Neural Repair: Methods and Protocols (Methods in Molecular Biology, 2616) 9781071629253, 9781071629260, 1071629255

Table of contents :
Preface
Contents
Contributors
Part I: Stroke Models and Surgical Interventions
Chapter 1: Rodent Stroke Models to Study Functional Recovery and Neural Repair
1 Introduction
2 Models of Ischemic Stroke for Chronic Outcomes
3 Concluding Remarks
References
Chapter 2: Subcortical White Matter Stroke in the Mouse: Inducing Injury and Tracking Cellular Proliferation
1 Introduction
2 Materials
2.1 EdU Administration
2.2 Induction of WMS
2.3 EdU Labeling
3 Methods
3.1 Administration of EdU
3.2 Induction of WMS
3.3 Labeling of EdU
4 Notes
References
Chapter 3: A Low-Budget Photothrombotic Rodent Stroke Model
1 Introduction
2 Materials
2.1 Animals
2.2 Equipment
2.3 Surgical Instruments
2.4 Reagents and Supplies
3 Methods
3.1 Aseptic Preparation of the Animal
3.2 Surgery
3.3 Post-Surgical Care
3.4 Further Experimentation and Outcome Measures
4 Notes
References
Chapter 4: Photothrombotic Model to Create an Infarct in the Hippocampus
1 Introduction
2 Materials
2.1 Photothrombotic Surgery
2.1.1 Materials and Reagents
2.1.2 Equipment
2.1.3 Laser Setup
2.2 2,3,5-Triphenyltetrazolium Chloride (TTC) Staining
2.2.1 Materials and Reagents
2.2.2 General Equipment
3 Methods
3.1 Hippocampal Photothrombosis
3.1.1 Preparation of Reagents and Work Area
3.1.2 Photothrombotic Surgery
3.2 TTC (2,3,5-Triphenyltetrazolium Chloride) Staining
3.2.1 Tissue Collection and Staining
3.2.2 Imaging of TTC-Stained Sections
4 Notes
References
Chapter 5: Bilateral Carotid Artery Stenosis and Cerebral Blood Flow Outcomes
1 Introduction
2 Materials
2.1 Equipment and Instruments
2.2 Supplies
3 Methods
3.1 Autoclave Surgery Materials
3.2 Surgical Preparation of Mice
3.3 Pre-BCAS Surgery, CBF Measurement
3.4 BCAS Surgery and Post-BCAS CBF Measurement
4 Notes
References
Chapter 6: Internal Carotid Artery Stenosis: A Surgical Mouse Model to Study Moyamoya Syndrome
1 Introduction
2 Materials
2.1 Animals and Surgical Supplies
3 Methods
3.1 Pre-ICAS Preparations
3.2 ICAS Procedure
4 Notes
References
Chapter 7: Modeling Distal Middle Cerebral Artery Occlusion in Neonatal Rodents with Magnetic Nanoparticles or Magnetized Red ...
1 Introduction
2 Materials
2.1 The SIMPLE Model
2.2 The SIMPLeR Model
3 Methods
3.1 SIMPLE Model
3.2 SIMPLeR Model
4 Notes
References
Part II: In Vivo and Post-mortem Imaging of Recovery Mechanisms
Chapter 8: In Vivo Imaging of the Structural Plasticity of Cortical Neurons After Stroke
1 Introduction
1.1 Optical Window on the Mouse Brain Cortex
1.2 In Vivo Two-Photon Imaging of Neuronal Plasticity
1.3 Photothrombotic Stroke
2 Materials
2.1 Optical Window on the Mouse Brain Cortex
2.2 In Vivo Two-Photon Imaging of Neuronal Plasticity
2.3 Photothrombotic Stroke
2.4 Image Processing and Analysis
3 Methods
3.1 Optical Window on the Mouse Brain Cortex
3.2 In Vivo Two-Photon Imaging of Neuronal Plasticity
3.3 Photothrombotic Stroke
3.4 Image Processing and Analysis
4 Notes
References
Chapter 9: Measurement of Uninterrupted Cerebral Blood Flow by Laser Speckle Contrast Imaging (LSCI) During the Mouse Middle C...
1 Introduction
2 Materials
2.1 Animals
2.2 Equipment for Modifying Bench Top
2.3 Laser Speckle Contrast Imaging System
2.4 Surgical Equipment and Supplies
3 Methods
3.1 Mounting the Inverted Laser Speckle Contrast Imaging System
3.2 Mounting the Head Frame to the Skull
3.3 Preparation of Skull to Improve Clarity for Imaging
3.4 Procedure for LSCI with the Middle Cerebral Artery Occlusion and Reperfusion (MCAo/R) Model
3.5 Modified Procedure for LSCI and MCAo/R Procedure with Aged Mice
4 Notes
References
Chapter 10: Multi-exposure Speckle Imaging for Quantitative Evaluation of Cortical Blood Flow
1 Introduction
2 Materials
2.1 Illumination Optics
2.2 Image Acquisition
2.3 Post-processing
3 Methods
3.1 System Setup
3.1.1 Optical Components
3.1.2 Electrical Components
3.2 Laser Alignment
3.3 MESI Calibration
3.4 MESI Acquisition
3.5 Shutdown
3.6 Data Processing
3.6.1 Calculating Speckle Contrast
3.6.2 Imaging and Extracting Quantitative Flow Information
4 Notes
References
Chapter 11: Wide-Field Optical Imaging in Mouse Models of Ischemic Stroke
1 Introduction
1.1 Animals and Imaging Contrasts, Housing, and Animal Preparation
1.2 Wide-Field Optical Imaging System: Overview
1.3 Imaging Protocol Considerations During Anesthesia
1.4 Imaging Protocols: Acclimation for Awake Imaging
1.5 Imaging Protocols: Evoked Responses
1.6 Imaging Protocols: Resting-State Imaging
1.7 WFOI Data Processing
1.8 WFOI Data Analysis
2 Materials
2.1 Animal Preparation for Serial Wide-Field Optical Imaging
2.2 Wide-Field Optical Imaging System
2.2.1 LEDS and Filters
2.2.2 Imaging System
2.2.3 Control Hardware
2.2.4 Head-Fixing Apparatus
2.3 Imaging Protocols: Anesthesia
2.4 Imaging Protocols: Acclimation for Awake
2.5 Imaging Protocols: Forepaw-Evoked Response Imaging
2.6 Imaging Protocols: Whisker-Evoked Response Imaging
2.7 Imaging Protocols: Resting State
2.8 WFOI Data Processing and Analysis
3 Methods
3.1 Animal Preparation for Serial Wide-Field Optical Imaging
3.2 Wide-Field Optical Imaging System
3.2.1 To Build the System of LEDs
3.2.2 Head Mounting Setup of the Imaging System (a Picture of Which Is Shown in Fig. 3)
3.3 Imaging Protocols: Anesthesia
3.4 Imaging Protocols: Acclimation for Awake Imaging
3.5 Imaging Protocols: Forepaw-Evoked Response Imaging
3.6 Imaging Protocols: Whisker-Evoked Response Imaging
3.7 Imaging Protocols: Resting-State Imaging
3.8 WFOI Processing
3.8.1 OIS Processing
3.8.2 Fluorescence Processing and Hemodynamic Correction
3.9 WFOI Analysis
4 Notes
References
Chapter 12: Post-mortem Magnetic Resonance Imaging of Degenerating and Reorganizing White Matter in Post-stroke Rodent Brain
1 Introduction
2 Materials
2.1 Transcardial Perfusion-Fixation
2.2 Preparation of Sample for Scanning
2.3 MR Scanning
3 Methods
3.1 Transcardial Perfusion-Fixation of Rat and Mouse Brains
3.2 Preparation of Sample for Scanning
3.3 MR Scanning
3.4 Analysis of Diffusion MRI Data
3.4.1 Pre-processing
3.4.2 Registration to Rodent Brain Atlas
3.4.3 Analysis of Diffusion Parameters
4 Notes
References
Untitled
Part III: Methods to Identify Molecular and Immune Mechanisms Supporting Recovery
Chapter 13: Quantitative Spatial Mapping of Axons Across Cortical Regions to Assess Axonal Sprouting After Stroke
1 Introduction
1.1 Overview of Methods
2 Materials
2.1 Axonal Tracer
2.2 Surgical Reagents
2.3 Surgical Equipment
2.4 Histology Reagents
2.5 Microscopy and Analysis
3 Methods
3.1 Animal Surgery
3.2 Histology
3.3 Microscopy and Semi-automated Axonal Tracing
3.4 Analysis
4 Notes
References
Chapter 14: Quantitative Evaluation of Cerebral Microhemorrhages in the Mouse Brain
1 Introduction
2 Materials
2.1 Mouse Brain Preparation
2.2 H & E Staining
2.3 Prussian Blue Staining
2.4 Cerebral Microhemorrhage Imaging and Quantification
3 Methods
3.1 Mouse Brain Preparation
3.2 Slide Preparation
3.3 H & E Staining
3.4 Prussian Blue Staining
3.5 Imaging and Quantification of H & E and Prussian Blue Staining
4 Notes
References
Chapter 15: In Vivo Evaluation of BBB Integrity in the Post-stroke Brain
1 Introduction
2 Materials
3 Methods
3.1 Preparation of Solutions
3.2 Quantitative Assays
3.2.1 Administration of Tracers, Cardiac Perfusion, and Serum and Brain Collection
3.2.2 Homogenization and Centrifugation
3.2.3 Fluorescence Measurement and Quantification
3.2.4 Example of Quantitation Calculation
3.3 Morphometric Assays
3.3.1 Section Slide Preparation
3.3.2 IHC-F Staining
4 Notes
References
Chapter 16: High-Resolution RNA Sequencing from PFA-Fixed Microscopy Sections
1 Introduction
2 Materials
2.1 Tissue Isolation
2.1.1 Dissection Equipment
2.1.2 General Equipment
2.1.3 Media and Reagents
2.2 Lysis and Reverse Cross-Link
2.3 Purify and Elute mRNA
2.3.1 General Equipment
2.3.2 Media and Reagents
2.4 RNA-seq with Smart-seq2
3 Methods
3.1 Brain Tissue Isolation via Microdissection
3.2 Lyse Tissue and Reverse Cross-Link
3.3 Purify and Elute mRNA
3.4 RNA-seq with Smart-seq2
4 Notes
References
Chapter 17: FACS to Identify Immune Subsets in Mouse Brain and Spleen
1 Introduction
2 Materials
2.1 Tissue Collection/Brain Perfusion
2.2 Spleen Processing
2.3 Brain Processing
2.4 Fluorescently Activated Cell Sorting (FACS Staining)
3 Methods
3.1 Tissue Collection
3.2 Spleen Processing
3.3 Brain Processing
3.4 FACS Staining of Mouse Brain and Spleen
3.5 Compensation Controls of Single Antibody Stain
3.5.1 Cells for Compensation
3.5.2 Beads for Compensation
4 Notes
References
Chapter 18: A Guide on Analyzing Flow Cytometry Data Using Clustering Methods and Nonlinear Dimensionality Reduction (tSNE or ...
1 Introduction
2 Materials
3 Methods
3.1 FlowAI
3.2 Gating Target Population and Concatenation
3.3 Using FlowSOM
3.4 Visualization in Two-Dimensional Space Using tSNE or UMAP
4 Notes
References
Chapter 19: Co-culturing Immune Cells and Mouse-Derived Mixed Cortical Cultures with Oxygen-Glucose Deprivation to In Vitro Si...
1 Introduction
2 Materials
2.1 Cell Culture
2.2 Oxygen-Glucose Deprivation
2.3 Cell Viability Assays
3 Methods
3.1 Cortex Isolation
3.2 Cell Culture
3.3 Co-culture MCC with B Cells
3.4 Assays for Cell Viability
3.4.1 General Cell Viability with the MTT Assay
3.4.2 Neuronal Health Assessment
4 Notes
References
Part IV: Behavioral Methods for Quantifying Functional Recovery
Chapter 20: Assessing Depression and Cognitive Impairment Following Stroke and Neurotrauma: Behavioral Methods for Quantifying...
1 Introduction
1.1 Barnes Maze Test
1.2 Novel Object Recognition Test (NORT)
1.3 Sucrose Preference Test Methods
1.4 Three-Chambered Sociability Approach Test/Social Interaction
1.5 Burrowing Test
2 Materials
2.1 Barnes Maze Test
2.2 Novel Object Recognition Test
2.3 Sucrose Preference Test Methods
2.4 Three-Chambered Sociability Approach Test/Social Interaction
2.5 Burrowing Test: Protocol for Mice
2.6 Burrowing Test: Protocol for Rats
3 Methods
3.1 Barnes Maze Test
3.1.1 Acclimation
3.1.2 Testing Spatial Learning and Memory
3.1.3 Data Analyses
3.2 Novel Object Recognition Test
3.2.1 Acclimation
3.2.2 Testing Novel Object Recognition
3.2.3 Data Analysis
3.3 Sucrose Preference Test Methods
3.3.1 Preparation
3.3.2 Acclimation
3.3.3 Testing Sucrose Preference
3.3.4 Data Analysis
3.4 Three-Chambered Sociability Approach Test/Social Interaction
3.4.1 Acclimation
3.4.2 Testing for Social Interaction
3.4.3 Data Analysis
3.5 Burrowing Test: Protocol for Mice
3.5.1 Preparation
3.5.2 Testing Burrowing (Baseline and Experiment)
3.5.3 Data Analysis
3.6 Burrowing Test: Protocol for Rats
3.6.1 Preparation
3.6.2 Acclimation
3.6.3 Testing Burrowing
3.6.4 Data Analysis
4 Notes
References
Chapter 21: Use of an Automated Mouse Touchscreen Platform for Quantification of Cognitive Deficits After Central Nervous Syst...
1 Introduction
2 Materials
3 Methods
3.1 Pre-experiment Preparation
3.1.1 Mouse Preparation
3.1.2 Food Restriction
3.1.3 Habituation to Reward
3.2 Daily Session Protocol
3.2.1 Mouse Habituation to Touchscreen Testing Environment
3.2.2 Chamber Setup: Loading Reward
3.2.3 Chamber Setup: Testing Touchscreen, Reward Magazine, and IR Beam
3.2.4 Chamber Setup: Load Session Schedules and Load Mice into Chambers
3.2.5 Chamber Setup: Between Subjects
3.2.6 Chamber and Room Cleaning: At End of Day
3.3 Pretraining
3.3.1 Preparing for Pretraining
3.3.2 Running Pretraining
3.3.3 Pretraining Data Collection
3.4 Paired Associates Learning (PAL)
3.4.1 PAL Task-Specific Training
3.4.2 PAL Test
3.4.3 PAL Data Collection
3.5 Location Discrimination Reversal (LDR)
3.5.1 LDR Train
3.5.2 LDR Test
3.5.3 LDR Data Collection
3.6 Autoshaping (AUTO)
3.6.1 AUTO-Specific Training
3.6.2 AUTO-Specific Testing
3.6.3 AUTO Data Collection
3.7 Extinction (EXT)
3.7.1 EXT-Specific Training (Acquisition of Stimulus-Response)
3.7.2 EXT-Specific Testing
3.7.3 EXT Data Collection
3.8 Troubleshooting
4 Notes
References
Chapter 22: Using Operant Reach Chambers to Assess Mouse Skilled Forelimb Use After Stroke
1 Introduction
2 Materials
3 Methods
3.1 Hardware Setup
3.2 Calibration and Testing (Initial Setup and Weekly)
3.3 Training-Group Phase (5 Consecutive Days, 1 Week)
3.4 Training: Individual Phase (5 Sessions per Week, up to 3 Weeks)
3.5 Individual Baseline Measurement (3 Consecutive Days)
3.6 Post-stroke Assessment and Post-stroke Rehabilitation
4 Notes
References
Chapter 23: The Finer Aspects of Grid-Walking and Cylinder Tests for Experimental Stroke Recovery Studies in Mice
1 Introduction
2 Materials
2.1 Animals
2.2 Equipment
2.3 Supplies
3 Methods
3.1 Preparation of Animals
3.2 Execution of the Grid-Walking and Cylinder Tests
3.3 Data Acquisition and Analysis
4 Notes
References
Chapter 24: Performing Enriched Environment Studies to Improve Functional Recovery
1 Introduction
2 Materials
2.1 Animals
2.2 Cages
2.2.1 Rat Cages
2.2.2 Mouse Cages
2.3 Components of Cages
2.3.1 Cage Interior
2.3.2 Food and Water Supply
3 Methods
3.1 Design of EE Experiments
3.2 Duration of the Experiments
3.3 Preparation of Cages
3.4 Animals´ Allocation into Cages
3.5 Daily Checkups
3.6 Assessment of Neurological Function After EE
3.6.1 Effects of Multimodal Stimulation by EE
3.6.2 Behavior Assessment
4 Notes
References
Part V: Neurotherapeutics and Functional Recovery
Chapter 25: Clinically Applicable Experimental Design and Considerations for Stroke Recovery Preclinical Studies
1 Introduction
2 Starting Considerations
3 Rehabilitation
4 Start of Therapy
5 Sequence and Pairing of Therapies
6 Dose and Duration of Therapy
7 Outcome Measures
8 Concluding Remarks
References
Chapter 26: Hydrogels and Nanoscaffolds for Long-Term Intraparenchymal Therapeutic Delivery After Stroke
1 Introduction
2 Materials
2.1 General Equipment
2.2 Chemicals and Reagents
2.2.1 Chitosan/β-Glycerophosphate Hydrogel
2.2.2 PVA-Tyramine Hydrogel
2.2.3 BioTime Hydrogel
2.2.4 Hydrogel Injection Surgery
3 Methods
3.1 Preparation of Hydrogels
3.1.1 Chitosan/β-Glycerophosphate Hydrogel
3.1.2 PVA-Tyramine Hydrogel
3.1.3 BioTime Hydrogel
3.2 Embedding of Therapeutics into Hydrogels
4 Notes
References
Chapter 27: Reverse Translation to Develop Post-stroke Therapeutic Interventions during Mechanical Thrombectomy: Lessons from ...
1 Introduction
2 BACTRAC Methods Relevant to Reverse Translation
2.1 Clinical Data Relevant to Identifying Mechanism(S) of Injury and Repair
2.2 Methods to Isolate Thrombus
2.3 Methods to Isolate Peri-Infarct and Systemic Arterial Blood Samples
2.4 Methods to Analyze Human Specimens Collected During Thrombectomy
3 Creating a BACTRAC-Relevant Animal Model of Stroke
3.1 Conclusion
4 Notes
References
Chapter 28: Methods to Study Drug Uptake at the Blood-Brain Barrier Following Experimental Ischemic Stroke: In Vitro and In Vi...
1 Introduction
2 Materials
2.1 Transwell Permeability Assay
2.2 Cellular Uptake Assay
2.3 In Situ Brain Perfusion
2.3.1 Perfusion Equipment and Surgical Instruments
2.3.2 Perfusion Media
3 Methods
3.1 Transwell Permeability Assay
3.2 Cellular Uptake Assay
3.3 In Situ Perfusion
3.3.1 Perfusion Preparation
3.3.2 Surgery and Perfusion
3.3.3 Data Analysis
4 Notes
References
Chapter 29: Gene Silencing in the Brain with siRNA to Promote Long-Term Post-Stroke Recovery
1 Introduction
2 Materials
2.1 siRNA Formulations
2.2 Materials for siRNA Delivery
2.3 Materials for Evaluating siRNA Efficiency and Toxicity
2.4 Materialsfor Infarct Assessment
3 Methods
4 Notes
References
Part VI: Models of Comorbidities
Chapter 30: Diabetic Rodent Models for Chronic Stroke Studies
1 Introduction
1.1 Overview of Animal Models of Diabetes
1.2 Spontaneous Diabetic Animal Models
1.2.1 AKITA Mice
1.2.2 Lepob/ob Mice and Leprdb/db Mice
1.2.3 KK Mice
1.2.4 BB Rats
1.2.5 OLETF Rats
2 Materials
2.1 Equipment
2.2 Personal Protective Equipment
2.3 Materials and Reagents
3 Methods
3.1 Preparation of Stock Citrate Buffer
3.2 Working Solution (STZ-Buffer)
3.3 Preparation of Streptozotocin Solution
3.4 Streptozotocin Injection
3.5 Follow-Up
4 Notes
References
Chapter 31: Use of Conventional Cigarette Smoking and E-Cigarette Vaping for Experimental Stroke Studies in Mice
1 Introduction
2 Materials
2.1 Animals
2.2 Conventional and E-Cigarettes
2.3 Smoking Equipment
2.4 Materials for Experimental Stroke
3 Methods
3.1 Generation of Cigarette Smoke and Chronic Animal Exposure
3.2 Induction of Stroke in Mice After Chronic Cigarette Smoke Exposure
3.3 Basic Outcome Measures
3.3.1 Behavioral Tests
3.3.2 Open-Field Test
3.3.3 Terminal Procedure
4 Notes
References
Chapter 32: Middle Cerebral Artery Occlusion in Aged Animal Model
1 Introduction
2 Materials
3 Methods
3.1 Preparation
3.2 Occlusion
3.3 Reperfusion
3.4 Postoperative Care
4 Notes
References
Chapter 33: Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models of Diabetes: Importance of Pre-op and Post...
1 Introduction
2 Materials
2.1 Animals
2.2 Diet
2.3 Chemicals
2.4 Tools
3 Methods
3.1 Diabetes Induction
3.2 Blood Clot Preparation
3.3 Nylon Suture
3.4 Pre-op Care and Preparation (Fig. 1, See Note 8)
3.5 Operation Procedures (See Note 10)
3.6 Post-op Care (Fig. 2)
3.7 Surgery Evaluation
4 Notes
References
Chapter 34: The DOCA-Salt Model of Hypertension for Studies of Cerebrovascular Function, Stroke, and Brain Health
1 Introduction
2 Materials
2.1 Pellet Preparation (See Notes 1 and 2)
2.2 Surgical Instruments and Setup (See Note 3)
2.3 Drinking Water
3 Methods
3.1 Pellet Preparation (If Using Pre-purchased Pellets, Proceed to 3.2)
3.2 Subcutaneous Implantation of DOCA Pellet
3.3 Post-Surgical Monitoring and Supply of 0.15 M NaCl in Drinking Water
4 Notes
References
Index

