Neurosci. 6, 825e831. Furuta, Y., Piston, D.W., Hogan, B.L., 1997. Bone morphogenetic proteins (BMPs) as regulators of dorsal forebrain development. Development 124, 2203e2212. Gaillard, A., Roger, M., 2000. Early commitment of embryonic neocortical cells to develop area-specific thalamic connections. Cerebr. Cortex 10, 443e453. Galceran, J., Miyashita-Lin, E.M., Devaney, E., Rubenstein, J.L., Grosschedl, R., 2000. Hippocampus development and generation of dentate gyrus granule cells is regulated by LEF1. Development 127, 469e482. Garel, S., Rubenstein, J.L., 2004. Intermediate targets in formation of topographic projections: inputs from the thalamocortical system. Trends Neurosci. 27, 533e539. Garel, S., Lopez-Bendito, G., 2014. Inputs from the thalamocortical system on axon pathfinding mechanisms. Curr. Opin. Neurobiol. 27, 143e150. Garel, S., Huffman, K.J., Rubenstein, J.L.R., 2003. Molecular regionalization of the neocortex is disrupted in Fgf8 hypomorphic mutants. Development 130, 1903e1914. Gerstmann, K., Pensold, D., Symmank, J., Khundadze, M., Hubner, C.A., Bolz, J., Zimmer, G., 2015. Thalamic afferents influence cortical progenitors via ephrin A5-EphA4 interactions. Development 142, 140e150. Ghosh, A., Shatz, C.J., 1993. A role for subplate neurons in the patterning of connections from thalamus to neocortex. Development 117, 1031e1047. Ghosh, A., Antonini, A., McConnell, S.K., Shatz, C.J., 1990. Requirement for subplate neurons in the formation of thalamocortical connections. Nature 347, 179e181. Golonzhka, O., Nord, A., Tang, P.L.F., Lindtner, S., Ypsilanti, A.R., Ferretti, E., Visel, A., Selleri, L., Rubenstein, J.L.R., 2015. Pbx regulates patterning of the cerebral cortex in progenitors and postmitotic neurons. Neuron 88, 1192e1207. Gorski, J.A., Talley, T., Qiu, M., Puelles, L., Rubenstein, J.L., Jones, K.R., 2002. Cortical excitatory neurons and glia, but not GABAergic neurons, are produced in the Emx1-expressing lineage. J. Neurosci. 22, 6309e6314. Grove, E.A., Tole, S., Limon, J., Yip, L., Ragsdale, C.W., 1998. The hem of the embryonic cerebral cortex is defined by the expression of multiple Wnt genes and is compromised in Gli3-deficient mice. Development 125, 2315e2325. Hamasaki, T., Leingartner, A., Ringstedt, T., O’Leary, D.D., 2004. EMX2 regulates sizes and positioning of the primary sensory and motor areas in neocortex by direct specification of cortical progenitors. Neuron 43, 359e372. Harada, H., Sato, T., Nakamura, H., 2016. Fgf8 signaling for development of the midbrain and hindbrain. Dev. Growth Differ. 58, 437e445. Herrmann, K., Antonini, A., Shatz, C.J., 1994. Ultrastructural evidence for synaptic interactions between thalamocortical axons and subplate neurons. Eur. J. Neurosci. 6, 1729e1742. Hevner, R.F., 2005. The cerebral cortex malformation in thanatophoric dysplasia: neuropathology and pathogenesis. Acta Neuropathol. 110, 208e221. Imaizumi, K., Fujimori, K., Ishii, S., Otomo, A., Hosoi, Y., Miyajima, H., Warita, H., Aoki, M., Hadano, S., Akamatsu, W., Okano, H., 2018. Rostrocaudal areal patterning of human PSC-derived cortical neurons by FGF8 signaling. eNeuro 5. Jones, W.D., Guadiana, S.M., Grove, E.A., 2019. A model of neocortical area patterning in the lissencephalic mouse may hold for larger gyrencephalic brains. J. Comp. Neurol. 527, 1461e1477. Kasberg, A.D., Brunskill, E.W., Steven Potter, S., 2013. SP8 regulates signaling centers during craniofacial development. Dev. Biol. 381, 312e323. Kawakami, Y., Esteban, C.R., Matsui, T., Rodriguez-Leon, J., Kato, S., Izpisua Belmonte, J.C., 2004. Sp8 and Sp9, two closely related buttonhead-like transcription factors, regulate Fgf8 expression and limb outgrowth in vertebrate embryos. Development 131, 4763e4774. Krubitzer, L., Stolzenberg, D.S., 2014. The evolutionary masquerade: genetic and epigenetic contributions to the neocortex. Curr. Opin. Neurobiol. 24, 157e165.

Area patterning of the mammalian neocortex Chapter | 3

67

Kudo, L.C., Karsten, S.L., Chen, J., Levitt, P., Geschwind, D.H., 2007. Genetic analysis of anterior posterior expression gradients in the developing mammalian forebrain. Cerebr. Cortex 17, 2108e2122. Lee, S.M., Tole, S., Grove, E., McMahon, A.P., 2000. A local Wnt-3a signal is required for development of the mammalian hippocampus. Development 127, 457e467. Li, H., Fertuzinhos, S., Mohns, E., Hnasko, T.S., Verhage, M., Edwards, R., Sestan, N., Crair, M.C., 2013. Laminar and columnar development of barrel cortex relies on thalamocortical neurotransmission. Neuron 79, 970e986. Liem Jr., K.F., Tremml, G., Jessell, T.M., 1997. A role for the roof plate and its resident TGFbeta-related proteins in neuronal patterning in the dorsal spinal cord. Cell 91, 127e138. Liem Jr., K.F., Tremml, G., Roelink, H., Jessell, T.M., 1995. Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 82, 969e979. Lopez-Bendito, G., Cautinat, A., Sanchez, J.A., Bielle, F., Flames, N., Garratt, A.N., Talmage, D.A., Role, L.W., Charnay, P., Marin, O., Garel, S., 2006. Tangential neuronal migration controls axon guidance: a role for neuregulin-1 in thalamocortical axon navigation. Cell 125, 127e142. Mangale, V.S., Hirokawa, K.E., Satyaki, P.R., Gokulchandran, N., Chikbire, S., Subramanian, L., Shetty, A.S., Martynoga, B., Paul, J., Mai, M.V., Li, Y., Flanagan, L.A., Tole, S., Monuki, E.S., 2008. Lhx2 selector activity specifies cortical identity and suppresses hippocampal organizer fate. Science 319, 304e309. Manuel, M., Georgala, P.A., Carr, C.B., Chanas, S., Kleinjan, D.A., Martynoga, B., Mason, J.O., Molinek, M., Pinson, J., Pratt, T., Quinn, J.C., Simpson, T.I., Tyas, D.A., van Heyningen, V., West, J.D., Price, D.J., 2007. Controlled overexpression of Pax6 in vivo negatively autoregulates the Pax6 locus, causing cell-autonomous defects of late cortical progenitor proliferation with little effect on cortical arealization. Development 134, 545e555. Manuel, M.N., Mi, D., Mason, J.O., Price, D.J., 2015. Regulation of cerebral cortical neurogenesis by the Pax6 transcription factor. Front. Cell. Neurosci. 9, 70. Martinez, S., Crossley, P.H., Cobos, I., Rubenstein, J.L., Martin, G.R., 1999. FGF8 induces formation of an ectopic isthmic organizer and isthmocerebellar development via a repressive effect on Otx2 expression. Development 126, 1189e1200. Matsui, A., Tran, M., Yoshida, A.C., Kikuchi, S.S., U M, Ogawa, M., shimogori, T., 2013. BTBD3 controls dendrite orientation toward active axons in mammalian neocortex. Science 342, 1114e1118. Meyers, E.N., Lewandoski, M., Martin, G.R., 1998. An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nat. Genet. 18, 136e141. Miyashita-Lin, E.M., Hevner, R., Wassarman, K.M., Martinez, S., Rubenstein, J.L., 1999. Early neocortical regionalization in the absence of thalamic innervation. Science 285, 906e909. Mizuno, H., Ikezoe, K., Nakazawa, S., Sato, T., Kitamura, K., Iwasato, T., 2018. Patchwork-type spontaneous activity in neonatal barrel cortex layer 4 transmitted via thalamocortical projections. Cell Rep. 22, 123e135. Moreno-Juan, V., Filipchuk, A., Anton-Bolanos, N., Mezzera, C., Gezelius, H., Andres, B., Rodriguez-Malmierca, L., Susin, R., Schaad, O., Iwasato, T., Schule, R., Rutlin, M., Nelson, S., Ducret, S., Valdeolmillos, M., Rijli, F.M., Lopez-Bendito, G., 2017. Prenatal thalamic waves regulate cortical area size prior to sensory processing. Nat. Commun. 8, 14172. Muzio, L., DiBenedetto, B., Stoykova, A., Boncinelli, E., Gruss, P., Mallamaci, A., 2002. Emx2 and Pax6 control regionalization of the pre-neuronogenic cortical primordium. Cerebr. Cortex 12, 129e139. Nakagawa, Y., Johnson, J.E., O’Leary, D.D., 1999. Graded and areal expression patterns of regulatory genes and cadherins in embryonic neocortex independent of thalamocortical input. J. Neurosci. 19, 10877e10885. Narboux-Neme, N., Goiame, R., Mattei, M.G., Cohen-Tannoudji, M., Wassef, M., 2012. Integration of H-2Z1, a somatosensory cortex-expressed transgene, interferes with the expression of the Satb1 and Tbc1d5 flanking genes and affects the differentiation of a subset of cortical interneurons. J. Neurosci. 32, 7287e7300. Nauta, W.J.H., Feirtag, M., 1986. Fundamental Neuroanatomy. W. H. Freeman and Company, New York. O’Leary, D.D., 1989. Do cortical areas emerge from a protocortex? Trends Neurosci. 12, 400e406. O’Leary, D.D., Stanfield, B.B., 1989. Selective elimination of axons extended by developing cortical neurons is dependent on regional locale: experiments utilizing fetal cortical transplants. J. Neurosci. 9, 2230e2246. O’Leary, D.D., Schlaggar, B.L., Tuttle, R., 1994. Specification of neocortical areas and thalamocortical connections. Annu. Rev. Neurosci. 17, 419e439. Ohkubo, Y., Chiang, C., Rubenstein, J.L., 2002. Coordinate regulation and synergistic actions of BMP4, SHH and FGF8 in the rostral prosencephalon regulate morphogenesis of the telencephalic and optic vesicles. Neuroscience 111, 1e17. Pallas, S.L., Sur, M., 1993. Visual projections induced into the auditory pathway of ferrets: II. Corticocortical connections of primary auditory cortex. J. Comp. Neurol. 337, 317e333. Parchure, A., Vyas, N., Mayor, S., 2018. Wnt and hedgehog: secretion of lipid-modified morphogens. Trends Cell Biol. 28, 157e170. Pattabiraman, K., Golonzhka, O., Lindtner, S., Nord, A.S., Taher, L., Hoch, R., Silberberg, S.N., Zhang, D., Chen, B., Zeng, H., Pennacchio, L.A., Puelles, L., Visel, A., Rubenstein, J.L., 2014. Transcriptional regulation of enhancers active in protodomains of the developing cerebral cortex. Neuron 82, 989e1003. Pennacchio, L.A., Ahituv, N., Moses, A.M., Prabhakar, S., Nobrega, M.A., Shoukry, M., Minovitsky, S., Dubchak, I., Holt, A., Lewis, K.D., PlajzerFrick, I., Akiyama, J., De Val, S., Afzal, V., Black, B.L., Couronne, O., Eisen, M.B., Visel, A., Rubin, E.M., 2006. In vivo enhancer analysis of human conserved non-coding sequences. Nature 444, 499e502. Pinon, M.C., Tuoc, T.C., Ashery-Padan, R., Molnar, Z., Stoykova, A., 2008. Altered molecular regionalization and normal thalamocortical connections in cortex-specific Pax6 knock-out mice. J. Neurosci. 28, 8724e8734. Powell, A.W., Sassa, T., Wu, Y., Tessier-Lavigne, M., Polleux, F., 2008. Topography of thalamic projections requires attractive and repulsive functions of Netrin-1 in the ventral telencephalon. PLoS Biol. 6, e116.

68

PART | I Induction and patterning of the CNS and PNS

Rakic, P., 1988. Specification of cerebral cortical areas. Science 241, 170e176. Rakic, P., Suner, I., Williams, R.W., 1991. A novel cytoarchitectonic area induced experimentally within the primate visual cortex. Proc. Natl. Acad. Sci. U.S.A. 88, 2083e2087. Roe, A.W., Pallas, S.L., Hahm, J.O., Sur, M., 1990. A map of visual space induced in primary auditory cortex. Science 250, 818e820. Roe, A.W., Pallas, S.L., Kwon, Y.H., Sur, M., 1992. Visual projections routed to the auditory pathway in ferrets: receptive fields of visual neurons in primary auditory cortex. J. Neurosci. 12, 3651e3664. Sansom, S.N., Hebert, J.M., Thammongkol, U., Smith, J., Nisbet, G., Surani, M.A., McConnell, S.K., Livesey, F.J., 2005. Genomic characterisation of a Fgf-regulated gradient-based neocortical protomap. Development 132, 3947e3961. Saulnier, A., Keruzore, M., De Clercq, S., Bar, I., Moers, V., Magnani, D., Walcher, T., Filippis, C., Kricha, S., Parlier, D., Viviani, L., Matson, C.K., Nakagawa, Y., Theil, T., Gotz, M., Mallamaci, A., Marine, J.C., Zarkower, D., Bellefroid, E.J., 2013. The doublesex homolog Dmrt5 is required for the development of the caudomedial cerebral cortex in mammals. Cerebr. Cortex 23, 2552e2567. Scearce-Levie, K., Roberson, E.D., Gerstein, H., Cholfin, J.A., Mandiyan, V.S., Shah, N.M., Rubenstein, J.L., Mucke, L., 2007. Abnormal social behaviors in mice lacking Fgf17. Genes Brain Behav. Schlaggar, B.L., O’Leary, D.D., 1991. Potential of visual cortex to develop an array of functional units unique to somatosensory cortex. Science 252, 1556e1560. Scholpp, S., Brand, M., 2004. Endocytosis controls spreading and effective signaling range of Fgf8 protein. Curr. Biol. 14, 1834e1841. Shanmugalingam, S., Houart, C., Picker, A., Reifers, F., Macdonald, R., Barth, A., Griffin, K., Brand, M., Wilson, S.W., 2000. Ace/Fgf8 is required for forebrain commissure formation and patterning of the telencephalon. Development 127, 2549e2561. Shimamura, K., Rubenstein, J.L.R., 1997. Inductive interactions direct early regionalization of the mouse forebrain. Development 124, 2709e2718. Shimogori, T., Grove, E.A., 2005. Fibroblast growth factor 8 regulates neocortical guidance of area-specific thalamic innervation. J. Neurosci. 25, 6550e6560. Sun, X., Meyers, E.N., Lewandoski, M., Martin, G.R., 1999. Targeted disruption of Fgf8 causes failure of cell migration in the gastrulating mouse embryo. Genes Dev. 13, 1834e1846. Sur, M., 1993. Cortical specification: microcircuits, perceptual identity, and an overall perspective. Perspect. Dev. Neurobiol. 1, 109e113. Sur, M., Rubenstein, J.L., 2005. Patterning and plasticity of the cerebral cortex. Science 310, 805e810. Sur, M., Garraghty, P.E., Roe, A.W., 1988. Experimentally induced visual projections into auditory thalamus and cortex. Science 242, 1437e1441. Sur, M., Pallas, S.L., Roe, A.W., 1990. Cross-modal plasticity in cortical development: differentiation and specification of sensory neocortex. Trends Neurosci. 13, 227e233. Tomassy, G.S., De Leonibus, E., Jabaudon, D., Lodato, S., Alfano, C., Mele, A., Macklis, J.D., Studer, M., 2010. Area-specific temporal control of corticospinal motor neuron differentiation by COUP-TFI. Proc. Natl. Acad. Sci. U.S.A. 107, 3576e3581. Toyoda, R., Assimacopoulos, S., Wilcoxon, J., Taylor, A., Feldman, P., Suzuki-Hirano, A., Shimogori, T., Grove, E.A., 2010. FGF8 acts as a classic diffusible morphogen to pattern the neocortex. Development 137, 3439e3448. Tronche, F., Kellendonk, C., Kretz, O., Gass, P., Anlag, K., Orban, P.C., Bock, R., Klein, R., Schutz, G., 1999. Disruption of the glucocorticoid receptor gene in the nervous system results in reduced anxiety. Nat. Genet. 23, 99e103. Tsai, S.Y., Tsai, M.J., 1997. Chick ovalbumin upstream promoter-transcription factors (COUP-TFs): coming of age. Endocr. Rev. 18, 229e240. Turing, A.M., 1952. The chemical basis of morphogenesis. Philos. Trans. R. Soc. Lond. B Biol. Sci. 237, 37e72. Vanderhaeghen, P., Polleux, F., 2004. Developmental mechanisms patterning thalamocortical projections: intrinsic, extrinsic and in between. Trends Neurosci. 27, 384e391. Visel, A., et al., 2013. A high-resolution enhancer atlas of the developing telencephalon. Cell. Vue, T.Y., Lee, M., Tan, Y.E., Werkhoven, Z., Wang, L., Nakagawa, Y., 2013. Thalamic control of neocortical area formation in mice. J. Neurosci. 33, 8442e8453. Wolpert, L., 1996. One hundred years of positional information. Trends Genet. 12, 359e364. Ypsilanti, A.R., Rubenstein, J.L., 2016. Transcriptional and epigenetic mechanisms of early cortical development: an examination of how Pax6 coordinates cortical development. J. Comp. Neurol. 524, 609e629. Yun, K., Potter, S., Rubenstein, J.L., 2001. Gsh2 and Pax6 play complementary roles in dorsoventral patterning of the mammalian telencephalon. Development 128, 193e205. Zembrzycki, A., Griesel, G., Stoykova, A., Mansouri, A., 2007. Genetic interplay between the transcription factors Sp8 and Emx2 in the patterning of the forebrain. Neural Dev. 2, 8. Zembrzycki, A., Chou, S.J., Ashery-Padan, R., Stoykova, A., O’Leary, D.D., 2013. Sensory cortex limits cortical maps and drives top-down plasticity in thalamocortical circuits. Nat. Neurosci. 16, 1060e1067. Zhou, L., Gall, D., Qu, Y., Prigogine, C., Cheron, G., Tissir, F., Schiffmann, S.N., Goffinet, A.M., 2010. Maturation of “neocortex isole” in vivo in mice. J. Neurosci. 30, 7928e7939. Zimmer, C., Lee, J., Griveau, A., Arber, S., Pierani, A., Garel, S., Guillemot, F., 2009. Role of Fgf8 signalling in the specification of rostral Cajal-Retzius cells. Development 137, 293e302.

Chapter 4

Patterning of thalamus Guillermina Lo´pez-Bendito* and Francisco J. Martini Instituto de Neurociencias de Alicante, Universidad Miguel Hernández-Consejo Superior de Investigaciones Científicas (UMH-CSIC), Sant Joan d’Alacant, Spain

Chapter outline 4.1. Introduction 4.2. Insights into diencephalic patterning 4.2.1. Columnar and neuromeric models 4.2.2. Morphologic segmentation of the diencephalon in the prosomeric model 4.2.3. Molecular regionalization of the diencephalon 4.2.3.1. Prosomere 1 4.2.3.2. Prosomere 2: the epithalamic domain 4.2.3.3. Prosomere 3

69 70 70 71 72 72 74 74

4.3. Prosomere 2: the thalamic domain 4.3.1. Cell lineages in the p2 alar plate 4.3.2. Signaling molecules during the initial patterning phase 4.3.2.1. Shh 4.3.2.2. Wnt 4.3.2.3. Fibroblast growth factor 4.3.3. Transcription factor control for neuronal identity List of acronyms and abbreviations References

75 75 76 77 77 78 79 82 83

4.1 Introduction The vertebrate diencephalon is part of the adult forebrain together with the hypothalamus, the retina, and the telencephalon. All these structures are derived from the embryonic prosencephalon. During the first stages of embryogenesis, the primary prosencephalon (forebrain) emerges as a vesicle in the rostralmost region of the neuroectoderm soon after neurulation. As development proceeds, morphological alterations of the forebrain due to evaginations, inward growth, and bending of the longitudinal axis start delineating its initial subdivisions. During gastrulation, the primary prosencephalon segregates, by the end of somitogenesis, into the secondary prosencephalon toward the anterior pole and the diencephalon toward the posterior pole. Adjacent to the diencephalon in the caudal direction lays the mesencephalon (midbrain) followed by the rhombencephalon (hindbrain) and spinal cord. The basic anteroposterior (AP) segments and dorsoventral (DV) domains of the brain primordium are established initially by primary organizers (the node, anterior visceral endoderm, and others) and subsequently refined by secondary organizers within the neuroectoderm. Among other molecules, the most relevant morphogenes influencing the diencephalic regionalization belong to the following gene families: sonic hedgehog (Shh), wingless-int (Wnt), fibroblast growth factor (Fgf), and bone morphogenetic protein (Bmp). In accordance with the prosomeric model, the early forebrain splits in relation to the underlying mesoderm into an epichordal domain (diencephalon) and a prechordal domain (secondary prosencephalon, including the hypothalamus). Defined by gene expression patterns and morphological landmarks, the diencephalon subsequently divides lengthwise into three segments. Following the prosomeric interpretation of the neural tube development, these three diencephalic segments or neuromeres are designated as prosomeres 1e3 (p1ep3) in caudorostral order. Along the longitudinal axis, discrete DV domains are identified in each of the prosomere (from dorsal to ventral): roof, alar, basal, and floor plates (Puelles and Rubenstein, 1993, 2003, 2015; Puelles, 2001; Puelles and Martinez, 2013).

* Senior author.

Patterning and Cell Type Specification in the Developing CNS and PNS. https://doi.org/10.1016/B978-0-12-814405-3.00004-7 Copyright © 2020 Elsevier Inc. All rights reserved.

69

70

PART | I Induction and patterning of the CNS and PNS

While the telencephalic derivatives of the forebrain comprise pallial (neocortex, hippocampus, piriform cortex, claustrum, pallial septum, and pallial amygdala) and subpallial structures (basal ganglia, subpallial septum, and subpallial amygdala), the diencephalic compartments develop into a collection of functional heterogeneous nuclei including the thalamic/prethalamic nuclear complex and the pretectum. The pretectum is the main derivative of p1 alar plate, and in adult humans, it is located rostral to the superior colliculus and subdivided into diverse nuclei that control oculomotor and visuomotor functions (Borostyánkoi-Baldauf and Herczeg, 2002). The basal plate of p1 gives rise to the prerubral tegmentum containing nuclei involved in motor control as the ventral tegmental area and substantia nigra, and the interstitial nucleus of Cajal, which are engaged in head orientation reflexes (García-López et al., 2004). The p2 alar plate develops into the epithalamus (habenula and pineal gland) and thalamus in mature brains. The habenula has been linked to reward signaling and depression among other functions, whereas the pineal gland participates in the regulation of the circadian rhythms (García-López et al., 2004; Hikosaka, 2010). The thalamus is a nuclear complex, anatomically the largest part of the diencephalon, representing the main relay station for sensory information to the overlaying telencephalon; it plays a part in motor control and association processing as well. The main structures derived from the p2 basal plate are the most anterior parts of the ventral tegmental area and the substantia nigra (García-López et al., 2004). Finally, in p3, the alar plate derivatives are the prethalamic nuclei (mainly the reticular thalamic nucleus, ventral lateral geniculate nucleus, subgeniculate nucleus, and zona incerta) and dorsally the prethalamic eminence (PThE) (García-López et al., 2004). The size and complexity of the numerous structures generated in the diencephalon varies among species; however, the molecular machinery underlying their formation is highly conserved through vertebrates as has been shown from studies undertaken in chick, frog, mouse, and fish. Therefore, the correct assembly of mature diencephalic circuits relies on a timely and spatially accurate, as well as largely conserved, network of genetic programs controlling cell proliferation, specification, migration, and differentiation into neuronal subtypes (Sena et al., 2016). Along this chapter we first examine general aspects of the patterning of the diencephalon and then we place emphasis on the largest anatomical structure derived from the developing diencephalon, the thalamus.

4.2 Insights into diencephalic patterning In the developing diencephalon, neural progenitors undergo proliferation in the ventricular epithelium along the third ventricle. As in other regions of the forebrain, diffusing signals released by nearby tissues unfold dedicated genetic programs that regionalize the prospective diencephalic territory. To achieve this transition from progenitors to mature neurons, proliferating cells acquire first positional identity by means of transcriptional regulation and signal transduction. In this way, the early patterning of the diencephalon along the anterioreposterior and dorsaleventral axes is established. The initial patterning defines then how, later in development, neurons populate specific nuclei building of the complex organization of the diencephalon. From the germinal epithelium, postmitotic cells migrate toward the mantle layer to aggregate in clusters that subsequently differentiate into domains and, later on, into nuclei (morphological and functional units). Within these nuclei, neurons have particular morphologies, neurotransmitters, and connectivity profiles that define their function and, collectively, the function of the brain structure that they belong to (Jones, 2012; Sherman and Guillery, 2009). In the following sections, we address the mechanisms involved in patterning and parcellation of the diencephalon. But first we need to define the conceptual paradigms underlying brain development to fit diencephalic developmental principles into a global framework and guide our understanding of how the brain is built.

4.2.1 Columnar and neuromeric models To unscramble the complex process of brain development, researchers in the last century looked for a basic program or paradigm that supplies a theoretical framework within which data are analyzed to formulate reliable conclusions (Puelles, 2001). A couple of essentially opposing models were schemed in pioneer neurodevelopment studies. Although proposed paradigms dissented in their fundamental principles, they coincided in a basic feature: brain development is orchestrated adjoining segments or modules of independent units. Therefore, a major quest of these paradigms was to identify to what extent brain development is a segmental process and, if this is the case, which mechanisms define and organize these building segments. In this sense, two opposing views arose to explain brain ontogenesis in terms of segments: the columnar and the neuromeric models. Generations of embryologists embraced either one or the other which caused conflicting data interpretation, especially regarding the diencephalic development. Although the columnar model prevailed for many years during the last century, the advent of modern molecular tools revaluated the neuromeric model turning into the current accepted conceptual paradigm for brain development.

Patterning of thalamus Chapter | 4

71

The initial neuromeric models were almost contemporaneous of the first columnar models; however, the latter ones and their evolved versions prevailed for decades during the 20th century. The columnar model postulated by Herrick in 1910 and lately extended (Alvarez-Bolado et al., 1995) holds that the basic units of forebrain organization are four adjacent longitudinal columns delimited by ventricular sulci and rostrally ending in the telencephalon. In the diencephalon, the resulting subdivisions were the epithalamus, dorsal thalamus, ventral thalamus, and hypothalamus. In the last decades, anatomical and molecular advances generated a plethora of new data that hardly fit into a paradigm where longitudinal segments are the basic units of brain compartmentalization. Instead, an axial organization stems from these results and, in addition, they are consistent with a length axis that follows the neural tube curvature and ends in the hypothalamus, rendering the telencephalon as an alar evagination and the hypothalamus as a basal component. Thus, among other facts, columnar models fail to allocate robust molecular data upraising then the formerly underestimated neuromeric model. From the 1980s onward, the neuromeric model gained acceptance and has progressively consolidated as the current paradigm that better encompasses forebrain development. In essence, the neuromeric model holds that the neural tube comprises a continuous and serial sequence of transverse units designated as neuromeres. Neuromeres are organized as follows (from rostral to caudal): in the forebrain, 1 large rostral unit (including hypothalamus, retinas, and telencephalon) and 3 diencephalic units (prosomere 3, including prethalamus; prosomere 2 including thalamus; prosomere 1 including pretectum); 1 unit in the midbrain (called mesomere), 11 units in the hindbrain (called rhombomeres), and the units that formed the spinal cord (called myelomeres). As previously mentioned and opposite to the columnar model, the neuromeric paradigm is consistent with a longitudinal axis that bends solidary with the neural tube, especially at the cephalic flexure (Fig. 4.1). Each neuromeric segment can be divided in longitudinal regions: roof, alar, basal, and floor plates, in DV order.

4.2.2 Morphologic segmentation of the diencephalon in the prosomeric model Classically, the primordial source of data to understand how the brain develops came from morphological studies. In the last decades, modern molecular techniques provided valuable insights into how brain regionalize because molecular patterns represent the primary cause underlying developmental mechanisms. Despite modern techniques’ advantages, the classic topological analysis gives usually important information regarding landmarks and positional meaning as expressed

P1

Mes

LV P2

Ctx

4V Str

P3 Poa

Hy Rb

SC

Anterior forebrain (telencephalon and hypothalamus) Diencephalon Mesencephalon Rhombencephalon Spinal cord

FIGURE 4.1 A schematic sagittal view of the brain according to the prosomeric model of the mouse around E12.5, based on Eurexpress Anatomy Atlas. The black dashed line marks the alar/basal boundary. Ventricles are labeled in gray. The prosomeric model holds that the diencephalon (colored in blue) is subdivided into three transverse segments. The caudalmost segment is prosomere 1 (P1) and corresponds to the pretectum. The rostralmost segment is prosomere 3 (P3) and corresponds to the prethalamus. Between P1 and P3, the prosomere 2 (P2) comprises the thalamus and epithalamus. Note that in a classic coronal section at the level of the diencephalon (vertical axis of the figure), rostral P3 would appear at the bottom of the slice equivalent to a ventral position in telencephalic levels, however, by no means corresponds to a ventral diencephalic region. 4V, forth ventricle; Ctx, cerebral cortex; Hy, hypothalamus; LV, lateral ventricle; Mes, mesencephalon; P1eP3, prosomeres 1e3; Poa, preoptic area; Rb, rhombencephalon; SC, spinal cord; Str, striatum.

72

PART | I Induction and patterning of the CNS and PNS

by Puelles (Puelles, 2001). A morphological standard to identify cerebral boundaries is the presence of constrictions, landmarks formed by two adjacent territories with differential growth. In this way, each constriction has a ventricular ridge and an external furrow in the wall. In experimental embryology, the chick embryo has been widely used as model organism to address brain segmentation. In the avian developing neural tube, the mesencephalon and primary prosencephalon are discernible by HH10. The primary prosencephalon emerges as a growing vescicle. Soon after (by stages HH11eHH12), an external furrow splits the primary prosencephalon in two morphological discontinuous halves: the secondary prosencephalon and the diencephalon (Vaage, 1969; Puelles et al., 1987; Trujillo et al., 2005). This constriction between telencephalic and diencephalic vesicles is mainly appreciated in the dorsal zone concurrent with the velum transversum, another landmark useful to delimit the anterior limit of the diencephalon. However, the telencephalic/diencephalic morphological boundary is not sharply defined probably because of complex anatomical transformations in this region caused by the extensive telencephalic growth (Lim and Golden, 2007). On the posterior pole of the diencephalon, a furrow sets the limit with the mesencephalon at the level of the cephalic flexure. Within the diencephalon, prospective prosomeres are not discernible yet at these stages because intradiencephalic boundaries are hardly visible. In consequence, the initial steps described so far have been designated as the “proneuromeric phase” of diencephalic morphogenesis (Puelles, 1987). Later on, between stages HH12 and HH13 in chick, circa E9.5 in mouse, the diencephalon splits into synencephalon (prospective p1) and parencephalon (prospective p2 and p3) initiating the second round of diencephalic morphogenesis, the “early neuromeric phase” that runs until HH18 (Puelles, 1987). A constriction segregating both territories surrounds the whole neural tube; however, it is more evident in the dorsal part as seen in horizontal sections. This limit corresponds to the line connecting the opticediencephalic angles and where the habenulopeduncular tract is found (Larsen et al., 2001; Suda et al., 2001). Finally, the “late neuromeric phase” unfolds around HH18eHH19, E10eE11 in mice, and its more remarkable feature is the transversal subdivision of the parencephalon into anterior and posterior domains. The parencephalon shows a diffuse delimitation marked by constrictions until the appearance of the zona limitans intrathalamica (ZLI), a transversal boundary formed by a strip of cells that widens toward the basal plate (Puelles, 1987). The formation of ZLI occurs near the region where the mesoderm divides into prechordal plate and notochord (Figdor and Stern, 1993; Shimamura et al., 1995). On the basis of morphology, the three diencephalic prosomeres are clearly ostensible by HH24 in chick embryos (E12.5 in mice).

4.2.3 Molecular regionalization of the diencephalon Anatomical landmarks were the original observations that started our current body of knowledge and, at the same time, inspired the formulation of conceptual frameworks for brain development, including the neuromeric/prosomeric model. Remarkably, in the last decades, a large number of reports provided valued data regarding molecular markers whose expression patterns delimit, consistently with the prosomeric model, divisions, and subdivisions of the diencephalon (Bulfone et al., 1993; Martinez-Ferre and Martinez, 2012; Price et al., 2012; Suzuki-Hirano et al., 2011). Moreover, the initial regionalization of the diencephalon at the molecular level, characterized by the expression of TF such as Otx2, Irx3, Six3, and Pax6 (Fig. 4.2), emerges before clear-cut morphological indicators become evident (Simeone et al., 1992; Robertshaw et al., 2013). Thus, the expression of specific genes in spatial and temporal domains from early stages allows a deeper exploration inside the mechanisms by which diencephalic development proceeds. As scarce data are available to track the development of the diencephalic basal plate, we mainly focus in the alar region and its derivatives. Below, the next three sections are devoted to summarize the molecular organization of prosomeres 1, 2 (only epithalamus), and 3; the thalamic derivative of p2 alar plate will be inspected afterward.

4.2.3.1 Prosomere 1 The main derivative from caudalmost prosomere (p1) is the pretectum. Residing in the alar plate, the pretectum in mature brains is an area involved in visual information processing and execution of visual reflexes. Based on a genoarchitectonic approach applying gene expression analysis to dissect different neural structures along AP and DV body axes, Ferran and colleagues identified over 30 distinct pretectal domains (Ferran et al., 2009). A number of studies showed that the posterior end of Pax6, Wnt8b, and EphA7 expression patterns among others (Li et al., 1994; Pera and Kessel, 1997; Nomura et al., 1998), together with the anterior limit of En2, En1, and Pax2 expression (Araki and Nakamura, 1999; Matsunaga et al., 2000), demarcate the pretectal/midbrain border from stage HH8-11 onward in birds. Fate-mapping studies corroborate these observations (García-López et al., 2004). Later in development, other gene

Patterning of thalamus Chapter | 4

73

Mes

PT

Th

ZLI PTh

En1

Lfng

Irx3

Fezf1

Six3

Otx2

Pax6

Emx2

Pax3

Wnt8b

Tel

FIGURE 4.2 Schematic representation of diencephalic and neighbor regions together with the differential expression of transcription factors during early diencephalic patterning. Mes, mesencephalon; PT, pretectum; PTh, prethalamus; Tel, telencephalon; Th, thalamus; ZLI, zona limitans intrathalamica.

markers display expression patterns in register with this border. From stages HH17 onward in chick embryos, Tcf4 and Lim1 are expressed in the caudal pretectum and a Pax7-negative stripe lies in the midbrain part of the boundary. Also, the anterior end of the expression pattern of Meis2 in the ventricular zone of the midbrain alar plate corresponds to this diencephalic/mesencephalic boundary (Ferran et al., 2007). The thalamo-pretectal boundary demarcates the pretectal anterior end and becomes detectable later than the pretectal/ midbrain boundary. Pax3 expression starts at the roof plate by HH13 in birds and spans throughout the midbrain and the caudal diencephalon with a sharp caudal boundary against p2 by HH17. Other gene markers useful to delimit this boundary are Wnt8b, L-fng, Meis1, and Ebf1 whose expression rostrally ends at this limit (Matsunaga et al., 2001; García-López et al., 2004; Ferran et al., 2007). Apart from prosomeric AP borders, the analysis of progenitor domains also made possible to distinguish secondary boundaries in the alar and roof plates of p1. Within the pretectum, internal transverse subdivisions consistent with previous morphological observations appear later to define AP parcels. In this sense, three main domains are contemplated: commissural pretectal, the precommissural pretectal, and the juxtacommissural pretectal domains. In birds, they are visible from stage HH17 onward by the differential expression of Pax7, Dmbx1, Dbx1, and Six3 among others. The same combinatorial code of transcription factors confirms the tripartite schema of the pretectum in mouse. The characterization of pretectal early regionalization using gene expression analysis has been extended in time and complexity up to nuclear levels in later stages (Ferran et al., 2007, 2008, and 2009). In the basal plate, the expression pattern of Nkx2.2 has a longitudinal band shape whose dorsal end marks the basal/alar limit of p1. Other gene expression patterns also delimit one or more DV domains. For example, Pax6, Ebf1, Six3, and Dbx1 signals are restricted to the alar plate of p1, respecting either basal/alar or alar/roof limit. This latter border separates the pretectum and the subcommissural organ, an area that differentiates entirely from p1 roof plate. In other cases, expression patterns respect one limit but overpass others; this is the case for Tcf4 and Dmbx1. They end at the roof/alar limit as well but, on the other side, extend ventrally into the basal plate. Within each AP pretectal segment, molecular markers also allow to distinguish subdivisions in the DV axis too. While five domains are acknowledged in the commissural pretectal domain (D1, D2, L1, L2, and V), only three are identified in the precommissural and in the juxtacommissural pretectal domains (D2, L, and V; Ferran et al., 2007) on the basis of differential gene expression (Pax6, Pax3, Pax7, Six3, Lim1, Dbx1, Ebf1, Zic1, and Emx1). Similar to the spinal cord, it is

74

PART | I Induction and patterning of the CNS and PNS

likely that the patterning mechanism involved in pretectal DV parcellation derives from the complementary effects of SHH diffusion from the floor plate against BMP and Wnt gradients from the roof plate (Ferran et al., 2007). Below the alar pretectal region, floor and basal plates correspond to the ventral tegmental domain of p1. Subpretectal tegmentum expresses Shh at the ventricular lining and Nkx2.2 at the ventricular zone and mantle. Other gene patterns depict p1 basal and roof plates; however, their barriers are not sharp enough. For example, Lim1, Six3, and Pax6-positive cells are found in p1 basal plate; however, their expression continues either caudally into the midbrain or rostrally into p2. Despite inadequate delimiting borders at first glance, their expression patterns do segregate territories because the shape of the labeling switches depending on the localization. For example, Pax6-positive cells of the diencephalic basal plate show a compact profile of expression rather than thin and elongated as in the mesencephalic side (Ferran et al., 2007).