Citation preview

Methods in Molecular Biology 2616

Vardan T. Karamyan Ann M. Stowe Editors

Neural Repair Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

Neural Repair Methods and Protocols

Edited by

Vardan T. Karamyan Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA; Department of Foundational Medical Studies, Oakland University William Beaumont School of Medicine, Rochester, MI, USA

Ann M. Stowe Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA

Editors Vardan T. Karamyan Department of Pharmaceutical Sciences School of Pharmacy Texas Tech University Health Sciences Center Amarillo, TX, USA

Ann M. Stowe Department of Neurology Department of Neuroscience The University of Kentucky Lexington, KY, USA

Department of Foundational Medical Studies Oakland University William Beaumont School of Medicine Rochester, MI, USA

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-2925-3 ISBN 978-1-0716-2926-0 (eBook) https://doi.org/10.1007/978-1-0716-2926-0 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover Illustration Caption: Maps of mean diffusivity (left), Fractional Anisotropy (middle) and an RGB display of the main eigenvector (right) following middle cerebral artery occlusion. See Chapter 12 for more details. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface This handbook explores a diverse range of topics related to neural repair and functional recovery following ischemic stroke. Techniques detailed in this book span from basic to emerging approaches used to evaluate hallmarks of neurorepair, including axonal remodeling and dendritic arborization, plasticity and functional connectivity, angiogenesis, neuroinflammation, and recovering cerebral blood flow. In addition, several chapters focus on rodent stroke models and post-stroke functional evaluation, clinically relevant therapeutic paradigms and co-morbidities, and pharmacotherapy and methods of delivery. The chapters are written as detailed, step-by-step laboratory protocols with complete lists of required materials and reagents, tips for avoiding recognized pitfalls, and troubleshooting issues. Chapters have been organized by method modalities, and most sections include a conceptual chapter to help with considerations for experimental design. Notably, many of the protocols are suitable for a larger community of researchers interested in neurorepair following other forms of CNS injury (e.g., TBI and spinal cord injury). Rochester, MI, USA Lexington, KY, USA

Vardan T. Karamyan Ann M. Stowe

v

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

STROKE MODELS AND SURGICAL INTERVENTIONS

1 Rodent Stroke Models to Study Functional Recovery and Neural Repair. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Daimen R. S. Britsch, Nausheen Syeara, Ann M. Stowe, and Vardan T. Karamyan 2 Subcortical White Matter Stroke in the Mouse: Inducing Injury and Tracking Cellular Proliferation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Miguel Alejandro Marin, Amy J. Gleichman, Andrew J. Brumm, and S. Thomas Carmichael 3 A Low-Budget Photothrombotic Rodent Stroke Model . . . . . . . . . . . . . . . . . . . . . Faisal F. Alamri, Serob T. Karamyan, and Vardan T. Karamyan 4 Photothrombotic Model to Create an Infarct in the Hippocampus. . . . . . . . . . . . Elena Blanco-Sua´rez 5 Bilateral Carotid Artery Stenosis and Cerebral Blood Flow Outcomes . . . . . . . . . Ifechukwude Joachim Biose and Gregory Jaye Bix 6 Internal Carotid Artery Stenosis: A Surgical Mouse Model to Study Moyamoya Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jill M. Roberts and Justin F. Fraser 7 Modeling Distal Middle Cerebral Artery Occlusion in Neonatal Rodents with Magnetic Nanoparticles or Magnetized Red Blood Cells . . . . . . . . . . . . . . . . Jie-Min Jia and Yuxiao Jin

PART II

v xi

3

13

21 29 39

47

55

IN VIVO AND POST-MORTEM IMAGING OF RECOVERY MECHANISMS

8 In Vivo Imaging of the Structural Plasticity of Cortical Neurons After Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Emilia Conti, Francesco Saverio Pavone, and Anna Letizia Allegra Mascaro 9 Measurement of Uninterrupted Cerebral Blood Flow by Laser Speckle Contrast Imaging (LSCI) During the Mouse Middle Cerebral Artery Occlusion Model by an Inverted LSCI Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sung-Ha Hong, Andrea Doan, and Sean P. Marrelli 10 Multi-exposure Speckle Imaging for Quantitative Evaluation of Cortical Blood Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adam Santorelli, Colin T. Sullender, and Andrew K. Dunn

vii

69

83

97

viii

11

12

Contents

Wide-Field Optical Imaging in Mouse Models of Ischemic Stroke . . . . . . . . . . . . 113 Jonah A. Padawer-Curry, Ryan M. Bowen, Anmol Jarang, Xiaodan Wang, Jin-Moo Lee, and Adam Q. Bauer Post-mortem Magnetic Resonance Imaging of Degenerating and Reorganizing White Matter in Post-stroke Rodent Brain. . . . . . . . . . . . . . . . . 153 Vera H. Wielenga, Rick M. Dijkhuizen, and Annette Van der Toorn

PART III 13

14

15 16

17

18

19

Quantitative Spatial Mapping of Axons Across Cortical Regions to Assess Axonal Sprouting After Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mary T. Joy, Samuel P. Bridges, and S. Thomas Carmichael Quantitative Evaluation of Cerebral Microhemorrhages in the Mouse Brain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rudy Chang and Rachita K. Sumbria In Vivo Evaluation of BBB Integrity in the Post-stroke Brain. . . . . . . . . . . . . . . . . Yong Zhang, Saeideh Nozohouri, and Thomas J. Abbruscato High-Resolution RNA Sequencing from PFA-Fixed Microscopy Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hao Ji, Simon Besson-Girard, Peter Androvic, Buket Bulut, Lu Liu, Yijing Wang, and Ozgun Gokce FACS to Identify Immune Subsets in Mouse Brain and Spleen . . . . . . . . . . . . . . . Mary K. Malone, Thomas A. Ujas, Katherine M. Cotter, Daimen R. S. Britsch, Jenny Lutshumba, Jadwiga Turchan-Cholewo, and Ann M. Stowe A Guide on Analyzing Flow Cytometry Data Using Clustering Methods and Nonlinear Dimensionality Reduction (tSNE or UMAP) . . . . . . . . . . . . . . . . . Thomas A. Ujas, Veronica Obregon-Perko, and Ann M. Stowe Co-culturing Immune Cells and Mouse-Derived Mixed Cortical Cultures with Oxygen-Glucose Deprivation to In Vitro Simulate Neuroinflammatory Interactions After Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thomas A. Ujas, Vanessa O. Torres, and Ann M. Stowe

PART IV 20

21

METHODS TO IDENTIFY MOLECULAR AND IMMUNE MECHANISMS SUPPORTING RECOVERY 171

181 191

205

213

231

251

BEHAVIORAL METHODS FOR QUANTIFYING FUNCTIONAL RECOVERY

Assessing Depression and Cognitive Impairment Following Stroke and Neurotrauma: Behavioral Methods for Quantifying Impairment and Functional Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Karienn A. de Souza, Michelle Hook, and Farida Sohrabji Use of an Automated Mouse Touchscreen Platform for Quantification of Cognitive Deficits After Central Nervous System Injury . . . . . . . . . . . . . . . . . . . 279 Katherine M. Cotter, Grace L. Bancroft, Harley A. Haas, Raymon Shi, Andrew N. Clarkson, Matthew E. Croxall, Ann M. Stowe, Sanghee Yun, and Amelia J. Eisch

Contents

ix

22

Using Operant Reach Chambers to Assess Mouse Skilled Forelimb Use After Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Dene Betz, April M. Becker, Katherine M. Cotter, Andrew M. Sloan, Ann M. Stowe, and Mark P. Goldberg 23 The Finer Aspects of Grid-Walking and Cylinder Tests for Experimental Stroke Recovery Studies in Mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Nausheen Syeara, Sounak Bagchi, Abdullah Al Shoyaib, Serob T. Karamyan, Faisal F. Alamri, and Vardan T. Karamyan 24 Performing Enriched Environment Studies to Improve Functional Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355 Daniela Talhada and Karsten Ruscher

PART V

NEUROTHERAPEUTICS AND FUNCTIONAL RECOVERY

25

Clinically Applicable Experimental Design and Considerations for Stroke Recovery Preclinical Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vardan T. Karamyan 26 Hydrogels and Nanoscaffolds for Long-Term Intraparenchymal Therapeutic Delivery After Stroke. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mozammel H. Bhuiyan, Josh Houlton, and Andrew N. Clarkson 27 Reverse Translation to Develop Post-stroke Therapeutic Interventions during Mechanical Thrombectomy: Lessons from the BACTRAC Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benton Maglinger, Jacqueline A. Frank, Justin F. Fraser, and Keith R. Pennypacker 28 Methods to Study Drug Uptake at the Blood-Brain Barrier Following Experimental Ischemic Stroke: In Vitro and In Vivo Approaches . . . . . . . . . . . . . Robert D. Betterton, Erica I. Williams, Kelsy L. Nilles, Thomas P. Davis, and Patrick T. Ronaldson 29 Gene Silencing in the Brain with siRNA to Promote Long-Term Post-Stroke Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bharath Chelluboina and Raghu Vemuganti

PART VI 30

369

379

391

403

419

MODELS OF COMORBIDITIES

Diabetic Rodent Models for Chronic Stroke Studies . . . . . . . . . . . . . . . . . . . . . . . . 429 Lea Julie Dalco and Kunjan R. Dave 31 Use of Conventional Cigarette Smoking and E-Cigarette Vaping for Experimental Stroke Studies in Mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 Salvatore Mancuso, Aditya Bhalerao, and Luca Cucullo 32 Middle Cerebral Artery Occlusion in Aged Animal Model . . . . . . . . . . . . . . . . . . . 453 Grant W. Goodman, Justin N. Nguyen, Frank W. Blixt, Michael E. Maniskas, Louise D. McCullough, and Anjali Chauhan

x

33

34

Contents

Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models of Diabetes: Importance of Pre-op and Post-op Care, Severity of Hyperglycemia, and Sex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 Weiguo Li, Yasir Abdul, and Adviye Ergul The DOCA-Salt Model of Hypertension for Studies of Cerebrovascular Function, Stroke, and Brain Health . . . . . . . . . . . . . . . . . . . . . . 481 T. Michael De Silva and Frank M. Faraci

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

489

Contributors THOMAS J. ABBRUSCATO • Department of Pharmaceutical Sciences, Center for Blood-Brain Barrier Research, Jerry. H. Hodge School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA YASIR ABDUL • Department of Pathology & Laboratory Medicine, Medical University of South Carolina, Charleston, SC, USA ABDULLAH AL SHOYAIB • Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA FAISAL F. ALAMRI • College of Sciences and Health Profession, King Saud bin Abdulaziz University for Health Sciences, Jeddah, Saudi Arabia; King Abdullah International Medical Research Center, Jeddah, Saudi Arabia ANNA LETIZIA ALLEGRA MASCARO • Neuroscience Institute, National Research Council, Pisa, Italy; European Laboratory for Non-Linear Spectroscopy, Sesto Fiorentino, Italy PETER ANDROVIC • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany SOUNAK BAGCHI • Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA GRACE L. BANCROFT • University of Pennsylvania, Philadelphia, PA, USA ADAM Q. BAUER • Department of Radiology, Washington University in St. Louis, St. Louis, MO, USA; Imaging Science PhD Program, Washington University in St. Louis, St. Louis, MO, USA; Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, USA APRIL M. BECKER • Department of Behavior Analysis, University of North Texas, Denton, TX, USA SIMON BESSON-GIRARD • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany ROBERT D. BETTERTON • Department of Pharmacology, College of Medicine, University of Arizona, Tucson, AZ, USA DENE BETZ • Department of Neurology, UT Health Sciences Center San Antonio, San Antonio, TX, USA ADITYA BHALERAO • Department of Biological and Biomedical Sciences, Oakland University, Rochester, MI, USA MOZAMMEL H. BHUIYAN • Department of Anatomy, Brain Health Research Centre and Brain Research New Zealand, University of Otago, Dunedin, New Zealand; Centre for Bioengineering and Nanomedicine, Faculty of Dentistry, University of Otago, Dunedin, New Zealand IFECHUKWUDE JOACHIM BIOSE • Clinical Neuroscience Research Center, Department of Neurosurgery, Tulane University School of Medicine, New Orleans, LA, USA GREGORY JAYE BIX • Clinical Neuroscience Research Center, Department of Neurosurgery, Tulane University School of Medicine, New Orleans, LA, USA ELENA BLANCO-SUA´REZ • Department of Neuroscience, Thomas Jefferson University, Philadelphia, PA, USA

xi

xii

Contributors

FRANK W. BLIXT • Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA RYAN M. BOWEN • Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, USA; Department of Neurology, Washington University in St. Louis, St. Louis, MO, USA SAMUEL P. BRIDGES • Department of Neurology, David Geffen School of Medicine at University of California – Los Angeles, Los Angeles, CA, USA DAIMEN R. S. BRITSCH • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA ANDREW J. BRUMM • Department of Neurology, David Geffen School of Medicine at University of California – Los Angeles, Los Angeles, CA, USA BUKET BULUT • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany S. THOMAS CARMICHAEL • Department of Neurology, David Geffen School of Medicine at University of California – Los Angeles, Los Angeles, CA, USA RUDY CHANG • Department of Biomedical and Pharmaceutical Sciences, School of Pharmacy, Chapman University, Irvine, CA, USA ANJALI CHAUHAN • Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA BHARATH CHELLUBOINA • Department of Neurological Surgery, University of Wisconsin, Madison, WI, USA ANDREW N. CLARKSON • Department of Anatomy, Brain Health Research Centre and Brain Research New Zealand, University of Otago, Dunedin, New Zealand EMILIA CONTI • Neuroscience Institute, National Research Council, Pisa, Italy; European Laboratory for Non-Linear Spectroscopy, Sesto Fiorentino, Italy KATHERINE M. COTTER • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA MATTHEW E. CROXALL • Lafayette Instrument Company, Lafayette, IN, USA LUCA CUCULLO • Department of Foundation Medical Studies, Oakland University, William Beaumont School of Medicine, Rochester, MI, USA LEA JULIE DALCO • Peritz Scheinberg Cerebral Vascular Disease Research Laboratories, Department of Neurology and Neuroscience Program, University of Miami Miller School of Medicine, Miami, FL, USA KUNJAN R. DAVE • Peritz Scheinberg Cerebral Vascular Disease Research Laboratories, Department of Neurology and Neuroscience Program, University of Miami Miller School of Medicine, Miami, FL, USA T. MICHAEL DE SILVA • Department of Microbiology, Anatomy, Physiology and Pharmacology, Centre for Cardiovascular Biology and Disease Research, School of Agriculture, Biomedicine and Environment, La Trobe University, Bundoora, VIC, Australia KARIENN A. DE SOUZA • Women’s Health in Neuroscience Program, Department of Neuroscience and Experimental Therapeutics, Texas A & M University College of Medicine, Bryan, TX, USA RICK M. DIJKHUIZEN • Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht & Utrecht University, Utrecht, The Netherlands

Contributors

xiii

ANDREA DOAN • Department of Neurology, University of Texas Health Science Center at Houston, Houston, TX, USA ANDREW K. DUNN • Department of Biomedical Engineering, University of Texas at Austin, Austin, TX, USA AMELIA J. EISCH • Department of Anesthesiology and Critical Care Medicine, Children’s Hospital of Philadelphia Research Institute, Philadelphia, PA, USA; Department of Neuroscience, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA ADVIYE ERGUL • Department of Pathology & Laboratory Medicine, Medical University of South Carolina, Charleston, SC, USA FRANK M. FARACI • Department of Internal Medicine, and Department of Neuroscience and Pharmacology, Francois M. Abboud Cardiovascular Center, University of Iowa Carver College of Medicine, Iowa City, IA, USA JACQUELINE A. FRANK • Department of Neurology, University of Kentucky, Lexington, KY, USA; Center for Advanced Translational Stroke Science, University of Kentucky, Lexington, KY, USA JUSTIN F. FRASER • Departments of Neurosurgery and Neuroscience, University of Kentucky, Lexington, KY, USA; Department of Neurology, University of Kentucky, Lexington, KY, USA; Center for Advanced Translational Stroke Science, University of Kentucky, Lexington, KY, USA; Department of Neurosurgery, University of Kentucky, Lexington, KY, USA; Department of Radiology, University of Kentucky, Lexington, KY, USA; Department of Neuroscience, University of Kentucky, Lexington, KY, USA AMY J. GLEICHMAN • Department of Neurology, David Geffen School of Medicine at University of California – Los Angeles, Los Angeles, CA, USA OZGUN GOKCE • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany; Munich Cluster for Systems Neurology (SyNergy), Munich, Germany MARK P. GOLDBERG • Department of Neurology, UT Health Sciences Center San Antonio, San Antonio, TX, USA GRANT W. GOODMAN • Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA HARLEY A. HAAS • University of Pennsylvania, Philadelphia, PA, USA SUNG-HA HONG • Department of Neurosurgery, University of Texas Health Science Center at Houston, Houston, TX, USA MICHELLE HOOK • Women’s Health in Neuroscience Program, Department of Neuroscience and Experimental Therapeutics, Texas A & M University College of Medicine, Bryan, TX, USA JOSH HOULTON • Department of Anatomy, Brain Health Research Centre and Brain Research New Zealand, University of Otago, Dunedin, New Zealand ANMOL JARANG • Department of Radiology, Washington University in St. Louis, St. Louis, MO, USA HAO JI • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany JIE-MIN JIA • Westlake Laboratory of Life Sciences and Biomedicine, School of Life Sciences, Westlake University, Hangzhou, China YUXIAO JIN • Westlake Laboratory of Life Sciences and Biomedicine, School of Life Sciences, Westlake University, Hangzhou, China

xiv

Contributors

MARY T. JOY • Department of Neurology, David Geffen School of Medicine at University of California – Los Angeles, Los Angeles, CA, USA SEROB T. KARAMYAN • Department of Pharmacology, Faculty of Pharmacy, Yerevan State Medical University, Yerevan, Armenia VARDAN T. KARAMYAN • Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA; Department of Foundational Medical Studies, Oakland University William Beaumont School of Medicine, Rochester, MI, USA JIN-MOO LEE • Department of Radiology, Washington University in St. Louis, St. Louis, MO, USA; Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, USA; Department of Neurology, Washington University in St. Louis, St. Louis, MO, USA WEIGUO LI • Department of Pathology & Laboratory Medicine, Medical University of South Carolina, Charleston, SC, USA LU LIU • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany JENNY LUTSHUMBA • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA BENTON MAGLINGER • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA MARY K. MALONE • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA SALVATORE MANCUSO • Department of Biological and Biomedical Sciences, Oakland University, Rochester, MI, USA MICHAEL E. MANISKAS • Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA MIGUEL ALEJANDRO MARIN • Department of Neurology, David Geffen School of Medicine at University of California – Los Angeles, Los Angeles, CA, USA SEAN P. MARRELLI • Department of Neurology, University of Texas Health Science Center at Houston, Houston, TX, USA LOUISE D. MCCULLOUGH • Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA JUSTIN N. NGUYEN • Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA KELSY L. NILLES • Department of Pharmacology, College of Medicine, University of Arizona, Tucson, AZ, USA SAEIDEH NOZOHOURI • Department of Pharmaceutical Sciences, Center for Blood-Brain Barrier Research, Jerry. H. Hodge School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA VERONICA OBREGON-PERKO • FlowJo, BD Life Sciences–Biosciences, Ashland, OR, USA JONAH A. PADAWER-CURRY • Department of Radiology, Washington University in St. Louis, St. Louis, MO, USA; Imaging Science PhD Program, Washington University in St. Louis, St. Louis, MO, USA FRANCESCO SAVERIO PAVONE • European Laboratory for Non-Linear Spectroscopy, Sesto Fiorentino, Italy; Department of Physics and Astronomy, University of Florence, Sesto Fiorentino, Italy; National Institute of Optics, National Research Council, Sesto Fiorentino, Italy