4.2.3.2 Prosomere 2: the epithalamic domain The epithalamus is the other structure derived from dorsal p2 together with the thalamus. In mature brains, the epithalamus holds the habenula and pineal body. The habenula connects the forebrain with midbrain and hindbrain structures. It is involved in reward and aversion processing among other functions. On the other hand, the pineal body is part of the brain machinery that controls circadian rhythms. Among the molecules that control epithalamic development, Fgf8 is a key factor governing the habenular and pineal specification because downregulation of Fgf8 leads to the reduction or lack of both structures. Wnt3a and Dbx1b have been demonstrated as important downstream effectors of Fgf8 to assure a correct habenular development. Dbx1b controls proliferation and differentiation of habenular progenitor cells (Martinez-Ferre and Martinez, 2009; Dean et al., 2014; Schmidt and Pasterkamp, 2017). Interestingly, thalamic fate specification in p2 alar plate is achieved partially by preventing the activation of the epithalamic genetic program. As explained below in Section 4.3.3, the suppression of Shh (from ZLI and basal plate) or Gbx2 (in the mantle zone) activity, both are thalamic fate regulators, generates that epithalamic signatures take over the presumptive thalamic territory (Chatterjee et al., 2012, 2014; Mallika et al., 2015). In the same line, in models where Pax6 is absent, Shh signaling is increased and the epithalamus is missing. During normal development, Shh signaling is kept out of the epithalamus, thanks to the persistent epithalamic expression of Pax6, which actually delimits the habenular/thalamic border. However, Pax6 profile of expression in earlier stages does not depict a habenular positive against a thalamic negative scenario. The earlier expression of Pax6 spreads widely through the alar diencephalon before ZLI formation. Only once ZLI is consolidated, Pax6 downregulation starts focally at the level of ZLI and progressively evolves toward both sides but with a stronger effect in the caudal direction, excluding the epithalamus.

4.2.3.3 Prosomere 3 The main alar structure of p3 is the prethalamus. Prethalamic-derived neurons contribute to populate the reticular thalamic nuclei, vLG, IGL, subgeniculate nucleus, and zona incerta (Martinez-Ferre and Martinez, 2012). Contrary to most of the p2-derived nuclei, prethalamic structures do not project to the cortex, instead they are connected to subcortical areas (suprachiasmatic nuclei, striatum, red nucleus, and substantia nigra) including other thalamic nuclei where they are able to modulate relay activity. Dorsal to the prethalamus opens up the PThE, another alar structure of p3. This region is functionally related to the contiguous habenular nuclei from the epithalamic region of p2 alar plate. In the prospective diencephalon, many molecular markers appear at the initial phase of p3 patterning. Some of them regard p3 borders bilaterally and others are shared with neighbor territories. In mice at E9.5, the developing p3 shows the expression of Wntb8, Emx2, Pax6, Otx2, Six3, Fezf1, L-fng, and Lrrn1 (Puelles and Martinez, 2013). Genes that are expressed in other prospective prosomeres but not in p3 are Pax3 and Irx3. At E12.5, the anterior prethalamic border is demarcated by the expression of Foxd1 and Gdf10. Both patterns are complemented by Foxg1 on the telencephalic side configurating a sharp boundary (Hatini et al., 1994; Herrera et al., 2004; Shimogori et al., 2010). The combined patterns of Foxd1 and Gdf10 together with Lhx9, Arx, and Olig2 demarcate the full extension of the prethalamic region. Moreover, different domains of gene expression in p3 can be outlined combining the latter three markers. Lhx9 and Gdf10 are expressed dorsally in the PThE that has a clear-cut border with the ventral expression domain of Arx. In the ventralmost domain, Arx is expressed in combination with Olig2. Other markers for PThE are calretinin, Cacna2d1, and Tbr1. Ventrally, toward the prethalamic domain, other markers are detected such as Fezf2, Lrrn1, Pax6, Six3, Dlx1/2/5/6, Sfrp2 (in the ventricular zone), Gsh2, and Gli3. Similar to the thalamic division into Th-c and Th-r domains, recently an AP compartmentalization of p3 alar plate in rostral and caudal domains has been proposed (PTh-r and PTh-c, respectively). While the former occupies a large proportion of the prethalamic region, PTh-c is a narrow band juxtaposed with the ZLI that is marked by the expression of Dlx1/2/5/6, Arx, Lhx1, and Dbx1.

Patterning of thalamus Chapter | 4

75

4.3 Prosomere 2: the thalamic domain Neurons in the mature thalamus are spatially arranged in clusters called nuclei. Each nucleus contains neurons that share similar properties. Many relevant physiological functions are to some extent processed by thalamic nuclei. Accordingly, a broad functional criterion classifies them into sensory, motor, and association nuclei. However, within each class, nuclei are quite diverse in terms of connectivity, molecular signatures, cell morphology, electrophysiological profile, and many other properties. For example, within sensory nuclei, the level of information processing complexity splits the group in two. On one hand, first-order sensory nuclei receive feedforward input from downstream structures and project to primary sensory cortices. On the other hand, higher-order sensory nuclei can obtain either feedforward or feedback input; this one comes from upstream structures. In addition, higher-order sensory nuclei project to primary sensory cortices and to other high-order processing centers as well, including cortical secondary and association areas. Moreover, within the first-order and also higher-order subclass, nuclei are sorted following the nature of information they process, i.e., visual, somatosensory, or auditory. Although more subdivisions can be mentioned, on the other hand, almost all thalamic nuclei share an important feature: their neurons use glutamate as the main neurotransmitter. Only a small number of thalamic nuclei are GABAergic, including IGL and vLG, and none of them project to the cortex. Altogether, thalamic nuclei are a varied but, at the same time, very well-organized assembly of neurons. The distinctive ordered heterogeneity of mature thalamus is based on the specification of different progenitor domains during early embryonic stages that subsequently differentiate into diverse subpopulations of neurons. These subpopulations constitute the building blocks that will populate thalamic nuclei. Soon after the formation of the ZLI (E10.5 in mice), cues from the initially patterned regions further subdivide the diencephalon into distinct progenitor domains. Based on the expression of molecular markers in the prospective thalamus, two domains are further identified known as rostral (or anterior) and caudal (or posterior) progenitor domains of the thalamus, Th-r and Th-c, respectively (Jeong et al., 2011; Suzuki-Hirano et al., 2011). In the caudal flank of the ZLI, the Th-r comprises a small area under the strong influence of Shh, whereas the Th-c is a larger area immediately posterior to Th-r (Fig. 4.3). Because exposing Th-r and Th-c to graded amounts of SHH and other signals leads to the activation of different genetic programs, it is expected that each progenitor domain originates different neuronal subtypes. Remarkably, each domain generates completely different subpopulations of thalamic neurons; Th-r-derived neurons acquire a GABAergic fate, whereas Th-c-derived neurons acquire a glutamatergic fate.

4.3.1 Cell lineages in the p2 alar plate As in other brain structures, lineage is a critical factor in thalamic development. To understand how different progenitor domains contribute to the formation of thalamic nuclei, it is important to know the lineage relationship between heterogeneous progenitor cells and the fate of their progeny of postmitotic cells. Genetic lineage tracing studies in mice demonstrated that the Th-r domain contributes to populate GABAergic thalamic nuclei, IGL and vLG nuclei, whereas cortical-projecting thalamic nuclei are mainly occupied by glutamatergic neurons derived from Th-c (Vue et al., 2007).

P1(dorsal) P3 (dorsal) Shh+ (ZLI) Shh+ (basal) Th-r Th-c ETh

FIGURE 4.3 Schematic representation of a sagittal view of the diencephalic in mouse embryos around E12.5 outlines the localizations of ZLI, alar prosomere 1 (green), alar prosomere 3 (orange), different progenitor domains within alar prosomere 2 (Th-r, blue; Th-c, light blue; epithalamus, red), Shhpositive basal diencephalon (lavender), and Shh-positive ZLI (purple). The area shaded in gray corresponds to territories flanking the diencephalon.

76

PART | I Induction and patterning of the CNS and PNS

Olig3 is expressed in both progenitor domains but other TFs could be used to discern between them. The Th-r is characterized by the expression of Ascl1 (also named Mash1), Tal1, and Nkx2.2. The Th-c is characterized by the expression of the proneural transcription factors Neurog1/2. While Neurog2 is expressed in the mantle zone, Neurog1 transcripts are detected only in the ventricular zone. Interactions between these TFs have been described as key steps in thalamic development. For example, Neurog2 downregulates Ascl1 preventing the inhibitory fate (Fode et al., 2000). Within the Th-c, Dbx1 is expressed gradually from high caudodorsal to low rostroventral, a pattern that complements that of Olig2. Importantly, Dbx1-derived progeny contributes to nuclei located caudal or dorsal and Olig2-derived progeny gives rise to anterior or ventral nuclei demonstrating that heterogeneity of progenitor cells is meaningful for the diversification of thalamic nuclei (Vue et al., 2007). Examples of the caudodorsal Dbx1-derived structures are central lateral (CL), mediodorsal (MD), and central medial (CM) nuclei. At E16.5, Olig2-derived cells express Sox2; they form principal sensory nuclei such as LP, dLG, MGv, and VP. In the last years, cell lineage analysis at the single-cell level has provided relevant information to understand how the thalamic nuclei are formed. Using MADM-based genetic lineage tracing, Wong and colleagues determined cell lineage of thalamic progenitor cells at the clonal level (Wong et al., 2018). They found that a dividing radial cell generates neurons that will be incorporated in multiple nuclei. Thalamic neurogenesis is characterized by a temporal-spatial pattern defined by a lateral-to-medial neurogenic gradient in which early-born neurons (E9.5eE11.5) generally contribute to more lateral nuclei (such as dLG and MGv), whereas latter neurons occupy more medial nuclei (such as CM, MD). However, this is not a rigid rule because in some cases members of the same lineage were found in both lateral and medial nuclei. Clustering analysis indicates that thalamic nuclei can be ontogenetically segregated on the basis of the distribution of cells from the same clone in different nuclei. This clustering analysis shows that there are groups of clones whose neurons are prompted to inhabit in certain nuclei while avoiding others. Then, thalamic nuclei can be sorted accordingly to the extent of clonal members’ dispersion. As clones are spatially organized, it follows that the spatial position of the nucleus defines the cluster they belong to. In general, the ontogenetic relationship dictates the spatial configuration of thalamic neurons in a rostrocaudal and DV pattern. Using a similar approach, Shi and colleagues obtained comparable results but came to different conclusions regarding the relationship between lineage and the organization of thalamic nuclei. They propose a segregation model that follows functionality, instead of anatomical position, in which inhibitory, cognition-related, sensory/motor first-order and sensory/ motor higher-order nuclei are ontogenetically discernible (Shi et al., 2017). Wong and colleagues showed that not only clones and nuclei can be sorted in clusters but also the location of original radial glial cells correlates with the nuclei that their progeny will populate. More rostral progenitors from Th-c contribute mainly to neurons destined to principal sensory nuclei, whereas more caudal clones contribute to caudodorsal nuclei. The fact that progenitor location predicts nuclear distribution corresponds with the gradient of gene expression observed across Th-c (Vue et al., 2007). Unexpectedly, in some cases of extremely rostral clones, progeny contributes concurrently to nuclei derived from both progenitor domains. For example, these neurons can be found in principal sensory nuclei derived from Th-c as well as in the vGL and IGL derived from Th-r. More research is needed to resolve this controversial result.

4.3.2 Signaling molecules during the initial patterning phase How are the different lineages established in the progenitor cells of the thalamus? At the beginning of forebrain development, the initial patterning process segregates the anterior neuroectoderm into two basic forebrain subdivisions (secondary prosencephalon and diencephalon). Transcriptional programs together with signaling mechanisms are responsible for the initial forebrain patterning. In the diencephalon, the main signaling molecules are Wnts, Shh, Bmps, and Fgfs (Wilson and Houart, 2004; Nakagawa and Shimogori, 2012). Diencephalic subdomains react differently to them indicating that there is a preceding acquisition of specific responsiveness to the diverse signaling cues. Many molecular markers have been shown to account for the different developmental pathways shown by diencephalic subdomains in response to signaling molecules. In general, the mutual confrontation of TF patterns of expression is a hallmark to establish different territories within the forebrain and its subdivisions. For example, among others, the complementary expression in chicken of Irx3 (caudal) and Six3 (rostral) is initially regulated by Wnt signaling and establishes two diencephalic subdomains (presumptive prethalamus and thalamus) with different competence for Shh signaling. In other words, presumptive prethalamic and thalamic cells embrace their different diencephalic identity by means of molecular mechanism downstream Shh signaling that, however, were differentially prespecified before Shh action (Kobayashi et al., 2002; Robertshaw et al., 2013). In this particular case, the prethalamic/thalamic boundary itself will turn into a territory that radiates Shh signaling because ZLI emerges in the intersection between Irx3 and Six3 patterns of expression.

Patterning of thalamus Chapter | 4

77

4.3.2.1 Shh Although initially expressed in nonneural tissues underlying the neural plate, Shh induces its own expression in the ventral midline determining the DV specification of the neuraxis. Besides its global function, in the diencephalon, Shh has multifaceted roles at different time points (Epstein, 2012). For example, early in development, Shh acts as an important mitogen; its absence reduces proliferation and survival of progenitor cells (Ishibashi and McMahon, 2002). Another key function of Shh relates to areal specification in the diencephalon; in this sense, Shh has been acknowledged central to establish the thalamic identity as a whole. In the prospective thalamus, Shh is not only expressed in the corresponding ventral part, as in other brain regions, but also in a transverse band that emerges at E10 in mice between p2 and p3, the ZLI. The combined action of basal and ZLI-derived Shh signaling is a determinant to specify diencephalic compartmentalization in vertebrates and, particularly in the thalamic region, determines the segregation of progenitor domains into Th-r and Th-c (Kiecker and Lumsden, 2004; Vue et al., 2009; Jeong et al., 2011). Central to the understanding of the role of Shh from ZLI is to address how this secondary organizer is formed. Two main homeodomain-containing TFs have been suggested to confer to the presumptive ZLI competence to induce Shh: Otx2 and Barhl2. A reduction or lack of these transcripts determines the lack of Shh expression in the ZLI. The molecular mechanism underlying these effects is based on the combined action of Otx2 and Barhl2 over Shh’s enhancers (Scholpp et al., 2007; Juraver-Geslin et al., 2014; Yao et al., 2016; Ding et al., 2017). Moreover, it seems that ZLI specification is initially induced by basal plateederived Shh and, at the same time, requires the interaction between the epichordal and prechordal diencephalic domains. It has been demonstrated that the permissive action of the interplay between a Fezf2/ Six3-positive prechordal domain and an Irx3-positive epichordal domain specifies prospective ZLI and indirectly induces Shh expression. In this regard, Fez genes have been proposed as direct or indirect repressors of caudal diencephalic genes in the rostral diencephalon during early stages (E9 in mice) (Kobayashi et al., 2002; Braun et al., 2003; Kiecker and Lumsden, 2004; Vieira et al., 2005; Hirata et al., 2006). In any case, a complete understanding of the genetic network governing the specification and expression of Shh in ZLI is still missing. Gli3 is a Shh-regulated repressor widely expressed in the alar diencephalon at early stages. However, Gli3 is locally downregulated in the prospective ZLI in chick embryos, a process mediated by Wnt signaling. Wnt8b activity (likely interacting to some extent with Fez, Otx2, Irx3, and others) mediates Gli3 inhibition preceding ZLI formation, a fact that allows the induction of the ZLI by basal plateederived Shh (Martinez-Ferre et al., 2013; Navarro-Garberi et al., 2016). Another important cue that helps to restrict Shh expression to the ZLI is Pax6. By means of cell-autonomous repression, Pax6 confines Shh expression to ZLI limits, a mechanism that keeps Shh activity at correct levels (Caballero et al., 2014). Actually, the repression is apparently mutual because Shh has also been considered to negatively regulate Pax6 (Robertshaw et al., 2013). As mentioned before, apart from its role in ZLI formation, Shh also participates in the specification of thalamic nuclei. Many studies from different species have demonstrated that the specific genetic programs induced by ZLI-derived and basal Shh correlate with fate adopted by neurons. In this case, the effect of Shh might be mediated by Fgf15, which regulates the expression of proneural TFs (Jones and Rubenstein, 2004; Suzuki-Hirano et al., 2011; Yuge et al., 2011; Martinez-Ferre et al., 2016). Neurons of thalamic nuclei derive at least from the two progenitor domains delineated by Shh, Th-r and Th-c. It seems that the graded Shh signaling from the ZLI and basal plate determines the formation of these two different domains. While the Th-r is a small region in the rostroventral zone of the thalamus exposed to high level of Shh, the Th-c is a large region in the caudodorsal zone that receives lower amounts of Shh (Hashimoto-Torii et al., 2003). Consistently, Th-r progenitors adopted a molecular program that resembles Th-c when Shh levels are reduced in the diencephalon (Jeong et al., 2011), and a more severe phenotype affecting also projecting neurons is unmasked when Shh is deleted from thalamus (Szabó et al., 2009; Vue et al., 2009).

4.3.2.2 Wnt Wingless (Wnt) signaling from outside the nervous system is crucial for early brain patterning. In chicken, Wnt signaling has been suggested to be involved in a dual role during thalamic development. At early stages, Wnt signals confer identity to the posterior forebrain. Once the basic regionalization of the diencephalon is achieved, Wnts participate in the segregation of diencephalic structures by inducing the expression of thalamic markers (Braun et al., 2003). In this process, the fates of the thalamus and prethalamus are segregated by means of the regulation of TFs expression such as Irx, Fez, and Six families (Kobayashi et al., 2002; Shimizu and Hibi, 2009). In mice, the absence of a member of the Wnt/b-catenin pathway, the Wnt co-receptor named Lrp6, leads to a thalamic compartment that shows histological and molecular abnormalities (Zhou et al., 2004). In this mutant mouse, the segmental boundary between the prethalamus and thalamus, the

78

PART | I Induction and patterning of the CNS and PNS

ZLI, was missing and its Shh signal lost. Also, the thalamic early markers (Gbx2, Tcf4) were reduced or switched to more differentiate state markers (Prox1 and ROR-alpha). The lack of ROR-alpha signal, which is expressed predominantly in primary sensory neurons (Nakagawa and O’Leary, 2003), correlates with the absence of thalamocortical projections. The defects described in the Lrp6-deficient model may come from abnormalities in the cell-autonomous function of Wnts or from an indirect effect through ZLI- or basal plateederived Shh. In any case, both pathways collaborate to induce thalamic development as seen in experiments where b-catenin activity was removed during thalamic neurogenesis in mice (E10.5eE12.5; Bluske et al., 2012). In this model, it is clear that Wnt signaling maintains Th-c thalamic identity but suppresses Th-c and prethalamic fates, a process that requires Neurog1/2. Which of the members of the Wnt family are responsible for this? At least, Wnt3, Wnt3a, and Wnt7b are expressed in the thalamus of mice at E10.5eE11.5. However, other Wnts may act together to activate Lef1/Tcf signals because their pattern of expression does not match with that of the downstream target gene Axin2 (Bluske et al., 2009). In the absence of b-catenin activity in the developing thalamus, an ectopic induction of prethalamic (Dlx2 for progenitors; Dlx5, Pax6, and Islet1 for postmitotic cells) and Th-r markers (Helt for progenitors; Gata2 and Gata3 for postmitotic cells) is induced in the Th-c in detriment of its own thalamic markers (Olig3 for progenitors; Gbx2, ROR-alpha, and Sox2 for postmitotic cells). Surprisingly, although specifies thalamic glutamatergic identity, Otx2 expression is not affected under these conditions, suggesting that its regulation is independent of Wnts. Shh from the ZLI and the basal plate could be mediating the induction of Th-r markers. However, reducing both Shh signal and b-catenin activity prevents Nkx2.2 expression but some other Th-r markers remain expressed suggesting a Shh-independent mechanism. Regarding the formation of ZLI, consistent information arises from experiments in chick embryos where Wnt signaling is crucial for the initial segregation between the thalamus and prethalamus. It has been described that Wntb8 signal initially spreads over the alar plate of p2 and p3. Progressively, it becomes restricted to the prospective ZLI, probably because of repression from flanking domains expressing Fez, Lrrn1, Otx2, and Irx3 just before the onset of Shh expression. Later ZLI cells coexpress Wnt8b and Shh, suggesting a positive interaction between both morphogenes. Indeed, Wnt signaling opens a permissive territory for the expression of Shh at the ZLI through repressing Gli3 expression, which negatively regulates Shh, and upregulating L-fng (Martinez-Ferre et al., 2013). Another member of the Wnt family that interacts with the Shh pathway is Wnt1, which is involved in patterning of the diencephalon, especially in setting the DV axis. Wnt1 is expressed in the dorsal midline of the posterior part of p2, p1, and mesencephalon (Martinez-Ferre et al., 2013). Bmp4 from the roof plate induces Wnt1 expression. Although Wnts are usually described as “dorsalizing” signals, Wnt1 seems to act as a “ventralizing” agent. The basal plate and the corresponding Shh expression are compressed when Wnt1 is absent and enlarged when Wnt1 is overexpressed (NavarroGarberi et al., 2016). In Wnt1 knockout mice, the expression of Shh in basal plate gets narrower and the dorsal progression through the alar plate that defines the ZLI seems to be compromised as well. On the other hand, gain-of-function experiments overexpressing Wnt1 show dorsal shift of the basal expression of Shh and a dorsal expansion of the Shh spike within ZLI reaching the level of the roof plate. A dorsal expansion of Shh expression has been also described when Fgf8 activity is reduced suggesting opposite effects of Wnt1 and Fgf8 in the regulation of Shh expression (Martinez-Ferre and Martinez, 2009). The position of alar/basal boundary may be mediated by the alar regulation of Gli3.

4.3.2.3 Fibroblast growth factor Fgf signaling has been described as an important agent during diencephalic development in concert to Shh and Wnt activity; however, its role needs to be fully scrutinized yet (Hagemann and Scholpp, 2012; Nakagawa and Shimogori, 2012). In mice, the Fgf8 morphogenetic gradient originated in the dorsalmost region of the ZLI influences AP patterning of the diencephalon because it controls diencephalic-specific genes and diencephalic dorsal regionalization. Overexpression of Fgf8 induces an expansion of the Th-r domain and a complementary shrinkage of Th-c (Kataoka and Shimogori, 2008). In the same line, interfering with Fgf signaling provokes the absence of Ascl1 subpopulation of GABAergic interneurons derived from the Th-r progenitor domain. On the contrary, enhancing Fgf activity shifts caudally the position of thalamic nuclei derived from the Th-r domain (VPM, dLGN, and others). Downregulation of Fgf8 expression causes a reduction of Gbx2 expression in Th-c that likely affects glutamatergic differentiation and also provokes an abnormal development of the habenula and pineal gland (Martinez-Ferre and Martinez, 2009; Schmidt and Pasterkamp, 2017). The competence of p2 and p1 territories to react to Fgf morphogenetic signals seems different from that of p3, highlighting the spatial segregation imposed by the ZLI and the role of combinatorial expression of TFs to guide progenitors toward different fates (Robertshaw et al., 2013).

Patterning of thalamus Chapter | 4

79

4.3.3 Transcription factor control for neuronal identity Although extracellular signals such as Shh, Fgfs, and Wnts have a prominent role in regionalization of the diencephalon, downstream these morphogenes, a large pool of TFs is required to regulate the formation of the heterogeneous thalamic nuclei (Fig. 4.4; Hagemann and Scholpp, 2012). Otx genes are key factors in brain regionalization and patterning by means of regulating cell identity and fate and may function as activator or repressor genes depending on the contextual molecular partners expressed in different territories. Otx2 is expressed in the ventricular zone of the diencephalon, and in mice, the inactivation of Otx2 from E10.5 onward generates a switch in the differentiation program of Th-c progenitors from a glutamatergic to GABAergic fate. Mice lacking Otx2 activate the thalamic expression of normally pretectal genes Lim1, Pax3, and Pax7 and repress the expression of Gbx2 in the mantle zone of the thalamus (Puelles et al., 2006). Otx2 imposes thalamic identity because its absence derepresses Pax3 and Pax7 expression in thalamic progenitors and later their postmitotic progeny becomes Lim1-positive cells. In the presence of Otx2, the normal terminal differentiation of these

PT

Th-c

Th-r ZLI

Olig3

Nkx2.2

Ascl1

Hes

Gsx2

Shh

Neurog1/2

Olig2

Dbx1

Sox2

Gbx2

Lhx2

Lhx9

Sox2

Neurog2

Dlx

Arx

Islet1

Gata2

Gata3

PTh

PT

Th-c

Th-r ZLI

GABA

Glut

Sox14

Six3

PTh

FIGURE 4.4 Diencephalic progenitor domains and transcription factors expressed in their ventricular zone by progenitor cells (top) and transcription factors expressed in early postmitotic cells (bottom). Excitatory and inhibitory fates are depicted with blue bars. PT, pretectum; PTh, prethalamus; Th-c, caudal thalamic progenitor domain; Th-r, rostral thalamic progenitor domain; ZLI, zona limitans intrathalamica.

80

PART | I Induction and patterning of the CNS and PNS

progenitors is onto glutamatergic neurons; however, VGlut2 in not detected when Otx2 in inactivated. In fact, neurons derived from Otx2-negative progenitors acquire a GABAergic fate suggesting that Otx2 in normal conditions suppress the differentiation program for inhibitory cells. Noteworthy, as mentioned before, these inhibitory neurons acquire a profile of gene expression similar to that of the pretectum, instead of those of the prethalamus (where Dlx1 and Dlx5 are expressed) or basal telencephalon. The acquisition of a GABAergic fate correlates with the gradual loss of Neurog2 from E12 and the corresponding activation of Mash1. Neurog2 is a glutamatergic differentiation factor widely expressed in the thalamus up to the lateral border occupied by postmitotic cells (Vue et al., 2007). Actually, glutamatergic differentiation, as well as GABAergic repression, is unfolded in those regions of the forebrain where Otx2 and Neurogs are co-expressed. The opposite scenario comes out in regions that only express Otx2 (pretectum) or none of them (basal telencephalon). Thalamic precursors that just exit the cell cycle, circa E10.5 in mice, express Gbx2, which is another important TF in thalamic development. Gbx2 is detected in postmitotic cells just below the pial surface. All thalamic precursors are derived from the Gbx2 lineage; however, its onset and offset are variable among the distinct thalamic nuclei (Chen et al., 2009; Li et al., 2012), suggesting a role for Gbx2 in nuclear organization within the thalamus. Some nuclei like VP and dLG show not only an early onset of expression at E10.5 but also a short duration because Gbx2 is downregulated soon after differentiation. An early onset is detected in MGv and LP as well, but Gbx2 expression persists into postnatal stages. Thus, the first wave of Gbx2 expression gives rise to most of the principal sensory nuclei. The second wave starts by E10.5eE11.5 and gives rise to many association nuclei in which Gbx2 expression is persistent (including MD, CL, paracentral, and CM nuclei) or transient (lateral, ventral medial basal nuclei). Finally, at E15.5, the third wave generates anteromedial nuclei, including anterior medial, paraventricular, and paratenial nuclei (Chen et al., 2009). Gbx2 seems to establish thalamic identity by repressing epithalamic fate. Deletion of Gbx2 provokes that presumptive thalamic cells acquire a partial habenular molecular identity, which is consistently followed by a disrupted cortical projection (Chen et al., 2009). The habenula and the thalamus are remarkably different structures regarding their function and connectivity. However, both progenitor pools share many molecular markers and only later, during neurogenesis, thalamic and habenular fates become distinguishable when genetic profiles of neural precursors are compared (Mallika et al., 2015). The fate switch in Gbx2-deficient mice is apparently mediated by derepressing Irx1 in the presumed Gbx2-expressing cellsdIrx1 is normally present throughout the p2 domain but is downregulated in the thalamic area when cells exit cell cycle and initiate Gbx2 expression. Subsequently, markers of epithalamic postmitotic neural precursors (Nrp2, Pou4f1, and Robo3) are ectopically expressed and thalamic markers dependent on Gbx2 are lost (including ROR-alpha, Chst1, Glr1, Cd47, Slc18a2, Gas7, Hs6st2, Epha3, Prokr2, and Id4). Gbx2 also regulates the axon responsiveness to guidance cues in cortical-projecting neurons of the thalamus. The dynamic temporal pattern of expression of Gbx2 may be related to a mechanism that topographically sorts thalamocortical axons from different nuclei. Gbx2 upregulates Lhx2 by repressing Lmo3, who is a potential negative regulator of Lhx2; at the same time, Lhx2 inhibits Lhx9 (Chatterjee et al., 2012). The regulation of the levels of Lhx2/Lhx9 is critical for Robo1/ Robo2 expression and thalamic development because Slit-Robo signaling exerts a key function in thalamocortical axon pathfinding (López-Bendito et al., 2007; Bielle et al., 2011). Actually, it has been shown that Lhx2 is dynamically regulated in different cortical-projecting neurons (Marcos-Mondéjar et al., 2012). Although detected in most thalamic neurons when they are born, Lhx2 is downregulated in some of them while migrating toward their final position. Specifically, Lhx2 shows weak levels of expression in VP and dLG nuclei and high levels in MGv cells (Marcos-Mondéjar et al., 2012; Nakagawa and O’Leary, 2001). Thus, specific levels of Gbx2, Lhx2, and Lhx9 may contribute to establish the spatial organization of thalamic projections required to properly connect with their corresponding cortical targets. Pax6 is another key factor in thalamic development. Initially expressed throughout the alar forebrain neuroepithelium, progenitor cells in the thalamus progressively downregulate Pax6 expression during neurogenesis (E10-E13) becoming undetectable in postmitotic thalamic neurons. Before fading away, Pax6 is expressed in a gradient through the thalamic progenitor layer from low levels in Th-r to high levels in Th-c. Prethalamic progenitors and postmitotic cells retain Pax6 expression; accordingly, these cells loss their prethalamic identity in Pax6-deficient mice. The role of Pax6 during thalamic development is manifold. The fate of regions rostral and caudal to the ZLI is defined before ZLI forms. Different combinations of TFs, including Pax6, demarcate prospective diencephalic compartments. While the expression of Irx3 and Pax6 prepatterns the neuroepithelium to become caudal diencephalic structures, Pax6 together with Six3 but not Irx3 demarcates the prospective prethalamus. The combinatorial expression of TFs confers specific competences to different diencephalic compartments that correlate with the asymmetric response they elicit to organizing signals such as Fgf8 and Shh (Kobayashi et al., 2002; Robertshaw et al., 2013). Pax6 also exerts an early function in diencephalic patterning by constraining cell-autonomously Shh expression to the ZLI. Then, there is an indirect contribution of Pax6 to proper AP regionalization of the diencephalon as explained before in this chapter. In the absence of Pax6, the ZLI and its corresponding Shh expression domain are enlarged causing

Patterning of thalamus Chapter | 4

81

misspecification in surrounding tissues (Pratt et al., 2000; Caballero et al., 2014). In the posterior side of the ZLI, the thalamus expands dorsally and the pretectum anteriorly at the expense of the epithalamus; actually, the habenula is lost (Chatterjee et al., 2014). Conversely, loss of Shh enlarges the habenula at the expense of the thalamus. Regarding the thalamic region, the main consequence of Pax6 deficits is the expansion of the Th-r marker Nkx2.2 expression over the Thc, a domain localized further away from the ZLI. Within Th-c, presumptive cortical-projecting neurons show a partial phenotype mixing rostral and caudal thalamic markers. Importantly, despite their partial mispatterning, they do extend axons toward extra-thalamic territories. However, they are misrouted toward the hypothalamus instead of targeting cerebral cortex (Kawano et al., 1999; Jones et al., 2002). This fact suggests that Pax6 is somehow implicated in mechanisms responsible for thalamocortical axon guidance through the telencephalon. Using chimeric models where normal cells surround cells lacking Pax6 in the thalamus, the level of expression of the Slit receptor Robo2 is reduced in mutant cells; however, this is not sufficient to cause misrouting toward the hypothalamus (a Slit-releasing zone) as in constitutive mutant mice because their axons target the cortex properly (Clegg et al., 2015). Thus, instead of an intrinsic effect, Pax6 may act indirectly on the different territories traversed by the thalamic axons allowing their proper pathfinding toward the cerebral cortex. The TFs bar homeobox-like 1 and 2 (Barhl1 and Barhl2) are expressed in the proliferative zone of p1, p2 (except in Th-r), ZLI, and prospective hypothalamus in the developing forebrain of mouse between E10 and E12 (Suzuki-Hirano et al., 2011). By E13, Barhl2 vanishes from prethalamic regions but remains high in thalamus and gradually fades through the pretectum. Because Pax6 expression profile precedes and almost complements that of Barhl2 in the diencephalon, it is suggested that Pax6 direct or indirectly represses Barhl2 expression (Parish et al., 2016). Thus, Barhl2 could mediate some of the proposed functions for Pax6, acting upstream of Shh could be critical for early steps in diencephalic patterning such as defining p1/p2 fates. Removing Barhl2 in mice pushes p2 progenitors into a partial p1 identity; p3 remains intact (Ding et al., 2017). Final cell division in progenitor cells is regulated by Hes/Her family of TFs and takes place between E10 and E15. In mice, Hes family has been demonstrated to repress proneural genes critical for thalamic nuclei development such as Neurogs and Ascl1 (Baek et al., 2006; Scholpp et al., 2009). In fact, high levels of Hes1 are found in boundaries (ZLI and isthmus), roof, and floor plate where neurogenesis is slow or absent. In the absence of members of the Hes family, ectopic neurogenesis is elicited by upregulation of proneural genes in nonboundary and boundary regions and also organizing centers are not properly formed. In Hes1/Hes3/Hes5 triple knockouts, Neurog2 and Ascl1 are found in regions presumptively negative for them. Ascl1 plays different roles in the compartments of the diencephalon. In p1 progenitors of the ventricular zone, Ascl1 induces the GABAergic pathway of differentiation by repressing Neurog2. Similarly, in the caudal domain of p2, Otx2 must repress Ascl1 to ensure the glutamatergic fate (Puelles et al., 2006). In the rostral domain of p2, Ascl1 participates in the molecular mechanism that establishes inhibitory fate but is not sufficient, probably because of the presence of members of the Hes family. In p3, Ascl1 promotes cell cycle exit to initiate the differentiation phase into GABAergic neurons. Closely related to Hes family, Helt is a transcription factor that has an expression pattern similar to Ascl1 and also is involved in regulating Th-r neuronal identity. Actually, Ascl1 and Helt cooperate to regulate thalamic progenitor specification establishing the Th-r identity. While Th-c identity takes over Th-r domains in Shh mutants, different results were obtained manipulating Ascl1 and Helt. Neither Ascl1 nor Helt single-knockout mice showed downregulation of TFs that characterized Th-r such as Tal2, Tal1, Gata2, Gata3, and Six3. Remarkably, p3 undergoes a GABAergic-to-glutamatergic switch. However, when both Ascl1 and Helt are absent, the Th-r developmental program steers into a more rostral differentiation pathway where progenitors express prethalamic markers such as Dlx2, Dlx5, and Arx at E12.5 (Song et al., 2015). Consistently, at E16.5, IGL and vLG cells derived from Th-r show a complete lack of expression Gata3 and Tal1, markers that otherwise are normally expressed in these nuclei. Thus, it has been proposed that the joint action of Ascl1 and Helt contributes to the specification of Th-r because their expression represses Dlx2 and Dlx5 repression, which means escaping from a prethalamic fate. If they are expressed, the Th-r differentiation sequence, which starts with the expression Gata2/Tal2 and then follows with Gata3/Tal1/Six3/Sox14/Gad1, is blocked. Conversely, loss of function of Dlx1 and Dlx2 transforms the prethalamus into a region similar to Th-r due to the lack of inhibition upon Gata2 (Delogu et al., 2012; Sellers et al., 2014). Gata2 seems to exert functions downstream Ascl1 and Helt, its expression is activated in early postmitotic GABAergic precursor cells in the ventricular zone of p1 and Th-r, and no expression is detected in p3 although Ascl1 is present. At E12.5, Gata2 is expressed in the ventricular zone of Th-r together with Tal2 and expands also to the intermediate and mantle zone; Gata3, Tal1, Sox14, and Gad1-positive cells reside only in the intermediate and mantle zone at this stage (Virolainen et al., 2012). The Th-c domain of p2 is absolutely devoid of any of these markers. At E18.5, Gata2 and Tal2 are downregulated in thalamic GABAergic nuclei, whereas Gata3 and Tal1 remain expressed overlapping with Nkx2.2 in IGL (in NPY and Penk1 subpopulations of GABAergic neurons) and the lateral vLG (Zhao et al., 2008; Virolainen et al., 2012).

82

PART | I Induction and patterning of the CNS and PNS

List of acronyms and abbreviations AP Anteroposterior Arx Aristaless-related gene Ascl1 Achaete-scute family bHLH transcription factor 1 Barhl1/2 BarH-like homeobox 1 and 2 Bmp Bone morphogenic protein Cacna2d1 Calcium voltage-gated channel auxiliary subunit alpha2delta 1 Dbx1 Developing brain homeobox 1 dLG Dorsal lateral nucleus of the thalamus Dlx1/2/5/6 Distal-less homeobox 1, 2, 5, and 6 DV Dorsoventral E9eE18 9 to 18 embryonic days postconception in mouse Ebf1 Early B cell factor 1 Emx1/2 Empty spiracles homeobox 1 and 2 Eph Ephrin Fezf1/2 FEZ family zinc finger 1 and 2 Fgf3/8/15 Fibroblast growth factor 3, 8, and 15 Foxd1 Forkhead box D1 Foxg1 Forkhead box G1 GABA amma-Aminobutyric acid Gad1 Glutamate decarboxylase 1 Gata2/3 GATA binding protein 2 and 3 Gbx2 Gastrulation brain homeobox 2 Gdf10 Growth differentiation factor 10, a member of the Bmp family Gli2/3 GLI family zinc finger 2 and 3 Gsh2 Genomic screen homeobox 2, currently known as GS homeobox 2 Helt Helt BHLH transcription factor Hes1 Hes family BHLH transcription factor 1 HH stages Hamburger Hamilton chronological stages commonly used in chick developmental staging IGL Intergeniculate leaflet Irx1/3: Iroquois homeobox 1 and 3 Islet1 ISL LIM homeobox 1 L-fng Lunatic fringe (Drosophila) homolog Lef Lymphoid enhancer binding factor 1 Lhx1/2/9 LIM homeobox 1, 2, and 9 LP Lateral posterior nucleus of the thalamus Lrp6 Low-density lipoprotein receptor-related protein 6 Lrrn1 Leucine-rich repeat neuronal 1 MADM Mosaic analysis with double markers Mash1 Mammalian achaete-scute homolog 1 MGv Medial geniculate body ventral subdivision Neurog1/2 Neurogenin 1 and 2 Nkx2.2 NK2 homeobox 2 NPY Neuropeptide Y Olig2/3 Oligodendrocyte transcription factor 2 and 3 Otx2 Orthodentricle homeobox 2 P1, p2, and p3 Prosomere 1, prosomere 2, and prosomere 3 Pax3/6 Paired box 3 and 6 Penk1 Proenkephalin 1 Prox1 Prospero homeobox 1 PTh-c Caudal progenitor domain of the prethalamus PTh-r Rostral progenitor domain of the prethalamus PThE Prethalamic eminence ROR-alpha Retinoic acids receptor-related orphan receptor Sfrp2 Secreted frizzled-related protein 2 Shh Sonic hedgehog Six3/6 SIX homeobox 3 and 6 Sox2/14 SRY-Box 2 and 1b

Patterning of thalamus Chapter | 4

83

Tal1/2 TAL BHLH transcription factor 1 and 2 Tbr1 T-Box, brain 1 Tcf4 Transcription factor 4 TF Transcription factor Th-c Caudal progenitor domain of the thalamus Th-r Rostral progenitor domain of the thalamus vGL Ventral geniculate nucleus VGlut2 Vesicular glutamate transporter 2 Wnt1/3/3a/7b/8b Wingless-int family ZLI Zona limitans intrathalamica

References Alvarez-Bolado, G., Rosenfeld, M.G., Swanson, L.W., 1995. Model of forebrain regionalization based on spatiotemporal patterns of POU-III homeobox gene expression, birthdates, and morphological features. J. Comp. Neurol. 355 (2), 237e295. https://doi.org/10.1002/cne.903550207. Araki, I., Nakamura, H., 1999. Engrailed defines the position of dorsal di-mesencephalic boundary by repressing diencephalic fate. Development 126 (22), 5127e5135. Baek, J.H., Hatakeyama, J., Sakamoto, S., Ohtsuka, T., Kageyama, R., 2006. Persistent and high levels of Hes1 expression regulate boundary formation in the developing central nervous system. Development 133 (13), 2467e2476. https://doi.org/10.1242/dev.02403. Bielle, F., Marcos-Mondéjar, P., Keita, M., Mailhes, C., Verney, C., Ba-Charvet, K.N., Tessier-Lavigne, M., López Bendito, G., Garel, S., 2011. Slit2 activity in the migration of guidepost neurons shapes thalamic projections during development and evolution. Neuron 69 (6), 1085e1098. https:// doi.org/10.1016/j.neuron.2011.02.026. Bluske, K.K., Kawakami, Y., Koyano-Nakagawa, N., Nakagawa, Y., 2009. Differential activity of Wnt/beta-catenin signaling in the embryonic mouse thalamus. Dev. Dynam. 238 (12), 3297e3309. https://doi.org/10.1002/dvdy.22167. Bluske, K.K., Yia Vue, T., Kawakami, Y., Taketo, M.M., Yoshikawa, K., Johnson, J.E., Nakagawa, Y., 2012. B-catenin signaling specifies progenitor cell identity in parallel with Shh signaling in the developing mammalian thalamus. Development 139 (15), 2692e2702. https://doi.org/10.1242/ dev.072314. Borostyánkoi-Baldauf, Z., Herczeg, L., 2002. Parcellation of the human pretectal complex: a chemoarchitectonic reappraisal. Neuroscience 110 (3), 527e540. Braun, M.M., Etheridge, A., Bernard, A., Robertson, C.P., Roelink, H., 2003. Wnt signaling is required at distinct stages of development for the induction of the posterior forebrain. Development 130 (23), 5579e5587. https://doi.org/10.1242/dev.00685. Bulfone, A., Frohman, M.A., Martin, G.R., Rubenstein, J.L., 1993. Spatially restricted expression of Dlx-1, Dlx-2 (Tes-1), Gbx-2, and Wnt- 3 in the embryonic day 12.5 mouse forebrain defines potential transverse and longitudinal segmental boundaries. J. Neurosci. 13 (7), 3155e3172. https:// doi.org/10.1523/JNEUROSCI.13-07-03155.1993. Caballero, I.M., Manuel, M.N., Molinek, M., Quintana-Urzainqui, I., Mi, D., Shimogori, T., Price, D.J., 2014. Cell-autonomous repression of Shh by transcription factor Pax6 regulates diencephalic patterning by controlling the central diencephalic organizer. Cell Rep. 8 (5), 1405e1418. https:// doi.org/10.1016/j.celrep.2014.07.051. Chatterjee, M., Guo, Q., Weber, S., Scholpp, S., James, Y.L., 2014. Pax6 regulates the formation of the habenular nuclei by controlling the temporospatial expression of Shh in the diencephalon in vertebrates. BMC Biol. 12 (1), 13. https://doi.org/10.1186/1741-7007-12-13. Chatterjee, M., Li, K., Chen, L., Xu, M., Guo, Q., Gan, L., James, Y.H.L., 2012. Gbx2 regulates thalamocortical axon guidance by modifying the LIM and Robo codes. Development 139 (24), 4633e4643. https://doi.org/10.1242/dev.086991. Chen, L., Guo, Q., James, Y.H.L., 2009. Transcription factor Gbx2 acts cell-nonautonomously to regulate the formation of lineage-restriction boundaries of the thalamus. Development 136 (8), 1317e1326. https://doi.org/10.1242/dev.030510. Clegg, J.M., Li, Z., Molinek, M., Caballero, I.M., Manuel, M.N., Price, D.J., 2015. Pax6 is required intrinsically by thalamic progenitors for the normal molecular patterning of thalamic neurons but not the growth and guidance of their axons. Neural Dev. 10 (1), 26. https://doi.org/10.1186/s13064-0150053-7. Dean, B.J., Erdogan, B., T Gamse, J., Wu, S.-Y., 2014. Dbx1b defines the dorsal habenular progenitor domain in the zebrafish epithalamus. Neural Dev. 9 (1), 20. https://doi.org/10.1186/1749-8104-9-20. Delogu, A., Sellers, K., Zagoraiou, L., Bocianowska-Zbrog, A., Mandal, S., Guimera, J., Rubenstein, J.L.R., Sugden, D., Jessell, T., Lumsden, A., 2012. Subcortical visual shell nuclei targeted by ipRGCs develop from a sox14þ-GABAergic progenitor and require Sox14 to regulate daily activity rhythms. Neuron 75 (4), 648e662. https://doi.org/10.1016/j.neuron.2012.06.013. Ding, Q., Zheng, D., Liang, G., 2017. Barhl2 determines the early patterning of the diencephalon by regulating Shh. Mol. Neurobiol. 54 (6), 4414e4420. https://doi.org/10.1007/s12035-016-0001-5. Epstein, D.J., 2012. Regulation of thalamic development by sonic hedgehog. Front. Neurosci. 6, 57. https://doi.org/10.3389/fnins.2012.00057. Ferran, J.L., Dutra de Oliveira, E., Merchán, P., Sandoval, J.E., Sánchez-Arrones, L., Martinez-de-la-Torre, M., Puelles, L., 2009. Genoarchitectonic profile of developing nuclear groups in the chicken pretectum. J. Comp. Neurol. 517 (4), 405e451. https://doi.org/10.1002/cne.22115. Ferran, J.L., Sánchez-Arrones, L., Sandoval, J.E., Puelles, L., 2007. A model of early molecular regionalization in the chicken embryonic pretectum. J. Comp. Neurol. 505 (4), 379e403. https://doi.org/10.1002/cne.21493.

84

PART | I Induction and patterning of the CNS and PNS

Ferran, J.L., Sánchez-Arrones, L., Bardet, S.M., Sandoval, J.E., Martínez-de-la-Torre, M., Puelles, L., 2008. Early pretectal gene expression pattern shows a conserved anteroposterior tripartition in mouse and chicken. Brain Res. Bull. 75 (2), 295e298. https://doi.org/10.1016/j.brainresbull.2007.10.039. Figdor, M.C., Stern, C.D., 1993. Segmental organization of embryonic diencephalon. Nature 363 (6430), 630e634. https://doi.org/10.1038/363630a0. Fode, C., Ma, Q., Casarosa, S., Ang, S.L., Anderson, D.J., Guillemot, F., 2000. A role for neural determination genes in specifying the dorsoventral identity of telencephalic neurons. Genes Dev. 14 (1), 67e80. García-López, R., Vieira, C., Echevarria, D., Martinez, S., 2004. Fate map of the diencephalon and the zona limitans at the 10-somites stage in chick embryos. Dev. Biol. 268 (2), 514e530. https://doi.org/10.1016/j.ydbio.2003.12.038. Hagemann, A.I.H., Scholpp, S., 2012. The tale of the three brothers - Shh, Wnt, and Fgf during development of the thalamus. Front. Neurosci. 6, 76. https://doi.org/10.3389/fnins.2012.00076. Hashimoto-Torii, K., Motoyama, J., Hui, C.-C., Kuroiwa, A., Nakafuku, M., Shimamura, K., 2003. Differential activities of sonic hedgehog mediated by Gli transcription factors define distinct neuronal subtypes in the dorsal thalamus. Mech. Dev. 120 (10), 1097e1111. Hatini, V., Tao, W., Lai, E., 1994. Expression of winged helix genes, BF-1 and BF-2, define adjacent domains within the developing forebrain and retina. J. Neurobiol. 25 (10), 1293e1309. https://doi.org/10.1002/neu.480251010. Herrera, E., Marcus, R., Li, S., Williams, S.E., Erskine, L., Lai, E., Mason, C., 2004. Foxd1 is required for proper formation of the optic chiasm. Development 131 (22), 5727e5739. https://doi.org/10.1242/dev.01431. Hikosaka, O., 2010. The habenula: from stress evasion to value-based decision-making. Nat. Rev. Neurosci. 11 (7), 503e513. https://doi.org/10.1038/ nrn2866. Hirata, T., Nakazawa, M., Muraoka, O., Nakayama, R., Suda, Y., Hibi, M., 2006. Zinc-finger genes Fez and Fez-like function in the establishment of diencephalon subdivisions. Development 133 (20), 3993e4004. https://doi.org/10.1242/dev.02585. Ishibashi, M., McMahon, A.P., 2002. A sonic hedgehog-dependent signaling relay regulates growth of diencephalic and mesencephalic primordia in the early mouse embryo. Development 129 (20), 4807e4819. Jeong, Y., Dolson, D.K., Waclaw, R.R., Matise, M.P., Sussel, L., Campbell, K., Kaestner, K.H., Epstein, D.J., 2011. Spatial and temporal requirements for sonic hedgehog in the regulation of thalamic interneuron identity. Development 138 (3), 531e541. https://doi.org/10.1242/dev.058917. Jones, E.G., 2012. In: Jones, E.G. (Ed.), The Thalamus. Springer Science & Business Media, Boston, MA. https://doi.org/10.1007/978-1-4615-1749-8. Jones, E.G., Rubenstein, J.L.R., 2004. Expression of regulatory genes during differentiation of thalamic nuclei in mouse and monkey. J. Comp. Neurol. 477 (1), 55e80. https://doi.org/10.1002/cne.20234. Jones, L., López Bendito, G., Gruss, P., Stoykova, A., Molnár, Z., 2002. Pax6 is required for the normal development of the forebrain axonal connections. Development 129 (21), 5041e5052. Juraver-Geslin, H.A., Gómez-Skarmeta, J.L., Durand, B.C., 2014. The conserved barH-like homeobox-2 gene Barhl2 acts downstream of orthodentricle-2 and together with iroquois-3 in establishment of the caudal forebrain signaling center induced by sonic hedgehog. Dev. Biol. 396 (1), 107e120. https://doi.org/10.1016/j.ydbio.2014.09.027. Kataoka, A., Shimogori, T., 2008. Fgf8 controls regional identity in the developing thalamus. Development 135 (17), 2873e2881. https://doi.org/10.1242/ dev.021618. Kawano, H., Fukuda, T., Kubo, K., Horie, M., Uyemura, K., Takeuchi, K., Osumi, N., Eto, K., Kawamura, K., 1999. Pax-6 is required for thalamocortical pathway formation in fetal rats. J. Comp. Neurol. 408 (2), 147e160. Kiecker, C., Lumsden, A., 2004. Hedgehog signaling from the ZLI regulates diencephalic regional identity. Nat. Neurosci. 7 (11), 1242e1249. https:// doi.org/10.1038/nn1338. Kobayashi, D., Kobayashi, M., Matsumoto, K., Ogura, T., Nakafuku, M., Shimamura, K., 2002. Early subdivisions in the neural plate define distinct competence for inductive signals. Development 129 (1), 83e93. Larsen, C.W., Zeltser, L.M., Lumsden, A., 2001. Boundary formation and compartition in the avian diencephalon. J. Neurosci. 21 (13), 4699e4711. Li, H.S., Yang, J.M., Jacobson, R.D., Pasko, D., Sundin, O., 1994. Pax-6 is first expressed in a region of ectoderm anterior to the early neural plate: implications for stepwise determination of the lens. Dev. Biol. 162 (1), 181e194. https://doi.org/10.1006/dbio.1994.1077. Li, K., Zhang, J., James, Y.H.L., 2012. Gbx2 plays an essential but transient role in the formation of thalamic nuclei. Edited by Zhang, X. PLoS One 7 (10), e47111. https://doi.org/10.1371/journal.pone.0047111. Lim, Y., Golden, J.A., 2007. Patterning the developing diencephalon. Brain Res. Rev. 53 (1), 17e26. https://doi.org/10.1016/j.brainresrev.2006.06.004. López-Bendito, G., Flames, N., Ma, L., Fouquet, C., Di Meglio, T., Chedotal, A., Tessier-Lavigne, M., Marin, O., 2007. Robo1 and Robo2 cooperate to control the guidance of major axonal tracts in the mammalian forebrain. J. Neurosci. 27 (13), 3395e3407. https://doi.org/10.1523/JNEUROSCI.460506.2007. Mallika, C., Guo, Q., James, Y.H.L., 2015. Gbx2 is essential for maintaining thalamic neuron identity and repressing habenular characters in the developing thalamus. Dev. Biol. 407 (1), 26e39. https://doi.org/10.1016/j.ydbio.2015.08.010. Marcos-Mondéjar, P., Peregrín, S., James, Y.L., Carlsson, L., Tole, S., López Bendito, G., 2012. The Lhx2 transcription factor controls thalamocortical axonal guidance by specific regulation of Robo1 and Robo2 receptors. J. Neurosci. 32 (13), 4372e4385. https://doi.org/10.1523/JNEUROSCI.585111.2012. Martinez-Ferre, A., Martinez, S., 2012. Molecular regionalization of the diencephalon. Front. Neurosci. 6 (May), 73. https://doi.org/10.3389/ fnins.2012.00073. Martinez-Ferre, A., Martinez, S., 2009. The development of the thalamic motor learning area is regulated by Fgf8 expression. J. Neurosci. 29 (42), 13389e13400. https://doi.org/10.1523/JNEUROSCI.2625-09.2009.

Patterning of thalamus Chapter | 4

85

Martinez-Ferre, A., Lloret-Quesada, C., Prakash, N., Wurst, W., Rubenstein, J.L.R., Martinez, S., 2016. Fgf15 regulates thalamic development by controlling the expression of proneural genes. Brain Struct. Funct. 221 (6), 3095e3109. https://doi.org/10.1007/s00429-015-1089-5. Martinez-Ferre, A., Navarro-Garberi, M., Bueno, C., Martinez, S., 2013. Wnt signal specifies the intrathalamic limit and its organizer properties by regulating Shh induction in the alar plate. J. Neurosci. 33 (9), 3967e3980. https://doi.org/10.1523/JNEUROSCI.0726-12.2013. Matsunaga, E., Araki, I., Nakamura, H., 2000. Pax6 defines the di-mesencephalic boundary by repressing En1 and Pax2. Development 127 (11), 2357e2365. Matsunaga, E., Araki, I., Nakamura, H., 2001. Role of Pax3/7 in the tectum regionalization. Development 128 (20), 4069e4077. Nakagawa, Y., O’Leary, D.D., 2001. Combinatorial expression patterns of LIM-homeodomain and other regulatory genes parcellate developing thalamus. J. Neurosci. 21 (8), 2711e2725. Nakagawa, Y., O’Leary, D.D.M., 2003. Dynamic patterned expression of orphan nuclear receptor genes RORalpha and RORbeta in developing mouse forebrain. Dev. Neurosci. 25 (2), 234e244. https://doi.org/10.1159/000072271. Nakagawa, Y., Shimogori, T., 2012. Diversity of thalamic progenitor cells and postmitotic neurons. Eur. J. Neurosci. 35 (10), 1554e1562. https://doi.org/ 10.1111/j.1460-9568.2012.08089.x. Navarro-Garberi, M., Bueno, C., Martinez, S., 2016. Wnt1 signal determines the patterning of the diencephalic dorso-ventral axis. Brain Struct. Funct. 221 (7), 3693e3708. https://doi.org/10.1007/s00429-015-1126-4. Nomura, T., Kawakami, A., Fujisawa, H., 1998. Correlation between tectum formation and expression of two PAX family genes, PAX7 and PAX6, in avian brains. Dev. Growth Differ. 40 (5), 485e495. Parish, E.V., Mason, J.O., Price, D.J., 2016. Expression of Barhl2 and its relationship with Pax6 expression in the forebrain of the mouse embryo. BMC Neurosci. 17 (1), 76. https://doi.org/10.1186/s12868-016-0311-6. Pera, E.M., Kessel, M., 1997. Patterning of the chick forebrain anlage by the prechordal plate. Development 124 (20), 4153e4162. Pratt, T., Vitalis, T., Warren, N., Edgar, J.M., Mason, J.O., Price, D.J., 2000. A role for Pax6 in the normal development of dorsal thalamus and its cortical connections. Development 127 (23), 5167e5178. Price, D.J., Clegg, J., Oliver Duocastella, X., Willshaw, D., Pratt, T., 2012. The importance of combinatorial gene expression in early mammalian thalamic patterning and thalamocortical axonal guidance. Front. Neurosci. 6, 37. https://doi.org/10.3389/fnins.2012.00037. Puelles, E., Acampora, D., Gogoi, R., Tuorto, F., Papalia, A., Guillemot, F., Ang, S.L., Simeone, A., 2006. Otx2 controls identity and fate of glutamatergic progenitors of the thalamus by repressing GABAergic differentiation. J. Neurosci. 26 (22), 5955e5964. https://doi.org/10.1523/JNEUROSCI.109706.2006. Puelles, L., 2001. Brain segmentation and forebrain development in amniotes. Brain Res. Bull. 55 (6), 695e710. Puelles, L., Amat, J.A., Martinez-de-la-Torre, M., 1987. Segment-related, mosaic neurogenetic pattern in the forebrain and mesencephalon of early chick embryos: I. Topography of AChE-positive neuroblasts up to stage HH18. J. Comp. Neurol. 266 (2), 247e268. https://doi.org/10.1002/ cne.902660210. Puelles, L., Martinez, S., 2013. Patterning of the Diencephalon. Patterning and Cell Type Specification in the Developing CNS and PNS. Elsevier. https:// doi.org/10.1016/b978-0-12-397265-1.00048-4. Puelles, L., Rubenstein, J.L., 1993. Expression patterns of homeobox and other putative regulatory genes in the embryonic mouse forebrain suggest a neuromeric organization. Trends Neurosci. 16 (11), 472e479. https://doi.org/10.1016/0166-2236(93)90080-6. Puelles, L., Rubenstein, J.L.R., 2015. A new scenario of hypothalamic organization: rationale of new hypotheses introduced in the updated prosomeric model. Front. Neuroanat. 9, 27. https://doi.org/10.3389/fnana.2015.00027. Puelles, L., Rubenstein, J.L.R., 2003. Forebrain gene expression domains and the evolving prosomeric model. Trends Neurosci. 26 (9), 469e476. https:// doi.org/10.1016/S0166-2236(03)00234-0. Robertshaw, E., Matsumoto, K., Lumsden, A., Kiecker, C., 2013. Irx3 and Pax6 establish differential competence for shh-mediated induction of GABAergic and glutamatergic neurons of the thalamus. Proc. Natl. Acad. Sci. U.S.A. 110 (41), E3919eE3926. https://doi.org/10.1073/ pnas.1304311110. Schmidt, E.R.E., Pasterkamp, R.J., 2017. The molecular mechanisms controlling morphogenesis and wiring of the habenula. Pharmacol. Biochem. Behav. 162, 29e37. https://doi.org/10.1016/j.pbb.2017.08.008. Scholpp, S., Delogu, A., Gilthorpe, J., Peukert, D., Schindler, S., Lumsden, A., 2009. Her6 regulates the neurogenetic gradient and neuronal identity in the thalamus. Proc. Natl. Acad. Sci. U.S.A. 106 (47), 19895e19900. https://doi.org/10.1073/pnas.0910894106. Scholpp, S., Foucher, I., Staudt, N., Peukert, D., Lumsden, A., Houart, C., 2007. Otx1l, Otx2 and Irx1b establish and position the ZLI in the diencephalon. Development 134 (17), 3167e3176. https://doi.org/10.1242/dev.001461. Sellers, K., Zyka, V., Lumsden, A.G., Delogu, A., 2014. Transcriptional control of GABAergic neuronal subtype identity in the thalamus. Neural Dev. 9 (1), 14. https://doi.org/10.1186/1749-8104-9-14. Sena, E., Feistel, K., C Durand, B., 2016. An evolutionarily conserved network mediates development of the zona limitans intrathalamica, a sonic hedgehog-secreting caudal forebrain signaling center. J. Dev. Biol. 4 (4). https://doi.org/10.3390/jdb4040031. Sherman, S.M., Guillery, R.W., 2009. Exploring the Thalamus and its Role in Cortical Function. MIT Press. Shi, W., Xianyu, A., Han, Z., Tang, X., Li, Z., Zhong, H., Mao, T., Huang, K., Shi, S.-H., 2017. Ontogenetic establishment of order-specific nuclear organization in the mammalian thalamus. Nat. Neurosci. 20 (4), 516e528. https://doi.org/10.1038/nn.4519. Shimamura, K., Hartigan, D.J., Martinez, S., Puelles, L., Rubenstein, J.L., 1995. Longitudinal organization of the anterior neural plate and neural tube. Development 121 (12), 3923e3933.

86

PART | I Induction and patterning of the CNS and PNS

Shimizu, T., Hibi, M., 2009. “Formation and patterning of the forebrain and olfactory system by zinc-finger genes Fezf1 and Fezf2. Dev. Growth Differ. 51 (3), 221e231. https://doi.org/10.1111/j.1440-169X.2009.01088.x. Shimogori, T., Lee, D.A., Miranda-Angulo, A., Yang, Y., Wang, H., Jiang, L., Yoshida, A.C., et al., 2010. A genomic atlas of mouse hypothalamic development. Nat. Neurosci. 13 (6), 767e775. https://doi.org/10.1038/nn.2545. Simeone, A., Acampora, D., Gulisano, M., Stornaiuolo, A., Boncinelli, E., 1992. Nested expression domains of four homeobox genes in developing rostral brain. Nature 358 (6388), 687e690. https://doi.org/10.1038/358687a0. Song, H., Lee, B., Pyun, D., Guimera, J., Son, Y., Yoon, J., Baek, K., Wurst, W., Jeong, Y., 2015. Ascl1 and Helt act combinatorially to specify thalamic neuronal identity by repressing Dlxs activation. Dev. Biol. 398 (2), 280e291. https://doi.org/10.1016/j.ydbio.2014.12.003. Suda, Y., Hossain, Z.M., Kobayashi, C., Hatano, O., Yoshida, M., Matsuo, I., Aizawa, S., 2001. Emx2 directs the development of diencephalon in cooperation with Otx2. Development 128 (13), 2433e2450. Suzuki-Hirano, A., Ogawa, M., Kataoka, A., Yoshida, A.C., Itoh, D., Ueno, M., Blackshaw, S., Shimogori, T., 2011. Dynamic spatiotemporal gene expression in embryonic mouse thalamus. J. Comp. Neurol. 519 (3), 528e543. https://doi.org/10.1002/cne.22531. Szabó, N.-E., Zhao, T., Zhou, X., Alvarez-Bolado, G., 2009. The role of sonic hedgehog of neural origin in thalamic differentiation in the mouse. J. Neurosci. 29 (8), 2453e2466. https://doi.org/10.1523/JNEUROSCI.4524-08.2009. Trujillo, C.M., Alonso, A., Delgado, A.C., Damas, C., 2005. The rostral and caudal boundaries of the diencephalon. Brain Res. Brain Res. Rev. 49 (2), 202e210. https://doi.org/10.1016/j.brainresrev.2005.01.002. Vaage, S., 1969. The segmentation of the primitive neural tube in chick embryos (Gallus Domesticus). A morphological, histochemical and autoradiographical investigation. Ergeb. Anat. Entwicklungsgesch. 41 (3), 3e87. Vieira, C., Garda, A.-L., Shimamura, K., Martinez, S., 2005. Thalamic development induced by Shh in the chick embryo. Dev. Biol. 284 (2), 351e363. https://doi.org/10.1016/j.ydbio.2005.05.031. Virolainen, S.-M., Achim, K., Peltopuro, P., Salminen, M., Partanen, J., 2012. Transcriptional regulatory mechanisms underlying the GABAergic neuron fate in different diencephalic prosomeres. Development 139 (20), 3795e3805. https://doi.org/10.1242/dev.075192. Vue, T.Y., Aaker, J., Taniguchi, A., Kazemzadeh, C., Skidmore, J.M., Martin, D.M., Martin, J.F., Treier, M., Nakagawa, Y., 2007. Characterization of progenitor domains in the developing mouse thalamus. J. Comp. Neurol. 505 (1), 73e91. https://doi.org/10.1002/cne.21467. Vue, Yia, T., Bluske, K., Amin, A., Yang, L.L., Koyano-Nakagawa, N., Bennett, N., Nakagawa, Y., 2009. Sonic hedgehog signaling controls thalamic progenitor identity and nuclei specification in mice. J. Neurosci. 29 (14), 4484e4497. https://doi.org/10.1523/JNEUROSCI.0656-09.2009. Wilson, S.W., Houart, C., 2004. Early steps in the development of the forebrain. Dev. Cell 6 (2), 167e181. https://doi.org/10.1016/s1534-5807(04)00027-9. Wong, S.Z.H., Scott, E.P., Mu, W., Guo, X., Borgenheimer, E., Freeman, M., Ming, G.-L., Wu, Q.-F., Song, H., Nakagawa, Y., 2018. In vivo clonal analysis reveals spatiotemporal regulation of thalamic nucleogenesis. PLoS Biol. 16 (4), e2005211. https://doi.org/10.1371/journal.pbio.2005211. Yao, Y., Minor, P.J., Zhao, Y.-T., Jeong, Y., Pani, A.M., King, A.N., Symmons, O., et al., 2016. Cis-regulatory architecture of a brain signaling center predates the origin of chordates. Nat. Genet. 48 (5), 575e580. https://doi.org/10.1038/ng.3542. Yuge, K., Kataoka, A., Yoshida, A.C., Itoh, D., Aggarwal, M., Mori, S., Blackshaw, S., Shimogori, T., 2011. Region-specific gene expression in early postnatal mouse thalamus. J. Comp. Neurol. 519 (3), 544e561. https://doi.org/10.1002/cne.22532. Zhao, T., Szabó, N., Ma, J., Luo, L., Zhou, X., Alvarez-Bolado, G., 2008. Genetic mapping of Foxb1-cell lineage shows migration from caudal diencephalon to telencephalon and lateral hypothalamus. Eur. J. Neurosci. 28 (10), 1941e1955. https://doi.org/10.1111/j.1460-9568.2008.06503.x. Zhou, C.-J., Pinson, K.I., Pleasure, S.J., 2004. Severe defects in dorsal thalamic development in low-density lipoprotein receptor-related protein- mutants. J. Neurosci. 24 (35), 7632e7639. https://doi.org/10.1523/JNEUROSCI.2123-04.2004.

Chapter 5

Midbrain patterning: polarity formation of the tectum, midbrain regionalization, and isthmus organizer Harukazu Nakamura Laboratory of Organ Morphogenesis, Graduate School of Life Sciences, Tohoku University, Aoba-ku, Sendai, Japan

Chapter outline 5.1. Introduction: brief description about midbrain 5.2. Tectum laminar formation 5.3. Optic tectum as a visual center for the lower vertebrate 5.3.1. Retinotectal projection in a retinotopic manner 5.3.2. Polarity formation in the optic tectum 5.4. Development of midbrain from the mesencephalic brain vesicle 5.4.1. Transcription factors that determine the midbrain 5.4.2. Midbrainehindbrain boundary formation 5.4.3. Diencephalonemesencephalon boundary formation 5.4.4. Dorsoventral patterning in the midbrain

87 89 89 89 91 93 93 94 95 96

5.5. Isthmus organizer 96 5.5.1. Isthmus emanates organizing signal 96 5.5.2. Competence of the neural tube to Fgf8 signaling is determined by preexisting transcription factors 98 5.5.3. Intracellular signal transduction 99 5.5.4. How tectum and cerebellum are organized by isthmus organizing signal? 100 5.6. Concluding remarks 101 List of abbreviations of genes and molecules 101 List of abbreviations (general) 101 Glossary 102 References 103

5.1 Introduction: brief description about midbrain In vertebrates, the central nervous system is originated from the neural tube. At the rostral part of the neural tube, three swellings, prosencephalon (forebrain), mesencephalon (midbrain), and rhombencephalon (hindbrain), are formed around neural tube closure. These swellings are called primary brain vesicles. The prosencephalon is subdivided into telencephalon and diencephalon, and the rhombencephalon is subdivided into metencephalon and myelencephalon, and the brain vesicles are now called secondary brain vesicles, which are the fundamental brain plan of the vertebrate. Telencephalon gives rise to the cerebral cortex and basal ganglion. In diencephalon, thalamus, hypothalamus, and pineal gland differentiate. Optic vesicle protrudes from the diencephalon and gives rise to the neuroretina. Dorsal part of the metencephalon gives rise to the cerebellum and the ventral part gives rise to pons. Myelencephalon gives rise to the medulla oblongata. Mesencephalic alar plate gives rise to the tectum, and the basal plate gives rise to the tegmentum. In birds, a pair of the optic tectum is conspicuous on the dorsal part of the midbrain. In mammals, we discern quadrigeminal bodies on the dorsal midbrain. The optic tectum in birds corresponds to the superior colliculus. Inferior colliculi, which are the relay of the auditory ascending pathway, could not be discerned. The torus semicircularis and the nucleus mesencephalicus lateralis are the relay for the ascending auditory pathway, receiving afferents through the lemniscus lateralis and sending ascending fibers bilaterally to the thalamic nucleus ovoidalis. So the torus semicircularis and nucleus mesencephalicus lateralis are thought to be correspondent to the inferior colliculus (Dubbeldam, 1998). The optic tectum is the primary visual center in birds and lower vertebrates. The tectum receives retinal fibers in retinotopic manner. Retinotectal projection will be

Patterning and Cell Type Specification in the Developing CNS and PNS. https://doi.org/10.1016/B978-0-12-814405-3.00005-9 Copyright © 2020 Elsevier Inc. All rights reserved.