Contributors

xv

KEITH R. PENNYPACKER • Department of Neurology, University of Kentucky, Lexington, KY, USA; Center for Advanced Translational Stroke Science, University of Kentucky, Lexington, KY, USA JILL M. ROBERTS • Departments of Neurosurgery and Neuroscience, University of Kentucky, Lexington, KY, USA PATRICK T. RONALDSON • Department of Pharmacology, College of Medicine, University of Arizona, Tucson, AZ, USA KARSTEN RUSCHER • Laboratory for Experimental Brain Research, Wallenberg Neuroscience Center, Lund University, Lund, Sweden ADAM SANTORELLI • Department of Biomedical Engineering, University of Texas at Austin, Austin, TX, USA RAYMON SHI • University of Pennsylvania, Philadelphia, PA, USA ANDREW M. SLOAN • Vulintus, LLC, Lafayette, CO, USA FARIDA SOHRABJI • Women’s Health in Neuroscience Program, Department of Neuroscience and Experimental Therapeutics, Texas A & M University College of Medicine, Bryan, TX, USA ANN M. STOWE • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA COLIN T. SULLENDER • Department of Biomedical Engineering, University of Texas at Austin, Austin, TX, USA RACHITA K. SUMBRIA • Department of Biomedical and Pharmaceutical Sciences, School of Pharmacy, Chapman University, Irvine, CA, USA; Department of Neurology, University of California, Irvine, CA, USA NAUSHEEN SYEARA • Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA DANIELA TALHADA • Laboratory for Experimental Brain Research, Wallenberg Neuroscience Center, Lund University, Lund, Sweden VANESSA O. TORRES • Denali Therapeutics, South San Francisco, CA, USA JADWIGA TURCHAN-CHOLEWO • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA THOMAS A. UJAS • Department of Neurology, Department of Neuroscience, The University of Kentucky, Lexington, KY, USA ANNETTE VAN DER TOORN • Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht & Utrecht University, Utrecht, The Netherlands RAGHU VEMUGANTI • Department of Neurological Surgery, University of Wisconsin, Madison, WI, USA; William S. Middleton Veterans Administration Hospital, Madison, WI, USA XIAODAN WANG • Department of Radiology, Washington University in St. Louis, St. Louis, MO, USA; Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, USA YIJING WANG • Systems Neuroscience Laboratory, Institute for Stroke and Dementia Research (ISD), Klinikum der Universit€ a t Mu¨nchen, Munich, Germany VERA H. WIELENGA • Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht & Utrecht University, Utrecht, The Netherlands ERICA I. WILLIAMS • Department of Pharmacology, College of Medicine, University of Arizona, Tucson, AZ, USA

xvi

Contributors

SANGHEE YUN • Department of Anesthesiology and Critical Care Medicine, Children’s Hospital of Philadelphia Research Institute, Philadelphia, PA, USA; Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA YONG ZHANG • Department of Pharmaceutical Sciences, Center for Blood-Brain Barrier Research, Jerry. H. Hodge School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX, USA

Part I Stroke Models and Surgical Interventions

Chapter 1 Rodent Stroke Models to Study Functional Recovery and Neural Repair Daimen R. S. Britsch, Nausheen Syeara, Ann M. Stowe, and Vardan T. Karamyan Abstract Rodent ischemic stroke models are essential research tools for studying this highly prevalent disease and represent a critical element in the translational pipeline for development of new therapies. The majority of ischemic stroke models have been developed to study the acute phase of the disease and neuroprotective strategies, but a subset of models is better suited for studying stroke recovery. Each model therefore has characteristics that lend itself to certain types of investigations and outcome measures, and it is important to consider both explicit and implicit details when designing experiments that utilize each model. The following chapter briefly summarizes the known aspects of the main rodent stroke models with emphasis on their clinical relevance and suitability for studying recovery and neural repair following stroke. Key words Chronic stroke, Functional impairment, Mouse, Rat, Post-stroke recovery

1

Introduction Stroke is the second leading cause of death worldwide, accounting for just over six million deaths in 2019 alone (https://www.who. int/news-room/fact-sheets/detail/the-top-10-causes-of-death). Despite this staggering statistic, most individuals survive their initial stroke beyond the first year, and approximately half of those survivors face the daunting prospect of lifelong dysfunction or disability [7, 39]. The majority of translational stroke trials looking at neuroprotection and neurorepair over the past several decades have, for the most part, failed to show clinical efficacy [11, 27]. To date the only FDA-approved treatments for acute ischemic stroke are thrombolysis and/or thrombectomy, limited to a small fraction of patients. Thus, the physical, emotional, and financial burden on stroke survivors and their families can be severe over the course of a lifetime.

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_1, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

3

4

Daimen R. S. Britsch et al.

Stroke research is a broad discipline with many models and methods employed. Preclinical animal models of stroke typically aim to replicate the pathophysiology of ischemic stroke since it comprises the majority of clinical cases. The importance of stroke research, particularly at chronic stages of recovery, is also evident. Ischemic stroke is commonly studied in murine models, but there are successful large animal models including non-human primate, canine, porcine, and ovine [19, 27]. This chapter will focus exclusively on murine models, discussing their strengths and limitations as they pertain to studying functional recovery after ischemic stroke. The intention is for researchers at all stages to gain a better understanding of contemporary ischemic stroke models and suitable applications.

2

Models of Ischemic Stroke for Chronic Outcomes Ischemic stroke cases in patients can vary in presentation and outcomes due to their heterogeneous nature. As such, no single rodent stroke model can completely portray the pathophysiology of human stroke. Each model has advantages and disadvantages, which depend on the scientific questions to be addressed. This discussion will cover the common, and some less common, models of ischemic stroke with regard to both mice and rats where applicable. Middle cerebral artery occlusion (MCAO) is the most prevalent model in ischemic stroke research because it is the most commonly affected vessel in human stroke [28, 45]. MCAO is an umbrella term covering a host of well-characterized subtypes to consider. Proximal MCAO is an occlusion at the origination of the MCA from the circle of Willis. The occlusion is accomplished by surgical insertion of a filament/suture through the common or external carotid artery, accessed via an incision on the ventral surface of the neck and carefully advanced until it reaches the origin of the MCA, and then secured in place. The filament/suture can be removed after a period of time to allow reperfusion and model transient ischemia or can be left in place to model permanent ischemia. Both transient and permanent proximal MCAO create a relatively large infarct and peri-infarct area across various brain regions like cortical and subcortical regions, substantia nigra, cervicomedullary junction, hippocampus, thalamus, and hypothalamus, which may result in deficits of heterogeneous functions like sensory, motor, autonomic, and cognition [28, 45]. Proximal MCAO as a model of ischemic stroke has several advantages. The surgical procedure is straightforward and has been adopted in many laboratories around the world, it allows modeling of transient and permanent ischemia, it does not require craniotomy, it has good lesion reproducibility, the resulting infarct

Rodent Stroke Models to Study Recovery and Neural Repair

5

and penumbra closely mimic the pathophysiology of human ischemic stroke, and the rapid reperfusion that occurs after removal of the occlusion closely mirrors outcomes of thrombectomy, bestowing clinical relevance. The model has been used to study ischemic cell death mechanisms, inflammatory pathways, blood–brain barrier injury, as well as majority of neuroprotective mechanisms [28, 45]. However, the MCAO model also suffers from a number of disadvantages including a requirement of advanced surgical skills and experience for consistent outcomes, the highly invasive nature of the procedure, significant mortality rate, induction of hyperthermia due to hypothalamic injury, and tendency to model more malignant infarctions with progressive brain swelling that mostly has lethal outcome in humans with large infarcts. The inclusion of laser Doppler or other imaging parameters to confirm a complete occlusion for the full time period (if transient) and reperfusion is invaluable for a priori inclusion/exclusion criteria and improves consistency between groups. Furthermore, threading the filament through the external carotid improves the rate of reperfusion, as the common carotid remains intact. This variation, however, requires even more advanced surgical skills and typically takes longer than the common carotid route. Distal MCAO (dMCAO) is an occlusion of the MCA at the level of the parietal cortex, past the lenticulostriate branches but before the first bifurcation of the MCA. This results in a smaller, more unilateral infarct and even more reduced peri-infarct region confined within the cortex. The surgical procedure generally requires a craniotomy and penetration of the dura to reach the target vessel. The occlusion itself can be accomplished by several approaches including electrocoagulation of the vessel, mechanical restriction via a wire, applied pressure, suture, or use of ferric chloride (FeCl3) [28, 45]. Electrocoagulation, either with or without subsequent severance of the vessel, is an approach used to model permanent ischemia. Mechanical restriction using microaneurysm clips, hooks, or ligatures can be used to mimic transient ischemia with reperfusion. These occlusions have a risk of damaging the vessel, and the generated infarct is somewhat more variable than the electrocoagulation method. FeCl3 can also be used to model transient or permanent ischemia based on the concentration used [47]. This approach is less invasive since dura is left intact; however it still requires a craniotomy. Importantly, another variation of dMCAO is simultaneous occlusion of the common carotid artery, unilaterally or bilaterally, in a temporary or permanent manner [23]. Pairing distal MCAO with hypoxia is yet another approach allowing to maximize cortical infraction and lower variability [13]. Advantages of the dMCAO include low mortality rate, lack of hypothalamic injury, and excellent lesion reproducibility, primarily confined to the cerebral cortex. However, the need of

6

Daimen R. S. Britsch et al.

craniotomy and disruption of the meninges coupled with limited peri-infarct region diminish the clinical relevance of this model. Post-MCAO functional recovery: In general, MCAO models of ischemic stroke can be applied to both mice and rats, although some recommend against using intraluminal MCAO in rodents due to the large infarcts and corresponding mortality rates. It is important to note here that the murine MCA does not supply blood to the primary motor cortex—unlike the human MCA—so the MCAO is not well suited to study motor deficits or motor recovery post-stroke [25, 47]. With regard to chronic outcomes and stroke recovery studies, the proximal MCAO model appears to be more suited for evaluation of cognitive, fear, mood, and sensory functions [14, 25, 34, 37]. Despite large infarction, motor function is usually spontaneously recovered within the first week(s) after stroke dependent on the strain of animal used and the duration of the transient occlusion [25]. Sensory and cognitive deficits are better modeled by the MCAO with regard to human stroke patients. Rodent studies have shown deficits in sensory function, cognition, and mood to be present up to 8–12 weeks after stroke [33, 46] while stroke-induced plasticity peaking 1–2 weeks poststroke [35]. In closing, it is notable that the first Stroke Recovery and Rehabilitation Roundtable (SRRR) [11] only moderately recommended the use of dMCAO model for stroke recovery studies of motor function in mice and rats and did not recommend the intraluminal suture MCAO model due to high morbidity and poor recovery. Photothrombosis-induced ischemia, aka photothrombotic (PT) model, is an irreversible, focal ischemia induced by peripheral administration of a photosensitive dye (most commonly Rose Bengal) and illumination of a preferred region of the cerebral cortex through the intact skull with cold light at a certain wavelength [4]. The irradiated dye produces reactive oxygen species that damage the endothelium initiating platelet activation and aggregation and eventually cortical ischemic lesion [51]. Advantages of the PT model include very low mortality rates, the least invasive nature of the technique compared to other stroke models, plus excellent lesion reproducibility due to stereotaxic targeting of the desired infarct location in most parts of the cerebral cortex. The main disadvantages of the model are occlusion of multiple blood vessels in the irradiated location, simultaneous induction of cytotoxic and vascular edema (instead of the former leading to the latter in human stroke), and presence of a much-limited penumbra and collateral blood flow. Notably, modifications of this technique allow targeting of individual brain vessels [48], induction of ischemia in subcortical regions [6], and implementation in awake/unanesthetized rodents [54]. Post-PT functional recovery: The most used version of this model is based on targeting of the primary motor cortex which

Rodent Stroke Models to Study Recovery and Neural Repair

7

leads to long-lasting (8–12 weeks) fine motor impairment [9] and has been fundamental for elucidation of brain repair mechanisms [1, 8, 9, 26], discovering new molecular pathways and potential therapeutic targets [24, 38], and testing of various therapeutic approaches [2, 3, 10, 32]. In fact, a recent study used a PT stroke to test the effects of short-chain fatty acids in drinking water on the long-term recovery of fine motor skills in mice following stroke, highlighting the use for testing neurotherapeutics [42]. While the PT model is rarely used for neuroprotection studies, based on SRRR recommendations, it is the first model of choice for mouse and the second model of choice for rat stroke recovery studies focusing on the motor function [11]. Endothelin-induced ischemia (ET-1) is a model of transient, focal ischemia induced by the application of endothelin-1 (ET-1), a potent vasoconstrictor peptide. ET-1 is applied to the brain surface or stereotaxically injected to desired region(s) of the brain (including subcortical areas) through a cranial window, and depending on the dose, it provides moderate control over the duration and severity of the ischemic event. The model was first developed in rats and deemed challenging in mice [17], but later modifications improved adaptation to mice [40, 49]. Advantages of the ET-1 model include low mortality rates and moderate invasiveness, the small infarct size, great precision and lesion reproducibility, and flexibility to produce infarction in most regions of the brain (though at the cost of a cranial window). However, like the PT model, infarction in this model is achieved by occlusion of multiple vessels with limited penumbra and edema formation, which diminishes clinical relevance. In addition, ET-1 is known to affect the function of neurons and astrocytesglial cells, and hence its effects go beyond vasoconstriction and may interfere with cell death and neural repair mechanisms in the post-stroke brain [43]. Post-ET-1 functional recovery: Because of these characteristics, the ET-1 stroke model has been used in both neuroprotection and neural repair studies; however, the technique targeting the primary motor cortex appears to be the most common version of this model used for studies focusing on stroke recovery. The latter produces sensorimotor impairment lasting for at least 2 weeks in mice [40, 49] and 4 weeks in rats [16, 44] and has been widely used for understanding the biology of neural repair [21, 22] and evaluation of experimental therapeutic approaches [5, 50]. Notably, based on SRRR recommendations, the ET-1 model is the first model of choice for stroke recovery studies of the motor function in rats, whereas it is not among the preferred mouse stroke models for such studies [11]. Embolic stroke models include approaches that allow induction of focal or multifocal (depending on the size and application route of the embolus), transient, or permanent ischemia in the brain and