87

88

PART | I Induction and patterning of the CNS and PNS

discussed later. In mammals, the visual center has moved to the occipital lobe of the cerebral cortex. The corresponding superior colliculus receives retinal fibers but most retinal fibers go to the lateral geniculate body. Fibers are relayed there and go to the occipital lobe in mammals. The superior colliculus functions as the center for orienting responses. One of the examples of the orienting responses is that we trace the moving object by turning our face to catch the image at the central fovea of the retina (Terashima, 2011). The optic tectum (Fig. 5.1) and superior colliculus consist of laminar structure (LaVail and Cowan, 1971a). The mature avian optic tectum consists of 7 main laminae and 10 sublaminae: stratum opticum (SO) through which retinal fibers run, stratum griseum et fibrosum superficiale (SGFS) which is a principal cell and fiber zone and consists of sublaminae aej, stratum griseum centrale (SGC) where principal efferent neurons of the tectum are located, stratum album centrale (SAC) which is a main efferent pathway of the tectum, stratum griseum periventriculare (SGP), stratum fibrosum periventriculare (SFP), and the ependymal layer (ventricular layer). SAC contains commissural fibers from the contralateral tectum and these fibers may end in SGP (LaVail and Cowan, 1971a). The number of the laminae and thickness of the tectum increases as development proceeds (Fig. 5.1) (LaVail and Cowan, 1971a; Omi et al., 2014). In chick, retinal fibers enter the tectum from the rostral pole and run the surface of the tectum (SO); they turn right into the tectum laminae and terminate at the retino-recipient laminae b, d, and f of the SGFS (Yamagata and Sanes, 1995). The lamina g of the SGFS may be the barrier for the retinal fibers and retinal fibers do not go through the laminae g of the SGFS (Sugiyama and Nakamura, 2003; Yamagata and Sanes, 1995). During development, some retinal fibers run through SAC and SFP, but these fibers are transient and degenerate around hatch (Omi et al., 2011). In the tegmentum of the mammalian midbrain, substantia nigra and red nucleus are notified. Neurons in substantia nigra in mammals contain melanin, so it looks black. Substantia nigra receives input from the striatum and sends in turn dopaminergic outputs to the striatum. Substantia nigra may be regarded as a part of striatum system and modifies the motor system. In birds, nucleus tegmenti pedunculopontinus is thought to correspond to the substantia nigra (Dubbeldam, 1998).

FIGURE 5.1 Tectum laminar formation. (AeC) Schematic drawing of the development of the tectum laminae, labeling of the tectal neuroepithelial cells by electroporation of lacZ expression plasmid vector showed that the cells that migrate after E5 split the layers that are formed by cells migrated before E5. The analysis unraveled that lamina II could be divided into upper and deeper laminae. The late migratory cells migrated into deeper lamina II (B), then form laminae hej of the SGFS (C). (D) Final laminae of the chicken optic tectum. (EeF) Clonal analysis of migration of tectal neurons. Electroporation of plasmid vector of RCAS-AP (Ｅ) and of RCAS-Grg4 (Ｆ) on virus-resistant embryos at E2 and fixed at E8. In virus-resistant embryos, misexpression is limited to the descendents of originally transfected cells. AP-positive cells are present throughout all laminae (E), but Grg4-expressing cells are present mainly in deeper lamina II (B). Scale bars: 100 mm. IeIV, laminae of the developing tectum; IId, deeper lamina II; IIu, upper lamina II; SAC, stratum album centrale; SFP, stratum fibrosum profundum; SGC, stratum griseum centrale; SGFS, stratum griseum et fibrosum superficiale; SGP, stratum griseum profundum; SO, stratum opticum; VL, ventricular layer. (AeC, E, and F) From Sugiyama, S., Nakamura, H., 2003. The role of Grg4 in tectal laminar formation. Development 130, 451e462; (D) From Omi, M., Nakamura, H., 2015. Engrailed and tectum development. Dev. Growth Differ. 57, 135e145.

Midbrain patterning Chapter | 5

89

Tegmenti pedunculopontinus contains dopaminergic neurons and have reciprocal fiber connections with the striatum. Red nucleus receives inputs from the cortex and the cerebellar nuclei and sends outputs to the olive and the spinal motor neuron. Red nucleus is also thought to be the modifier of the motor system. Nucleus of the oculomotor nerve and the accessory nucleus of the oculomotor (EdingereWestphal nucleus in mammals) nerve are also important nuclei in the tegmentum (Dubbeldam, 1998). In this chapter, I focus on the development of the optic tectum and retinotectal projection. How the optic tectum is defined and the mechanisms of the polarity formation of the tectum for the basis of retinotectal projection are described. I also focus on the signal from the isthmus (midbrainehindbrain boundary) and its intracellular transduction pathway for the midbrain and hindbrain definement is discussed.

5.2 Tectum laminar formation Postmitotic cell migration and layer formation in the chick optic tectum were studied autoradiographically (LaVail and Cowan, 1971b). It was shown that the tectum is formed by three cell migration waves. The first wave (between E3eE5, embryonic day 3eembryonic day 5) forms the deep layers (SGC and SGP), the second (E4eE7), the superficial layers (laminae aeg of the SGFS), and the third wave (E6eE8) forms the middle layers (laminae hej). Migration pattern in the tectum is different from that in the mammalian neocortex, where insideeout type migration occurs. Consistent results were obtained by labeling tectal premigratory cells by lacZ by in ovo electroporation (Sugiyama and Nakamura, 2003). It was shown that the destination of postmitotic neurons that migrate before and after E5 is different. The late migratory neurons split the early migratory neurons and form laminae hej of the SGFS (Fig. 5.1AeC). Grg4 (Groucho-related gene) belongs to the Gro/Grg/TLE family and functions as a transcriptional repressor downstream of the Notch or Wnt signaling. It represses Wnt signaling by interacting with Tcf. Grg4 is expressed strongly in E2 mesencephalon, then diminishes and disappears at E4. Grg4 expression then reappears in the tectal neuroepithelium at E5 (Sugiyama and Nakamura, 2003). Misexpression by the plasmids that contains Grg4-retrovirus DNA by electroporation was carried out to show if Grg4 is related to cell migration. By electroporation with retrovirus plasmid vector on virus-resistant chick embryos, only the descendants of the originally transfected cells express Grg4, which allows us clonal analysis. By electroporation on virus-sensitive chick embryos, virus from the originally transfected cells infect the adjacent cells and massive expression of Grg4 could be obtained. When the tectal premigratory cells were labeled clonally at E2 with lacZ, the progeny spanned all the layers of the tectum. On the other hand, clonal misexpression of Grg4 in tectal premigratory cells showed that Grg4-misexpressing cells preferentially migrated to the layer that would form laminae hej of the SGFS (Fig. 5.1E and F). Grg4-misexpressing cells may have acquired the property of late migrating cells. After massive misexpression of Grg4, laminae hej of SGFS enlarged and lamina g disappeared in the Grg4-misexpressing region. Treatment with Grg4 morpholino antisense oligonucleotide or misexpression of N-terminal region of Grg4, which had been shown to work as the dominant negative to Grg4 exerted effects opposite to misexpression of Grg4, that is, decrease of the layer that would form laminae hej of SGFS. Both gain and loss of function of Grg4 experiments support the idea that Grg4 instructs the tectal postmitotic cells to follow the late migratory pathway (Sugiyama and Nakamura, 2003). Retinal fibers terminate in the laminae b, d, and f of the SGFS, and they do not penetrate the lamina g (LaVail and Cowan, 1971a; Sugiyama and Nakamura, 2003; Yamagata and Sanes, 1995). In the Grg4-misexpressing regions, where lamina g of the SGFS was disrupted, retinal axons were seen to extend deeper than g. Refinement of retinal projections was also disturbed in the Grg4-misexpressing region. It has been proposed that synchronized firing of the neighboring fibers plays an important role in refining the retinal projection. For activity-dependent refinement in the visual cortex in mice, it was suggested that maturation of GABAergic inhibitory neurons is important (Fagiolini and Hensch, 2000). Indeed, in the tectum, parvalbumin, which is expressed in GABAergic neurons, is expressed in lamina g. It still awaits further study, but the results suggest that the lamina g of SGFS seems to contribute to fine tunings of the retinal projection.

5.3 Optic tectum as a visual center for the lower vertebrate 5.3.1 Retinotectal projection in a retinotopic manner Optic tectum is a visual center in lower vertebrates and receives retinal fibers in a retinotopic manner. In chick, the tectum is also a visual center. Retinal ganglion cells at the nasal and temporal of the retina project to the caudal and rostral part of the tectum, respectively (Fig. 5.2A). Retinal ganglion cells at the dorsal and ventral part of the retina project to the ventral and dorsal part of the retina, respectively (Crossland and Uchwat, 1979). The image on the retina is projected invertedly on the tectum. Since the image on the retina is inverted one by the lens, the outer object is projected as it is on the tectum.

90

PART | I Induction and patterning of the CNS and PNS

FIGURE 5.2 Polarity formation of the tectum and retinotectal projection. (A) normal; (B) rotation of the midbrain at E1.5; (C) heterotopic transplantation of the alar plate of the midbrain to the diencephalon at E1.5; (D) heterotopic transplantation of the E1.5 alar plate of the midbrain to the E3 diencephalon; and (E) misexpression of the En by retrovirus vector. Left column E1.5 except for (D), central column: E4, right column: retinotectal projection at E15. In normal development (A), En2 (light blue) is expressed in a gradient in the midbrain, caudal high and rostral low. Then in the tectum ephrin-A2 and ephrin-A5 are expressed in a gradient (blue), caudal high and rostral low. In the temporal eye, EphA3 is expressed in a gradient (purple). The temporal retinal fibers, which express highly EphA3, are repelled by ephrin-A2 and ephrin-A5 and project to the rostral tectum. Nasal retinal fibers, which do not express EphA3, could reach to the caudal tectum and project there. After the rotation of the alar plate of the midbrain at E1.5 (B), En2 expression is regulated as the host pattern byE4, and the retinotectal projection is just as the normal pattern. After transplantation of the E1.5 alar plate of the midbrain to the E1.5 diencephalon, En2 expression in the ectopic tectum becomes mirror image to the host tectum, and the retinal fiber projection to the ectopic tecta is also mirror image to the host. After transplantation of the E1.5 alar plate of the midbrain to the E3 diencephalon (D), En2 expression is kept as the original pattern, and the nasal retinal fibers projected to the caudal part of the ectopic tecta. Temporal retinal fibers could not enter the ectopic tecta. After misexpression of En1 or En2, ephrin-A2 and ephrin-A5 expression is induced corresponding to the En-misexpressing site, and the temporal retinal fibers were repelled by ectopically expressed ephrin-A2 and ephrin-A5. Nasal retinal fibers extended to the caudal tectum and projected there. Nasal retinal fibers send branches to the En2-expressing sites. C, caudal; di, diencephalon; hind, hindbrain; isth, isthmus; mid, midbrain; N, nasal; R, rostral; T, temporal; tect, tectum; tect-ect, ectopic tectum.

The visual system has served as a model to study neural circuit formation, and axon pathfinding since the system could be manipulated experimentally. I will review brief history of the study of retinotectal projection, and the axon pathfinding mechanism. The pioneer work on retinotectal projection was carried out by Sperry (1943). He cut the optic nerve of a frog and rotated the eye 180 degrees. The frog can regenerate optic nerve. The eye-rotated frog behaved as if the outside world is projected to the visual center 180 degrees rotated; when a bait is in front of the frog, the frog jumped posteriorly, and when a bait is at the right side, the frog jumped to the left side. Sperry proposed chemoaffinity theory for the proper retinotectal projection (Sperry, 1963). He proposed that growing retinal fibers and the tectal cells carry individual cytochemical tags by which they are distinguish one from another so that the retinal ganglion cell projects to the proper site on the tectum. Researchers were attracted by this theory and tried to find such molecules. Bonhoeffer’s group prepared alternating stripes of rostral and caudal tectal membrane fragments and examined behavior of the nasal and temporal retinal

Midbrain patterning Chapter | 5

91

fibers in vitro. Temporal retinal fibers extended only on the membrane fragments of the rostral tectum, target of the temporal retinal ganglion cells, while the nasal retinal fibers extended on rostral and membrane fragments of both rostral and caudal tectum (Walter et al., 1987b). Next, they prepared the membrane stripe after heat treatment (Walter et al., 1987a). On the stripe prepared after heat treatment of membrane fragments of the caudal tectum, temporal retinal axons abolished the preference to extend on the rostral membrane fragments. They interpreted these results that the temporal retinal fibers preferentially extend on the rostral membrane fragments because the temporal retinal fibers are repelled by the caudal tectal membrane fragments. This idea well explains the retinotectal projection formation. The temporal retinal fibers are repelled by the caudal tectum and project on the rostral tectum, while nasal retinal fibers can extend through rostral tectum to the caudal tectum. Bonhoeffer’s group finally purified glycosylphosphatidylinositol-anchored glycoprotein and cloned its cDNA, which belongs to the ligands for receptor tyrosine kinase Eph subfamily (Drescher et al., 1995). They first named it as RAGS (repulsive axon guidance signal) and later renamed as ephrin-A5. It was expressed in the tectum in a caudorostral gradient. Ephrin-A5 induced collapse and repelled both temporal retinal axons. Through the search for Eph receptor ligand family, Flanagan’s group cloned ligand ELF-1 and the receptor Mek4, which were renamed as ephrin-A2 and EphA3, respectively (Cheng and Flanagan, 1994). They showed that ephrin-A2 is expressed in the tectum as caudorostral gradient, and that EphA3 is expressed in the retina as temporo-nasal gradient (Cheng et al., 1995). Since the ephrin-A2 and EphA3 expression is complementary in view of retinotectal projection, it was proposed that ephrin-A2eEphA3 ligand receptor system may be positional labels for map development. Nakamoto et al. (1996) showed that ephrin-A2 acted as a repellent axon guidance factor. They also showed that temporal retinal fibers avoided ephrin-A2eexpressing site, which was produced by retroviral vector. These results strongly suggested that graded expression of ephrin-A2, ephrin-A5 in the tectum, and EphA3 in the retina is responsible for the formation of precise retinotectal projection along rostrocaudal axis of the tectum (Fig. 5.2A). Mutant mice lacking ephrin-A5, ephrin-A2, and double mutant of ephrin-A2 and ephrin-A5 were generated (Frisén et al., 1998; Feldheim et al., 2004). In either ephrin-A2 or ephrin-A5 null mutant mice, temporal retinal fibers made terminal arborization at their correct target site as well as at more caudal site (Frisén et al., 1998; Feldheim et al., 2004). In addition, retinal axons overshot the superior colliculus and extended into the inferior colliculus (Frisén et al., 1998). In double homozygous mutant mice of ephrin-A2 and ephrin-A5, anteroposterior order of retinocollicular projection was almost but not completely lost (Feldheim et al., 2000). Temporal and nasal retinal fibers made multiple terminations widely on the superior colliculus. Feldheim et al. (2004) adopted axon competition to explain for the nasal retinal fibers terminate at the rostral colliculus in double mutant mice. In normal situation, nasal retinal fibers cannot terminate at the rostral colliculus where temporal retinal fibers terminate. In double mutant mice, many temporal retinal fibers bypass the rostral colliculus so nasal retinal fibers can terminate there because nasal retinal fibers can find space for termination. For the map formation along dorsoventral axis, it has been suggested that ephrin-B and EphBs system is acting. In the retina, EphB receptors are expressed as ventral high to dorsal low gradient, and ephrin-B1 is expressed as medial (dorsal) high to lateral (ventral low) (McLaughlin et al., 2003). Primary retinal axons lack DV ordering along the LM tectal axis, but they send interstitial branches. Ephrin-B1 acts as an attractant for interstitial branches from the lower ephrin-B1 expression site than their target site so that the interstitial branches are attracted by their target. Ephrin-B1 acts as repellent from the higher ephrin-B1 expression site. These interstitial branches are repelled by higher ephrin-B1 at more medial site, and go toward the target. These interstitial branches rest and become terminal arborizations. McLaughlin et al. (2003) showed this controversial ephrin-B1 and EphBs action by misexpressing ephrin-B1 in the chick tectum and in null mutation mice of EphB.

5.3.2 Polarity formation in the optic tectum I have shown that the tectum is polarized along rostrocaudal and mediolateral axes. It is very interesting how the polarity is established. I will show the rostrocaudal polarity formation of the tectum. Engrailed gene, homologous to the Drosophila segment polarity gene, was cloned (Joyner et al., 1985), then it was shown that mouse has two engrailed, En1 and En2 (Joyner et al., 1985). It was shown by immunohistochemistry that En is expressed in the early stage of developing mesencephalon of zebrafish, chick, and mouse (Patel et al., 1989). Since En expression was in a gradient, caudal high and rostral low (Fig. 5.2), researchers expected that En would play an important role in tectum rostrocaudal polarity formation. In the chick tectum, we recognize rostrocaudal polarity in cytoarchitectonic development. The rostral tectum differentiates faster than the caudal and has thicker wall and more layers than the caudal (Itasaki et al., 1991; LaVail and Cowan, 1971a). When the quail mesencephalic alar plate was transplanted into the chick by reverse rostrocaudal axis, the rostrocaudal polarity of the transplant was regulated as the host pattern (Fig. 5.2B) (Itasaki et al., 1991). En expression,

92

PART | I Induction and patterning of the CNS and PNS

cytoarchitectonic development, and retinotectal projection were the same as the control (Ichijo et al., 1990; Itasaki et al., 1991; Martinez and Alvarado-Mallart, 1990). The rostral part of the chimeric tectum, which had been originally caudal, developed as the rostral tectum and received the temporal retinal fibers (Fig. 5.2B). The caudal part of the chimeric tectum received the nasal retinal fibers. These results indicated that En expression pattern is related to the rostrocaudal polarity of the tectum and also indicated that the rostrocaudal polarity is determined by the signal emanated from the neighboring tissue. The alar plate of the chick mesencephalon differentiates into the tectum when transplanted into the diencephalon ectopically (Alvarado-Mallart and Sotelo, 1984; Nakamura, 1990). But En expression pattern and cytoarchitectonic development in the transplant were reversed from their original pattern, that is, En expression and cytoarchitecture in the transplant became mirror image to the host (Fig. 5.2C) (Itasaki et al., 1991). Then, retinal projection pattern on the ectopic tectum produced in the diencephalon by heterotopic transplantation at E1.5 was examined. The tectum rotates between E7 and E12 so that the tectal rostrocaudal axis accords with the ventrodorsal axis of the body, and the rostral end of the tectum comes to face the optic tract (Fig. 5.2C and D) (Itasaki and Nakamura, 1992). The ectopic tectum on the diencephalon rotated in a reverse direction so that the caudal end of the ectopic tectum came to face the optic tract (Fig. 5.2C and D) (Itasaki and Nakamura, 1992). Rotation of the tectum proper and the ectopic tectum, which is a result of transplantation at E1.5, resulted in that the pole where En expression had been weak faced the optic tract (Fig. 5.2C). The nasal retinal fibers entered the ectopic tectum from the caudal pole, where En expression had been weak, and projected near the rostral pole. The temporal retinal fibers projected near the caudal pole (Fig. 5.2C). The retinal fiber projection pattern also became the mirror image to the host. When E1.5 mesencephalic alar plate was transplanted into the E3 diencephalon, the transplant kept its original En expression patterndcaudal high and rostral low (Fig. 5.2D). The temporal retinal fibers could not enter the ectopic tectum, and the nasal retinal fibers projected just after entering the ectopic tectum, where En expression had been high (Fig. 5.2D) (Itasaki and Nakamura, 1992). The study of En expression and retinotectal projection in the ectopic tectum indicated that En conferred the caudal characteristics of the tectum. Then the role of En on rostrocaudal polarity formation was studied in chick embryos by misexpression of En by retroviral vector (Itasaki and Nakamura, 1996; Friedman and O’Leary, 1996). Patchy En2 misexpression islands were made after infection of retrovirus containing En1 and En2. On En1 or En2 misexpressed-tecta, nasal retinal fibers extended on the tectum toward their proper target site, but they did not make or made weak terminal arborizations at the proper target site (Fig. 5.2E). Nasal retinal fibers made arborizations at the En-expressing sites on the way to their proper target site (Fig. 5.2E) (Itasaki and Nakamura, 1996). Temporal retinal fibers were repelled at the En-expressing site (Fig. 5.2E), and in some times, temporal retinal fibers could not enter the tectum (Itasaki and Nakamura, 1996; Friedman and O’Leary, 1996). These results strongly indicated that En confers caudal characteristics on the tectum. Both papers showed that En induces some factors to attract nasal retinal fibers. It was shown that both En1 and En2 induce ephrin-A2 and ephrin-A5 (Logan et al., 1996; Shigetani et al., 1997). Prochiantz and his colleagues have greatly contributed in the function of En. En is an homeobox-containing transcription factor, but they showed that En could be secreted and has transcription and noncell-autonomous activities (Prochiantz and Joliot, 2003). They also showed in vitro that En2 directly attracts nasal retinal fibers (Brunet et al., 2005). En2 is expressed in chick tecta in a caudorostral gradient, caudal high and rostral low at E10 (Omi et al., 2014). In the caudal tecta, En2 is expressed in laminae gej of the SGFS, but in the rostral, En2 expression was hardly recognized (Omi et al., 2014). Since transposon mediated genome integration system, and Tet-on and Tet-off system in chick embryos were developed, long-term expression of the transgene, and turn-on and turn-off of the transgene at desired time became possible (Hilgers et al., 2005; Sato et al., 2007; Watanabe et al., 2007). En2 misexpression by in ovo electroporation in the mesencephalon was carried out by this system (Omi et al., 2014). As En2 misexpression alters the rostrocaudal polarity of the tectum (Itasaki and Nakamura, 1996), En2 expression was induced from E8.5 by administration of Dox every 24 h, and fixed at E13.5. In the tecta where En2 was continuously expressed from E8.5, En2-expressing cells were not found in the surface layer, but found in the layers deeper than lamina g of the SGFS. Time-course analysis by fixing the tectum after 7, 12, and 24 h after Dox (Doxycycline, Tetracycline analog) administration showed that at 7 h after Dox administration En2-expressing cells were found in the superficial layers, but at 12 h after Dox administration En2-expressing cells in the superficial layers reduced in number and disappeared at 24 h after Dox administration (Fig. 5.3) (Omi et al., 2014). Time-lapse analysis of behavior of the En-expressing cells in slice culture revealed that En2-expressing cells migrated back from the superficial layers toward the middle layers where En2 is strongly expressed endogenously (Omi et al., 2014). Since the retinal fibers pass the superficial layer of the tectum, it is hard to think that En directly participates in retinotectal map formation. En may be responsible for neural cell migration and positioning in the tectum rather than direct contribution to the retinotectal map formation.

Midbrain patterning Chapter | 5

(A)

(B)

(C)

(D)

(E)

(F)

93

FIGURE 5.3 En2-misexpressing cells cannot stay in the superficial layers. Transfection of Tet-on system plasmids was carried out by electroporation at E1.5, and Dox was administered from E8.5, when the transfected cells had reached the superficial layer. (A, C, and E, Control) Electroporation of enhanced green fluorescent protein (EGFP) expression vector. (B, D, and F) Electroporation of En2-EGFP expression vector. At 7 h after Dox administration, which is just after the induction of En2 misexpression because it takes about 6 h to induce expression of the gene of interest by Tet-system, En2-misexpressing cells are found in the superficial layers (B) as in the case of EGFP transfection (A). At 12 h after Dox administration, En2-expressing cells were reduced in number in superficial layers (D), and 24 h after Dox administration, En2-misexpressing cells are not found in the superficial layers (F). In the control side (EGFP transfection), EGFP-expressing cells in the superficial layers are increased in number after 24 h after Dox administration. aef, g, h, i: laminae of the SGFS. After Omi, M., Harada, H., Watanabe, Y., Funahashi, J., Nakamura, H., 2014. Role of En2 in the tectal laminar formation of chick embryos. Development 141, 2131e2138.

5.4 Development of midbrain from the mesencephalic brain vesicle 5.4.1 Transcription factors that determine the midbrain In Fig. 5.4, expression of some representative genes in the brain vesicle around stage 10 chick embryos is shown. En1 and En2 are expressed in the isthmus and the mesencephalon. In En1 knockout mice, the tectum and the cerebellum are hypomorphic (Wurst et al., 1994). En2 knockout mice show subtle cerebellar phenotype (Joyner et al., 1991). En1 expression commences earlier than that of En2 and covers the whole of the mesencephalon. Its expression then regresses and is confined to the isthmic region. Knockin of En2 into En1 knockout mice showed that En2 can substitute for En1 in the midbrainehindbrain region (Hanks et al., 1995). Onset of expression and effects of mutation indicates that En1 plays more crucial role in mesencephalic fate determination than En2. Pax2 and Pax5 belong to the same Pax subfamily (Gruss and Walther, 1992). Homozygous zebrafish mutants of Pax2.1 [no isthmus (noi)] lack the isthmic constriction, cerebellum, and optic tectum (Pfeffer et al., 1998). Pax2 mutant mice show various phenotypes ranging from a deletion of the cerebellum and the posterior midbrain to exencephaly at the midbrain level (Favor et al., 1996). Pax5 mutant mice show a rather more subtle phenotypedreduction of the posterior midbrain and abnormal foliation of the cerebellum (Urbánek et al., 1997). Pax2 and Pax5 double knockout mice completely lack a mesencephalon. Expression of Pax2 begins earlier than expression of Pax5 in mouse and chick embryos (Okafuji et al., 1999; Urbanek et al., 1997). In chick embryos Pax2 expression commences at four somite stage and covers the whole of the mesencephalon (Okafuji et al., 1999). By contrast, expression of Pax5 in chick embryos begins around 10 somite stage with high levels in the isthmic region (Funahashi et al., 1999). Time course of the expression of Pax2 and Pax5 indicates that Pax2 plays more crucial role in mesencephalon development. Indeed, it was shown that Pax5 was expressed under the Pax2 locus, and that Pax2 regulates Pax5 expression at the midbrainehindbrain boundary (Pfeffer et al., 2000).

94

PART | I Induction and patterning of the CNS and PNS

FIGURE 5.4 Schematic drawing of brain vesicles. Transcription factors expressed in the brain vesicles, and the secondary organizers, isthmus and anterior neural ridge, are indicated. Fgf8 signal from the isthmus activates RaseERK signaling pathway to organize the cerebellum. The mesemetencephalon boundary is determined by repressive interaction between Otx2 and Gbx2. Fgf8 expression is induced at the boundary of Otx2 and Gbx2 expression overlapping with Gbx2 expression. The diemesencephalon boundary is determined by repressive interaction between Pax6 and En1/ Pax7. The region where En1, Pax2, and Otx2 are expressed acquires mesencephalic characteristics, and the addition of Pax7 makes optic tectum. Interaction among Fgf8, Limx1b, and Wnt1 determines the Fgf8 expression at the isthmus abutting against the Wnt1 expression. pros, prosencephalon; mes, mesencephalon; rhomb, rhombencephalon; tel, telencephalon; di, diencephalon; met, metencephalon; myel, myelencephalon. Modified after Nakamura, H., 2001a. Regionalisation of the optic tectum: combination of the gene expression that defines the tectum. Trends Neurosci. 24, 32e39; Nakamura, H., 2001b. Regionalisation and polarity formation of the optic tectum. Prog. Neurobiol. 65, 473e488; Nakamura, H., Katahira, T., Matsunaga, E., Sato, T., 2005. Isthmus organizer for midbrain and hindbrain development. Brain Res. Rev. 49, 120e126; Nakamura, H., Sato, T., Suzuki-Hirano, A., 2008. Isthmus organizer for mesencephalon and metencephalon. Dev. Growth Differ. 50, S113eS118.

Otx2 is expressed from very early stage of development in the rostral part of the embryo and plays a role in placing the anterior visceral endoderm, which functions as the head organizer in mice, at right place (Kimura et al., 2000), then Otx2 is expressed in the prosencephalon and mesencephalon (Simeone et al., 1992). There are two Otx genes in mice, Otx1 and Otx2 (Simeone et al., 1992). Otx1 null mice survive but show abnormal corticogenesis and epilepsy (Acampora et al., 1996), and Otx2 null mice lack forebrain and midbrain and rostral hindbrain due to a failure of the anterior neural plate during gastrulation (Acampora et al., 1995; Ang et al., 1996; Matsuo et al., 1995). Otx1þ/ and Otx2þ/ mice show failure in development of the mesencephalon and caudal diencephalon and anterior shift in Gbx2, Fgf 8, and Wnt1 expression resulting in anterior extension of the metencephalon (Suda et al., 1997). Now it is accepted that Otx2 plays a crucial role in regionalization and Otx1 participate in corticogenesis. The posterior border of the Otx2 expression domain does not correspond to the posterior border of the mesencephalic swelling at around stage 10 (Millet et al., 1996). Homotopic transplantation between quail and chick showed that the posterior border of the Otx2 expression domain corresponded to the posterior border of the tectum. Posterior part of the mesencephalic swelling where Otx2 is not expressed differentiated into the cerebellum. This gave a clear explanation to the former experimental results that caudal part of the mesencephalic swelling differentiated into the cerebellum (Martinez and Alvarado-Mallart, 1989). The fate of the brain vesicle is determined by the combination of transcription factors expressed (Nakamura et al., 2005). Mesencephalon is the region where En1, Pax2, and Otx2 are expressed, and addition of Pax3/7 expression in the alar plate of the mesencephalon leads to the tectum development (Fig. 5.4). Isthmus functions as the secondary organizer for the midbrain and hindbrain by regulating the transcription factors. I will review how we came to these conclusions.

5.4.2 Midbrainehindbrain boundary formation Otx2 and Gbx2 are expressed from the very early stage of vertebrate development (Simeone et al., 1992; Shamim and Mason, 1998; Niss and Leutz, 1998). Otx2 is expressed in the prosencephalon and mesencephalon (Simeone et al., 1992), and Gbx 2 is expressed in the metencephalic region (Fig. 5.3) (Simeone et al., 1992; Shamim and Mason, 1998).

Midbrain patterning Chapter | 5

95

The expression domain of Otx 2 and Gbx 2 makes a clear border at the isthmus (Shamim and Mason, 1998; Broccoli et al., 1999; Millet et al., 1999; Katahira et al., 2000). Otx2 knockout mice lack forebrain, and midbrain and anterior hindbrain (Acampora et al., 1995; Ang et al., 1996; Matsuo et al., 1995). On the other hand, Gbx2 knockout mice lack r1er3 (rhombomere 1erhombomere 3) structure and show abnormalities in the isthmus and caudal extension of the midbrain (Wassarman et al., 1997). At 4e6 somite stage, the expression domains of Otx2, Fgf8, and Wnt1 are expanded posteriorly in Gbx2 knockout mice, which also indicate caudal extension of the midbrain (Millet et al., 1999). From the expression pattern of Otx2 and Gbx2 and the phenotype of knockout mice of these genes, it was assumed that Otx2 and Gbx2 play important roles in regionalization of the midbrain and hindbrain. To prove the assumption, misexpression of Otx2 in the metencephalon and that of Gbx2 in the mesencephalon was carried out (Broccoli et al., 1999; Millet et al., 1999; Katahira et al., 2000). Otx2 misexpression in r1 driven by En1 promoter in mice resulted in enlargement and caudal shift of the inferior colliculi and lack of the anterior part of the cerebellum (Broccoli et al., 1999). In chick, Otx2 misexpression by in ovo electroporation changed the fate of the alar plate of the metencephalon to form the tectum. In both mice and chick, Gbx2 expression was repressed by Otx2 in the metencephalon. In addition, endogenous Fgf8 expression was repressed by Otx2, but Fgf8 expression was induced ectopically in the metencephalon. In chick, Fgf8 expression was induced in a ring just outside of the Otx2 misexpression site (Katahira et al., 2000), and in mice Fgf8 expression shifted caudally (Broccoli et al., 1999). Given that Gbx2 is expressed in the metencephalon, Fgf8 was induced at the boundary of Otx2 and Gbx2 expression site. Gbx2 misexpression in chick embryos resulted in an anterior shift of the caudal limit of the tectum, and repression of Otx2 by Gbx2. Fgf8 expression was induced at the periphery of the Gbx2 misexpression site in the mesencephalon (Katahira et al., 2000). Since Otx2 is expressed in the midbrain, Fgf8 was induced at the boundary of Otx2 and Gbx2 expression site. Gbx2 misexpression in the posterior mesencephalon driven by Wnt1 promoter in mice exerted similar effects; repression of Otx2 by Gbx2 in the mesencephalon, expansion of the hindbrain, and reduction of the midbrain. Fgf8 expression shifted rostrally. From the studies of Otx2 and Gbx2 misexpression together with the studies in Otx2 and Gbx2 knockout mice, it became accepted that midbrainehindbrain boundary is established by repressive interaction between Otx2 and Gbx2 (Fig. 5.4). These studies also indicated that Fgf8 expression is induced at the interface of Otx2 and Gbx2 expression overlapping with Gbx2 expression. This notion was confirmed by transplantation experiments. When r1 was transplanted in the diencephalon, where Otx2 is expressed, Fgf8 expression was induced (Hidalgo-Sánchez et al., 1999; Irving and Mason, 1999).

5.4.3 Diencephalonemesencephalon boundary formation Pax6 is expressed in the prosencephalon and is essential for the development of the diencephalon (Fig. 5.4) (Walther and Gruss, 1991; Stoykova et al., 1996; Grindley et al., 1997; Mastick et al., 1997). Pax6 mutant mice, Sey, show fate change of the pretectum, the caudal part of the diencephalon, to the mesencephalon (Mastick et al., 1997). En1, En2, Pax2, or Pax5 misexpression in the diencephalon repressed Pax6 expression and changed the fate of the alar plate of the diencephalon to the tectum (Araki and Nakamura, 1999; Funahashi et al., 1999; Okafuji et al., 1999). In other words, diencephalone mesencephalic boundary shifted anteriorly. En, Fgf8, Pax2, and Pax5 are in a positive feedback loop for their expression so that misexpression of any of these genes in the diencephalon acts on the feedback loop resulting in induction of the optic tectum in the diencephalon (Araki and Nakamura, 1999; Crossley et al., 1996; Funahashi et al., 1999; Okafuji et al., 1999; Sato et al., 2001). En contains transcription repressor domain, EH1, which interacts with a corepressor Groucho. Misexpression of a mutant En2, which bears mutation in EH1 domain so that it cannot interact with Groucho, did not repress Pax6 expression and did not cause anterior shift of the diencephalonemesencephalic boundary (Araki and Nakamura, 1999). Chimeric En2, in which EH1 domain was substituted with transcription activator VP16, exerted the opposite effects to the wild-type En. These results indicated that En directly repressed Pax6, but turning on the positive feedback loop to induce Fgf8, Pax2, and Pax5 needs some intervening molecule (Araki and Nakamura, 1999). Pax6 misexpression in the mesencephalon repressed En1, Pax2, and other mesencephalon-related gene expression and caused posterior shift of the dimesencephalic boundary (Matsunaga et al., 2000). Chimeric transcription activator Pax6, Pax6-VP16, exerted more severe effects than wild-type Pax6, and chimeric transcription repressor, Pax6-ENR, which is a chimera of Pax6 and En repressor domain EH1, exerted opposite effects to the wild-type Pax6 (Matsunaga et al., 2000). So it was suggested that Pax6 functions transcription activator at the diencephalonemesencephalic region, and repression of En1 and Pax2 is indirect effects. Chick Grg4 is expressed at first in the anterior neural fold. The expression localizes from the posterior diencephalon to the mesencephalon by stage 10. In Drosophila, Groucho inhibits engrailed expression in wing disc. Thus, Grg4 was paid

96

PART | I Induction and patterning of the CNS and PNS

attention as Grg4 is expressed in the diencephalonemesencephalic region. Misexpression of Grg4 by in ovo electroporation in chick embryos resulted in repression of isthmus-related genes such as En2, Pax5, and Fgf8 and resulted in posterior extension of the Pax6 expression domain (Sugiyama et al., 2000). Consequently, the diencephalone mesencephalic boundary shifted to the posterior, diencephalic cytoarchitecture extended posteriorly. Grg proteins mediate transcriptional repression as homo- or heterotetramers, and the N-terminal region of Grg proteins is important for tetramerization (Chen and Courey, 2000). Misexpression of N-terminal region of Grg4 exerted reverse effects to the wild-type Grg4. Grg works downstream of Notch or Wnt signaling (Chen and Courey, 2000). Misexpression of constitutive active Notch did not affect En2 expression, but misexpression of constitutive active b-catenin, which transduces Wnt signaling, antagonized Grg4 activity. Because Tcf1 and Tcf3 are expressed in the diencephalonemesencephalic region, it was proposed that Grg4 antagonizes Wnt signaling by binding to either Tcf1 or Tcf3 in the dimesencephalic region (Sugiyama et al., 2000). These studies indicated that the dimesencephalic boundary is established by repressive interaction between Pax6 and En1/Pax2 (Fig. 5.4). Grg4 may also participate in diencephalonemesencephalic boundary formation by repressing Pax2 and En1 expression.

5.4.4 Dorsoventral patterning in the midbrain The tectum and tegmentum differentiates from the alar plate and basal plate, respectively. Pax3 and Pax7 are expressed near the roof plate in chick embryos around 10 somite stage (E1.5) (Matsunaga et al., 2001) and cover the tectal primordium around stage 17 (E2.5). Misexpression of Pax3 or Pax7 extended the tectum territory ventrally. Lim1 is expressed in two stripes in the ventral mesencephalon and is thought as a tegmentum marker. By misexpression of Pax3 or Pax7, one of Lim1 stripes was disappeared. Shh is a secreted protein and is expressed in the floor plate of the neural tube. Dorsoventral patterning of the neural tube occurs depending on the concentration of Shh (Echelard et al., 1993; Ericson et al., 1996; Roelink et al., 1995). In the midbrain, Shh is also expressed in the floor plate (Nomura and Fujisawa, 2000; Watanabe and Nakamura, 2000). Transplantation of the floor plate of quail embryos or Shh-expressing fibroblast cells into the alar plate of the chick midbrain changed the fate of the surrounding alar plate cells into the tegmentum (Nomura and Fujisawa, 2000). After misexpression of Shh in the midbrain by in ovo electroporation in chick embryos (Watanabe and Nakamura, 2000), tectal swelling could not be detected. Instead, the area of the dopaminergic neurons extended dorsally, and the number of motor neurons increased; many motor nerve fibers came out of the dorsal part of the midbrain. The expression of Pax7, a tectum-related molecule, was suppressed, and the expression of Hnf 3b, floor plateerelated gene, and of Ptc, target of Shh expanded dorsally. Lim1/2 expression, tegmentum-related molecule, was also extended dorsally. Thus, it was concluded that Shh promotes the midbrain cells to become the tegmentum.

5.5 Isthmus organizer 5.5.1 Isthmus emanates organizing signal Since the quailechick chimera method to pursue the transplanted quail cells in host chick embryos was introduced (Le Douarin, 1973), repertory of neural crest cell differentiation has been elucidated. This method was applied to study the fate determination of brain vesicles (Alvarado-Mallart and Sotelo, 1984; Nakamura et al., 1986). Mesencephalon and metencephalon were shown to differentiate as their original fate when transplanted at around 10 somite stage to the ectopic site; transplanted alar plate of the mesencephalon into the diencephalon kept its original fate and differentiated into the tectum ectopically (Alvarado-Mallart and Sotelo, 1984; Nakamura, 1990). Alar plate of the metencephalon also kept its original fate and differentiated into the cerebellum (Nakamura, 1990). Only the alar plate of the diencephalon could change its fate and differentiate into the tectum when it was transplanted in the posterior part of the mesencephalon (Nakamura et al., 1986, 1988, 1991; Nakamura and Itasaki, 1992). When the diencephalic alar plate was transplanted into the posterior mesencephalon, En expression was induced in the transplant, and it differentiated into the tectum. Since the fate change occurred only when the diencephalic alar plate was transplanted near the isthmus, it was natural to think that the isthmus emanates some signal to regulate the neighboring tissue. Martinez et al. (1991) transplanted the isthmus into the diencephalon, isthmus induced En expression in the adjacent diencephalon, and the En-induced host region differentiated into the tectum. They also transplanted the isthmus into the rhombomere region, there the isthmus induced En and the En-induced region differentiated into the cerebellum (Martínez et al., 1995). Thus, the isthmus (midbrainehindbrain boundary) has been accepted as the organizing center for the tectum and cerebellum.

Midbrain patterning Chapter | 5

97

Then researchers pursued the organizing molecule. Crossley and Martin (1995) paid attention to Fgf8, which, they showed, was expressed in the isthmus. They inserted an Fgf8-soaked bead in the diencephalon and showed that Fgf8-soaked bead acted just as transplantation of the isthmus (Crossley et al., 1996). Fgf8 induced En and Wnt1 in the diencephalon, and En-induced region differentiated into the tectum. The results suggested that Fgf8 is the organizing molecule for the tectum. Subsequent gain-of-function experiments in chick and mouse embryos have all suggested that Fgf8 is the isthmus organizing molecule (Liu et al., 1999; Martinez et al., 1999; Shamim et al., 1999). Loss of function of Fgf8 in mice and zebrafish supported this idea because Fgf8-deleted mice and zebrafish showed deletion of midbrain and cerebellum (Brand et al., 1996; Meyers et al., 1998). Fgf17 and Fgf18, which belong to the same Fgf subfamily to Fgf8, are also expressed in the mesencephalon and metencephalon (Maruoka et al., 1998; Xu et al., 2000). The biochemical properties of Fgf8b, 17b, and 18 appear to be similar based on in vitro assays (Xu et al., 1999). Fgf17 and Fgf18 expression covers wider area, and expression commences later than that of Fgf8. Bead implantation experiment showed that Fgf8 induced Fgf17 and Fgf18 (Liu et al., 2003; Reifers et al., 2000). Pax2 and Fgf8 are required for Fgf17 expression in the isthmus (Reifers et al., 2000). Expression of both genes is rapidly lost in Fgf8 mesencephalon/r1-specific conditional mutants (Chi et al., 2003). Studies of gain and loss of function of these molecules suggest that Fgf8 is the main organizing molecule, and Fgf8, Fgf17, and 18 coordinately function as organizing molecules (Sato et al., 2004). Wnt1 is expressed in a ring at the most caudal part of the mesencephalon abutting the expression of Fgf8 (McMahon and Bradley, 1990). Wnt1 knockout mice show defects in the midbrain and hindbrain (McMahon and Bradley, 1990). So Wnt1 was a candidate of the isthmus organizing signal. But now, it is accepted that Wnt1 functions to maintain Fgf8 expression in the isthmus region since Wnt signaling could not induce Fgf8 signaling in the isthmus (Canning et al., 2007). When r1 (excluding the Wnt1-expression ring) was excised, Fgf8 was expressed there, but Fgf8 expression was lost after overnight culture (Canning et al., 2007). Addition of LiCl, which constitutively activates Wnt signaling, could not recover Fgf8 expression in the culture (Canning et al., 2007). Limx1b, which is expressed in the mesencephalon overlapping with Wnt1 and could induce Wnt1 expression, repressed Fgf8 expression cell autonomously, but induced Fgf8 noncell autonomously via Wnt1. Since Limx1b represses Fgf8 expression, clear border of Limx1b/Wnt1 and Fgf8 expression is made (Fig. 5.4). Thus, it was suggested that interaction among Fgf8, Limx1b, and Wnt1 contributes to patterning the Fgf8 expression (Fig. 5.4) (Matsunaga et al., 2002). Studies in chick embryos greatly enhanced our understanding in the role of Fgf8 on mesencephalon and metencephalon development. It was shown that Fgf8a and Fgf8b are expressed in the isthmus among several splice isoforms of Fgf8 (Sato et al., 2001). The difference between Fgf8a and Fgf8b is very subtle: Fgf8b contains 33 additional base pairs (Crossley and Martin, 1995). But misexpression by electroporation exerted drastically different effects on mesencephalon and metencephalon development (Sato et al., 2001). Misexpression of Fgf8a did not affect the mesencephalon- and metencephalon-related gene expression patterns, Otx2, Gbx2, and Irx2, except for En2. En2 expression was upregulated in the mesencephalon and in the diencephalon, and consequently transformation of the presumptive diencephalon to the mesencephalon occurred (Fig. 5.5A) (Sato et al., 2001). In contrast, Fgf8b misexpression resulted in repression of Otx2 expression in the mesencephalon, and induction of Gbx2 and Irx2 expression in the mesencephalon. As a result, fate change of the alar plate of the mesencephalon from the tectum to the cerebellum occurred (Fig. 5.5BeD) (Sato et al., 2001). Provoked by the report that Fgf8b has stronger transforming activity than Fgf8a (MacArthur et al., 1995), Sato presumed that the difference of the activity of Fgf8a and Fgf8b in the tectum and cerebellum development is due to difference in signal intensity. He carried out semiquantitative experiment, that is, to electroporate Fgf8b expression vector at different concentration (1.0, 0.1, 0.01, 0.001, and 0.0001 mg/mL). Electroporation at 1.0 and 0.1 mg/mL changed the fate of the mesencephalon to cerebellum, but electroporation at 0.001 mg/mL exerted Fgf8a-type effects, and that at 0.0001 mg/mL exerted no effects. Structural analysis by surface plasmon resonance supported the assumption that difference in signal intensity causees difference in organizing activity (Olsen et al., 2006). They showed that Fgf8b binds more intimately to the c isoforms of the Fgf receptor 1e3 (FGFR1-3) owing to the additional 11 amino acids, phenylalanine 32 (F32) being most important for binding. Mutation of F32 to alanine (F32A) reduced the affinity to FGFR, and the mutation converted the action of Fgf8b to that of Fgf8a (Olsen et al., 2006). Cotransfection of Otx2 and Fgf8b expression vector showed that Otx2 increased Fgf8b signal threshold for hindbrain differentiation (Sato et al., 2001). Preexisting transcription factor may be related to the competence of the tissue to respond to the signal. This result well explains why the posterior part of the midbrain that receives strong Fgf8 signal differentiates as the midbrain.

98

PART | I Induction and patterning of the CNS and PNS

(A)

(E)

(I)

(B)

(F)

(C)

(D)

(G)

(J)

(H)

(K)

FIGURE 5.5 Effects of Fgf8 and its signal transduction. (A) Misexpression of Fgf8a. Fate change of the diencephalon to midbrain occurs, and the tectum at the experimental side (right) extends toward diencephalon. (BeD) Misexpression of Fgf8b. Fate change of the midbrain to hindbrain occurs, and the cerebellum differentiates in place of the tectum (cer-ect). In the ectopically differentiated cerebellum external granular layer (arrows) and the Purkinje cells (arrow heads on D) are differentiated. (EeH) Effects of Fgf8a (F), Fgf8b (G), and 1/100 diluted Fgf8b (H) on ERK phosphorylation. Transfection region is indicated in (E). ERK is activated corresponding to the transfection site after Fgf8b misexpression (G). Ectopic activation of ERK is seen only in the diencephalon after misexpression of both Fgf8a and 1/100 dilution of Fgf8b, which exerts similar gross morphological effects. (IeK) Disruption of RaseERK signaling by dominant negative form of Ras (RasS17N, a serine-to-asparagine mutation at residue 17). ERK activity is downregulated (compare arrows on I, left-hand side and right-hand side are the control and the experimental side, respectively), and the tectum is formed in place of the cerebellum (arrow on J and tect. Ect on K). Modified after Sato, T., Araki, I., Nakamura, H., 2001. Inductive signal and tissue responsiveness defining the tectum and the cerebellum. Development 128, 2466e2469; Sato, T., Nakamura, H., 2004. The Fgf8 signal causes cerebellar differentiation by activating Ras-ERK signaling pathway. Development 131, 4275e4285; Nakamura, H., Katahira, T., Matsunaga, E., Sato, T., 2005. Isthmus organizer for midbrain and hindbrain development. Brain Res. Rev. 49, 120e126.

5.5.2 Competence of the neural tube to Fgf8 signaling is determined by preexisting transcription factors I mentioned that transplantation of the isthmus into the diencephalon induces En2, and that the En2-induced host diencephalon differentiates into the tectum. But the competence of the diencephalon to differentiate into the tectum is limited to prosomeres 1 and 2 (Puelles and Rubenstein, 2003). Between prosomeres 2 and 3, zona limitans interthalamica (ZLI) exists. Transplantation of the En2-expressing isthmus region into the prosomere 3 just rostral to the ZLI induced En2 in the P2 but not in the P3 (Bloch-Gallego et al., 1996). Shimamura and his colleagues gave a clear answer to this phenomenon (Kobayashi et al., 2002). They found that transcription factors Six3 and Irx3 are expressed rostral and caudal to the ZLI, respectively. Fgf8 induces En2 in the Irx3-expressing region and the tectum differentiates there. Fgf8 is expressed in the anterior neural ridge (ANR) (Fig. 5.4). Fgf8 from the ANR induces Bf1 in the Six3-expressing region and the telencephalon differentiates there. Misexpression of Irx3 in the P3 transformed this region to differentiate into the tectum (Kobayashi et al., 2002). They also showed that Irx3 and Si3 are mutually repressing their expression. The study of Kobayashi et al. (2002) showed that competence of the tissue to the organizing signal is determined by preexisting transcription factors. In response to the Fgf8 from the isthmus, due to Otx2 expression posterior part of the midbrain differentiates into the tectum under the strong Fgf8 signal.

Midbrain patterning Chapter | 5

99

FIGURE 5.6 Fgf8 signaling for the specification of the midbrain and hindbrain. Otx2 and Gbx2 are expressed from a very early stage of development, and Fgf8 expression is induced at the border of Otx2 and Gbx2 expression. Fgf8 signaling induces En1 and Pax2 expression where Otx2 is expressed to specify the midbrain development. Transducing pathway for the midbrain specification has not been elucidated. Fgf8 activates RaseERK signaling pathway and specifies the hindbrain development. Negative regulators for the RaseERK signaling, Sprouty2, Sef and Mkp3, are induced, and differential activation level of ERK is formed in the hindbrain. The region where ERK activity is strongly repressed is specified as r1 by Irx2, and the region where ERK activity is weakly repressed is specified as r0 by Pea3. When RaseERK pathway is disrupted, fate change of the hindbrain occurs, and the tectum differentiates in place of the cerebellum. Modified after Harada, H., Sato, T., Nakamura, H., 2016. Fgf8 signaling for development of the midbrain and hindbrain. Dev. Growth Differ. 58, 437e445.

5.5.3 Intracellular signal transduction Now I move to the signal transduction of Fgf8. Fgf is received by a receptor tyrosin kinase, and the signal is transduced through several signaling cascade (Ornitz and Ito, 2015). RaseERK signaling pathway was focused for the cerebellar differentiation by Fgf8b (Sato and Nakamura, 2004). ERK activation could be detected immunohistochemically with an anti-diphosphorylated ERK (dpERK) antibody. In the normal chick embryos, ERK was activated in the isthmus corresponding to Fgf8 expression. Both Fgf8a and Fgf8b misexpression activated ERK ectopically, but in a different fashion. Fgf8b activated ERK corresponding to the Fgf8b misexpression (Fig. 5.5G). Fgf8a activated ERK only in the diencephalon (Fig. 5.5F). It is very interesting that by Fgf8b misexpression at 0.01 mg/mL, which exerted Fgf8a-type effects, ERK activation pattern was also same as that after Fgf8a misexpression (Fig. 5.5G and H). Disruption of RaseERK signaling by dominant negative form of Ras (DN-Ras) was carried out (Fig. 5.5I) (Sato and Nakamura, 2004). Misexpression of DN-Ras changed the property of the metencephalon to that of the mesencephalon, Gbx2 expression in the metencephalon was repressed, and Otx2 expression was induced in the metencephalon. Finally, a tectum differentiated in place of the cerebellum (Fig. 5.5J and K) (Sato and Nakamura, 2004). It was concluded that strong Fgf8 signal activates RaseERK signaling pathway to induce cerebellar differentiation (Fig. 5.6). Fgf8eRaseERK signaling is very strong, and activation of Fgf8eRaseERK signaling is needed for short window of developmental process so that ERK activity should be quickly downregulated (reviewed in Mason et al., 2006). Fgf8 itself induces negative regulators for RaseERK signaling in the isthmus; Sprouty2 and Sef expression is induced through Fgf8eRaseERK signaling (Echevarria et al., 2005; Suzuki-Hirano et al., 2005) and Mkp3 is induced through Fgf8ePI3KeAKT signaling (Echevarria et al., 2005). Sprouty2 may be alternatively induced by nuclear-translocated Fgf8

100 PART | I Induction and patterning of the CNS and PNS

itself (Suzuki et al., 2012). ERK is phosphorylated (activated) widely in the midbrain and hindbrain at 8 and 9 somite stage in chick embryos, but ERK phosphorylation level is still high in the midbrain but ERK phosphorylation level becomes low by 12 somite stage in the hindbrain (Sato and Nakamura, 2004). Sprouty2 has stronger repressive activity to ERK than Mkp3 (Suzuki-Hirano et al., 2010), and Sprouty2 misexpression in the midbrain/hindbrain resulted in tectum differentiation in place of the cerebellum (Suzuki-Hirano et al., 2005). Suzuki-Hirano carefully examined ERK activation in the midbrain after Fgf8b misexpression by electroporation, which forces fate change of the midbrain to the hindbrain. She noticed that ERK was phosphorylated just after electroporation kept high till 15 h after electroporation, and that ERK became dephosphorylated by 18 h after electroporation. She then raised a question what happens if ERK is continuously phosphorylated by repressing the activity of the negative regulators. To answer the question she electroporated Fgf8b and DN-Sprouty2 expression vectors in the midbrain, by which ERK phosphorylation level was kept high even after 18 h after electroporation. The fate change of the midbrain to the hindbrain did not occur (Suzuki-Hirano et al., 2010). She then turned off DN-Sprouty2 by Tet-off system (Hilgers et al., 2005; Watanabe et al., 2007) after electroporation of Fgf8b and DN-Sprouty2 expression vectors. By turning off DN-Sprouty2, ERK phosphorylation level became weak in the midbrain by 18 h after electroporation, and fate change of the midbrain to the hindbrain occurred (Suzuki-Hirano et al., 2010). Comparison of the phosphorylation level of ERK in the midbrainehindbrain region after DN-Sprouty2 and DNMkp3 misexpression showed that Sprouty2 has a stronger activity than Mkp3 (Suzuki-Hirano et al., 2010). After misexpression of Fgf8b and DNMkp3, ERK phosphorylation level in the midbrain at 18 h after electroporation kept high but lower than that after Fgf8b and DN-Sprouty2 misexpression, which is because Sprouty2 may have been induced there. In the transfected midbrain region, trochlear neurons differentiated so that it was concluded that the property of the midbrain changed to that of the isthmus (rhombomere 0, r0) (Suzuki-Hirano et al., 2010). In normal development of the chick hindbrain, ERK is activated in r0er1 at the 10 somite stage, but in the r1 ERK activity is downregulated by 13 somite stage (Suzuki-Hirano et al., 2010). Thus, after activation of ERK in the hindbrain, activity may be downregulated differentially: the region where ERK activity is weakly kept may differentiate into the isthmus (r0) and the region where ERK is strongly downregulated may acquire the property as r1 and the cerebellum differentiates (Suzuki-Hirano et al., 2010). It was concluded that strong Fgf8 signal activates RaseERK signaling pathway to organize cerebellar differentiation. But activation of ERK should be downregulated for the cerebellar differentiation, and according to differential ERK activity set by negative regulators of RaseERK signaling, isthmus and cerebellum may differentiate.

5.5.4 How tectum and cerebellum are organized by isthmus organizing signal? FgfeRaseERK signaling eventually regulates gene expression through modifications of transcription factors by activated ERK (Tsang and Dawid, 2004). As a downstream of RaseERK pathway, Pea3 subfamily of Ets-type transcription factors was suggested (Tsang and Dawid, 2004; Raible and Brand, 2011). Pea3 subfamily consists of Pea3, Erm, and Er81. In the isthmus, Pea3 and Erm are expressed in zebrafish, and their expression was induced by ectopically misexpressed Fgf8 and was repressed by FGFR inhibitor SU5402 or by Sprouty4 (Roehl and Nüsslein-Volhard, 2001; Raible and Brand, 2011). In chick embryos, Pea3 is expressed in the midbrainehindbrain region, and its expression was regulated by RaseERK pathway; disruption of RaseERK signaling by DN-Ras repressed Pea3 expression (Harada et al., 2015). Gain and loss of function of Pea3 exerted similar effects as Fgf8b and DN-Ras misexpression on Otx2, Gbx2, and Fgf8 expression, respectively. Pea3 misexpression resulted in repression of Otx2 in the midbrain, and rostral extension of Gbx2 and Fgf8 expression. Repression of Pea3 function by misexpression of a chimeric molecule of engrailed repressor domain EH1 and Pea3 (eh1-Pea3) resulted in upregulation of Otx2 in the hindbrain and repression of Gbx2 and Fgf8 in the hindbrain (Harada et al., 2015). From the effects on downstream gene expression, it was expected that Pea3 would function to specify hindbrain at the downstream of ERK, so that Pea3 misexpression would induce cerebellum in the midbrain (Harada et al., 2015). But the cerebellum was not differentiated after Pea3 misexpression in the midbrain region. Instead, isthmus (rhombomere 0, r0) structure extended rostrally in case of Pea3 misexpression; trochlear neurons differentiated rostral to the proper trochlear nucleus. Repression of Pea3 function by eh1-Pea3 resulted in caudal extension of the midbrain; oculomotor neurons differentiated caudal to the oculomotor nucleus. These results indicated that Pea3 specifies isthmus (r0) under RaseERK pathway (Fig. 5.6) (Harada et al., 2015). Recently, it was reported that bHLH transcription factor Atoh1 is expressed in the rostral part of the hindbrain, and that Atoh1-expressing region would give rise to isthmic nuclei in mouse and chick (Green et al., 2014). Experimental results suggest that Pea3 may be responsible for the specification of Atoh1-expressing region. For the specification of the r1 region, Irx2 may be essential (Fig. 5.6). Fgf8b could ectopically induce Irx2 in the midbrain, but Fgf8a could not (Sato et al., 2001). Consequently, Fgf8a could not change the fate of the midbrain. Comisexpession of Fgf8a and Irx2 could induce the ectopic cerebellum in the midbrain region. Thus, Irx2 expression may be induced through RaseERK signaling. Irx2 could induce Fgf8 and Gbx2 expression in the midbrain and convert the fate of this region from the tectum to the cerebellum (Matsumoto et al., 2004).

Midbrain patterning Chapter | 5

101

In hindbrain region (r0 and r1), RaseERK pathway is activated by strong Fgf8 signal. Differential ERK activation level is created by negative regulators of the RaseERK signaling. ERK is deactivated in the r1 region, where Irx2 is activated and the cerebellum differentiates. In the r0, ERK activity is weakened, but remains. There Pea3 is activated and differentiates as the isthmus.

5.6 Concluding remarks Here, I focused on midbrain especially on the optic tectum. Since the tectum is the visual center in lower vertebrates and receives retinal fibers in a retinotopic manner, I mentioned how the polarity of the tectum for the precise retinotopic projection is made, and the molecular mechanisms of retinotopic projection. Through classic transplantation experiment, midbrainehindbrain boundary, isthmus, was shown to function as the secondary organizer for the midbrain and hindbrain, and Fgf8 was identified as the organizing signal. I reviewed these process and signal transduction of Fgf8. I myself found that alar plate of the diencephalon could change its fate to differentiate into the tectum when transplanted in the posterior midbrain. I also found that the rostrocaudal polarity of the tectum is plastic by rotation of the midbrain. I was very much struck when the isthmus emanates organizing signal, which explains the fate change of the diencephalon at the posterior midbrain and the plasticity of the rostrocaudal axis of the midbrain.

List of abbreviations of genes and molecules Atoh1 Atonal homology 1 Bf Brain factor bHLH Basic helix loop helix EH1 Engrailed homology 1 En Engrailed Eph Named for its expression in erythropoietin-producing human hepatocellular carcinoma cell line ERK Extracellular signal-regulated kinase Erm Ets-related molecule Ets E26 transformation specific Fgf Fibroblast growth factor FGFR Fgf receptor Gbx Gastrulation brain homeobox Grg Groucho-related gene Gro Drosophila groucho Hnf Hepatocyte nuclear factors Irx Iroquois homeobox protein Lim Lin-1, Islet-1, Mec-3 MAPK Mitogen-activated protein kinase MKP MAP kinase phosphatase Mkp MAP kinase phosphatase otx Orthodenticle homolog pax Paired box gene Pea3 Polyomavirus enhancer activator PI3K Phosphatidylinositol-3 kinase Ptc Patched Ras Rat sarcoma Sef Similar expression for fgf Shh Sonic hedgehog Tcf T-cell factor TLE Transducin-like enhancer of split Wnt Drosophila wingless þ int-1

List of abbreviations (general) ANR Anterior neural ridge AVE Anterior visceral endoderm DN Dominant negative Dox Doxycycline DV Dorsoventral E2 Embryonic day 2

102 PART | I Induction and patterning of the CNS and PNS

E2, E10 Embryonic day 2, embryonic day 10 ELF Eph ligand family GABA Gamma aminobutyric acid GPI Glycosylphosphatidylinositol LM Lateromedial P1, P2 Prosomere 1, prosomere 2 r0, r1 Rhombomere 1, rhombomere 2 RAGS Repulsive axon guidance signal SAC Stratum album centrale SFP Stratum fibrosum periventriculare SGC Stratum griseum centrale SGFS Stratum griseum et fibrosum superficiale SGP Stratum periventriculare SO Stratum opticum Tet Tetracycline ZLI Zona limitans interthalamica

Glossary In ovo electroporation Chick embryos have been served as model animals for experimental embryology since embryos are easily manipulated during embryogenesis. In particular, transplantation between chick and quail embryos have enabled close monitoring of the fate of long migrating cells during development based on the difference in distribution of heterochromatin in the nucleus between chick and quail (Le Douarin, 1982, 2008). However, production of transgenic chicken and gene targeting has been difficult in the chick, because the embryos remain in the oviduct and uterus for the first 24 h of embryogenesis, which has limited analyses of gene function, and led to a gradual decrease in interest in developmental model.In ovo electroporation was developed as a new gene transfer method to chick embryos first by Muramatsu and his colleagues (Muramatsu et al., 1997). This is a very simple method: we have only to prepare an expression vector by inserting cDNA into an appropriate expression plasmid, locally apply this plasmid to a chick embryo in ovo, and charge rectangular electric pulses. Small pores are made on the membrane by an electric field, and through the pores, expression vectors can enter the cells. If the strength of the pulse length and duration are appropriate, pores are closed after removal of the field. Then DNA enters the nucleus and is transcribed under the enhancer and promoter of the vector. We revised this method to efficiently and constantly introduce a gene of interest in the neural tube of chick embryos (Funahashi et al., 1999). To transfect the neural tube, we only insert expression plasmids into central canal and charge a 10e25 V, 50 ms/s rectangular pulse several times. Since DNA is negatively charged, transfection occurs at the anode side of the wall of the neural tube. Cathode side can be served as the control. By this method, translation product of the transfected DNA is observed only 2 h after electroporation (Funahashi et al., 1999).For loss of function of a gene of interest, electroporation of shRNA (short hairpin RNA) expression vector is effective (Katahira and Nakamura, 2003). Short hairpin RNA interferes with the target RNA (Svoboda et al., 2001). This technique allows us to knockdown the gene of interest at the desired place and time in chick embryos.Transposon-mediated gene transfer in chick by electroporation was developed by Takahashi’s group (Sato et al., 2007). They adopted Tol2 transposon, which was isolated from Japanese medaka fish. Genes in the transposon vector are integrated into the genome in the presence of transposase so that we can get long-term expression of the transgene. In addition, combination of the Tet-on or Tet-off system and transposon system made it possible to control the timing of expression of a transferred gene for a longer period (Hilgers et al., 2005; Sato et al., 2007; Watanabe et al., 2007). Now electroporation system has become a routine method in the study of developmental biology, and chick embryos were revived as model animals for the study of developmental biology. Prosomere Metameric pattern in the vertebrate caudal prosencephalon is called prosomere. Prosomeres 1 to 3 (P1 to P3) are proposed based on the gene expression pattern, structure, and function (Puelles and Rubenstein, 2003). P1 differentiates into the pretectum, P2 into the habenula and thalamus, and P3 into the prethalamus and prethalamic eminence (Puelles and Rubenstein, 2003). Rhombomere Segmentation in the hindbrain of the vertebrates is called rhombomere. Rhombomeres consist of compartments, that is, the cells do not move across the boundary of the rhombomere (Fraser et al., 1990). Hox gene is expressed rhombomere specifically and combination of Hox proteins expressed defines the characteristics of the rhombomere (Hunt et al., 1991). For detail see Chapter 9. Tet-on, Tet-off system This system enables expression of the gene of interest by administrating tetracyclin. The transcriptional activators, reverse tet-controlled transcriptional activator (rtTA) and tet-controlled transcriptional activator (tTA), act on the cis-element promoter, tetracycline responsive element (TRE). rtTA could be used for Tet-on, since it binds TRE only in the presence of doxycycline (an analog of tetracycline; Dox) and activates transcription of the TRE-driven gene. tTA could be used for Tet-off, since it binds to the TRE and activates the TRE-driven gene in the absence of Dox. Whereas in the presence of Dox, tTA is released from TRE, leading to an inactivation of the TRE-driven gene (Watanabe et al., 2007). Transposon system Transposons are genetic elements that move from one locus in the genome to another. The Tol2 transposable element, which was originally found in medaka (Koga et al., 1996), was shown to be capable of undergoing transposition in chick embryos (Sato et al., 2007). When a DNA plasmid that contains a transposon construct carrying a gene expression cassette is introduced into vertebrate cells with the transposase activity, the transposon construct is excised from the plasmid, and the cassette is subsequently integrated into the host genome (Kawakami, 2005). Electroporation of a gene cassette cloned in the Tol2 transposon vector together with transposase expression vector has permitted stable expression of the gene in the chicken cells.Through combination of transposon and Tet-on or Tet-off system, we can conditionally express gene of interest at desired site at desired developmental period.

Midbrain patterning Chapter | 5

103

References Acampora, D., Mazan, S., Lallemand, Y., Avantaggiato, V., Maury, M., Simeone, A., Brület, P., 1995. Forebrain and midbrain regions are deleted in Otx2/ mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121, 3279e3290. Acampora, D., Mazan, S., Avantaggiato, V., Barone, P., Tuorto, F., Lallemand, Y., Brület, P., Simeone, A., 1996. Epilepsy and brain abnormalities in mice lacking Otx1 gene. Nat. Genet. 14, 218e222. Alvarado-Mallart, R.M., Sotelo, C., 1984. Homotopic and heterotopic transplantations of quail tectal primordia in chick embryos: organization of the retinotectal projections in the chimeric embryos. Dev. Biol. 103, 378e398. Ang, S.L., Jin, O., Rhinn, M., Daigle, N., Stevenson, L., Rossant, J., 1996. A targeted mouse Otx2 mutation leads to severe defects in gastrulation and formation of axial mesoderm and to deletion of rostral brain. Development 122, 243e252. Araki, I., Nakamura, H., 1999. Engrailed defines the position of dorsal diemesencephalic boundary by repressing diencephalic fate. Development 126, 5127e5135. Bloch-Gallego, E., Millet, S., Alvarado-Mallart, R.M., 1996. Further observations on the susceptibility of diencephalic prosomeres to En-2 induction and on the resulting histogenetic capabilities. Mech. Dev. 58, 51e63. Brand, M., Heisenberg, C.P., Jiang, Y.J., Beuchle, D., Lun, K., Furutani-Seiki, M., Granato, M., Haffter, P., Hammerschmidt, M., Kane, D.A., Kelsh, R.N., Mullins, M.C., Odenthal, J., van Eeden, F.J., Nüsslein-Volhard, C., 1996. Mutations in zebrafish genes affecting the formation of the boundary between midbrain and hindbrain. Development 123, 179e190. Broccoli, V., Boncinelli, E., Wurst, W., 1999. The caudal limit of Otx2 expression positions the isthmic organizer. Nature 401, 164e168. Brunet, I., Wein, C., Piper, M., Trembleau, A., Volovitch, M., Harris, W., Prochiantz, A., Holt, C., 2005. The transcription factor Engrailed-2 guides retinal axons. Nature 438, 94e98. Canning, C.A., Lee, L., Irving, C., Mason, I., Jones, C.M., 2007. Sustained interactive Wnt and FGF signaling is required to maintain isthmic identity. Dev. Biol. 305, 276e286. Chen, G., Courey, A.J., 2000. Groucho/TLE family proteins and transcriptional repression. Gene 249, 1e16. Cheng, H.J., Flanagan, J.G., 1994. Identification and cloning of ELF-1, a developmentally expressed ligand for the Mek4 and Sek receptor tyrosine kinases. Cell 79, 157e168. Cheng, H.J., Nakamoto, M., Bergemann, A.D., Flanagan, J.G., 1995. Complementary gradients in expression and binding of ELF-1 and Mek4 in development of the topographic retinotectal projection map. Cell 82, 371e381. Chi, C.L., Martinez, S., Wurst, W., Martin, G.R., 2003. The isthmic organizer signal FGF8 is required for cell survival in the prospective midbrain and cerebellum. Development 130, 2633e2644. Crossland, W.J., Uchwat, C.J., 1979. Topographic projections of the retina and optic tectum upon the ventral lateral geniculate nucleus in the chick. J. Comp. Neurol. 185, 87e106. Crossley, P.H., Martin, G.R., 1995. The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439e451. Crossley, P.H., Martinez, S., Martin, G.R., 1996. Midbrain development induced by FGF8 in the chick embryo. Nature 380, 66e68. Drescher, U., Kremoser, C., Handwerker, C., Löschinger, J., Noda, M., Bonhoeffer, F., 1995. In vitro guidance of retinal ganglion cell axons by RAGS, a 25 kDa tectal protein related to ligands for Eph receptor tyrosine kinases. Cell 82, 359e370. Dubbeldam, J.L., 1998. Mesencephalon. In: The Central Nervous System of Vertebrates, vol. 3. Springer-Verlag Berlin Heidelberg, pp. 1568e1580. Echelard, Y., Epstein, D.J., St-Jacques, B., Shen, L., Mohler, J., McMahon, J.A., McMahon, A.P., 1993. Sonic Hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75, 1417e1430. Echevarria, D., Martinez, S., Marques, S., Lucas-Teixeira, V., Belo, J.A., 2005. Mkp3 is a negative feedback modulator of Fgf8 signaling in the mammalian isthmic organizer. Dev. Biol. 277, 114e128. Ericson, J., Morton, S., Kawakami, A., Roelink, H., Jessell, T.M., 1996. Two crucial periods of sonic hedgehog signaling required for the specification of motor neuron identity. Cell 87, 661e673. Fagiolini, M., Hensch, T.K., 2000. Inhibitory threshold for critical-period activation in primary visual cortex. Nature 404, 183e186. Favor, J., Sandulache, R., Neuhäuser-Klaus, A., Pretsch, W., Chatterjee, B., Senft, E., Wurst, W., Blanquet, V., Grimes, P., Spörle, R., Schughart, K., 1996. The mouse Pax2(1Neu) mutation is identical to a human PAX2 mutation in a family with renalecoloboma syndrome and results in developmental defects of the brain, ear, eye, and kidney. Proc. Natl. Acad. Sci. U.S.A. 93, 13870e13875. Feldheim, D.A., Kim, Y.I., Bergemann, A.D., Frisén, J., Barbacid, M., Flanagan, J.G., 2000. Genetic analysis of ephrin-A2 and ephrin-A5 shows their requirement in multiple aspects of retinocollicular mapping. Neuron 25, 563e574. Feldheim, D.A., Nakamoto, M., Osterfield, M., Gale, N.W., DeChiara, T.M., Rohatgi, R., Yancopoulos, G.D., Flanagan, J.G., 2004. Loss-of-function analysis of EphA receptors in retinotectal mapping. J. Neurosci. 24, 2542e2550. Fraser, S., Keynes, R., Lumsden, A., 1990. Segmentation in the chick embryo hindbrain is defined by cell lineage restrictions. Nature 344, 431e435. Friedman, G.C., O’Leary, D.D., 1996. Retroviral misexpression of engrailed genes in the chick optic tectum perturbs the topographic targeting of retinal axons. J. Neurosci. 16, 5498e5509. Frisén, J., Yates, P.A., McLaughlin, T., Friedman, G.C., O’Leary, D.D., Barbacid, M., 1998. Ephrin-A5 (AL-1/RAGS) is essential for proper retinal axon guidance and topographic mapping in the mammalian visual system. Neuron 20, 235e243. Funahashi, J.-I., Okafuji, T., Ohuchi, H., Noji, S., Tanaka, H., Nakamura, H., 1999. Role of Pax-5 in the regulation of a mid-hindbrain organizer’s activity. Dev. Growth Differ. 41, 59e72. Green, M.J., Myat, A.M., Emmenegger, B.A., Wechsler-Reya, R.J., Wilson, L.J., Wingate, R.J., 2014. Independently specified Atoh1 domains define novel developmental compartments in rhombomere 1. Development 141, 389e398.