8

Daimen R. S. Britsch et al.

allow targeting cortical and subcortical regions as well as some areas of the white matter. Thromboembolic models of stroke induce transient ischemia via injection of blood clots created from subject blood samples, but can also refer to injection of thrombin to initiate clotting in a vessel, usually the MCA [36]. The primary advantages of the thromboembolic models include that they most closely mimic human embolic stroke and are best suited to study spontaneous and recombinant tPA-induced thrombolysis. While similar to humans, there is a great deal of variability in infarct size and location, as well as spontaneous lysis of the clot (partial or complete) in thromboembolic models. These features may be disadvantageous for experimental studies as they affect both the extent and duration of the ischemic injury which may introduce wide variation in recovery of treated animals. Other embolic (non-thromboembolic) models of stroke induce permanent ischemia via administration of artificial clot-like particles. There are various materials that have proven successful in inducing embolic stroke in rodents [45], which depending on their particle size can be broadly broken down into two groups. Microsphere-based models use 20–50 μm particles that target the MCA or internal carotid artery (ICA). These create multifocal, heterogeneous infarcts throughout the brain that take up to 24 h to fully develop [31]. The macrosphere-based models use approximately 300–400 μm particles to target MCA or ICA and result in somewhat more localized infarcts that often resemble that of the proximal MCAO model [15]. Advantages and disadvantages of non-thromboembolic models are comparable to that of the thromboembolic models with an exception that spontaneous lysis of the embolus is absent and that spheres cause simultaneous embolization in multiple vessels with poor directability, leading to rapid formation of vasogenic edema [53]. It is important to note that in recent years, two new approaches have been developed to induce embolic stroke using magnetic nanoparticles (SIMPLE, stroke induced by magnetic nanoparticles) and magnetic nanoparticle-coated erythrocytes (SIMPLeR, stroke induced by magnetic nanoparticles coated red blood cells) [18]. In these models, magnetic particles are introduced into the systemic circulation, and a magnetic field is applied through the intact skull (in case of pups) or CCA to cause embolization. The main advantages of these models are directability of the particles to induce focal cerebral ischemia with high precision and the ability to have controllable reperfusion [18]. Post-embolic functional recovery: It is notable that none of the embolic stroke models is recommended by the SRRR for stroke recovery studies [11], although thromboembolic and magnetic nanoparticle-based models may be well suited to study the molecular and cellular mechanisms of spontaneous recovery and will

Rodent Stroke Models to Study Recovery and Neural Repair

9

hopefully be used in future studies to establish relevance and efficacy. NOS inhibitor-induced ischemia is a model of focal ischemia induced by stereotaxic injection of L-NIO, a non-selective inhibitor of all nitric oxide synthases (NOS), which results in potent vasoconstriction. The procedure requires craniotomy and is sometimes accompanied by additional occlusion of a larger blood vessel [52]. L-NIO models are traditionally associated with white matter stroke (WMS [30]) but have been successful in targeting areas of the striatum and cortex as well [52]. WMS models are increasingly clinically relevant as they are tied to “silent stroke” phenotypes, depression, and vascular dementia [29]. Depending on the targeted area and the measures used, L-NIO shows persistent motor deficits through 5–12 weeks post-stroke and CNS plasticity and regeneration occurring through 14-week timepoints [12, 41]. Post-L-NIO functional recovery: Advantages of the L-NIO stroke model include moderate clinical relevance, high precision and reproducibility of the lesions, and a straightforward and relatively quick surgical procedure. Disadvantages include simultaneous occlusion of multiple vessels (similar to the ET-1 model), potential inhibition of NOS in non-endothelial cells and its relatively short history of use in the stroke research field. Perhaps because of these features, the L-NIO stroke model is not among top rodent models recommended by the SSSR for motor recovery studies [11].

3

Concluding Remarks Rodent models of ischemic stroke are indispensable research tools for studying various facets of stroke, including long-term functional recovery. By recognizing that ischemic stroke is a complex and heterogeneous disorder in patients, it is easy to accept that no one animal model can fully mimic all aspects of the human disease. Being cognizant about advantages and disadvantages of the individual models is critically important for designing experiments and addressing clinically relevant questions in rodent stroke models. While the majority of these models have been used to study pathophysiology of acute ischemic stroke and neuroprotective strategies, a few models are more suited for stroke recovery studies. Importantly, suitability of models for stroke recovery is largely based on experimental studies focusing on motor function, whereas less is known about which models are best suited for recovery studies of other functions (e.g., cognition, vision, etc.) [20]. Use of proper models will remain critical for the continued success of the field to study the biology of brain repair, discover new molecular pathways and potential therapeutic targets, and test new therapeutic approaches.

10

Daimen R. S. Britsch et al.

References 1. Al Shoyaib A, Alamri FF, Biggers A et al (2021) Delayed exercise-induced upregulation of Angiogenic proteins and recovery of motor function after photothrombotic stroke in mice. Neuroscience 461:57–71 2. Al Shoyaib A, Alamri FF, Syeara N et al (2021) The effect of histone deacetylase inhibitors panobinostat or entinostat on motor recovery in mice after ischemic stroke. NeuroMolecular Med 23:471–484 3. Alamri FF, Al Shoyaib A, Syeara N et al (2021) Delayed atomoxetine or fluoxetine treatment coupled with limited voluntary running promotes motor recovery in mice after ischemic stroke. Neural Regen Res 16:1244–1251 4. Alamri FF, Karamyan ST, Karamyan VT (2023) A low budget, photothrombotic rodent stroke model. In: Karamyan VT, Stowe AM (eds) Neural repair. Springer Nature 5. Bell JA, Wolke ML, Ortez RC et al (2015) Training intensity affects motor rehabilitation efficacy following unilateral ischemic insult of the sensorimotor cortex in C57BL/6 mice. Neurorehabil Neural Repair 29:590–598 6. Blanco-Sua´rez E (2023) Photothrombotic model to create an infarct in the hippocampus. In: Karamyan VT, Stowe AM (eds) Neural repair. Springer Nature 7. Bronnum-Hansen H, Davidsen M, Thorvaldsen P et al (2001) Long-term survival and causes of death after stroke. Stroke 32:2131– 2136 8. Clarkson AN, Huang BS, Macisaac SE et al (2010) Reducing excessive GABA-mediated tonic inhibition promotes functional recovery after stroke. Nature 468:305–309 9. Clarkson AN, Lopez-Valdes HE, Overman JJ et al (2013) Multimodal examination of structural and functional remapping in the mouse photothrombotic stroke model. J Cereb Blood Flow Metab 33:716–723 10. Cook DJ, Nguyen C, Chun HN et al (2017) Hydrogel-delivered brain-derived neurotrophic factor promotes tissue repair and recovery after stroke. J Cereb Blood Flow Metab 37: 1030–1045 11. Corbett D, Carmichael ST, Murphy TH et al (2017) Enhancing the alignment of the preclinical and clinical stroke recovery research pipeline: consensus-based core recommendations from the stroke recovery and rehabilitation roundtable translational working group. Int J Stroke 12:462–471 12. Dingman AL, Rodgers KM, Dietz RM et al (2019) Oligodendrocyte progenitor cell

proliferation and fate after white matter stroke in juvenile and adult mice. Dev Neurosci 40:1– 16 13. Doyle KP, Fathali N, Siddiqui MR et al (2012) Distal hypoxic stroke: a new mouse model of stroke with high throughput, low variability and a quantifiable functional deficit. J Neurosci Methods 207:31–40 14. Eldahshan W, Sayed MA, Awad ME et al (2021) Stimulation of angiotensin II receptor 2 preserves cognitive function and is associated with an enhanced cerebral vascular density after stroke. Vasc Pharmacol 141:106904 15. Gerriets T, Li F, Silva MD et al (2003) The macrosphere model: evaluation of a new stroke model for permanent middle cerebral artery occlusion in rats. J Neurosci Methods 122: 201–211 16. Gilmour G, Iversen SD, O’neill MF et al (2004) The effects of intracortical endothelin1 injections on skilled forelimb use: implications for modelling recovery of function after stroke. Behav Brain Res 150:171–183 17. Horie N, Maag AL, Hamilton SA et al (2008) Mouse model of focal cerebral ischemia using endothelin-1. J Neurosci Methods 173:286– 290 18. Jia J-M, Jin Y (2023) Modeling distal middle cerebral artery occlusion in neonatal rodents with magnetic nanoparticles or magnetized red blood cells. In: Karamyan VT, Stowe AM (eds) Neural repair. Springer Nature 19. Kaiser EE, West FD (2020) Large animal ischemic stroke models: replicating human stroke pathophysiology. Neural Regen Res 15:1377– 1387 20. Karamyan VT (2023) Clinically applicable experimental design and considerations for stroke recovery preclinical studies. In: Karamyan VT, Stowe AM (eds) Neural repair. Springer Nature 21. Kerr AL, Wolke ML, Bell JA et al (2013) Poststroke protection from maladaptive effects of learning with the non-paretic forelimb by bimanual home cage experience in C57BL/6 mice. Behav Brain Res 252:180–187 22. Kim SY, Hsu JE, Husbands LC et al (2018) Coordinated plasticity of synapses and astrocytes underlies practice-driven functional vicariation in Peri-infarct motor cortex. J Neurosci 38:93–107 23. Lee B, Clarke D, Al Ahmad A et al (2011) Perlecan domain V is neuroprotective and proangiogenic following ischemic stroke in rodents. J Clin Invest 121:3005–3023

Rodent Stroke Models to Study Recovery and Neural Repair 24. Li S, Nie EH, Yin Y et al (2015) GDF10 is a signal for axonal sprouting and functional recovery after stroke. Nat Neurosci 18:1737– 1745 25. Liu F, Schafer DP, Mccullough LD (2009) TTC, fluoro-Jade B and NeuN staining confirm evolving phases of infarction induced by middle cerebral artery occlusion. J Neurosci Methods 179:1–8 26. Liu Z, Li Y, Cui Y et al (2014) Beneficial effects of gfap/vimentin reactive astrocytes for axonal remodeling and motor behavioral recovery in mice after stroke. Glia 62:2022–2033 27. Lourbopoulos A, Mourouzis I, Xinaris C et al (2021) Translational block in stroke: a constructive and out-of-the-box reappraisal. Front Neurosci 15:652403 28. Macrae IM (2011) Preclinical stroke research– advantages and disadvantages of the most common rodent models of focal ischaemia. Br J Pharmacol 164:1062–1078 29. Marin MA, Carmichael ST (2018) Stroke in CNS white matter: models and mechanisms. Neurosci Lett 684:193–199 30. Marin MA, Gleichman AJ, Brumm AJ et al (2023) Subcortical white matter stroke in the mouse: inducing injury and tracking cellular proliferation. In: Karamyan VT, Stowe AM (eds) Neural repair. Springer Nature 31. Mayzel-Oreg O, Omae T, Kazemi M et al (2004) Microsphere-induced embolic stroke: an MRI study. Magn Reson Med 51:1232– 1238 32. Minnerup J, Kim JB, Schmidt A et al (2011) Effects of neural progenitor cells on sensorimotor recovery and endogenous repair mechanisms after photothrombotic stroke. Stroke 42: 1757–1763 33. Modo M, Stroemer RP, Tang E et al (2000) Neurological sequelae and long-term behavioural assessment of rats with transient middle cerebral artery occlusion. J Neurosci Methods 104:99–109 34. Montgomery KM, Hook M, Sohrabji F (2023) Assessing depression and cognitive impairment following stroke and neurotrauma. In: Karamyan VT, Stowe AM (eds) Neural repair. Springer Nature 35. Ohab JJ, Fleming S, Blesch A et al (2006) A neurovascular niche for neurogenesis after stroke. J Neurosci 26:13007–13016 36. Orset C, Macrez R, Young AR et al (2007) Mouse model of in situ thromboembolic stroke and reperfusion. Stroke 38:2771–2778 37. Ortega SB, Torres VO, Latchney SE et al (2020) B cells migrate into remote brain areas and support neurogenesis and functional

11

recovery after focal stroke in mice. Proc Natl Acad Sci U S A 117:4983–4993 38. Overman JJ, Clarkson AN, Wanner IB et al (2012) A role for ephrin-A5 in axonal sprouting, recovery, and activity-dependent plasticity after stroke. Proc Natl Acad Sci U S A 109: E2230–E2239 39. Romain G, Mariet AS, Jooste V et al (2020) Long-term relative survival after stroke: the dijon stroke registry. Neuroepidemiology 54: 498–505 40. Roome RB, Bartlett RF, Jeffers M et al (2014) A reproducible Endothelin-1 model of forelimb motor cortex stroke in the mouse. J Neurosci Methods 233:34–44 41. Rosenzweig S, Carmichael ST (2013) Age-dependent exacerbation of white matter stroke outcomes: a role for oxidative damage and inflammatory mediators. Stroke 44:2579– 2586 42. Sadler R, Cramer JV, Heindl S et al (2020) Short-chain fatty acids improve poststroke recovery via immunological mechanisms. J Neurosci 40:1162–1173 43. Schinelli S (2006) Pharmacology and physiopathology of the brain endothelin system: an overview. Curr Med Chem 13:627–638 44. Soleman S, Yip P, Leasure JL et al (2010) Sustained sensorimotor impairments after endothelin-1 induced focal cerebral ischemia (stroke) in aged rats. Exp Neurol 222:13–24 45. Sommer CJ (2017) Ischemic stroke: experimental models and reality. Acta Neuropathol 133:245–261 46. Stroemer P, Patel S, Hope A et al (2009) The neural stem cell line CTX0E03 promotes behavioral recovery and endogenous neurogenesis after experimental stroke in a dosedependent fashion. Neurorehabil Neural Repair 23:895–909 47. Syeara N, Alamri FF, Jayaraman S et al (2020) Motor deficit in the mouse ferric chlorideinduced distal middle cerebral artery occlusion model of stroke. Behav Brain Res 380:112418 48. Talley Watts L, Zheng W, Garling RJ et al (2015) Rose Bengal Photothrombosis by confocal optical imaging in vivo: a model of single vessel stroke. J Vis Exp 100:e52794 49. Tennant KA, Jones TA (2009) Sensorimotor behavioral effects of endothelin-1 induced small cortical infarcts in C57BL/6 mice. J Neurosci Methods 181:18–26 50. Tennant KA, Kerr AL, Adkins DL et al (2015) Age-dependent reorganization of peri-infarct premotor cortex with task-specific rehabilitative training in mice. Neurorehabil Neural Repair 29:193–202

12

Daimen R. S. Britsch et al.