104 PART | I Induction and patterning of the CNS and PNS

Grindley, J.C., Hargett, L.K., Hill, R.E., Ross, A., Hogan, B.L., 1997. Disruption of PAX6 function in mice homozygous for the Pax6Sey-1Neu mutation produces abnormalities in the early development and regionalization of the diencephalon. Mech. Dev. 64, 111e126. Gruss, P., Walther, C., 1992. Pax in development. Cell 69, 719e722. Hanks, M., Wurst, W., Anson-Cartwright, L., Auerbach, A.B., Joyner, A.L., 1995. Rescue of the En-1 mutant phenotype by replacement of En-1 with En-2. Science 269, 679e682. Harada, H., Omi, M., Sato, T., Nakamura, H., 2015. Pea3 determines the isthmus region at the downstream of Fgf8-Ras-ERK signaling pathway. Dev. Growth Differ. 57, 657e666. Harada, H., Sato, T., Nakamura, H., 2016. Fgf8 signaling for development of the midbrain and hindbrain. Dev. Growth Differ. 58, 437e445. Hidalgo-Sánchez, M., Simeone, A., Alvarado-Mallart, R.-M., 1999. Fgf8 and Gbx2 induction concomitant with Otx2 repression is correlated with midbrainehindbrain fate of caudal prosencephalon. Development 126, 3191e3203. Hilgers, V., Pourquié, O., Dubrulle, J., 2005. In vivo analysis of mRNA stability using the Tet-Off system in the chicken embryo. Dev. Biol. 284, 292e300. Hunt, P., Whiting, J., Muchamore, I., Marshall, H., Krumlauf, R., 1991. Homeobox genes and models for patterning the hindbrain and branchial arches. Dev. Suppl. 187e196. Ichijo, H., Fujita, S., Matsuno, T., Nakamura, H., 1990. Rotation of the tectal primordium reveals plasticity of target recognition in retino-tectal projection. Development 110, 331e342. Irving, C., Mason, I., 1999. Regeneration of isthmic tissue is the result of a specific and direct interaction between rhombomere 1 and midbrain. Development 126, 3981e3989. Itasaki, N., Nakamura, H., 1992. Rostrocaudal polarity of tectum in birds: correlation of en gradient and topographic order in retinotectal projection. Neuron 8, 787e798. Itasaki, N., Nakamura, H., 1996. A role of gradient en expression in positional specification on the optic tectum. Neuron 16, 55e62. Itasaki, N., Ichijo, H., Hama, C., Matsuno, T., Nakamura, H., 1991. Establishment of rostrocaudal polarity in tectal primordium: engrailed expression and subsequent tectal polarity. Development 113, 1133e1144. Joyner, A.L., Lebo, R.V., Kan, Y.W., Tjian, R., Cox, D.R., Martin, G.R., 1985. Comparative chromosome mapping of a conserved homoeo box region in mouse and human. Nature 314, 173e175. Joyner, A.L., Herrup, K., Auerbach, A., Davis, C.A., Rossant, J., 1991. Subtle cerebellar phenotype in mice homozygous for a targeted deletion of the En2 homeobox. Science 251, 1239e1243. Katahira, T., Nakamura, H., 2003. Gene silencing in chick embryos with a vector-based small interfering RNA system. Dev. Growth Differ. 45, 361e367. Katahira, T., Sato, T., Sugiyama, S., Okafuji, T., Araki, I., Funahashi, J., Nakamura, H., 2000. Interaction between Otx2 and Gbx2 defines the organizing center for the optic tectum. Mech. Dev. 91, 43e52. Kawakami, K., 2005. Transposon tools and methods in zebrafish. Dev. Dynam. 234, 244e254. Kimura, C., Yoshinaga, K., Tian, E., Suzuki, M., Aizawa, S., Matsuo, I., 2000. Visceral endoderm mediates forebrain development by suppressing posteriorizing Signals. Dev. Biol. 225, 304e321. Kobayashi, E., Kobayashi, M., Matsumoto, K., Ogura, T., Nakafuku, M., Shimamura, K., 2002. Early subdivisions in the neural plate define distinct competence for inductive signals. Development 129, 83e93. Koga, A., Suzuki, M., Inagaki, H., Bessho, Y., Hori, H., 1996. Transposable element in fish. Nature 383, 30. LaVail, J.H., Cowan, W.M., 1971a. The development of the chick optic tectum. I. Normal morphology and cytoarchitectonic development. Brain Res. 28, 391e419. LaVail, J.H., Cowan, W.M., 1971b. The development of the chick optic tectum. II. Autoradiographic studies. Brain Res. 28, 421e441. Le Douarin, N.M., 1973. A biological cell labeling technique and its use in experimental embryology. Dev. Biol. 30, 217e222. Le Douarin, N., 1982. The Neural Crest. Cambridge University Press, Cambridge. Le Douarin, N.M., 2008. Developmental patterning deciphered in avian chimeras. Dev. Growth Differ. 50, S11eS28. Liu, A., Losos, K., Joyner, A.L., 1999. FGF8 can activate Gbx2 and transform regions of the rostral mouse brain into a hindbrain fate. Development 126, 4827e4838. Liu, A., Li, J.Y.H., Bromleigh, C., Lao, Z., Niswander, L.A., Joyner, A.L., 2003. FGF17 and FGF18 have different midbrain regulatory properties from FGF8b or activated FGF receptors. Development 130, 6175e6185. Logan, C., Wizenmann, A., Drescher, U., Monschau, B., Bonhoeffer, F., Lumsden, A., 1996. Rostral optic tectum acquires caudal characteristics following ectopic Engrailed expression. Curr. Biol. 6, 1006e1014. MacArthur, C.A., Lawshé, A., Shankar, D.B., Heikinheimo, M., Shackleford, G.M., 1995. FGF-8 isoforms differ in NIH3T3 cell transforming potential. Cell Growth Differ. 6, 817e825. Martinez, S., Alvarado-Mallart, R.-M., 1989. Rostral cerebellum originates from caudal portion of the so-called ‘mesencephalic’vesicle: a study using chick/quail chimeras. Eur. J. Neurosci. 1, 549e560. Martinez, S., Alvarado-Mallart, R.M., 1990. Expression of the homeobox Chick-en gene in chick/quail chimeras with inverted mes-metencephalic grafts. Dev. Biol. 139, 432e436. Martinez, S., Wassef, M., Alvarado-Mallart, R.M., 1991. Induction of a mesencephalic phenotype in the 2-day-old chick prosencephalon is preceded by the early expression of the homeobox gene en. Neuron 6, 971e981. Martínez, S., Marín, F., Nieto, M.A., Puelles, L., 1995. Induction of ectopic engrailed expression and fate change in avian rhombomeres: intersegmental boundaries as barriers. Mech. Dev. 51, 289e303.

Midbrain patterning Chapter | 5

105

Martinez, S., Crossley, P.H., Cobos, I., Rubenstein, J.L., Martin, G.R., 1999. FGF8 induces formation of an ectopic isthmic organizer and isthmocerebellar development via a repressive effect on Otx2 expression. Development 126, 1189e1200. Maruoka, Y., Ohbayashi, N., Hoshikawa, M., Itoh, N., Hogan, B.L.M., Furuta, Y., 1998. Comparison of the expression of three highly related genes, Fgf8, Fgf17 and Fgf18, in the mouse embryo. Mech. Dev. 74, 175e177. Mason, J.M., Morrison, D.J., Basson, M.A., Licht, J.D., 2006. Sprouty proteins: multifaceted negative-feedback regulators of receptor tyrosine kinase signaling. Trends Cell Biol. 16, 45e54. Mastick, G.S., Davis, N.M., Andrew, G.L., Easter Jr., S.S., 1997. Pax-6 functions in boundary formation and axon guidance in the embryonic mouse forebrain. Development 124, 1985e1997. Matsumoto, K., Nishihara, S., Kamimura, M., Shiraishi, T., Otoguro, T., Uehara, M., Maeda, Y., Ogura, K., Lumsden, A., Ogura, T., 2004. The prepattern transcription factor Irx2, a target of the FGF8/MAP kinase cascade, is involved in cerebellum formation. Nat. Neurosci. 7, 605e612. Matsunaga, E., Araki, I., Nakamura, H., 2000. Pax6 defines the diemesencephalic boundary by repressing En1 and Pax2. Development 127, 2357e2365. Matsunaga, E., Araki, I., Nakamura, H., 2001. Role of Pax3/7 in the tectum regionalization. Development 128, 4069e4077. Matsunaga, E., Katahira, T., Nakamura, H., 2002. Role of Lmx1b and Wnt1 in mesencephalon and metencephalon development. Development 129, 5269e5277. Matsuo, I., Kuratani, S., Kimura, C., Takeda, N., Aizawa, S., 1995. Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev. 9, 2646e2658. McLaughlin, T., Hindges, R., Yates, P.A., O’Leary, D.D.M., 2003. Bifunctional action of ephrin-B1 as a repellent and attractant to control bidirectional branch extension in dorsaleventral retinotopic mapping. Development 130, 2407e2418. McMahon, A.P., Bradley, A., 1990. The Wnt-1 (int-1) proto-oncogene is required for development of a large region of the mouse brain. Cell 62, 1073e1085. Meyers, E.N., Lewandoski, M., Martin, G.R., 1998. An Fgf8 mutant allelic series generated by Cre- and Flp-meddiated recombination. Nat. Genet. 18, 136e141. Millet, S., Bloch-Gallego, E., Simeone, A., Alvarado-Mallart, R.-M., 1996. The caudal limit of Otx2 gene expression as a marker of the midbrain/ hindbrain boundary: a study using in situ hybridization and chick/quail homotopic grafts. Development 122, 3785e3797. Millet, S., Campbell, K., Epstein, D.J., Losos, K., Harris, E., Joyner, A.L., 1999. A role for Gbx2 in repression of Otx2 and positioning the mid/hindbrain organizer. Nature 401, 161e164. Muramatsu, T., Mizutani, Y., Ohmori, Y., Okumura, J., 1997. Comparison of three nonviral transfection methods for foreign gene expression in early chicken embryos in ovo. Biochem. Biophys. Res. Commun. 230, 376e380. Nakamoto, M., Cheng, H.J., Friedman, G.C., McLaughlin, T., Hansen, M.J., Yoon, C.H., O’Leary, D.D., Flanagan, J.G., 1996. Topographically specific effects of ELF-1 on retinal axon guidance in vitro and retinal axon mapping in vivo. Cell 86, 755e766. Nakamura, H., 1990. Do CNS anlagen have plasticity in differentiation? Analysis in quail-chick-chimera. Brain Res. 511, 122e128. Nakamura, H., 2001a. Regionalisation of the optic tectum: combination of the gene expression that defines the tectum. Trends Neurosci. 24, 32e39. Nakamura, H., 2001b. Regionalisation and polarity formation of the optic tectum. Prog. Neurobiol. 65, 473e488. Nakamura, H., Nakano, K.E., Igawa, H.H., Takagi, S., Fujisawa, H., 1986. Plasticity and rigidity of differentiation of brain vesicles studied in quailechick chimeras. Cell Differ. 19, 187e193. Nakamura, H., Takagi, S., Tsuji, T., Matsui, K.A., Fujisawa, H., 1988. The Prosencephalon has the capacity to differentiate into the optic tectum: analysis by chick-specific monoclonal antibodies in quail-chick-chimeric brains. Dev. Growth Differ. 30, 717e725. Nakamura, H., Matsui, K.A., Takagi, S., Fujisawa, H., 1991. Projection of the retinal ganglion cells to the tectum differentiated from the prosencephalon. Neurosci. Res. 11, 189e197. Nakamura, H., Itasaki, N., 1992. Expression of en in the prosencephalon heterotopically transplanted into the mesencephalon. Develop. Growth Differ. 34, 387e391. Nakamura, H., Katahira, T., Matsunaga, E., Sato, T., 2005. Isthmus organizer for midbrain and hindbrain development. Brain Res. Rev. 49, 120e126. Nakamura, H., Sato, T., Suzuki-Hirano, A., 2008. Isthmus organizer for mesencephalon and metencephalon. Dev. Growth Differ. 50, S113eS118. Niss, K., Leutz, A., 1998. Expression of the homeobox gene GBX2 during chicken development. Mech. Dev. 76, 151e155. Nomura, T., Fujisawa, H., 2000. Alteration of the retinotectal projection map by the graft of mesencephalic floor plate or sonic hedgehog. Development 127, 1899e1910. Okafuji, T., Funahashi, J., Nakamura, H., 1999. Roles of Pax-2 in initiation of the chick tectal development. Dev. Brain Res. 116, 41e49. Olsen, S.K., Li, J.Y.H., Bromleigh, C., Eliseenkova, A.V., Ibrahimi, O.A., Lao, Z., Zhang, F., Linhardt, R.J., Joyner, A.L., Mohammadi, M., 2006. Structural basis by which alternative splicing modulates the organizer activity of FGF8 in the brain. Genes Dev. 20, 185e198. Omi, M., Harada, H., Nakamura, H., 2011. Identification of retinotectal projection pathway in the deep tectal laminae in the chick. J. Comp. Neurol. 519, 2615e2621. Omi, M., Harada, H., Watanabe, Funahashi, J., Nakamura, H., 2014. Role of En2 in the tectal laminar formation of chick embryos. Development 141, 2131e2138. Omi, M., Nakamura, H., 2015. Engrailed and tectum development. Dev. Growth Differ. 57, 135e145. Ornitz, D.M., Ito, N., 2015. The fibroblast growth factor signaling pathway. Wiley Interdiscip. Rev. Dev. Biol. 4, 215e266. Patel, N.H., Martin-Blanco, E., Coleman, K.G., Poole, S.J., Ellis, M.C., Kornberg, T.B., Goodman, C.S., 1989. Expression of engrailed proteins in arthropods, annelids, and chordates. Cell 58, 55e68.

106 PART | I Induction and patterning of the CNS and PNS

Pfeffer, P.L., Gerster, T., Lun, K., Brand, M., Busslinger, M., 1998. Characterization of three novel members of the zebrafish Pax2/5/8 family: dependecy of Pax5 and Pax8 expression on the Pax2.1 (noi ) function. Development 125, 3063e3074. Pfeffer, P.L., Bouchard, M., Busslinger, M., 2000. Pax2 and homeodomain proteins cooperatively regulates a 435 bp enhancer of the mouse Pax5 gene at the midbrainehindbrain boundary. Development 127, 1017e1028. Prochiantz, A., Joliot, A., 2003. Can transcription factors function as cell-cell signalling molecules? Nat. Rev. Mol. Cell Biol. 4, 814e819. Puelles, L., Rubenstein, J.L., 2003. Forebrain gene expression domains and the evolving prosomeric model. Trends Neurosci. 26, 469e476. Raible, F., Brand, M., 2011. Tight transcriptional control of the ETS domain factors Erm and Pea3 by Fgf signaling during early zebrafish development. Mech. Dev. 107, 105e117. Reifers, F., Adams, J., Mason, I.J., Schulte-Merkerm, S., Brand, M., 2000. Overlapping and distinct functions provided by fgf17, a new zebrafish member of the Fgf8/17/18 subgroup of Fgfs. Mech. Dev. 99, 39e49. Roehl, H., Nüsslein-Volhard, C., 2001. Zebrafish pea3 and erm are general targets of FGF8 signaling. Curr. Biol. 11, 503e507. Roelink, H., Porter, J.A., Chiang, C., Tanabe, Y., Chang, D.T., Beauchy, P.A., Jessell, T.M., 1995. Floor plate and motor neuron induction by different concentrations of the amino-terminal cleavage product of Sonic Hedgehog autoproteolysis. Cell 81, 445e455. Sato, T., Nakamura, H., 2004. The Fgf8 signal causes cerebellar differentiation by activating Ras-ERK signaling pathway. Development 131, 4275e4285. Sato, T., Araki, I., Nakamura, H., 2001. Inductive signal and tissue responsiveness defining the tectum and the cerebellum. Development 128, 2466e2469. Sato, T., Joyner, A.L., Nakamura, H., 2004. How does Fgf signaling from the isthmic organizer induce midbrain and cerebellum development? Dev. Growth Differ. 46, 487e494. Sato, Y., Kasai, T., Nakagawa, S., Tanabe, K., Watanabe, T., Kawakami, K., Takahashi, Y., 2007. Stable integration and conditional expression of electroporated transgenes in chicken embryos. Dev. Biol. 305, 616e624. Shamim, H., Mason, I., 1998. Expression of Gbx-2 during early development of the chick embryo. Mech. Dev. 76, 157e159. Shamim, H., Mahmood, R., Logan, C., Doherty, P., Lumsden, A., Mason, I., 1999. Sequential roles for Fgf4, En1 and Fgf8 in specification and regionalisation of the midbrain. Development 126, 945e959. Shigetani, Y., Funahashi, J., Nakamura, H., 1997. En-2 regulates the expression of the ligands for Eph type tyrosine kinases in chick embryonic tectum. Neurosci. Res. 27, 211e217. Simeone, A., Acampora, D., Gulisano, M., Stornaiuolo, A., Boncinelli, E., 1992. Nested expression domains of four homeobox genes in developing rostral brain. Nature 358, 687e690. Sperry, R.W., 1943. Visuomotor coordination in the newt (Triturus viridescens) after regeneration of the optic nerve. J. Comp. Neurol. 79, 33e55. Sperry, R.W., 1963. Chemoaffinity in the orderly growth of nerve fiber patterns and connections. Proc. Natl. Acad. Sci. U.S.A. 50, 703e710. Stoykova, A., Fritsch, R., Walther, C., Gruss, P., 1996. Forebrain patterning defects in Small eye mutant mice. Development 122, 3453e3465. Suda, Y., Matsuo, I., Aizawa, S., 1997. Cooperation between Otx1 and Otx2 genes in developmental patterning of rostral brain. Mech. Dev. 69, 125e141. Sugiyama, S., Nakamura, H., 2003. The role of Grg4 in tectal laminar formation. Development 130, 451e462. Sugiyama, S., Funahashi, J., Nakamura, H., 2000. Antagonizing activity of chick Grg4 against tectum-organizing activity. Dev. Biol. 221, 168e180. Suzuki, A., Harada, H., Nakamura, H., 2012. Nuclear translocation of FGF8 and its implication to induce Sprouty2. Dev. Growth Differ. 54, 463e473. Suzuki-Hirano, A., Sato, T., Nakamura, H., 2005. Regulation of isthmic Fgf8 signaling by Sprouty2. Development 132, 257e265. Suzuki-Hirano, A., Harada, H., Sato, T., Nakamura, H., 2010. Activation of Ras-ERK pathway by Fgf8 and its down regulation by Sprouty2 for the isthmus organizing activity. Dev. Biol. 337, 284e293. Svoboda, P., Stein, P., Schultz, R.M., 2001. RNAi in mouse oocytes and preimplantation embryos: effectiveness of hairpin dsRNA. Biochem. Biophys. Res. Commun. 287, 1099e1104. Terashima, T., 2011. Kobe Note of Human Neuroanatomy (in Japanese). Kinpodo, Kyoto. Tsang, M., Dawid, I.B., 2004. Promotion and attenuation of FGF signaling through Ras-MAPK pathway. Sci. STKE 2004 (228), pe17. Urbánek, P., Fetka, I., Meisler, M.H., Busslinger, M., 1997. Cooperation of Pax2 and Pax5 in midbrain and cerebellum development. Proc. Natl. Acad. Sci. U.S.A. 94, 5703e5708. Walter, J., Henke-Fahle, S., Bonhoeffer, F., 1987a. Avoidance of posterior tectal membranes by temporal retinal axons. Development 101, 909e913. Walter, J., Kern-Veits, B., Huf, J., Stolze, B., Bonhoeffer, F., 1987b. Recognition of position-specific properties of tectal cell membranes by retinal axons in vitro. Development 101, 685e696. Walther, C., Gruss, P., 1991. Pax-6, a murine paired box gene, is expressed in the developing CNS. Development 113, 1435e1449. Wassarman, K.M., Lewandoski, M., Campbell, K., Joyner, A.L., Rubenstein, J.L.R., Martinez, S., Martin, G.R., 1997. Specification of the anterior hindbrain and establishment of a normal mid/hindbrain organizer is dependent on Gbx2 gene function. Development 124, 2923e2934. Watanabe, Y., Nakamura, H., 2000. Control of chick tectum territory along dorsoventral axis by sonic hedgehog. Development 127, 1131e1140. Watanabe, T., Saito, D., Tanabe, K., Suetsugu, R., Nakaya, Y., Nakagawa, S., Takahashi, Y., 2007. Tet-on inducible system combined with in ovo electroporation dissects multiple roles of genes in somitogenesis of chicken embryos. Dev. Biol. 305, 625e636. Wurst, W., Auerbach, A.B., Joyner, A.L., 1994. Multiple developmental defects in Engrailed-1 mutant mice: an early midehindbrain deletion and patterning defects in forelimbs and sternum. Development 120, 2065e2075. Xu, J., Lawshe, A., MacArthur, C.A., Ornitz, D.M., 1999. Genomic structure, mapping, activity and expression of fibroblast growth factor 17. Mech. Dev. 83, 165e178. Xu, J., Liu, Z., Ornitz, D.M., 2000. Temporal and spatial gradients of Fgf8 and Fgf17 regulate proliferation and differentiation of midline cerebellar structures. Development 127, 1833e1843. Yamagata, M., Sanes, J.R., 1995. Lamina-specific cues guide outgrowth and arborization of retinal axons in the optic tectum. Development 121 (1), 189e200.

Chapter 6

Cerebellar patterning Ludovic Galas1, Yutaro Komuro2, Nobuhiko Ohno3, 4, David Vaudry1, 5 and Hitoshi Komuro6 1

Normandie University, UNIROUEN, INSERM, PRIMACEN, Mont-Saint-Aignan, France; 2Department of Neurology, David Geffen School of

Medicine, University of California, Los Angeles, Los Angeles, CA, United States; 3Department of Anatomy, Division of Histology and Cell Biology, Jichi Medical University, Shimotsuke-Shi, Tochigi, Japan; 4Division of Ultrastructural Research, National Institute for Physiological Sciences, Okazaki, Aichi, Japan; 5Normandie University, UNIROUEN, INSERM, U1239, DC2N, Mont-Saint-Aignan, France; 6Department of Neuroscience, Yale University School of Medicine, New Haven, CT, United States

Chapter outline 6.1. Introduction 107 6.2. Early formation of cerebellum 108 6.2.1. Morphogenetic aspect of first steps of cerebellar formation 108 6.2.2. Molecular mechanisms underlying initial formation of cerebellum 108 6.3. Three types of cerebellar patterning in adult mammals 111 6.3.1. Cerebellar anterioreposterior patterning 111 6.3.1.1. Lobes 111 6.3.1.2. Lobules (IeX) 111 6.3.1.3. Functional roles of lobes 111 6.3.2. Cerebellar medialelateral patterning 111 6.3.2.1. Parasagittal zones 111 6.3.2.2. Parasagittal stripes 112 6.3.2.3. Correspondence between parasagittal zones and parasagittal stripes 113 6.3.2.4. Functional roles of parasagittal zones and stripes 114 6.3.3. Cerebellar outereinner patterning 114 6.3.3.1. The molecular layer 114 6.3.3.2. The Purkinje cell layer 114 6.3.3.3. The granular layer 114 6.3.3.4. The white matter 115 6.3.3.5. The cerebellar nuclei 115 6.3.3.6. Roles of cerebellar outereinner patterning 115 6.4. Formation of cerebellar patterning 115

6.4.1. Formation of cerebellar anterioreposterior patterning 6.4.1.1. Formation of lobes and lobules 6.4.1.2. Cellular mechanisms underlying the formation of lobes and lobules 6.4.2. Formation of cerebellar medialelateral patterning 6.4.2.1. Formation of parasagittal zones 6.4.2.2. Cellular and molecular mechanisms underlying the formation of parasagittal zones 6.4.2.3. Formation of parasagittal stripes 6.4.2.4. Critical roles of Purkinje cell birth date in the formation of embryonic and adult parasagittal stripes and parasagittal zones 6.4.3. Formation of cerebellar outereinner patterning 6.4.3.1. Formation of the molecular layer 6.4.3.2. Formation of the Purkinje cell layer 6.4.3.3. Formation of the granular layer 6.4.3.4. Formation of the white matter and the cerebellar nuclei 6.4.3.5. Mechanisms underlying the control of neuronal migration 6.4.3.6. The deficits of neuronal migration by exposure to toxic substances and natural environmental factors result in abnormal O-I patterning References

115 115 116 118 118

118 119

119 120 121 123 123 125 126

129 131

6.1 Introduction The adult mammalian cerebellum is composed of three cortical layers (the molecular layer, ML; the Purkinje cell layer, PCL; the granular layer, GL), white matter (WM), and three centrally located nuclei (fastigial nucleus, interposed nucleus, dentate nucleus). In its medialelateral (M-L) extent, the cerebellar cortex is divided into three longitudinal regions: vermis, paravermis, and hemisphere. Each of these regions is folded into lobules (lobule I to lobule X).

Patterning and Cell Type Specification in the Developing CNS and PNS. https://doi.org/10.1016/B978-0-12-814405-3.00006-0 Copyright © 2020 Elsevier Inc. All rights reserved.

107

108 PART | I Induction and patterning of the CNS and PNS

Based on gross morphological features, neuronal cytoarchitecture, distinct neuronal connections, and region-specific expression of genes, three types of cerebellar patterning (anterioreposterior [A-P] patterning; M-L patterning; outere inner [O-I] patterning) are recognized in the adult mammalian cerebellum (Fig. 6.1AeC). Each cerebellar patterning plays a crucial role in a variety of cerebellar functions (such as motor control and sensory information processing). The disruption of each cerebellar patterning results in abnormal behavior (such as ataxia) and cognitive defects. During embryonic and early postnatal development, each cerebellar patterning is formed in separate cellular processes that interact with each other. Cerebellar A-P patterning is characterized by 10 distinct lobules at the vermis of adult cerebellum along the A-P axis (Fig. 6.1A). The formation of cerebellar A-P patterning starts when four shallow fissures begin to form in the vermis to produce the five cardinal lobes by approximately embryonic day 17 (wE17). Subsequently, cardinal lobes undergo extensive outgrowth and subdivision into lobules, culminating in the formation of 10 distinct lobules at the vermis along the A-P axis during early postnatal development. The formation of A-P patterning is completed by approximately postnatal day 15 (wP15) (Fig. 6.1A). Cerebellar M-L patterning is distinguished by the region-specific neural connections between the inferior olive (IO), cerebellar cortex, and cerebellar nuclei (CN) (parasagittal zones) and by the region-specific expression of genes in the cerebellar cortex (especially Purkinje cells) (parasagittal stripes) (Fig. 6.1B). In the embryonic cerebellum, the formation of parasagittal zones starts when climbing fibers from the IO make synaptic contact with immature Purkinje cells and concurrently Purkinje cell axons make synaptic contact with neurons in the CN and vestibular nuclei. The formation of parasagittal stripes starts when subsets of Purkinje cell clusters exhibit differential expression of multiple molecules and genes and array symmetrically along the M-L axis on each side of the midline during the period of E14eE17. Cerebellar O-I patterning is defined by the cortical layer- and nucleus-dependent allocation of cerebellar neurons, which results from cell typeespecific migration of immature neurons during embryonic and early postnatal development (Fig. 6.1C). In the developing cerebellum, all glutamatergic neurons come from the upper rhombic lip (uRL), whereas all GABAergic neurons come from the ventricular zone (VZ) of the cerebellar plate. Each type of cerebellar neuron exhibits a unique migration pathway to reach their final destination and settle themselves in genetically predetermined positions within the developing cerebellum. The cell typeespecific allocation of cerebellar neurons results in the formation of cerebellar layers and nuclei along O-I axis. By wP20, the formation of the cerebellar O-I pattern is complete. In this chapter, we first will describe the characteristic features of the three different types of cerebellar patterning (A-P, M-L, and O-I patterning) in the adult cerebellum. Second, we will review the process of the formation of each cerebellar patterning during embryonic and early postnatal development.

6.2 Early formation of cerebellum 6.2.1 Morphogenetic aspect of first steps of cerebellar formation During early embryonic development, the neural tube develops three vesicles, which are designated as the prosencephalon (forebrain), the mesencephalon (midbrain), and the rhombencephalon (hindbrain), at its anterior end. Subsequently, the prosencephalon is divided into the diencephalon and telencephalon, while the rhombencephalon is divided into the pons, cerebellum, and medulla oblongata. The initial formation of the cerebellum starts at wE9 in mice, when the hindbrain begins expressing distinct transcription factors such as Gbx2 and Otx2 (Fig. 6.2A1), which delineate the midbrain/hindbrain boundary. Following hindbrain specification, by wE9.5, spatial coordinates are set along the rostralecaudal axis of rhombomere 1 (r1), possibly in response to signals from the isthmic organizer (IsO) located between r1 and the mesencephalon (Fig. 6.2A1). The dorsal r1 rotates w90 degrees between wE9.5 and wE12.5. The cells that are initially located at the most caudal part of r1 locate at the most lateral part of the cerebellar primordia after the w90 degrees rotation (Fig. 6.2A1 and A2). Because there is little cell mixing along the rostralecaudal axis, cells maintain their relative positions before and after the rotation, and thus the spatial coordinates set at E9.5 along the rostralecaudal axis are converted to M-L coordinates in the E12.5 wing-like cerebellar primordia (Fig. 6.2A2) (Sgaier et al., 2005). A new A-P axis is set at wE12.5. Thereafter, the medial part of the cerebellar primordia preferentially expands to produce a cylindrical cerebellum by wE16.5.

6.2.2 Molecular mechanisms underlying initial formation of cerebellum The topographical boundary between the midbrain and the hindbrain is the IsO that secretes fibroblast growth factor 8 (Fgf8) (Fig. 6.2A1) (Crossley and Martin, 1995; Lee et al., 1997). The earliest molecular event for IsO specification is the differential expression of Otx2 (transcription factor) in the caudal mesencephalon and Gbx2 (transcription factor) in the rostral rhombencephalon (Fig. 6.2A1) (Millet et al., 1996; Martinez et al., 2013). The caudal limit of Otx2 expression and the rostral

Formation of cerebellar cytoarchitecture Chapter | 6

109

FIGURE 6.1 Schematic representation showing three cerebellar patternings (anterioreposterior patterning [A-P], medialelateral [M-L] patterning, and outereinner patterning [O-I]) in mammals. (A) Two micrographs show A-P patterning in the cerebellum of postnatal (P) 10-day-old mouse (CD-1). (A1) Dorsal view of cerebellum (indicated by white dotted line). (A2) Parasagittal view of cerebellum. The section in A2 was obtained from the cerebellum shown in A1 (indicated by the blue box). In A1 and A2, yellow and white asterisks indicate the rostral and caudal direction, respectively. In A1e2 and B2, roman numerals (IeX) indicate lobules. (B) Schematic drawings showing M-L patterning in adult cerebellum. (B1) Parasagittal zones (AeD2) in adult cerebellum identified by olivocerebellar and corticonuclear projections. (B2) Parasagittal stripes in adult cerebellum recognized by the expression of zebrin II in Purkinje cells. Blue areas represent zebrin IIepositive regions in adult cerebellum, while white areas represent zebrin-negative regions. (C) Schematic representation showing O-I patterning in adult cerebellum. Three cerebellar cortical layers and cerebellar nuclei are arranged in an outer (pial surface) to inner (white matter) manner. Distinct types of neurons are located in the four cortical layers (ML, PCL, GL, and WM) and the cerebellar nuclei in the adult cerebellum, suggesting that specific allocation of neurons plays a critical role in the formation of cortical layers and nuclei through directed neuronal migration during early development. b, basket cell; cbl, cerebellum; cbr, cerebrum; eP, excitatory projection neuron; G, Golgi cell; g, granule cell; GL, granular layer; ic, inferior colliculus; iI, inhibitory interneuron; iP, inhibitory projection neuron; L, Lugaro cell; mo, medulla oblongata; ML, molecular layer; P, Purkinje cell; PCL, Purkinje cell layer; s, stellate cell; sc, superior colliculus; u, unipolar brush cell; WM, white matter.

110 PART | I Induction and patterning of the CNS and PNS

FIGURE 6.2 Schematic representation showing the formation of cerebellar primordium during early embryonic development. (A) Altering the orientation of the rostralecaudal axis of cerebellar anlages to the medialelateral axis results in the initial formation of cerebellar primordium. (A1) Cerebellar anlage is located in rhombomere 1 (r1) that lies between isthmic organizer (IsO) and r2 at wE9.5. Several transcription factors and extracellular proteins are expressed in the mesencephalon, IsO, and rhombomere 1 in a region-specific manner at wE9.5. (A2) Both sides of cerebellar anlages rotate, fuse at the midline, and form cerebellar primordium at wE12.5. As a result of alteration of orientation of cerebellar anlages, the rostral part (marked by yellow asterisks in A1) of r1 transforms to the medial part (marked by yellow asterisks in A2) of cerebellar primordium. Also, the caudal part (marked by white asterisks in A1) of r1 transforms to the lateral part (marked by white asterisks in A2) of cerebellar primordium. (B) The possible interaction (induction and inhibition) of genes expressed in the mesencephalon (presented by green fonts), IsO (presented by blue fonts), and rhombomere 1 (presented by red fonts), which are thought to play a key role in the development of cerebellar primordium. Large and small fonts represent high and low expression levels of genes, respectively. 4V, fourth ventricle; CbP, cerebellar primordium; FB, forebrain; Md, midbrain; Mes, mesencephalon; MO, medulla oblongata; r1er4, rhombomere 1erhombomere 4; VM, velum medullaris.

limit of Gbx2 expression mark the IsO (the mid-hindbrain boundary). Otx2 and Gbx2 play key roles in the formation of the midbrain-hindbrain boundary (Wassarman et al., 1997). For instance, increases or a posterior shift in the expression of Otx2 or a decrease in Gbx2 causes the midbrain-hindbrain boundary to shift caudally, while decreases in Otx2 or increases or anterior shifts in Gbx2 cause the midbrain-hindbrain boundary to shift rostrally (Nakamura et al., 2005). At wE8.5 in mice, Fgf8 expression is first activated in the IsO. In loss-of-function experiments, inactivation of Fgf8 leads to loss of the entire tectum and cerebellum (Chi et al., 2003). The FGF8 signal acts in the IsO in concert with other signaling molecules, such as WNT1 (extracellular protein) (Fig. 6.2B). The morphogenetic activity of the IsO is a consequence of a specific spatiotemporal expression of molecular signals, which regulate the specification and structural development of the mesencephalon and rhombencephalon. Alterations of Fgf8 and Otx2 gene expression lead to massive disruption of the mid-hindbrain neural territory. A decreasing gradient of FGF8 protein concentrations in the IsO and the r1 is fundamental for cell survival and cerebellar development. Engrailed family En1 and En2 (transcription factors) also play a critical role in the early formation of the mesencephalic tectum and cerebellar primordia. Mouse En1 mutants lack most of the tectum and cerebellum and die at birth. Mouse En2 mutants are viable with a smaller cerebellum and have defects in cerebellar foliation (Hanks et al., 1995). During the formation of cerebellar primordia, the relative amounts of dorsalizing factors and ventralizing factors play essential roles in the formation of the dorsal-ventral patterning. The most important dorsalizing factors are proteins belonging to the bone morphogenic protein (BMP) family that is produced by the nonneuronal ectoderm of the roof plate (Chizhikov, 2006). The most important ventralizing factor is sonic hedgehog (Shh), a signaling molecule that emanates from the prechordal plate and floor plate (Wurst and Bally-Cuif, 2001). Along the dorsal-ventral axis, the rostral hindbrain is divided into the pons (ventral side) and the cerebellum (dorsal side).