51. Uzdensky AB (2018) Photothrombotic stroke as a model of ischemic stroke. Transl Stroke Res 9:437–451 52. Van Slooten AR, Sun Y, Clarkson AN et al (2015) L-NIO as a novel mechanism for inducing focal cerebral ischemia in the adult rat brain. J Neurosci Methods 245:44–57

53. Wilmes FJ, Garcia JH, Conger KA et al (1983) Mechanisms of blood-brain barrier breakdown after microembolization of the cat’s brain. J Neuropathol Exp Neurol 42:421–438 54. Yu CL, Zhou H, Chai AP et al (2015) Wholescale neurobehavioral assessments of photothrombotic ischemia in freely moving mice. J Neurosci Methods 239:100–107

Chapter 2 Subcortical White Matter Stroke in the Mouse: Inducing Injury and Tracking Cellular Proliferation Miguel Alejandro Marin, Amy J. Gleichman, Andrew J. Brumm, and S. Thomas Carmichael Abstract Here, we describe a method for inducing subcortical white matter stroke in mice, as well as tracking cellular proliferation through drinking water administration of EdU and ex vivo labeling. Key words White matter stroke, Vasoconstriction, N5-(1-iminoethyl)-L-ornithine HCL, 2’-Deoxy5-ethynyluridine, EdU, Cell proliferation

1

Introduction Subcortical white matter stroke (WMS) is a progressive disorder which is characterized by the formation of ischemic lesions along white matter tracts in the central nervous system (CNS). In WMS the initial infarction is often asymptomatic; however, as the disease progresses, patients begin to experience severe motor and cognitive decline [1, 5–7]. Despite the prevalence of WMS—by the age of 80, most individuals will have WMS lesions—little is known about the etiology of this disease [2]. Thus, modeling WMS in mice is critical toward the development of viable therapies. WMS induces high levels of proliferation in multiple cell types in both the ischemic core and the surrounding spared tissue. Tracking this proliferation in an inexpensive and minimally invasive way allows investigators to probe the dynamics of the cellular response to WMS. This can be attained through drinking water administration of the thymidine analogue 2′-deoxy-5-ethynyluridine (EdU), which will incorporate into the DNA of dividing cells. EdU can then be detected through a copper-catalyzed reaction between EdU and a fluorescent azide [8]. In this article, we will outline a protocol for inducing WMS in adult mice [3, 4] in combination

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_2, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

13

14

Miguel Alejandro Marin et al.

with tracking cellular proliferation through administration of EdU in drinking water followed by ex vivo labeling.

2

Materials

2.1 EdU Administration

1. EdU (2’-Deoxy-5-ethynyluridine) (see Note 1). 2. Sulfamethoxazole-trimethoprim antibiotic suspension (per 5 mL: 200 mg sulfamethoxazole, 40 mg trimethoprim), cherry flavored (TMS) (see Note 2). 3. Light-protected water bottles (see Note 3).

2.2 Induction of WMS

1. Vasoconstrictor (L-NIO-HCL, N5-(1-iminoethyl)-L-ornithine): Dissolve L-NIO-HCL in sterile 0.9% saline at a concentration of 27 mg/mL. Aliquot and store at -20 °C (see Note 4). 2. 0.5 mm glass capillary tubes. 3. Vertical pipette puller. 4. Gel-loading tips. 5. Delivery of L-NIO-HCL: Picospritzer II – Intracellular Microinjection system (Parker Instrumentation) set at a pressure of 20 psi (see Note 5). 6. Dissection microscope fitted with a calibrated eye piece reticle (see Note 6). 7. Small animal stereotaxic instrument fitted with glass pipette holder attached to injection arm, digital stereotaxic control panel, and mouse gas anesthesia head holder/nose cone. 8. Infrared warming pad with controller and rectal temperature probe. 9. Isoflurane vaporizer anesthesia system with induction chamber. 10. Pneumatic dental drill with 1/8 Carbide Burr. 11. Dry sterilizer. 12. Autoclaved surgical tools: fine forceps and fine scissors. 13. 0.9% sterile saline. 14. Sterile cotton tipped applicators. 15. Ophthalmic ointment. 16. Electric razor. 17. Tissue adhesive. 18. 70% ethanol and Betadine surgical scrub. 19. Heating pad for postoperative care.

White Matter Stroke: Induction and Cell Proliferation

2.3

EdU Labeling

15

All solutions are in ultrapure water unless otherwise noted. 1. Phosphate-buffered saline. 2. Triton X-100, 0.5%. 3. Tris-buffered saline, pH 7.6, 132 mM. 4. CuSO4, 100 mM. 5. Sulfo-fluorophore azides, 2 μM, in DMSO (see Note 7). 6. Sodium ascorbate, 500 mM (see Note 8).

3

Methods

3.1 Administration of EdU

1. Prepare EdU stock solution: 5 mg/mL in drinking water. Shake at 37 °C until fully dissolved while keeping protected from light (see Note 9). 2. Prepare light-protected EdU water bottles: 200 μg/mL EdU, 1:100 TMS (see Note 10). 3. Replace untreated water in cage with EdU water bottle (see Note 11). 4. Change EdU water every 2 days for duration of labeling period (see Note 12).

3.2 Induction of WMS

1. Prior to surgery, dry sterilize autoclaved surgical tools. 2. Pull glass capillary tubes to a final distal diameter of 15-25 μm (see Note 13). 3. Backload pipette with 2-5 μL of L-NIO-HCL using suitable autoclaved gel loading tips (see Note 14); keep on ice (see Note 15). 4. Adjust surgical arm to 45° posterior to anterior (see Note 16). 5. Induce anesthesia by placing mouse in induction chamber with isoflurane vaporizer set at 5 L/min until animal fails to respond to toe pinch. 6. Move mouse to the stereotaxic instrument and maintain anesthetic state with isoflurane vaporizer, 1.5–2 L/min. Occasionally check depth of anesthesia with toe pinch. 7. Insert temperature probe and maintain animal’s body heat at 37 °C. 8. Apply ophthalmic ointment to both eyes with sterile cotton tipped applicator. 9. Prepare surgical field by first shaving the scalp with an electric razor and then sterilize scalp with three alternating sets of 70% ethanol and Betadine surgical scrubs. 10. Make 1.5 cm midline scalp incision to expose skull.

16

Miguel Alejandro Marin et al.

Fig. 1 (a) Location of craniotomy for WMS injections; measurements relative to bregma. (b) Schematic of 45° injection angle to minimize damage to motor cortex. (Created with BioRender.com) Table 1 L-NIO-HCL injection coordinates Injection

Anterior/posterior (A/P)

Medial/lateral (M/L)

Dorsal/ventral (D/V)

1

0.75

0.96

2.25

2

1.00

0.96

2.20

3

1.25

0.96

2.15

11. Dry skull with sterile cotton tip applicators. 12. Mark bregma with a permanent marker. 13. Drill a ~ 1.5 mm elliptical craniotomy posterior to bregma and left of midline (Fig. 1a) (see Note 17). 14. Maintain moisture of surgical field with occasional application of sterile physiological saline. 15. Install pulled glass pipette with L-NIO-HCL to the injection arm of the stereotaxic apparatus (Fig. 1b) (see Note 18). 16. Three 200 nL injections of L-NIO-HCL will be administered using coordinates in Table 1 (see Note 19). 17. Zero A/P and M/L coordinates by aligning the tip of the glass pipette with bregma. 18. Move injection arm to A/P corresponding to injection site 1.

and

M/L

coordinates

19. Lower glass pipette to the surface of the cortex and zero dorsal/ventral coordinates. 20. Slowly lower the glass pipette to D/V coordinates corresponding to injection site 1.

White Matter Stroke: Induction and Cell Proliferation

17

21. Inject 200 nL of L-NIO-HCL with a pulse duration of 10–30 ms set on the Picospritzer II. 22. Upon completion of the injection, maintain the glass pipette at the injection site for 5 min to prevent reflux of L-NIO-HCL up the glass pipette. 23. Slowly withdraw pipette and repeat steps 17–22 for injection sites 2 and 3. 24. Upon final injection, remove the pipette and seal the scalp incision with tissue adhesive. 25. Move mouse to a home cage for postoperative care and heating (see Note 20). 3.3

Labeling of EdU

All incubations take place at room temperature, on a shaker. 1. Rinse free-floating brain sections from EdU-treated mice in PBS (see Note 21), 3 × 5 min. 2. Permeabilize the sections for 30 min in PBS + 0.5% Triton X-100 (see Note 22). 3. Prepare sodium ascorbate solution (500 mM, in water). 4. Rinse sections in PBS + 0.1% Triton X-100, 5 min. 5. Prepare EdU detection cocktail, 500 μL/animal. Add reagents in the order listed (final concentration): TBS (100 mM), CuSO4 (4 mM), sulfo-Cy azide (2 μM), sodium ascorbate (100 mM) (see Note 23); see Table 2 for example volumes. Protect from light. 6. Incubate sections in EdU detection cocktail, 30 min. Protect sections from light from this point forward. 7. Rinse sections in PBS, 3 × 5 min. 8. Proceed with blocking and antibody labeling (Fig. 2) (see Note 24).

Table 2 EdU cocktail volumes Total volume (μL)

500

1000

2000

3000

4000

5000

TBS, 132 mM, pH 7.6

378.75

757.5

1515

2272.5

3030

3787.5

120

160

200

CuSO4, 100 mM Sulfo-Cy azide, 2 mM Sodium ascorbate, 500 mM

20 1.25 100

40 2.5 200

80 5 400

7.5 600

10 800

12.5 1000

18

Miguel Alejandro Marin et al.

Fig. 2 Example of EdU labeling (sulfo-Cy5 azide, white) in WMS, 7 days after stroke, combined with a vascular stain (CD31, red). Scale bar: 200 μm

4

Notes 1. While there are multiple EdU suppliers, we have found Biosynth-Carbosynth to be the most cost-effective. 2. An antibiotic is required post-surgery; we have found that including cherry-flavored TMS in the EdU drinking water as opposed to less-sweet antibiotics promotes increased consumption by the mice and therefore more robust labeling of newborn cells. If antibiotic is not necessary or desired, sweeten the water with 5 mM saccharine to encourage consumption. 3. Both EdU and TMS are light-sensitive and should be protected from light in amber- or red-colored water bottles; alternatively, water bottles can be wrapped in aluminum foil, although mice will tend to scratch away the bottom layer of foil. 4. Repeated freeze-thaws decrease the efficacy of L-NIO-HCL. Aliquot volumes should be based on need for individual experiments. Dispose of remaining thawed L-NIO-HCL following surgery. 5. Our protocol is optimized for the Picospritzer II which has since been replaced by a new model Picospritzer III from Parker Instrumentation (part number 052-0500-900). 6. The volume of L-NIO-HCL injection is determined with the eye piece reticle by aligning the meniscus of the L-NIO-HCL in the micropipette with the lines of the reticle. The meniscus should be displaced by 0.200 mm3 (1 mm length in a 0.5 mm diameter pipette, yielding 200 nL), for each injection coordinate.

White Matter Stroke: Induction and Cell Proliferation

19

7. There are multiple suppliers of fluorophore-conjugated azides; we have found the Lumiprobe sulfo-Cy3 and Cy5 azides to be highly effective. 8. Sodium ascorbate oxidizes quickly and should be made fresh for each use. Other stock solutions for EdU labeling can be made in advance and stored at 4 °C. 9. Use drinking water for preparation of EdU stock solution as well as final water bottles. EdU stock solutions can be prepared in advance and frozen at -20 °C; however, this typically causes EdU to fall out of solution. In this case, re-dissolve by shaking at 37 °C for 30–60 min while keeping protected from light. 10. We have found that 125 mL of water is sufficient for one cage with 4–5 mice for 2 days while minimizing EdU waste. Therefore, a bottle will include 5 mL of 5 mg/mL EdU, 1.25 mL TMS, and 118.75 mL drinking water. 11. For robust labeling of all stroke-induced proliferation, replace untreated water with EdU water 12–24 h before stroke. 12. The duration of the EdU administration reflects the experimental design; in this protocol, we euthanized animals at 7 days post-stroke and maintained EdU water until euthanasia. 13. To minimize damage to overlying cortex, a long-tapered micropipette is used to deliver L-NIO-HCL into subcortical white matter. 14. Eppendorf femtotip microloader tips work well to backload pulled glass pipettes, although other very long gel-loading tips will also work. Allow pipettes to rest for several minutes after loading to allow liquid to wick to the end of the pipette tip. 15. Given the temperature-sensitive nature of L-NIO, both the single-use aliquot of L-NIO and the filled glass pipettes should be kept on ice until the point of delivery. Be careful to avoid touching the tip of the pipette to hard surfaces; it is very fragile. 16. L-NIO is targeted to subcortical white matter ventral to the motor cortex. To ensure minimal damage to motor cortex, the glass pipette is delivered into the brain at a 45° angle. 17. Given the angle of injection, centering the craniotomy around (A/P -0.75, M/L 0.96) yields a window that accommodates all three injections. 18. Before loading onto Picospritzer, gently flick the pipette if necessary to dislodge remaining bubble(s). 19. These coordinates were optimized for 3–4-month-old male C57Bl6J mice. The use of other strains of mice or older/ younger and female mice may require adjusting injection coordinates.

20

Miguel Alejandro Marin et al.

20. To minimize overheating, place half the home cage on at heating pad set to medium. This allows mice to freely navigate between hot and cold regions of the cage. Maintain the cage on the heating pad until all animals are awake and mobile. 21. Because EdU labeling is dependent on small molecules that diffuse easily, it is compatible with tissue sections of any thickness, including whole brain. 22. If combining EdU labeling with antibody staining that requires antigen retrieval, it is possible to include the antigen retrieval step prior to permeabilization. 23. EdU labeling solution should be used within 30 min of preparation. 24. EdU labeling can also be performed after antibody staining. References 1. Baezner H, Blahak C, Poggesi A et al (2008) Association of gait and balance disorders with age-related white matter changes: the LADIS study. Neurology 70:935–942. https://doi. org/10.1212/01.wnl.0000305959.46197.e6 2. de Leeuw F-E, de Groot JC, Achten, et al. (2001) Prevalence of cerebral white matter lesions in elderly people: a population based magnetic resonance imaging study. The rotterdam scan study. J Neurology Neurosurg Psychiatry 70:9–14. https://doi.org/10.1136/jnnp. 70.1.9 3. Nunez S, Doroudchi MM, Gleichman AJ et al (2016) A versatile murine model of subcortical white matter stroke for the study of axonal degeneration and white matter neurobiology. J Vis Exp 109:e53404. https://doi.org/10. 3791/53404 4. Rosenzweig S, Carmichael ST (2013) Age-dependent exacerbation of white matter stroke outcomes. Stroke 44:2579–2586. https://doi.org/10.1161/strokeaha.113. 001796

5. Soumare´ A, Elbaz A, Zhu Y, Maillard P et al (2009) White matter lesions volume and motor performances in the elderly. Ann Neurol 65: 706–715. https://doi.org/10.1002/ana. 21674 6. Vermeer SE, Longstreth WT, Koudstaal PJ (2007) Silent brain infarcts: a systematic review. Lancet Neurol 6:611–619. https://doi.org/10. 1016/s1474-4422(07)70170-9 7. Whitman GT, Tang T, Lin A, Baloh RW (2001) A prospective study of cerebral white matter abnormalities in older people with gait dysfunction. Neurology 57:990–994. https://doi.org/ 10.1212/wnl.57.6.990 8. Zeng C, Pan F, Jones LA, Lim MM et al (2010) Evaluation of 5-ethynyl-2′-deoxyuridine staining as a sensitive and reliable method for studying cell proliferation in the adult nervous system. Brain Res 1319:21–32. https://doi.org/10. 1016/j.brainres.2009.12.092

Chapter 3 A Low-Budget Photothrombotic Rodent Stroke Model Faisal F. Alamri, Serob T. Karamyan, and Vardan T. Karamyan Abstract A number of animal stroke models have been developed and used over the years to study the pathological mechanisms of this disorder and develop new therapies. Among them, the photothrombotic model of ischemic stroke has been central in various studies focusing on understanding of the basic biology of neural repair, identification and validation of key molecular targets involved in post-stroke recovery, and preclinical testing of various therapeutic approaches. To facilitate uniformity among various experimental groups using this expert-recommended mouse model of choice for stroke recovery studies, in this chapter we describe in detail a low-budget technique to induce photothrombosis in the mouse primary motor cortex. Additionally, we provide tips for conducting this procedure in other cerebral cortical regions of the mouse brain and in rats. Key words Photothrombosis, Mouse permanent occlusion stroke model, Stroke recovery, Neurorecovery, Preclinical stroke study

1

Introduction Animal stroke models are an important part of the discovery research focusing on the identification and understanding of pathological mechanisms and development of new therapies [6]. Among them, rodent ischemic stroke models are based on transient or permanent occlusion of a vessel(s), supplying blood to a region of the brain, through various techniques ranging from noninvasive to very invasive procedures [19]. These models differ in their severity and localization of stroke, the associated functional impairment, some aspects of the injury mechanisms, and biology of repair and broadly represent the human pathology, which itself is highly heterogeneous. One model that is expert-recommended for stroke recovery studies in rodents and has received considerable attention in recent years is the photothrombotic stroke model [10]. The concept of photothrombosis was originally proposed by Rosenblum and El-Sabban in 1977 [18] and further developed by Watson and

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_3, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

21

22

Faisal F. Alamri et al.

colleagues in 1985 [24]. This model of ischemic stroke produces brain injury by affecting a group of pial and/or intraparenchymal blood vessels at a predetermined region. The procedure consists of three main steps: (1) preparation of the animal and brain area for stroke (usually surface of the skull); (2) systemic administration of an anionic xanthene dye (usually rose bengal), which does not cross the blood-brain barrier and remains in the bloodstream; and (3) photoexcitation of the dye in the selected brain region, to induce generation of reactive oxygen species which react and harm the microvascular endothelium leading to platelet aggregation and thrombosis, followed by interruption of the blood supply and brain infarction [21]. The model is noninvasive (no craniotomy or major incisions) and technically simple and has low mortality rate (< 10%), and the resulting stroke is accompanied by little or no reperfusion and long-lasting functional deficits [4, 8]. The main disadvantages of this model are the simultaneous occlusion of multiple vessels (instead of one occurring in humans), concurrent induction of cytotoxic and vasogenic edema (in humans cytotoxic edema develops into vasogenic), and the confined penumbra region which leads to a small border of peri-infarct tissue (i.e., the hub of neural repair mechanisms, bigger in other stroke models) [6]. Recognizing that there is no “gold standard” model that is most relevant to human stroke, the photothrombotic stroke is the model of choice for recovery studies in rodents focusing on biology of neuronal repair and development of new therapies [10]. In this chapter, we describe a low-budget technique to induce photothrombosis in the mouse primary motor cortex and provide tips for conducting the procedure in other cerebral cortical regions and in rats. Notably, another chapter of this book details a technique to induce photothrombosis in hippocampus and other subcortical regions [5].

2

Materials

2.1

Animals

2.2

Equipment

Three to 18-month-old, male and female CD-1 or C57BL/6 mice (see Note 1). 1. Stereomicroscope. 2. Isoflurane vaporizer. 3. Small animal surgical platform. 4. Body temperature probe with a feedback-regulated heating pad and lamp. 5. Fiber-optic illuminator equipped with a reflector halogen bulb for cold light, silicone-sheathed fiber-optic cable, and 90°,

A Low-Budget Photothrombotic Rodent Stroke Model

23

stainless steel fiber-optic pipet probe (with 2.16 mm fiber diameter). 6. Three-joint magnetic base holder. 7. Glass bead sterilizer. 8. Electric clipper. 9. Small animal scale. 10. Small animal recovery chamber. 11. Timer. 2.3 Surgical Instruments

1. Dumont tweezers (curved thin tips, two pairs). 2. Sterile surgical blades. 3. Olsen-Hegar needle holder/scissors. 4. Microdissecting forceps (serrated, full curve).

2.4 Reagents and Supplies

1. Isoflurane, USP. 2. Cotton swabs (autoclaved). 3. Rose bengal (95% dye content). 4. Saline for injection, USP. 5. Chlorhexidine scrub. 6. Ethyl alcohol (70%). 7. Hydrogen peroxide solution, USP (3%). 8. Ophthalmic lubricant. 9. First aid antibiotic pain relieving ointment. 10. Tape (1 cm width). 11. Sterile drape. 12. Sterile gauze. 13. Sterile nylon sutures (with reverse cutting needle, singlearmed). 14. Sterile single use syringe (with 31 gauge, 5 mm needle).

3

Methods

3.1 Aseptic Preparation of the Animal

1. The mouse is anesthetized with isoflurane, and hair is carefully removed from top of the head using an electric clipper (see Notes 2 and 3). 2. The mouse is placed on the surgical platform (experimental setup shown in Fig. 1) to continue anesthesia, and the ophthalmic lubricant is applied to the eyes (see Note 4). 3. The mouse is positioned and secured with a tape to keep the head level.

24

Faisal F. Alamri et al.

Fig. 1 Illustration of the experimental setup for the main equipment used for the photothrombotic model of stroke in rodents

4. The bare skin of the head is thoroughly cleansed with chlorhexidine scrub followed by rinsing with 70% ethyl alcohol using autoclaved cotton swabs. This step is repeated twice (see Note 5). 5. A sterile drape is used to cover the surgical site and animal (see Note 6). 3.2

Surgery

1. After confirming a constant respiratory rate and lack of response to noxious stimuli (e.g., toe pinch), an incision is made through the skin (~1.5 cm) in the head midline using a sterile surgical blade. 2. Using autoclaved Dumont tweezers, the edges of the incision are pulled in opposite directions to expose the skull (1.0–1.5 cm opening, most in the hemisphere where infarction should occur). 3. The connective tissue covering the skull is removed using the tweezers, followed by cleaning the surface with an autoclaved cotton swab wetted in hydrogen peroxide (in a sterile microtube; see Note 7). 4. Using a sterile, single-use syringe rose bengal solution (8 mg/ mL in saline for injection, prepared in a sterile microtube) is administered intraperitoneally at 10 mL/kg volume. Timer is started for 5 min. 5. The surface of the skull is cleaned and dried using autoclaved cotton swabs, and Bregma 0 is located under the microscope.

A Low-Budget Photothrombotic Rodent Stroke Model

25

6. The fiber-optic pipet probe (thoroughly cleaned with 70% ethyl alcohol) is placed 1 mm lateral to midline and Bregma 0 in the hemisphere where the infarction is desired (see Notes 8 and 9). 7. Five minutes after rose bengal administration, the cold light is exposed to the brain through the intact scull at the highest intensity of the lamp (see Note 10). 8. After 15 min illumination (see Note 11), the probe is removed, and the incision is closed by sutures in a simple interrupted pattern (using the sterile nylon suture with a needle and the autoclaved microdissecting forceps and needle holder/scissors; see Note 12). 9. The wound is gently cleaned by chlorhexidine and dried, and a thin layer of the first aid antibiotic pain relieving ointment is applied using an autoclaved cotton swab. 10. Sham controls are subjected to the same steps except for light illumination. 3.3 Post-Surgical Care

1. The mouse is placed in the small animal recovery chamber (~37 °C, with no bedding) for 1.5–2 h, followed by transfer to a new cage with access to fresh water and food (see Notes 13 and 14). 2. During the first several hours after the surgery, especially while at the recovery chamber, the mouse is monitored for uneventful recovery from anesthesia (usually takes 75% in the MCA territory is expected. 18. Covering animal’s body with a light cloth (even a folded Kimwipes) can help prevent heat loss and reduce the amount of heat needed from the heating pad and thus variation in body temperature. 19. If a bilateral drop in CBF occurs after filament removal, hemorrhagic stroke may have resulted from the filament perforating the ICA in the circle of Willis region. In our lab, bilateral drop in CBF with occluder removal is among the exclusion criteria for our stroke studies. 20. Recording during the euthanasia period provides the LSCI values for true zero CBF. 21. We found that the aged mouse is very susceptible to death depending on the duration of CCA clamping during MCAo procedure. Minimization of the time that the CCA is clamped can improve the survival of aged mice during experimental stroke procedure. 22. We find that minimizing the isoflurane concentration (providing adequate but not excess anesthesia) improves survival during the procedure.

Acknowledgments This work was supported by American Heart Association Postdoctoral Fellowship (POST, SH) and Brain Aneurysm Foundation funded Research Grant (SH), NIH R56NS120709 (SPM), NIH NS094280 (SPM), and NIH NS096186 (SPM). References 1. Campbell BCV, De Silva DA, Macleod MR et al (2019) Ischaemic stroke. Nat Rev Dis Primers 5:70 2. Virani SS, Alonso A, Aparicio HJ et al (2021) Heart disease and stroke statistics-2021 update: a report from the American Heart Association. Circulation 143:e254–e743 3. El Amki M, Wegener S (2017) Improving cerebral blood flow after arterial recanalization: a novel therapeutic strategy in stroke. Int J Mol Sci 18:2669

4. Fasipe TA, Hong SH, Da Q et al (2018) Extracellular Vimentin/VWF (von Willebrand factor) interaction contributes to VWF string formation and stroke pathology. Stroke 49: 2536–2540 5. Ayata C, Dunn AK, Gursoy OY et al (2004) Laser speckle flowmetry for the study of cerebrovascular physiology in normal and ischemic mouse cortex. J Cereb Blood Flow Metab 24: 744–755

96

Sung-Ha Hong et al.