Formation of cerebellar cytoarchitecture Chapter | 6

111

6.3 Three types of cerebellar patterning in adult mammals 6.3.1 Cerebellar anterioreposterior patterning The cerebellar A-P patterning is recognized as the ordered alignment of lobes and lobules along the A-P axis.

6.3.1.1 Lobes Based on aspects of cerebellar formation and function, the adult mammalian cerebellum can be divided into three lobes (anterior lobe, posterior lobe, and flocculonodular lobe) along the A-P axis. The anterior lobe (also called the paleocerebellum) lies rostral to the primary fissure and is comprised of lobules IeV. The posterior lobe (also called the neocerebellum) lies between the primary fissure and the posterolateral fissure and is comprised of lobules VIeIX. The posterior lobe is phylogenetically the newest portion of the cerebellum. The flocculonodular lobe (also called the archicerebellum) lies caudal to the posterolateral fissure and is comprised of lobule X. The flocculonodular lobe is phylogenetically the oldest part of the cerebellum.

6.3.1.2 Lobules (IeX) The adult cerebellum is further divided into lobules along the A-P axis by five deep fissures (primary fissure, posterior superior fissure, horizontal fissure, prepyramidal fissure, and posterolateral fissure) and by the folia (Fig. 6.1A1 and A2). There are 10 primary lobules in the vermis and 8 primary lobules in the hemispheres. Each primary lobule is further divided into secondary and tertiary sublobules, depending on the degree of foliation. The consistent presentation of lobules among different species suggests a conserved evolutionary genetic mechanism underlying foliation.

6.3.1.3 Functional roles of lobes Each lobe plays a distinct function in controlling body movement and position. The anterior lobe receives extensive inputs from the dorsal columns of the spinal cord, the trigeminal nerves, and visual and auditory systems and connects to the fastigial nuclei and the interposed nuclei in the CN via Purkinje cell axons. The fastigial nuclei in turn connect to the reticular formation, which can affect muscle tone and crude movement via the reticulospinal tract. The interposed nuclei connect primarily to the contralateral red nucleus. This nucleus is the origin for the rubrospinal tract that mainly influences limb flexor muscles. Therefore, the anterior lobe can influence both muscle tone and coordination of the extremities. The posterior lobe receives the vast majority of its input from the pontine nuclei, which receive input from the majority of the cerebral cortex via corticopontine fibers, and connects to the dentate nucleus in the CN via Purkinje cell axons. The dentate nucleus in turn projects to the ventral lateral nucleus of the thalamus which relays back to the cerebral cortex. The posterior lobe is involved in planning movement that is about to occur and also has purely cognitive functions. Lesions in the posterior lobe primarily affect skilled voluntary and associated movements. Ataxia is a synergic disturbance associated with lesions of the posterior lobe. The flocculonodular lobe receives vestibular input from both the semicircular canals and vestibular nuclei and projects fibers back to the medial and lateral vestibular (LV) nuclei. Therefore, the flocculonodular lobe plays a crucial role in regulating the vestibular system. The most important function of the flocculonodular lobe is to allow adaptation to vestibular damage. Although the effects of damage to the vestibular system are severe, with time, these injuries can be well compensated. This compensation does not happen if the flocculonodular lobe is also injured.

6.3.2 Cerebellar medialelateral patterning The cerebellar M-L patterning is recognized by parasagittal zones of neuronal network between the IOePurkinje cellseCN and by parasagittal stripes in the expression of specific genes and molecules along the M-L axis.

6.3.2.1 Parasagittal zones Analysis of the climbing fiber projections (olivocerebellar projections) and Purkinje cell axon projections (corticonuclear projections) reveals parasagittal zones (A, B, C1e3, and D1e2) along the M-L axis in the adult cerebellum (Fig. 6.3) (Sugihara and Shinoda, 2004; Voogd and Ruigrok, 2004; Voogd, 2011). In the A zone, which presents in the medial portion of the vermis, Purkinje cells receive input from climbing fibers that originate from the caudal medial accessory olive (cMAO) and Purkinje cell axons project to the fastigial nucleus (Fast). In the B zone, which presents in the lateral portion of the vermis and restricted anterior and posterior parts of the cerebellum, Purkinje cells receive input from

112 PART | I Induction and patterning of the CNS and PNS

FIGURE 6.3 Schematic representation showing the parasagittal zones (AeD2) along the medialelateral axis in adult cerebellum, which are recognized by distinct regionespecific olivocerebellar and corticonuclear projection. Red-dotted lines represent olivocerebellar projection (i.e., climbing fibers) and green-dotted lines represent corticonuclear projection (Purkinje cell axons). aIP, anterior interposed nucleus; cDAO, caudal part of dorsal accessory olive; cDent, caudolateral portion of lateral dentate nuclei; cMAO, caudal part of medial accessory olive; dlPO, dorsal lamina of the principal olive; Fast, fastigial nucleus; LV, lateral vestibular nucleus; pIP, posterior interposed nucleus; rDAO, rostral part of dorsal accessory olive; rDent, rostromedial portion of lateral dentate nuclei; rMAO, rostral part of medial accessory olive; vlPO, ventral lamina of the principal olive.

climbing fibers that originate from the caudal dorsal accessory olive (cDAO), and Purkinje cell axons project to the lateral vestibular nucleus (LV) nucleus. In the C1 and C3 zones, which present in the restricted anterior and posterior parts of the hemisphere, Purkinje cells receive input from climbing fibers that originate from the rostral dorsal accessory olive (rDAO), and Purkinje cell axons project to the anterior interposed nucleus (globose nucleus, aIP). In the C2 zone, which is located between the C1 zone and the C3 zone, Purkinje cells receive input from climbing fibers that originate from the rostral medial accessory olive (rMAO), and Purkinje cell axons project to the posterior interposed nucleus (emboliform nucleus, pIP). In the D1 zone, which presents in the lateral hemisphere, Purkinje cells receive input from climbing fibers that originate from the ventral lamina of the principal olive (vlPO), and Purkinje cell axons project to the rostromedial and caudal portion of the lateral dentate nuclei (cDent). In the D2 zone, which presents in the most lateral hemisphere, Purkinje cells receive input from climbing fibers that originate from the dorsal lamina of the principal olive (dlPO), and Purkinje cell axons project to the rostromedial portion of the lateral dentate nuclei (rDent).

6.3.2.2 Parasagittal stripes The expression of late-onset marker genes and proteins reveals the parasagittal strips in the adult mammalian cerebellum along the M-L axis (Fig. 6.4A) (Dastjerdi et al., 2012). To date, the most examined parasagittal stripe marker of the adult

Formation of cerebellar cytoarchitecture Chapter | 6

113

FIGURE 6.4 Parasagittal stripes recognized by the expression of cellular markers in embryonic and adult mouse cerebellum. (A) Schematic drawings representing the parasagittal stripes (blue areas) recognized by zebrin II expression on Purkinje cells in adult mouse cerebellum. Roman numerals (IVeIX) indicate lobules. (B) Selective co-expression of cellular markers in zebrin-positive and zebrin-negative parasagittal stripes in adult mouse cerebellum. Zebrin IIepositive regions express cellular markers (such as PLCb3, EAAT4, GABABR2, and NCS1), while zebrin IIenegative regions express other cellular markers (such as PCLb4, mGluR1b, MAP1A, neuroplastin, and neurogranin). (C) Developmental alterations of the expression of early-onset makers (C1) in embryonic/early postnatal cerebellum and late-onset makers (C2) in adult cerebellum.

cerebellum is zebrin II (also known as aldolase C) (Ahn et al., 1994). Zebrin II is expressed in parasagittal stripes of Purkinje cells in lobules IeV and lobule VIIIeIX, while most of the Purkinje cells in lobules VIeVII and lobule X uniformly express zebrin II (Fig. 6.4A) (Sugihara and Quy, 2007). Interestingly, lobules that uniformly express zebrin II express the small heat shock protein Hsp25 in Purkinje cells in a striking array of parasagittal stripes (Armstrong et al., 2000, 2001). Recent study shows that the levels of zebrin II gene expression vary across the stripes in which it is expressed (Fujita et al., 2014). The general expression pattern of zebrin II in adult cerebellum is highly consistent and conserved across mammals and birds (Apps and Hawkes, 2009). Interestingly, various late-onset genes and molecular markers, including phospholipase Cb3 (PLCb3), excitatory amino acid transporter 4 (EAAT4), GABA B receptor subtype 2 (GABABR2), and neuronal calcium sensor 1 (NCS1), are expressed in the zebrin IIepositive stripes in the adult cerebellum (Fig. 6.4B) (Dehnes et al., 1998). Conversely, the other late-onset genes and molecular markers, including phospholipase Cb4 (PLCb4), metabotropic glutamate receptor 1b (mGluR1b), microtubule-associate protein 1A (MAP1A), neuroplastin, and neurogranin, are expressed in the zebrin IIenegative stripes in the adult cerebellum (Fig. 6.4B) (Sarna et al., 2006). The expression of late-onset genes and molecular markers is not only limited to Purkinje cells in the adult cerebellum. Some late-onset genes and molecular markers are heterogeneously expressed by subsets of unipolar brush cells, Golgi cells, and granule cells in the adult cerebellum (Consalez and Hawkes, 2012). Furthermore, subsets of cerebellar afferents (such as climbing fibers and mossy fibers) also expressed late-onset genes and molecular markers in the adult cerebellum. Climbing fibers in the zebrin IIepositive stripes intensely express glutamate transporter VGLUT2 (Paukert et al., 2010). Moreover, the terminals of climbing fibers and mossy fibers arising from different sources align with parasagittal stripes of various late-onset genes and molecular markers (including zebrin II) (Sotelo and Chedotal, 2005; Cerminara et al., 2013). Parasagittal zones in which Purkinje cells are zebrin IIepositive receive climbing fiber input from the regions of IO dominated by descending inputs (Voogd et al., 2003; Sugihara and Shinoda, 2004). In contrast, parasagittal zones in which Purkinje cells are zebrin IIenegative receive climbing fiber input from the regions of IO that receive predominantly peripheral inputs (Voogd et al., 2003; Sugihara and Shinoda, 2004).

6.3.2.3 Correspondence between parasagittal zones and parasagittal stripes The comparison of the locations of parasagittal zones (identified by olivocerebellar projections and corticonuclear projections) and parasagittal stripes (identified by the expression of late-onset genes and molecular markers, such as zebrin II)

114 PART | I Induction and patterning of the CNS and PNS

reveals a close correspondence between the two sagittal patterns (Chen et al., 1996; Paradies et al., 1996; Hallem et al., 1999; Sugihara and Shinoda, 2004; Voogd and Ruigrok, 2004). In fact, the topography of zebrin II expression pattern corresponds to that of the olivary projections to the cerebellar cortex (Sugihara and Quy, 2007). In adult mouse cerebellum, Purkinje cells of the B, C1, and C3 zones are zebrin-negative, while Purkinje cells of the C2, D1, and D2 zones are zebrinpositive (Sugihara and Quy, 2007). The A zone is a composite of zebrin-positive and zebrin-negative areas.

6.3.2.4 Functional roles of parasagittal zones and stripes Accumulated evidence suggests the roles of parasagittal zones and stripes in modulating electrical activity in cerebellar cortical neurons (in particular Purkinje cells). First, EAAT4 expresses zebrin IIepositive Purkinje cells (in zebrin-positive parasagittal stripes) and acts to limit the duration of action of glutamate at the synapses between climbing fibers and Purkinje cells and between parallel fibers and Purkinje cells (Nagao et al., 1997). EAAT4 also limits the diffusion of glutamate extrasynaptic receptors and receptors in nearby synapses (Brasnjo and Otis, 2001). As a result, the ability of complex spikes in Purkinje cells to induce synaptic plasticity (long-term depression of parallel fiber synaptic efficacy) is reduced in zebrin IIepositive parasagittal stripes (Wadiche and Jahr, 2005). Second, Purkinje cells in zebrin IIenegative parasagittal stripes exhibit significantly higher firing rates of simple spikes than those in zebrin IIepositive parasagittal stripes (Xiao et al., 2014). Furthermore, Purkinje cells in zebrin IIepositive parasagittal stripes exhibit large irregularity in simple spike firing (Xiao et al., 2014). Third, climbing fibers in zebrin IIepositive parasagittal stripes release more glutamate than those in zebrin IIenegative parasagittal stripes, resulting that the complex spikes of Purkinje cells in zebrin IIepositive parasagittal stripes have a longer duration and are composed of more spikelets (Zhou et al., 2014). Fourth, there is a huge difference in the ability of parallel fiber activity to drive the activity of stellate cells and basket cells in zebrin IIepositive and zebrin IIenegative parasagittal stripes (Gao et al., 2006).

6.3.3 Cerebellar outereinner patterning There are three cortical layers (the ML, the PCL, and the GL) and the WM (including three nuclei) in the adult cerebellum along the O-I axis. Each layer and nucleus is comprised of distinct neuronal cell types and neuronal processes (Fig. 6.1C).

6.3.3.1 The molecular layer The ML is the top layer of the cerebellar cortices and is located under the pial surface (Fig. 6.1C). The ML contains stellate cells, basket cells, dendritic arbors of Purkinje cells, parallel fibers (granule cell axons), and climbing fibers. Both basket cells and stellate cells are GABAergic interneurons and receive an excitatory synaptic input from parallel fibers, and their axons make an inhibitory synapse with Purkinje cell dendrites (in the case of stellate cells) and Purkinje cell somata (in the case of basket cells). Both basket cells and stellate cells are scattered throughout the ML, but there is a bias in their allocation. The majority of stellate cells are located at the upper half of the ML, while the majority of basket cells are located at the lower half. Purkinje cell dendritic arbors lie in two-dimensional planes, with neighboring Purkinje cell arbors in parallel planes, and receive excitatory synaptic input from climbing fibers originating from the contralateral side of the IO in the medulla oblongata. Each parallel fiber runs orthogonally through these arbors and forms excitatory synapses with Purkinje cell dendrites.

6.3.3.2 The Purkinje cell layer The PCL is located between the ML and the GL (Fig. 6.1C). The PCL is comprised of large somata of Purkinje cells, which align like dominos stacked one in front of the other. Purkinje cells are GABAergic inhibitory projection neurons and provide the sole output of the cerebellar cortex to the CN and vestibular nuclei in the medulla oblongata. Purkinje cells receive an excitatory synaptic input from climbing fibers and parallel fibers and an inhibitory synaptic input from axons of both stellate cells and basket cells.

6.3.3.3 The granular layer The GL is located between the PCL and the WM (Fig. 6.1C). The GL is comprised of glutamatergic granule cells, glutamatergic unipolar brush cells, GABAergic Golgi cells, GABAergic Lugaro cells, and mossy fibers (including mossy rosettes). The GL in the adult mammal cerebellum is densely packed with the small somata (diameter, w8 mm) of granule cells. Granule cells, which are the most abundant type of neurons in the brain, are glutamatergic interneurons and have four to five short dendrites emitted from the somata. Granule cell dendrites have a claw-like shape at the end, which surround mossy fiber terminals (mossy rosettes). The complex of dendritic terminals of granule cells, mossy rosettes, and axonal

Formation of cerebellar cytoarchitecture Chapter | 6

115

terminals of Golgi cells forms the glomerulus in the GL. Granule cells send their ascending axons toward the ML through the PCL. After entering the ML, granule cell axons bifurcate and become parallel fibers (T-shaped axons), which make excitatory synapse with the spines of Purkinje cell dendrites, basket cells, stellate cells, and Golgi cell dendrites in the ML. Granule cells receive an excitatory synaptic input from mossy fibers and unipolar brush cells and an inhibitory synaptic input from Golgi cell axons. Golgi cells are GABAergic inhibitory interneurons scattered in regions throughout the GL and send their long multiple dendrites to the ML. Golgi cells receive excitatory synaptic input from mossy fibers and parallel fibers, and their axons form inhibitory synapses with granule cells (at the site of cerebellar glomerulus) and unipolar brush cells. Unipolar brush cells, which are abundant in lobules IX and X, are glutamatergic interneurons and receive excitatory synaptic input from mossy fibers and inhibitory synaptic input from Golgi cells (Mugnaini and Floris, 1994; Mugnaini et al., 2011). The axons of unipolar brush cells make an excitatory synapse with granule cells and other unipolar brush cells. Lugaro cells are fusiform GABAergic inhibitory interneurons located just beneath the PCL of the adult cerebellum (Melik-Musyan and Fanardzhyan, 2004). Lugaro cells receive inhibitory input from the collateral axons of Purkinje cells and send their axons to basket cells and stellate cells in the ML (Laine and Axelrad, 1998).

6.3.3.4 The white matter The WM is located under the GL (Fig. 6.1C) and is comprised of mossy fibers, climbing fibers, and Purkinje cell axons. Mossy fibers originate from many regions in the brain and spinal cord (including the basilar pontine nuclei, vestibular nuclei, lateral reticular nuclei, and external cuneate nucleus), and the terminal of each axon is anatomically recognizable as rosettes located within the GL in the adult cerebellum. Fasciculated bundles of mossy fibers enter the cerebellum through the superior, middle, or inferior cerebellar peduncles. In the WM, myelinated mossy fibers bifurcate repeatedly. Depending on the source of the mossy fibers, their termination within the cerebellum can be predominantly ipsilateral or contralateral and is restricted to particular lobules (Shinoda et al., 2000). Climbing fibers originating from the IO pass through the WM to reach the PCL and the ML. In the WM, myelinated climbing fibers give off several fine collaterals that radiate into the GL. The main stem of climbing fibers pursues a straight trajectory through the WM, but bends at right angles to enter the GL (Shinoda et al., 2000). The myelinated axons of Purkinje cells pass through the WM and terminate within the three CN (fastigial nucleus, interposed nucleus, dentate nucleus) and the vestibular nuclei in the medulla oblongata.

6.3.3.5 The cerebellar nuclei The CN are embedded in the WM and comprise three nuclei (fastigial nucleus, interposed nucleus, dentate nucleus). In the cerebellum of some mammals, the interposed nucleus separates two nuclei (emboliform nucleus and globose nucleus). Each CN is composed of glutamatergic excitatory projection neurons, GABAergic inhibitory projection neurons (nucleoolivary neurons), and GABAergic inhibitory interneurons, which receive excitatory inputs from axon collaterals of mossy fibers and climbing fibers, and inhibitory inputs from Purkinje cell axons. Glutamatergic excitatory projection neurons send their axons to multiple brain regions (including red nucleus, nucleus reticularis tegmenti pontis, spinal cord, and oculomotor nuclei) (Asanuma et al., 1983), while GABAergic inhibitory projection neurons send their axons to the IO.

6.3.3.6 Roles of cerebellar outereinner patterning The cerebellar O-I patterning plays essential roles in performing cerebellar function. The deficits in the formation of cerebellar O-I patterning are involved in the ectopic neurons, the abnormal neuronal connections, and the improper differentiation of neurons, leading to a variety of neurological disorders (such as cerebellar hypoplasia) in adult mammals (including human and mouse) (Hong et al., 2000; Basson and Wingate, 2013). For example, weaver mice, which develop abnormal cerebellar O-I patterning due to cell death of granule cells through a single amino acid mutation in a G proteinecoupled, inwardly rectifying Kþ channel (GIRK2), exhibit uneven weave to its gate, ataxia, mild locomotor hyperactivity, and tonic seizures (Harkins and Fox, 2002).

6.4 Formation of cerebellar patterning 6.4.1 Formation of cerebellar anterioreposterior patterning 6.4.1.1 Formation of lobes and lobules Cerebellar A-P patterning arises through a dynamic series of structural changes and can be divided into two stages. The first stage is the formation of cardinal lobes during embryonic development and the second stage is the formation of lobules during early postnatal development. The surface of the embryonic cerebellum is initially smooth at wE15 (Fig. 6.5A).

116 PART | I Induction and patterning of the CNS and PNS

FIGURE 6.5 Schematic representation showing foliation and formation of lobes and lobules during embryonic and early postnatal development, which results in the establishment of cerebellar anterioreposterior patterning. (A) Schematic drawing of a parasagittal section of mouse cerebellum at wE15. The cerebellar surface is smooth. (B) Schematic drawing of a parasagittal section of mouse cerebellum at wE18. The initial formation of four primary fissures (preculminate fissure, primary fissure, secondary fissure, and posterolateral fissure) divides the cerebellum into the five cardinal lobes (anterobasal lobe, anterodorsal lobe, central lobe, posterior lobe, and inferior lobe). (C) Schematic drawing of a parasagittal section at the vermis of mouse cerebellum at wP15. The formation of six secondary fissures (declival fissure, intercrural fissure, intraculminate fissure, intercrural fissure, prepyramidal fissure, uvular fissure) divides the cerebellum into 10 lobules (IeX) at the vermis of the cerebellum along the anterioreposterior axis. Lobules are indicated by roman numerals (IeX). Abl, anterobasal lobe; Adl, anterodorsal lobe; Cel, central lobe; dcl, declival fissure; EGL, external granule layer; icl, intraculminate fissure; ict, intercrural fissure; Inl, inferior lobe; pct, precentral fissure; pcu, preculminate fissure; po, posterolateral fissure; Pol, posterior lobe; ppy, prepyramidal fissure; pr, primary fissure; se, secondary fissure; uvu, uvular fissure; VZ, ventricular zone of cerebellar plate.

By wE17 in mice, four shallow fissures (preculminate fissure, primary fissure, secondary fissure, and posterolateral fissure) begin to form in the vermis to produce the five cardinal lobes (anterobasal lobe, anterodorsal lobe, central lobe, posterior lobe, and inferior lobe) (Fig. 6.5B). After birth, the cardinal lobes undergo extensive outgrowth and the development of six secondary fissures (declival fissure, intercrural fissure, intraculminate fissure, precentral fissure, prepyramidal fissure, uvular fissure) that divides the cardinal lobes into 10 lobules (IeX) at the vermis of the cerebellum along the A-P axis (Fig. 6.5C). Lobule formation is completed by wP15. During the formation of lobes and lobules, the rostrocaudal length of the cerebellum increases w25-fold, while the width of the vermis increases only w1.5-fold (Gallagher et al., 1998). The formation of lobes and lobules maximizes the number of neurons and neuronal processes in the cerebellar cortical layers (Sillitoe and Joyner, 2007; Joyner et al., 2017).

6.4.1.2 Cellular mechanisms underlying the formation of lobes and lobules The formation of lobes and lobules during embryonic and early postnatal development depends on fissure formation. To date, the mechanisms underlying fissure formation remain obscure, but possible scenarios have been proposed. One scenario emphasizes the role of Purkinje cells in fissure formation. Based on this scenario, Purkinje cells anchor the cerebellar cortex to the underlying WM via their axons at the positions that define the base of fissures (Altman and Bayer, 1997). Although this is an attractive idea, there are currently not many experimental results that support this idea. Alternatively, granule cell precursor proliferation may play a role in the formation of fissure (Sudarov and Joyner, 2007). First, the reduction of the proliferation of granule cell precursors by x-irradiation and methylazoxymethanol results in the decrease in the number and size of lobules, suggesting that the expansion of external granular layer (EGL) by active proliferation of granule cell precursors is essential for the formation of lobes and lobules (Doughty et al., 1998a). Second, alterations in the proliferation of granule cell precursors by modifying the levels of Shh signaling affect fissure formation (Corrales et al., 2004, 2006) The decrease of Shh levels results in the inhibition of fissure formation, while the increase of Shh levels results in the formation of extra fissures. Third, the prolonged proliferation of granule cell precursors due to hypothyroidism results in the formation of extra fissures (Lauder et al., 1974). Fourth, based on the detailed analysis of proliferation and behavior of granule cell precursors, Sudarov and Joyner (2007) proposed a new model for the initial formation of fissures during the embryonic development of mouse cerebellum. The model is composed of four cellular steps. The first step is that the cerebellar surface is smooth, and both the EGL and the Purkinje cell plate (PCP) have uniform widths under the pia mater at wE15 in the mouse embryos (Fig. 6.6A). The second step is that the cerebellar surface is still smooth at wE16.5, but granule cell precursors in the EGL start to locally increase in number through enhanced proliferation and invade inwardly, resulting in a local increase in the width of EGL (Fig. 6.6B). As a result of

Formation of cerebellar cytoarchitecture Chapter | 6

117

FIGURE 6.6 Schematic representation showing possible cellular steps for initial formation of primary fissures in embryonic cerebellum, leading to the formation of cardinal lobes. (A) The cerebellar surface is smooth at wE15. Both the EGL and the PCP have uniform widths. (B) Although the cerebellar surface is still smooth at wE16.5, granule cell precursors start to locally increase in the number and invade inwardly, resulting in a local increase in the width of EGL. As a result of inward invasion of granule cell precursors, the PCP also locally moves inwardly. (C) At wE17.5, granule cell precursors continue to increase their proliferation and invade inwardly. At the inward invasion sites, granule cell precursors change their shape from spherical to vertical elongated. The PCP further moves inwardly without altering width. Importantly, at the inward invasion sites of granule cell precursors, the pial surface is dented, which is the initial sign of fissure formation. (D) As vertically elongated granule cell precursors further invade inwardly, the pial surface is dented deeper at wE18.5, resulting in the formation of the fissure. At the site of developing fissures, the processes of Bergmann glia cells change the orientation from vertical to oblique. These alterations of the orientation of Bergmann glial cell processes affect the direction of granule cell migration in the molecular layer later because granule cells migrate along the Bergmann glial cell processes. In (C and D), red asterisks indicate the sites of the initial formation of primary fissure. B, Bergmann glial cells; EGL, external granular layer; g, granule cell precursors, P, Purkinje cells; PCP, Purkinje cell plate; PM, pia mater.

118 PART | I Induction and patterning of the CNS and PNS

inward invasion of granule cell precursors, Purkinje cells in the PCP also locally move inwardly (Fig. 6.6B). The third step is that at the inward invasion sites (marked by a red asterisk in Fig. 6.6C), granule cell precursors change their shape from spherical to vertical elongated at wE17.5. Purkinje cells in the PCP further move inwardly without altering width (Fig. 6.6C). At the inward invasion sites of granule cell precursors, the pial surface is dented, which is the initial sign of fissure formation (Fig. 6.6C). The fourth step is that as vertically elongated granule cell precursors further invade inwardly, the pial surface is dented deeper at wE18.5, resulting in the initial formation of the fissure (marked by a red asterisk in Fig. 6.6D). The PCP also further moves inwardly without altering width and becomes the U-shape (Fig. 6.6D). At the site of developing fissures, the processes of Bergmann glial cells change the orientation from vertical to oblique, resulting in fan-like organization (Fig. 6.6D). These alterations of the orientation of Bergmann glial cell processes can affect the direction of the migration of postmitotic granule cell migration in the ML after birth because granule cells migrate along the Bergmann glial cell processes (Rakic, 1971). These changes in the direction of granule cell migration in the ML around the base of the fissure result in the regional differences in the width of the internal granular layer (IGL). The width of the IGL is thin at the base of each fissure in both the postnatal and adult cerebellum. The timing of fissure formation also affects the growth of lobes and lobules. En1/2 homeobox genes have been found to be crucial for the production of the distinct medial and lateral foliation patterns in mammalian cerebellum (Orvis et al., 2012). Interestingly, in the mutant mice, the loss of En2 results in the premature formation of the prepyramidal fissure (nonprincipal fissure), which is responsible for the separation of lobule VII and lobule VIII (Sudarov and Joyner, 2007). Conversely, the loss of En2 results in the delayed formation of the secondary fissure (principal fissure), which is responsible for the separation of lobule VIII and lobule IX (Sudarov and Joyner, 2007). These alterations of the timing of fissure formation affect the size, shape, and position of the formation of the intervening lobule VIII (Sudarov and Joyner, 2007).

6.4.2 Formation of cerebellar medialelateral patterning 6.4.2.1 Formation of parasagittal zones The formation of parasagittal zones (A, B, C1e3, and D1e2) depends on the development of region-specific connections between the IO and Purkinje cells through the climbing fiber projections (olivocerebellar projections) and between Purkinje cells and CN through the Purkinje cell axon projection (corticonuclear projections) during embryonic and early postnatal development (Fig. 6.3). The IO nucleus is subdivided into three major subnucleus: the principal olivary nucleus, which projects mainly to the cerebellar hemisphere (D1e2 zones); the dorsal accessory olivary nucleus, which projects to the vermis and hemisphere (B, C1, and C3 zones); and the medial accessory olivary nucleus, which projects to the vermis and paravermis (A and C2 zones). During embryonic development, the inferior olivary axons cross the midline and enter the contralateral cerebellum through the inferior cerebellar peduncle. The inferior olivary axons reach the cerebellum in mice at wE14 (Rahimi-Balaei et al., 2015). After entering the cerebellum, the inferior olivary axons give off thick and thin branches. The thick branches form the climbing fibers, which terminate in the cerebellar cortex and synapse directly with Purkinje cells. The thin branches terminate in the CN and synapse directly with both excitatory and inhibitory neurons in CN. At wE16, the first synapses between the climbing fibers and Purkinje cells are detected (Mason et al., 1990; Morara et al., 2001; Kita et al., 2015). At the same time (wE16), climbing fibers are organized into rudimentary parasagittal zones (Paradies and Eisenman, 1993; Paradies et al., 1996). There is a long-standing hypothesis that climbing fibers immediately recognize the patterned architectures of Purkinje cell clusters upon their arrival in the cerebellar cortex (Chedotal and Sotelo, 1992; Paradies et al., 1996). Early in postnatal development, each Purkinje cell is innervated by multiple climbing fibers. Climbing fiber innervation is subsequently reduced leaving a single strong excitatory synapse in an activity-dependent fashion (Mariani and Changeux, 1981; Paradies and Eisenman, 1993). Mono-innervation of the climbing fibers to the Purkinje cells occurs at wP21 (Watanabe and Kano, 2011). Purkinje cells start to extend their axons as early as wE12.5 (Miyata et al., 2010). Purkinje cell axons reach the CN and the vestibular nuclei at wE14.5 (Sillitoe et al., 2009). In the embryonic cerebellum, the growing axons of Purkinje cells project primarily to cerebellar nucleus according to their M-L positions, resulting in the formation of the initial parasagittal zone (Sillitoe et al., 2009). Also, the projection of Purkinje cell axons to the CN (corticonuclear projections) respects the topographic organization defined by afferent terminals (Chung et al., 2009; Sugihara, 2011).

6.4.2.2 Cellular and molecular mechanisms underlying the formation of parasagittal zones The parasagittal zones result from the formation of the region-specific neuronal connections between climbing fibers (axons of inferior olivary neurons) and Purkinje cells and between Purkinje cell axons and CN neurons. It has been shown that various types of ligands (i.e., semaphorins, netrins, ephrins, and slits) and its receptors, which are known to play

Formation of cerebellar cytoarchitecture Chapter | 6

119

crucial roles in axonal guidance in the developing brain, are expressed in the embryonic cerebellum during the formation of olivocerebellar projections and corticonuclear projections (Beckinghausen and Sillitoe, 2018). Among them, the roles of EphA (receptor)eephrin-A (ligand) signaling in the formation of olivocerebellar projections present in detail (Nishida et al., 2002). In the embryonic chicken, EphA3, EphA5, and EphA6 are differentially expressed in neurons occupied in different areas of the IO. Likewise, ephrin-A2, ephrin-A3, and ephrin-A5 are differentially expressed in Purkinje cells located at different M-L zones. The expression of EphAs in the IO and the expression of ephrin-As in the cerebellum are inversely correlated, and ephrin-As inhibit the outgrowth of EphAs-expressing IO axons (climbing fibers) in an areaspecific manner. When ephrin-A2 is overexpressed in the cerebellum, the olivocerebellar projection is disrupted. Conversely, overexpression of a truncated EphA3 receptor in the cerebellum reduces endogenous ligand activity to undetectable levels and causes aberrant olivocerebellar projection, with high receptor axons invading high ligand domains. In vitro, ephrin-A2 inhibits the outgrowth of EphAs-expressing IO axons (climbing fibers) in a region-specific manner. Collectively, these results suggest that spatially accurate interaction between EphAs (expressed in climbing fibers) and ephrin-As (expressed in Purkinje cells) is essential for the formation of olivocerebellar projections, which is recognized as parasagittal zones. Interestingly, the region-specific expression of ephrin-A2 and ephrin-A5 is detected in the developing CN (Nishida et al., 2002), and Purkinje cells express EphAs (Beckinghausen and Sillitoe, 2018), suggesting that the EphAeephrin-A interaction also plays a critical role in the formation of region-specific corticonuclear projections (parasagittal zones) between Purkinje cell axons and CN neurons.

6.4.2.3 Formation of parasagittal stripes The M-L axisedependent expression of genes and proteins in cerebellar neurons (in particular Purkinje cells) reveals the parasagittal stripes in the embryonic cerebellum as early as at wE14 (Herrup and Kuemerle, 1997). The embryonic parasagittal stripes are identified by the striped expression of genes and proteins that are called “early-onset markers” (Fig. 6.4C1) (Millen et al., 1995), whereas the adult parasagittal stripes are identified by other genes and proteins, which are called “late-onset markers” (Fig. 6.4C2) (Herrup and Kuemerle, 1997). Starting at wE18, the embryonic parasagittal stripes recognized by early-onset markers begin to disperse and disappear during the first week of postnatal development (Millen et al., 1995). Thereafter, the adult parasagittal stripes recognized by late-onset markers begin to appear at 2e 3 weeks old and continuously express during an entire adult life (Tano et al., 1992). The adult cerebellum is composed of more complex parasagittal stripes than the embryonic cerebellum (Herrup and Kuemerle, 1997). It has been difficult to determine how parasagittal stripes are established in the embryonic cerebellum and how embryonic parasagittal stripes and adult parasagittal stripes are related. To date, molecular mechanisms underlying the formation of parasagittal stripes remain to be determined. Homeobox transcription factors, engrailed1 (En1) and engrailed2 (En2), may be possible candidates for controlling the formation of parasagittal stripes (Sillitoe et al., 2008). As described in previous sections, it has been shown that En1/2 homeobox genes play crucial roles for the production of the distinct medial and lateral foliation patterns in mammalian cerebellum (Sudarov and Joyner, 2007; Orvis et al., 2012). Joyner and her colleagues found that in En1/2 mutant mice, the complementary expression of zebrin II and heat shock protein 25 (Hsp25) changes in unison in the adult cerebellum. Interestingly, in En1/2 mutant mice with almost normal foliation, parasagittal stripes recognized by the expression of zebrin II are severely disrupted, suggesting that En1/2 regulates the foliation and the formation of parasagittal stripes in different cell types (Sillitoe et al., 2008).

6.4.2.4 Critical roles of Purkinje cell birth date in the formation of embryonic and adult parasagittal stripes and parasagittal zones Using an adenoviral vector, Hashimoto and his colleagues demonstrate that Purkinje cell birth date plays crucial roles in the formation of parasagittal stripes and parasagittal zones (Hashimoto and Mikoshiba, 2003; Namba et al., 2011). First, Purkinje cells, which share the same birth date, form parasagittal stripes along the M-L axis in the embryonic cerebellum (Fig. 6.7AeD) (Hashimoto and Mikoshiba, 2003). These results suggest that Purkinje cells are fated to form specific parasagittal stripes after their birth between E10.5 and E12.5. Second, in the embryonic cerebellum, there is a correlation between parasagittal stripes recognized by Purkinje cell birth date and parasagittal stripes recognized by the expression of early-onset genes and proteins such as EN2, Wnt7b, and EphA4 (Fig. 6.7AeG) (Hashimoto and Mikoshiba, 2003). Third, parasagittal stripes recognized by Purkinje cell birth date do not change during development from the embryonic stage to the adult stage (Hashimoto and Mikoshiba, 2003). Fourth, in the adult cerebellum, parasagittal stripes recognized by Purkinje cell birth date strikingly correlate with parasagittal stripes recognized by zebrin II (late-onset gene) zones (Namba et al., 2011). Fifth, in the adult cerebellum, parasagittal stripes recognized by Purkinje cell birth date are identical to the topographic pattern of olivocerebellar climbing fiber projections (parasagittal zones) (Namba et al., 2011).

120 PART | I Induction and patterning of the CNS and PNS

FIGURE 6.7 Schematic representation showing a key role of Purkinje cell birth date in the formation of parasagittal stripes in the embryonic (E18.5) mouse cerebellum. (A) The relative medialelateral positions of birth dateedependent Purkinje cell clusters are defined and named from 1 to 8. Dotted lines indicate the boundaries of medialelateral positions of Purkinje cell clusters. (BeD) At E18.5, the distributions of Purkinje cell clusters born on E10.5 (B), E11.5 (C), and E12.5 (D) are analyzed in eight longitudinal compartments (shown in A) along the medialelateral axis. Each birth dateerelated Purkinje cell cluster is located to a specific subset of compartments. (E) E18.5 mouse cerebella exhibit parasagittal strips recognized by the expression of EphA4. (F and G) Parasagittal expression of En2 (F) and Wnt7b (G) in E18.5 mouse cerebella recognized by whole-mount in situ hybridization.

6.4.3 Formation of cerebellar outereinner patterning Four layers (the ML, the PCL, the GL, and the WM) and CN are found in the adult cerebellum along the O-I axis (Fig. 6.1C). Each layer and nucleus is comprised of distinct type of neurons and neuronal processes. The formation of cerebellar O-I patterning results from the migration of immature neurons during embryonic and early postnatal development (Fig. 6.8) (Galas et al., 2017). The cerebellum consists of 10 different types of neurons (i.e., granule cells, Purkinje

Formation of cerebellar cytoarchitecture Chapter | 6

121

FIGURE 6.8 The 3D-reconstruction of a migrating granule cell and a migrating stellate/basket cell and a Bergmann glial cell in postnatal 10-day-old mouse (CD-1) cerebellum. The images are reconstructed from serial EM images with the use of Fiji. Two different views of the 3D-reconstruction images are presented in A1 and A2, respectively. In the IGL, granule cells migrate toward the bottom of the IGL, while basket/stellate cells migrate toward the ML. BG, Bergmann glial cell; g, granule cell; IGL, internal granular layer; LP, leading process; MGC, migrating granule cell; ML, molecular layer; MS/ BC, migrating stellate/basket cell; P, Purkinje cell; PCL, Purkinje cell layer; TP, trailing process.

cells, basket cells, stellate cells, Golgi cells, unipolar brush cells, Lugaro cells, CN excitatory projection neurons, CN inhibitory projection neurons, and CN inhibitory interneurons), which use either glutamate or GABA as neurotransmitter. The cerebellar primordia have two distinct progenitor zones (the VZ and the uRL). The VZ expresses the bHLH factor Ptf1a, and the uRL expresses the bHLH factor Math1. Loss-of-function and fate-mapping experiments demonstrate that Ptf1a and Math1 are not only useful markers for each cerebellar progenitor zone but also essential for the generation of specified progenitors within their respective germinal zones. In the absence of Ptf1a, the VZ fails to generate all GABAergic cerebellar neurons (Hoshino et al., 2005). Similarly, Math1-positive cells in the uRL give rise to all glutamatergic neurons (Wang et al., 2005). During cerebellar development, different types of neurons are produced at different timepoints. The first neurons to be generated are CN inhibitory projection neurons, CN excitatory projection neurons, and Purkinje cells, which are produced from E10 to E13 in mice. The next neurons to be produced are Golgi cells, basket cells, stellate cells, CN inhibitory interneurons, Lugaro cells, and unipolar brush cells from E13 to P8. The last neurons to be produced are granule cells from P0 to P15. As a result of neuronal migration, dendritic growth, and axonal invasion, each cortical layer and nucleus is formed, leading to the formation of cerebellar O-I patterning. In the following sections, we will describe the process of the formation of each cerebellar layer and nucleus one by one.

6.4.3.1 Formation of the molecular layer The formation of the ML results from (1) the migration of stellate cells and basket cells, (2) the formation of parallel fibers (granule cell axons), and (3) the growth of Purkinje cell dendrites. (a) Migration of stellate cells and basket cells: In studies of their migration, stellate cells and basket cells are often called stellate/basket cells, and the migration of both cells is analyzed as a single cell type. This is because migrating basket cells and stellate cells are morphologically indistinguishable. During embryonic development, the progenitors of stellate/basket cells, which originate from the VZ, first migrate outwardly to reach the prospective WM and then continue to divide in the prospective WM as well as in the folial WM (1e2 in Fig. 6.9A). After final mitosis, stellate/basket cells migrate from the deep WM through the folial WM, the IGL, and the PCL, to the ML (their final destination) during early postnatal development (Cameron et al., 2009) (Fig. 6.8 and 3e5 in Fig. 6.9A). Upon entering the ML, stellate/basket cells migrate radially toward the top of the ML for w7.4 h (6 in Fig. 6.9A). After reaching the top of the ML, the stellate/basket cells turn and change their orientation from vertical to horizontal (7 in Fig. 6.9A). Then, stellate/basket cells migrate tangentially in the rostrocaudal direction (perpendicular to the direction of the extension of the parallel fibers) at the top of the ML for w16.8 h. Thereafter, stellate/basket cells turn and change orientation from horizontal to vertical (8 in Fig. 6.9A). Stellate/basket cells then migrate radially within the ML from the top to the bottom and vice versa for w19.3 h (9 in Fig. 6.9A). After prolonged radial migration within the ML, stellate/basket cells again turn and change their orientation from vertical to horizontal in the middle of the

122 PART | I Induction and patterning of the CNS and PNS

FIGURE 6.9 Schematic representation showing the migration of stellate/basket cells, Purkinje cells, and granule cells in the embryonic and early postnatal cerebellum. (A) Stellate/basket cell migration. (B) Purkinje cell migration. (C) Granule cell migration. In (AeC), numbers (1e10) in each figure represent the order of migration steps. DF, differentiating field; dWM, deep white matter; EGL, external granular layer; IGL, internal granular layer; IeX, lobule Ielobule X; ML, molecular layer; PCL, Purkinje cell layer; pWM, prospective white matter; RG, radial glial cells; RLS, rostral rhombic lip migratory stream; uRL, upper rhombic lip; VZ, ventricular zone of cerebellar plate.

ML (10 in Fig. 6.9A). Subsequently, stellate/basket cells migrate tangentially in a rostrocaudal direction for w10.2 h in the middle of the ML. Then, the majority of stellate/basket cells stop and complete their migration within the ML (Cameron et al., 2009). Although little is known about why basket cells and stellate cells loiter in the ML for an extended period of time (w54 h), the prolonged migration of basket cells and stellate cells in the ML may be a prerequisite for dispersing the cells and for proper allocation. (b) Formation of parallel fibers (granule cell axons): The process of parallel fiber formation at the EGL-ML border of the early postnatal mouse cerebellum has been described in detail (Komuro et al., 2001). At the EGL-ML border, tangentially migrating granule cells, which have two long horizontal processes emitting from opposite sides of the soma, start to extend a vertical process from the ventral side of the soma into the ML. Then, the nucleus and surrounding cytoplasm of the granule cell soma enter into the short vertical process descending into the ML. It takes w30 min for the completion of translocation of the nucleus and surrounding cytoplasm from the horizontal orientation to the vertical orientation. After changing nuclear orientation, the granule cell soma radially migrates toward the bottom of the ML (5e6 in Fig. 6.9C). As a result of the somal translocation within the vertically oriented leading process, granule cells develop a thin trailing process connected with two horizontal processes (immature parallel fibers). Therefore, two horizontal processes emitting from each side of the granule cell soma at the EGL-ML border transform into future parallel fibers (Komuro et al., 2001).

Formation of cerebellar cytoarchitecture Chapter | 6

123

The majority of parallel fibers develop from the two preexisting horizontal processes of tangentially migrating granule cells at the EGL-ML border, but there is another mechanism for the formation of parallel fibers. During the initiation of radial migration at the bottom of the EGL, the tip of horizontally extended leading processes of granule cells turns toward the ML, which is followed by the soma (Komuro et al., 2001; Kumada et al., 2009). As a result, the horizontal trailing process of granule cells becomes one side of the parallel fibers. Subsequently, granule cells develop a new small process at the rear part of the vertically elongated soma (Komuro et al., 2001). The new process extends horizontally toward the opposite direction of the extension of the horizontal trailing process and becomes the other side of the parallel fiber. Interestingly, early-generated parallel fibers lie at the bottom of the ML (near the PCL), while late-generated parallel fibers lie at the top of the ML (near the EGL). There is an inside-outside order in the accumulation of parallel fibers during early-postnatal development. (c) Growth of Purkinje cell dendrites: After the completion of migration, there is a continuous outgrowth of new stem dendrites emerging from all aspects of the Purkinje cell soma (Armengol and Sotelo, 1991). Between P3 and P6 in rats, multiple processes emitting from the soma progressively disappear. Then, during the period of P6eP10, Purkinje cells exhibit an explosive outgrowth of perisomatic protrusions emerging in all directions from the soma. Purkinje cells remain polarized with a short single or double apical dendritic cup, terminated by growth cones for the formation of the eventual dendritic tree. During the formation of these dendrites, there is a simultaneous withdrawal of the long somatic process and the formation of a single stem dendrite, which grows into the ML and splits into many branches, from which tertiary and distal spiny branchlets grow. From P10 to P15, the growth of the dendritic arbor occurs mainly in its lateral domain, where adult width is achieved by P13. From P15 on, there is a change in the orientation of the plane of growth from transverse to vertical. The extension in height of the dendritic field reaches its adult size by P30. The large dendrite arbors of Purkinje cells form nearly two-dimensional layers in the ML.

6.4.3.2 Formation of the Purkinje cell layer The formation of the PCL depends on the migration of Purkinje cells and the subsequent formation of a single, horizontal alignment of Purkinje cells. (a) Migration of Purkinje cells: The progenitors of Purkinje cells actively proliferate in the VZ of the cerebellar plate from E10 to E13 in mice (Yuasa et al., 1991). After final mitosis, Purkinje cells migrate radially toward the cortical surface from E13 to E17 in mice and complete their migration directly beneath the rostral rhombic lip migratory stream (RLS) early and the EGL later (1e4 in Fig. 6.9B). The migration of Purkinje cells is guided by contact with the processes of radial glial cells. Ultrastructural analysis reveals that the presence of puncta and macula adhaerentia in the contact region between migrating Purkinje cells and radial glial processes (Yuasa et al., 1996). Regional differences in the migratory process are evident: the final settlement of the Purkinje cells proceeds earlier in the lateral and posterior parts of the embryonic cerebellum, exhibiting L-M and A-P diminishing sequences (Yuasa et al., 1991). The EGL signaling (likely mediated by reelin) appears to be crucial to terminate Purkinje cell migration. (b) Formation of a single, horizontal line alignment of Purkinje cells: In the adult cerebellum, the large somata of Purkinje cells are aligned like stacked dominos in the PCL. By the time of birth, all Purkinje cell clusters (PCP) occupy their position between the EGL and the IGL, although the characteristic monolayer is not attained until P4eP5 in mice. Although the cellular mechanisms underlying the formation of monolayer arrangement of Purkinje cells remain to be determined, there is a possible scenario. The massive proliferation of granule cell precursors at the top of the EGL leads to a large increase in the external surface area (pial surface) of the cerebellum and the folding into lobules and folia, most prominently in the rostralecaudal direction. This increase in surface area may cause the spreading of Purkinje cells into a monolayer.

6.4.3.3 Formation of the granular layer The formation of the GL results from the migration of four different types of neurons (granule cells, Golgi cells, unipolar brush cells, and Lugaro cells) into the IGL. (a) Migration of granule cells: In the uRL of the mouse embryo, granule cell progenitors begin to proliferate by E10. From wE12.5, granule cell precursors migrate tangentially to cover the superficial zone of the embryonic cerebellum (1e4 in Fig. 6.9C). By E15 in the mouse embryo, granule cell precursors have covered most of the cerebellar surface, following L-M and A-P paths. The cell layer occupied by granule cell precursors is called the EGL. After clonal expansion in the superficial half of the EGL, granule cell precursors begin to produce postmitotic granule cells.

124 PART | I Induction and patterning of the CNS and PNS

During early postnatal development, coincident with the extension of two uneven horizontal processes oriented parallel to the longitudinal axis of the folium, postmitotic granule cells start to migrate tangentially in the direction of the larger process (5 in Fig. 6.9C) (Komuro et al., 2001; Komuro and Yacubova, 2003). Interestingly, their morphology and the speed of cell movement change systematically with their position within the EGL. The speed of tangential cell movement is fastest (w14.8 mm/h) in the middle of the EGL, when cells have two short horizontal processes. As granule cells elongate their somata and extend longer horizontal processes at the bottom of the EGL, they move at a reduced rate (w12.6 mm/h). At the interface of the EGL and ML where cells migrate tangentially at the slowest rate (w4.1 mm/ h), their somata become round and then begin to extend couples of the descending processes into the ML. Then, granule cell nuclei and surrounding cytoplasm enter into the short vertical process descending into the ML. After the completion of translocation of the nucleus and surrounding cytoplasm from the horizontally extended process to the vertical process, granule cell somata quickly move toward the bottom of the ML (6 in Fig. 6.9C) (Komuro et al., 2001). In the ML, migrating granule cells have a vertically elongated soma, a thin trailing process, and a more voluminous leading process. In the ML, migrating granule cells are attached to the surface of Bergmann glial fibers (6 in Fig. 6.9C), suggesting that granule cells move along Bergmann glial fibers during the entire translocation across the ML (Rakic, 1971). The speed of granule cell migration in the ML depends on the age of the cerebellum: the average speed increases systematically during the period of the first and second postnatal weeks (Komuro and Rakic, 1995). Consequently, granule cells traverse the developing ML within a relatively constant time period despite the doubling in width of the ML. At the bottom of the ML, granule cell somata move toward the PCL while the length of their leading process gradually decreases. Once granule cell somata enter the PCL, their shapes abruptly transform from vertically elongated spindles to spheres (7 in Fig. 6.9C) (Komuro and Rakic, 1998a). The rounded somata significantly slow their speed of movement, which stops completely in the PCL. The rounded somata remain stationary in the PCL for w2 h. Highly motile lamellipodia develop at the distal portion of the leading process, which penetrates the IGL, although the leading process does not exhibit a net extension in length. After a prolonged stationary period, granule cells in the PCL begin to reextend their somata and leading processes. During this transformation, granule cells gradually accelerate the rate of their migration and cross the PCLeIGL border. In the IGL, which becomes the GL after the completion of cerebellar development, the spindle-shaped granule cells migrate toward the WM at a speed comparable to that recorded for granule cells migrating along Bergmann glial fibers within the ML (Komuro and Rakic, 1995, 1998a; Komuro et al., 2001). The long axis of the granule cell soma remains oriented perpendicular to the PCLeIGL boundary line during this radial migration (Fig. 6.8 and 8 in Fig. 6.9C). Once the tip of a leading process approaches the IGLeWM border, granule cell somata become rounded. Granule cells then slow their migration and stop their movement near the IGLeWM border (9 in Fig. 6.9C) (Komuro et al., 2001; Kumada and Komuro, 2004). In the P10 mouse cerebellum, the average transit time of granule cells is 25.0 h in the EGL, 9.8 h in the ML, 5.2 h in the PCL, and 11.1 h to attain their final position in the IGL. Therefore, granule cells move from the top of the EGL through the ML and the PCL to their final position in the bottom of the IGL within about 2 days (average, 51 h) after the initiation of their tangential migration in the middle of the EGL. (b) Migration of Golgi cells: Golgi cell precursors, which are produced in the VZ of the embryonic cerebellum, first migrate outwardly to reach the prospective WM and then continue to divide in the prospective WM as well as in the folial WM during their migration (1e3 in Fig. 6.10A) (Zhang and Goldman, 1996; Weisheit et al., 2006). Thereafter, postmitotic Golgi cells migrate from the deep WM through the folial WM to the IGL (4 in Fig. 6.10A) (Zhang and Goldman, 1996; Maricich and Herrup, 1999). The generation of Golgi cells has not been completed by P4 (Zhang and Goldman, 1996). (c) Migration of unipolar brush cells: Two subtypes of unipolar brush cells have been identified: one subtype expresses calretinin and the other expresses metabotropic glutamate receptor 1a (Kalinichenko and Okhotin, 2005). Unipolar brush cell progenitors, which express Pax6 and Math1, actively proliferate in the uRL during the late embryonic and perinatal periods in mice. The fate commitment and subtype specification of unipolar brush cells probably occur in the uRL, either during or immediately after neurogenesis (Englund et al., 2006). After final cell division, unipolar brush cells traverse a novel pathway to their final destination (the IGL). First, unipolar brush cells exit the uRL via short, narrow channel between the developing cerebellar cortex and the VZ (1ae2a in Fig. 6.10B). This channel leads to the prospective WM, where unipolar brush cells disperse widely and proceed to the IGL (3ae4a in Fig. 6.10B) (Englund et al., 2006). The majority of unipolar brush cells reach the IGL by P10, where they continue to mature throughout the first postnatal month in mice. In addition, some of the unipolar brush cells migrate rostrally along the VZ toward the brain stem to enter cochlear nucleus (additional location of unipolar brush cells in the adult brain) (1be4b in Fig. 6.10B) (Englund et al., 2006).

Formation of cerebellar cytoarchitecture Chapter | 6

125

FIGURE 6.10 Schematic representation showing the migration of Golgi cells, unipolar brush cells, and Lugaro cells in the embryonic and early postnatal cerebellum. (A) Golgi cell migration. (B) Unipolar brush cell migration. (C) Lugaro cell migration. In (AeC), numbers (1e4) in each figure represent the order of migration steps. Each left figure represents embryonic development and each right figure represents early postnatal development. DF, differentiating field; dWM, deep white matter; EGL, external granular layer; IGL, internal granular layer; IeX, lobule Ielobule X; pWM, prospective white matter; RG, radial glial cells; uRL, upper rhombic lip; VZ, ventricular zone of cerebellar plate.

(d) Migration of Lugaro cells: Lugaro cell precursors migrate from the VZ to the prospective WM (1e2 in Fig. 6.10C) and then continuously proliferate in the deep WM (3 in Fig. 6.10C) (Leto et al., 2012). After final mitosis, Lugaro cells migrate through the folial WM to their final position, the top of the IGL (3e4 in Fig. 6.10C).

6.4.3.4 Formation of the white matter and the cerebellar nuclei The formation of the WM results from the invasion of mossy fibers and climbing fibers and the development of Purkinje cell axons, while the formation of the CN results from the migration of glutamatergic CN excitatory projection neurons, GABAergic CN inhibitory projection neurons (nucleo-olivary neurons), and GABAergic CN inhibitory interneurons.

126 PART | I Induction and patterning of the CNS and PNS

(a) Invasion of mossy fibers, climbing fibers, and Purkinje cell axons: Mossy fibers from the vestibular ganglion are the first mossy fibers to arrive in the cerebellum and are present in the cerebellar anlage by E13 in mice (Ashwell and Zhang, 1992). The next mossy fiber subsets to arrive are from the vestibular nuclei and spinal cord at E15. At P0, mossy fibers originating from the lateral reticular nucleus and the pontine nuclei reach the cerebellum. In contrast with mossy fibers, climbing fibers originate solely from the IO in the medulla oblongata of the brainstem. Climbing fibers cross the midline in the brainstem, enter the cerebellum through the inferior cerebellar peduncle, and terminate contralaterally within the cerebellum. In mice, the first climbing fibers arrive in the cerebellum at E14 (Ashwell and Zhang, 1992; Rahimi-Balaei et al., 2015). The Purkinje cell axons pass through the IGL, enter the WM, and terminate within the CN or the vestibular nuclei in the brain stem. In mice, immature Purkinje cells start to extend their axon-like fibers by E13.5 (Miyata et al., 2010). (b) Migration of glutamatergic CN excitatory projection neurons: In the embryonic cerebellum, glutamatergic CN excitatory projection neurons migrate rostrally from the uRL to the nuclear transitory zone (NTZ), a transient cell mass which is located just below the pial surface at the rostral end of the cerebellar plate (1e4 in Fig. 6.11A) (Fink et al., 2006). The migration pathway from the uRL to the NTZ is known as the rostral rhombic lip migratory stream (RLS). During later stages of embryonic development, the RLS is replaced by the EGL and the NTZ is subsequently portioned and organized into distinct CN (Wang et al., 2005). (c) Migration of GABAergic CN inhibitory projection neurons (nucleo-olivary neurons): Progenitors of GABAergic CN inhibitory projection neurons actively proliferate in the VZ of cerebellar plate from E10 to E13. After final mitosis, GABAergic CN inhibitory projection neurons migrate radially toward cortical surface and complete their migration within the NTZ (1e2 in Fig. 6.11B). (d) Migration of GABAergic CN inhibitory interneurons: The precursors of GABAergic CN inhibitory interneurons first migrate radially from the VZ to the intermediate zone (1e2 in Fig. 6.11C) and then continue to proliferate as they migrate through the prospective WM (3 in Fig. 6.11C). Thereafter, GABAergic CN inhibitory interneurons settle in the CN (4 in Fig. 6.11C) (Zhang and Goldman, 1996; Maricich and Herrup, 1999).

6.4.3.5 Mechanisms underlying the control of neuronal migration As described in previous sections, each cerebellar neuron exhibits cell typeespecific and cortical layerespecific migration. To understand the mechanisms underlying the migration of cerebellar neurons, we will focus on the studies examining the cellular mechanisms of granule cell migration. This is because granule cell migration has been extensively examined for the last four decades, and it has been shown that cellular and molecular mechanisms underlying granule cell migration are utilized in other neurons with minor modifications (Komuro and Rakic, 1998b; Komuro and Kumada, 2005; Botia et al., 2007; Jiang et al., 2008; Raoult et al., 2014; Komuro et al., 2015; Galas et al., 2017). In this section, we will review the role of intracellular and extracellular molecules in controlling granule cell migration one by one. (a) Ca2þ channels: In the early 1990s, the combined use of acute cerebellar slice preparations and pharmacological tools revealed the role of voltage-gated Ca2þ channels, especially the N-type Ca2þ channel, in granule cell migration (Komuro and Rakic, 1992). Granule cells at the middle and the bottom of the EGL start to express N-type Ca2þ channels before the initiation of their migration. The blockade of N-type Ca2þ channel activity by a specific antagonist significantly reduces the speed of granule cell migration in the ML, suggesting that the Ca2þ influx through the Ntype Ca2þ channels plays a role in controlling the speed of granule cell migration. (b) NMDA (N-methyl-D-aspartate) receptors: The presence of spontaneous activity of the NMDA receptors on the surface of migrating cerebellar granule cells has been confirmed by patch-clamp analysis (Rossi and Slater, 1993). The frequency of the spontaneous NMDA receptorecoupled channel activity is low in the middle and the bottom of the EGL, but high in the ML. Migrating granule cells co-express the NR1 and NR2A or NR2B subunits of the NMDA receptor, whereas postmigratory cells in the IGL express the NR1 and NR2C types (Farrant et al., 1994). Blocking NMDA receptor activity with its antagonists significantly decreases the speed of granule cell movement in the ML (Komuro and Rakic, 1993). The role of the NMDA receptor in granule cell migration is further supported by evidence that changes in Mg2þ or glycine concentration affect the speed of granule cell movement. (c) Ca2þ spikes: The use of Ca2þ indicator dyes reveals that granule cells exhibit a distinct pattern of transient Ca2þ elevations as they migrate in different cortical layers (Kumada and Komuro, 2004). The changes in the frequency of intracellular Ca2þ transients of granule cell somata along the migratory pathway are as follows: In the EGL: At the top of the EGL, granule cell precursors exhibit the transient elevations of intracellular Ca2þ levels in their somata with a low frequency (average frequency, 8.3/h). The intervals of occurrences are regular and the amplitude is

Formation of cerebellar cytoarchitecture Chapter | 6

127

FIGURE 6.11 Schematic representation showing the migration of CN excitatory projection neurons, CN inhibitory projection neurons, and CN inhibitory interneurons in the embryonic cerebellum. (A) Migration of CN excitatory projection neurons. (B) Migration of CN inhibitory projection neurons. (C) Migration of CN inhibitory interneurons. In (AeC), numbers (1e4) in each figure represent the order of migration steps. AbL, anterobasal lobe; AdL, anterodorsal lobe; CeL, central lobe; CN, cerebellar nuclei; DF, differentiating field; EGL, external granular layer; InL, inferior lobe; IZ, intermediate zone; NTZ, nuclear transitory zone; PoL, posterior lobe; pWM, prospective white matter; RG, radial glial cells; RLS, rostral rhombic lip migratory stream; uRL, upper rhombic lip; VZ, ventricular zone of cerebellar plate.

uniform. Concomitant with the initiation of tangential migration at the middle of the EGL, postmitotic granule cells significantly increase the frequency of Ca2þ transients (20.9/h). The Ca2þ transients gradually decrease in number at the bottom of the EGL (15.9/h) and the EGL-ML border (12.8/h), and the rhythm becomes irregular, containing short, silent periods. In the ML: Once granule cells enter the ML, the cells slightly increase the numbers of Ca2þ transients (15.1/h at the top of the ML and 17.2/h at the middle). At the bottom of the ML, the Ca2þ transient frequency gradually decreases to 12.2/h and the amplitudes of Ca2þ transients become variable. In the PCL: Upon entering the PCL, granule cells significantly reduce the frequency of Ca2þ transients with long, silent periods and also decrease the amplitude of Ca2þ transients. The average frequencies of Ca2þ transients are 7.3/h at the top of the PCL and 6.9/h at the bottom.

128 PART | I Induction and patterning of the CNS and PNS

In the IGL: At the top of the IGL, granule cells significantly increase the Ca2þ transient frequency (15.1/h), although the rhythms are irregular and the amplitudes are variable. As the granule cells traverse the middle of the IGL, the frequency of Ca2þ transients gradually decreases to 9.3/h and the amplitude becomes smaller. At the bottom of the IGL, the Ca2þ transients disappear or significantly decrease in frequency (2.4/h). The changes in the frequency of Ca2þ transients in the granule cell somata along the migratory pathway positively correlate with the changes in the speed of cell movement (correlation coefficient, 0.85) (Kumada and Komuro, 2004), suggesting that the frequency of the Ca2þ transient may be one of the factors which control the alterations of granule cell migration in a cortical layerespecific manner. At their final destination of migration in the IGL, granule cells completely lose the Ca2þ transients or significantly reduce the frequency (Kumada and Komuro, 2004), suggesting that the loss of Ca2þ transients may be prerequisite for completing granule cell migration at their final destination. The role of the loss of Ca2þ transients in the completion of migration is supported by the experiments examining the effects of alterations of the Ca2þ transient frequency on granule cell migration at the bottom of the IGL. The inhibition of the Ca2þ signaling by decreasing Ca2þ influx or internal Ca2þ release results in a significant reduction of the Ca2þ transient frequency and a slowdown of granule cell movement at the bottom of the IGL (Kumada and Komuro, 2004; Komuro and Kumada, 2005). In contrast, stimulating the Ca2þ signaling by enhancing the internal Ca2þ release significantly increases the Ca2þ transient frequency and accelerates granule cell movement at the bottom of the IGL. These results indicate that the loss of Ca2þ transients may trigger molecular cascades leading to the completion of granule cell migration. (d) Somatostatin (SST): Somatostatin, a neuropeptide, has two bioactive products, somatostatin-14 (SST-14) and somatostatin-28 (SST-28), which are differing cleavage products of prosomatostatin. Both SST-14 and SST-28 bind to all five somatostatin receptors (SSTRs). Numerous brain regions, including the cerebral cortex and cerebellum, exhibit high levels of somatostatin and its receptor expression early in development, followed by a decrease to adult levels (Yacubova and Komuro, 2003). In the early postnatal cerebellum, postmitotic granule cells express all five types of SSTRs before an initiation of migration, although differentiated granule cells in the adult do not express the receptors. High levels of somatostatin are present along the migratory route of granule cells and in their final destination (Yacubova and Komuro, 2002). SST-14 is present in Purkinje cells, Golgi cells, and climbing fibers, and SST28 is present in Golgi cells and mossy fiber terminals. The application of SST-14 or SST-28 significantly increases the speed of granule cell movement in the EGL, slightly decreases the speed in the ML, and significantly decreases the speed in the IGL (Yacubova and Komuro, 2002). In contrast, the application of a somatostatin antagonist (AC178,335) significantly decreases the speed of granule cell migration in the EGL, slightly increases the speed in the ML, and significantly increases the speed in the IGL. These results indicate that somatostatin accelerates the tangential movement of granule cells near the birthplace within the EGL, but significantly slows down the radial movement near their final destination within the IGL. (e) Pituitary adenylate cyclase-activating polypeptide (PACAP): PACAP, a member of the secretin/glucagon/vasoactive intestinal polypeptide family, is known to control physiological functions of a wide range of cells (Vaudry et al., 2000). PACAP has two bioactive products, PACAP38 and PACAP27. PACAP27 is the N-terminal 27-amino acid sequence of PACAP38. There are three types of PACAP receptors (PAC1, VPAC1, and VPAC2) (Vaudry et al., 2000). There is a unique pattern of endogenous PACAP expression in the developing cerebellum: PACAP is present sporadically in the bottom of the ML, expressed intensively in the PCL, and dispersedly throughout the IGL (Cameron et al., 2007), indicating that endogenous PACAP is highly expressed in the route of granule cell migration in specific cerebellar cortical layers. Furthermore, granule cells and its precursors express high levels of PAC1 receptors in the EGL of the developing cerebellum (Falluel-Morel et al., 2005; Botia et al., 2007). Granule cells and its precursors also express VPAC1 receptors, but at a much lower level than the PAC1 receptors, but do not express VPAC2 receptors. The application of exogenous PACAP38 significantly slows down the radial migration of granule cells in the ML, suggesting that PACAP acts on granule cell migration as a “brake” (stop signal) for cell movement (Cameron et al., 2007). Surprisingly, the effect of exogenous PACAP38 and a potent PACAP antagonist (PACAP638) on granule cell migration varies among each cortical layer (Cameron et al., 2007). For example, the application of exogenous PACAP38 reduces granule cell motility by 62% in the EGL, 62% at the top of the ML, 52% at the bottom of the ML, 8% in the PCL, and 5% at the top of the IGL. Conversely, the application of PACAP6-38 reduces granule cell motility by 10% in the EGL and 1% at the top of the ML but increases by 19% at the bottom of the ML, 68% in the PCL, and 1% at the top of the IGL. These results demonstrate that blocking the activation of PACAP receptors significantly increases the speed of granule cell migration in the PCL, suggesting that the slowdown of the granule cell migration observed in the PCL is caused by endogenous PACAP through the activation of its receptors. (f) Brain-derived neurotrophic factor (BDNF): Granule cells express BDNF and its high affinity receptor (TrkB). In BDNF/ mice, granule cells exhibit impaired migration, and the application of BDNF accelerates movement,

Formation of cerebellar cytoarchitecture Chapter | 6

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o) (p)

129

suggesting that BDNF secreted by granule cells controls their migration through the activation of TrkB receptors in a paracrine manner (Borghesani et al., 2002). Neurotrophin-3 (NT-3): NT-3 mRNA expression in the EGL increases during the period of granule cell migration. Migrating granule cells express the NT-3 receptor (TrkC). The application of NT-3 reduces the thickness of the EGL without cell death, suggesting that NT-3 accelerates the exit of granule cells from the EGL (Doughty et al., 1998b). Neuregulin (NRG): Granule cells express NRG, and glial cells express its receptor, erbB4. Blockade of the glial erbB4 receptors impairs granule cell migration along the glial fibers, indicating that NRG and erbB4 are essential for glia-associated granule cell migration (Rio et al., 1997). Stromal cell-derived factor 1a (SDF-1a) and CXCR4: SDF-1a is present in the pial membrane, while granule cells express its receptor CXCR4. In SDF-1a- or CXCR4-knockout mice, granule cells prematurely migrate away from the EGL and are found ectopically outside the EGL. Because SDF-1a induces chemotactic responses in granule cell precursors, SDF-1a and CXCR4 play a role in retaining granule cells in the EGL (Ma et al., 1998). Ephrin-B2 and EphB receptors: Ephrin-B2 and its receptor (EphB2) are expressed in the EGL. The chemoattractant effect of SDF-1a to granule cells is inhibited by soluble EphB2 receptor through reverse signaling of ephrin-B2, suggesting that when granule cells are ready to migrate out from the EGL, they lose responsiveness to SDF-1a (Lu et al., 2001). Tissue plasminogen activator (tPA): Granule cells both secrete tPA and bind tPA to their cell surface (Raoult et al., 2011). It has been shown that tPA facilitates the migration of both granule cells and basket/stellate cells in the ML through degradation of the extracellular matrix (Raoult et al., 2014). In tPA/ mice, granule cells migrate in the ML at a reduced speed. The application of tPA inhibitor decelerates granule cell migration in the ML, suggesting that tPA is required for maintaining the speed of granule cell migration in the ML (Friedman and Seeds, 1995). Platelet-activating factor (PAF): PAF has been implicated in the human neuronal migration disorder MillereDieker lissencephaly. Indeed, application of the nonhydrolyzable PAF receptor agonist yields a dose-dependent decrease in granule cell migration (Bix and Clark, 1998). Cyclin-dependent kinase 5 (Cdk5): In the cdk5/-cdk5þ/þ chimeric mice, many cdk5/ granule cells stay in the ML and fail to reach the IGL. The overexpression of a dominant-negative form of Cdk5 disrupts radial migration of granule cells, suggesting that cdk5 plays a critical role in glia-associated radial migration of granule cells in ML (Ohshima et al., 1999). Cyclic AMP (cAMP) and cyclic GMP (cGMP): cAMP and cGMP signaling control granule cell migration in opposing ways. Stimulation of cAMP activity decelerates granule cell migration, whereas stimulation of cGMP activity accelerates migration. Conversely, inhibition of cAMP activity accelerates granule cell migration, whereas inhibition of cGMP activity decelerates migration (Komuro et al., 2015). Mitogen-activated protein kinase (MAPK): MAPK signaling is a downstream target for extracellular molecules such as BDNF, NT-3, SDF-1a, and Cdk5, which control granule cell migration (Gunn-Moore et al., 1997). Semaphorin6A (Sema6A) and plexin-A2: Migrating granule cells in the EGL express Sema6A and its receptor (plexin-A2). In both Sema6A and plexin-A2 knockout mice, granule cells fail to enter the ML because of a defect in nucleusecentrosome coupling at the beginning of the radial migration (Renaud and Chédotal, 2014).

6.4.3.6 The deficits of neuronal migration by exposure to toxic substances and natural environmental factors result in abnormal O-I patterning Exposure to toxic foods, hazardous material, and even natural environments (such as light) during gestation and lactation results in abnormal migration of immature neurons, leading to defects in the formation of cerebellar cortical layers (O-I patterning). In this section, we will review how toxic foods, hazardous material, and even natural environments adversely affect the migration of immature neurons. (a) Alcohol: To date, alcohol is the most common chemical teratogen causing malformation and mental deficiency in humans. Prolonged exposure to alcohol during gestation and lactation correlates results in abnormal development in newborns, which is called as “fetal alcohol spectrum disorders” (FASDs). The spectrum of alcohol’s teratogenic effects spans a wide continuum that includes growth deficiency, central nervous dysfunction, craniofacial anomalies, and pathologic organ and skeletal conditions (Riley and McGee, 2005; Welch-Carre, 2005). Alcohol is the most common preventable cause of birth defects and the leading cause of mental retardation ahead of Down syndrome and cerebral palsy. The most devastating consequences of alcohol exposure are its effect on the brain (Guerri, 2002). Several aspects of the developmental program are involved in the alcohol-induced malformation of the brain. Among them, the most striking abnormalities appear to involve the impairment of neuronal cell migration (Kumada et al., 2006). Children with FASDs

130 PART | I Induction and patterning of the CNS and PNS

show neurological signs associated with cerebellar damage such as delayed motor development, problems with fine tasks, and ataxia (Coffin et al., 2005; Manzardo et al., 2005). The most vulnerable period of cerebellar development in humans is during the third trimester (Clarren, 1986). The equivalent time of development in mice is during the early postnatal period (Kornguth et al., 1979). Using the early postnatal mouse as model system, Kumada and his colleagues demonstrate the effects of alcohol on the migration of cerebellar granule cells (Kumada et al., 2006, 2010). The administration of ethanol immediately slows down the tangential migration of granule cells in the EGL of P10 mouse in a dose-dependent manner. For example, 10 mM ethanol (equivalent to blood ethanol level