6. Boas DA, Dunn AK (2010) Laser speckle contrast imaging in biomedical optics. J Biomed Opt 15:011109 7. Dunn AK, Bolay H, Moskowitz MA et al (2001) Dynamic imaging of cerebral blood flow using laser speckle. J Cereb Blood Flow Metab 21:195–201 8. Ponticorvo A, Dunn AK (2010) How to build a Laser Speckle Contrast Imaging (LSCI) system to monitor blood flow. J Vis Exp 9. Richards LM, Kazmi SM, Davis JL et al (2013) Low-cost laser speckle contrast imaging of blood flow using a webcam. Biomed Opt Express 4:2269–2283 10. Belayev L, Alonso OF, Busto R et al (1996) Middle cerebral artery occlusion in the rat by intraluminal suture. Neurological and pathological evaluation of an improved model. Stroke 27:1616–1622. discussion 1623 11. Belayev L, Busto R, Zhao W et al (1999) Middle cerebral artery occlusion in the mouse by intraluminal suture coated with poly-L-lysine: neurological and histological validation. Brain Res 833:181–190

12. Chiang T, Messing RO, Chou WH (2011) Mouse model of middle cerebral artery occlusion. J Vis Exp 48:2761 13. Hong SH, Hong JH, Lahey MT et al (2021) A low-cost mouse cage warming system provides improved intra-ischemic and post-ischemic body temperature control - application for reducing variability in experimental stroke studies. J Neurosci Methods 360:109228 14. Longa EZ, Weinstein PR, Carlson S et al (1989) Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 20: 84–91 15. Roy-O’reilly M, Mccullough LD (2018) Age and sex are critical factors in ischemic stroke pathology. Endocrinology 159:3120–3131 16. Fisher M, Feuerstein G, Howells DW et al (2009) Update of the stroke therapy academic industry roundtable preclinical recommendations. Stroke 40:2244–2250 17. Kahle MP, Bix GJ (2012) Successfully climbing the "STAIRs": surmounting failed translation of experimental ischemic stroke treatments. Stroke Res Treat 2012:374098

Chapter 10 Multi-exposure Speckle Imaging for Quantitative Evaluation of Cortical Blood Flow Adam Santorelli, Colin T. Sullender, and Andrew K. Dunn Abstract Laser speckle contrast imaging (LSCI) is a label-free optical imaging technique that can quantify flow dynamics across an entire image. Multi-exposure speckle imaging (MESI) is an extension of LSCI that allows for reproducible and quantifiable measurements of flow. MESI has the potential to provide quantitative cerebral blood flow information in both preclinical and clinical applications; in fact, MESI can be extended to resolve the flow dynamics in any exposed tissue. A MESI system can be divided into three primary components: (i) the illumination optics, consisting of the optical source and a method of modulating and gating the illumination intensity; (ii) the collection optics, consisting of a high-speed camera that can be triggered and gated to match the pulsed illumination; and finally (iii) post-processing hardware and software to extract the flow information from the recorded raw intensity images. In the following protocol, we offer a guide to design, operate, and test a MESI system. Key words Cerebral blood flow, Optical imaging, Laser speckle contrast imaging, Multi-exposure speckle imaging, Neurosurgery, Intraoperative imaging

1

Introduction Monitoring cerebral blood flow (CBF) is crucial to the success of neurosurgical interventions and neuroscience applications [1– 3]. Currently, imaging techniques such as indocyanine green angiography (ICGA) and digital subtraction angiography (DSA) are used to monitor CBF intraoperatively; however, these techniques suffer from requiring contrast agents, a disruption to a surgical procedure if used intraoperatively, and require radiation exposure in the case of DSA [4, 5]. Laser speckle contrast imaging (LSCI) is a full-field, label-free, optical imaging technique that can provide continuous maps of blood flow; thus, it can be used in an extensive number of applications across neuroscience, dermatology, dentistry, and ophthalmology [6, 7]. Additionally, LSCI has been shown to be a powerful research tool for preclinical stroke studies, primarily with rodent

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_10, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

97

98

Adam Santorelli et al.

models, helping to inform neurosurgical interventions and broadening neuroscience advances. These studies have covered research topics such as monitoring CBF during ischemic stroke induction [8], chronic monitoring of the vasculature remodeling after stroke induction [9–11], investigating the effects of isoflurane-based vasodilation [12], and studying the impact of obesity on CBF [13]. There have also been advancements in research to integrate speckle contrast measurements within a surgical microscope for intraoperative use [1, 14]. Laser speckle is the random interference pattern produced when coherent light scatters from a rough surface or inhomogeneous medium. When a sample contains scattering particles in motion, the speckle will vary temporally, creating a speckle pattern. When integrated over the exposure time of the camera, this speckle pattern will encode information about the underlying particle motion. Thus, a measure of flow can be obtained by quantifying the spatial or temporal statistics of the speckle pattern. This speckle pattern requires calculating the local speckle contrast, K, defined as σ K¼ , ð1Þ hI i where σ is the standard deviation and hIi is the average intensity over a sliding window of pixels (e.g., 7  7) of the imaged data. These speckle contrast values are only indicative of the amount of motion in a sample and are not proportional to particle speed or volumetric flow. To determine a measure of flow in the sample, it is necessary to relate the speckle contrast value to the correlation time constant (τc) of the speckle autocorrelation function, which is inversely related to the speed of the scatters, although the exact nature of this relationship varies with the portion of the vasculature sampled by the detected light [15]. That is, the inverse correlation time (ICT), defined as 1/τc, is related to the flow of scatters (in this application the flow of red blood cells). The relationship between speckle contrast and ICT can be defined as follows:  2x   1 þ 2x e , ð2Þ K ðT , τc Þ2 ¼ β 2x 2 where x ¼ T/τc, β is a constant instrumentation factor, T is the exposure time of the camera, and τc is the correlation time constant. However, LSCI suffers from several drawbacks, namely, a high dependency upon instrumentation, it does not account for the effect of static scatterers that are present in actual tissue, and it does not account for noise [16]. Due to these limitations, LSCI is typically limited to measurements of the relative changes in blood flow within a single subject during a single experiment, and it is challenging to provide consistent, reproducible, quantitative measures of flow.

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow

99

Multi-exposure speckle imaging (MESI) was developed as an extension of LSCI to address these issues. The primary goal of MESI is to provide a more robust estimate of the correlation time constant (τc) that is repeatable and quantifiable. MESI requires collecting LSCI images over a wide range of camera exposure times to properly sample the underlying flow dynamics. In this protocol, and in many of our lab’s experiments, 15 MESI images spanning three decades of exposure times (50 μs–80 ms) are chosen [10, 12, 15, 16]; however, it is not mandatory, and in fact, some published work from our lab has investigated using a select fewer number of exposure times [17]. While the complexity of the MESI hardware has been challenging to integrate into a clinical setting [1], MESI has been used for a wide array of in vivo mouse studies. MESI is based upon a more rigorous dynamic light scattering model that accounts for static scattering events and non-ideal conditions, producing the MESI equation [12]   2x  1 þ 2x e x  1 þ x 2 2e K ðT , τc Þ ¼ β ρ þ 4ρð1  ρÞ þν 2x 2 x2 ð3Þ where x ¼ T/τc, ρ is the ratio of intensity undergoing dynamic scattering events to the total intensity, ν is the noise arising from experimental noise and due to simplifying assumptions made in the model, β is a constant instrumentation factor, T is the camera exposure time, and τc is the correlation time constant. By collecting this sequence of images at various values of T, it is possible to fully resolve the shape of K2 and properly solve for τc. Solving for τc from the experimental speckle contrast values requires implementing a fitting procedure to solve for β, ρ, ν, and τc in Eq. 3. This process is done on a pixel-by-pixel basis. Thus, MESI can create flow maps over the entire region of interest (ROI) that is being imaged. A typical MESI imaging system involves three major components: illumination optics, image acquisition and the collection optics, and post-processing tools. The illumination optics include both optical and electrical components. A stable laser source, that can be directed toward the sample under test, and a fast, triggerable camera are required. Furthermore, MESI requires precise control over both the camera exposure time and the laser intensity. Varying the exposure time alone would result in shot noise overwhelming the speckle signal at longer exposure times because of differences in intensity. However, by modulating the illuminating laser light, the average intensity of the images can be held constant. Optical devices such as acousto-optic modulators (AOMs) have historically been used for precise control over laser intensity and duration, but manual laser diode modulation via the laser controller is also possible. A multi-function I/O device that can communicate with a

100

Adam Santorelli et al.

computer, the AOM, and the camera is required for these control electronics. The procedure in this protocol can be summarized as follows: first, set up and test the optical and electrical components, respectively. Then ensure that the sample under test is properly illuminated by verifying the laser alignment. A MESI calibration is required prior to the MESI acquisition phase. Finally, once the raw images have been recorded, the data must be processed by first computing the speckle contrast images and finally creating the ICT maps that provide quantitative flow information.

2

Materials

2.1 Illumination Optics

1. Illumination source: A wavelength-stabilized laser diode (ideally in the 600–850 nm wavelength range), capable of generating a decent amount of power to ensure sufficient light at the shortest exposure time, is the ideal choice of illumination for MESI (e.g., the 300 mW LD785-SEV300 from Thorlabs, Inc.). If a bare diode laser is used, items 2–4 are required, but if a complete laser diode system is used, items 2–4 are not required. 2. Temperature-controlled housing: Mount the laser diode in a laser diode mount with an integrated temperature controller. 3. Temperature controller: For stability and repeatability, use a temperature controller (TEC) to set and maintain the laser diode operating temperature. 4. Laser diode controller: For the stability and repeatability of the illumination light, drive the laser diode with a constant current, controlled by a laser diode controller (LDC) that interacts with the laser diode mounting. 5. Isolation: Use a free space optical isolator to minimize back reflections that interfere with single frequency performance. 6. Fiber patch cable (optional): Couple the emitted light into a fiber-optic cable to correct any beam shape irregularities (see Note 1). 7. Light modulation: Use a free space AOM (e.g., AOMO 3100125, Gooch and Housego) and an RF driver (97-03307-34, Gooch and Housego) to modulate the intensity of the collimated laser light, ensuring the successful implementation of MESI. Use an iris to isolate the first-order diffracted light from the free space AOM. 8. Control electronics: Use a multifunction I/O device (USB-6363, National Instruments Corp.), referred to as the data acquisition hardware (DAQ), to produce the camera exposure trigger signals and AOM modulation voltages. The drivers

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow

101

and associated libraries (NI-DAQmx library) for the DAQ will need to be installed. 9. Mirrors: Use steering mirrors as needed throughout the system to direct light, ensure maximum power output, and proper sample illumination (see Note 2). 10. Power: Use a DC power supply to power the RF driver. 2.2 Image Acquisition

1. Collection optics: Install a pair of camera lenses to image the scattered light and direct the light to the camera sensor. A common configuration is to couple two camera lenses (e.g., 50 mm Nikon DSLR lenses) face to face and set their foci to infinity in a macro-lens arrangement. 2. Filter: Place a bandpass filter, centered around your laser operating wavelength (785 nm in our system), along the collection optics path to remove any background noise (white light) from the collected light. 3. Camera: Use a high-speed CMOS camera that can be triggered from an external source (such as the acA1920-155um Basler AG). The camera can be monochromatic as only pixel intensity is required. 4. Image collection software: Install camera software that allows for image collection, for example, the pylon Camera Software Suite for Basler cameras (see Note 3). 5. Light intensity calibration: Determine the control the voltages supplied to the AOM so that light intensity levels are balanced at each exposure time (see Note 4). The voltage for each exposure duration should be recorded. 6. Camera exposure time gating: Define the chosen exposure times for MESI. These exposure times will be needed by the DAQ, the image collection software, and the host computer to properly gate the camera and modulate the AOM to the appropriate exposure times (see Note 4 for details about the gating process, and see Note 5 about the choice of camera exposure times).

2.3

Post-processing

1. Computing power: Use a workstation (desktop or laptop) with ample storage and strong computing capabilities. Speckle imaging records large amounts of data and requires fast processing to quickly produce flow images. 2. Processing software: While many signal and image processing and computing platforms exist and can be used, the general approach to solve the MESI equation is via nonlinear least squares curve fitting.

102

3 3.1

Adam Santorelli et al.

Methods For ease of testing and design, it is highly suggested to first build the system on an optical table or optical breadboard and test with controllable microfluidic channels or flow phantoms, prior to attempting in vivo imaging. Additionally, follow all laser safety guidelines that apply for the specific class of the laser chosen (proper PPE and signage).

System Setup

1. Use Fig. 1 as a reference schematic for the layout and connectivity of the optical components.

3.1.1 Optical Components

2. Connect the temperature controller (TEC) and the laser diode controller (LDC) to the mounted laser diode. Operate the LDC at the suggested operating current from the laser diode data sheet. The TEC should be set and active to stabilize the diode at a consistent, pre-determined, temperature (such as 25  C). 3. Align the light output from the laser diode with the isolator. 4. Mount all steering mirrors with posts on the optical table. Additionally, it is suggested to mount the fiber coupling lenses on translational stages. Use the steering mirrors and the translational aspheric lens to couple the laser light into the single mode fiber (see Note 3 for details on maximizing power into the AOM).

M

M

M

M

CMOS

RF Driver

AOM

Isolator

L1

L4

L2

M

F1 L3

785nm

0th

1st

M

M A

Fig. 1 A schematic detailing the connections for the optical components. The 785 nm laser diode is connected to the isolator, and mirrors (marked M in the figure) and a set of aspheric lenses (L1 and L2) are used to couple the light into a fiber-optic patch cable to obtain a Gaussian beam and to re-collimate the output from the single mode fiber. After passing through the AOM, the first-order diffraction is isolated using an iris (A) and is then relayed to obliquely illuminate the sample under test. Two lenses, L3 and L4, are used to collect the light, with a bandpass filter, F1, placed along the collection optics path. This filtered collected light is imaged onto the CMOS camera and contains the raw pixel intensity image data

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow

Computer

103

+24V DC

NI USB-6363 Multifunction I/O Device

Analog Modulation

/Dev1/ao0 USB 3.0

RF Driver

/Dev1/ao1 /Dev1/ai7

AOM

Camera Trigger

Breakout Box

CMOS Exposure Active

+5V DC

USB 3.0

Fig. 2 A schematic of the connections required for the electronics. The DAQ serves as a communication hub between the workstation (lab desktop or clinical laptop) and the RF driver and the camera, generating the necessary voltages and trigger signals to modulate the laser light and the camera exposure times. The CMOS camera is connected to the workstation and transfers the image data back to the image collection software. A breakout box is used to supply a reference voltage to the DAQ

5. Use the second set of steering mirrors to direct the re-collimated light output from the end of the fiber into the AOM. 6. Use a third set of steering mirrors to direct light into the iris and isolate the first-order diffraction from the AOM. 7. Relay this first-order diffraction light to obliquely illuminate the sample under test. 3.1.2 Electrical Components

1. Use Fig. 2 to provide an overview of the connections between the various electrical components. 2. Connect the DC power supply to the RF driver. The model suggested in this chapter requires 24 V DC. If using a different AOM/RF driver, please see the manufacturer data sheet. 3. Connect the DAQ to the workstation via USB 3.0. 4. Connect the first analog output of the DAQ (/Dev1/ao0) to the RF driver of the AOM. 5. Connect the second analog output of the DAQ (/Dev1/ao1) to the camera trigger input using a 6-pin I/O plug. 6. Connect the camera to the workstation using a high-speed USB cable (such as USB 3.0).

3.2

Laser Alignment

Prior to any image acquisition, it is suggested to verify that the field of view (FOV) is properly illuminated. This section gives instructions on how to properly illuminate the imaging FOV. 1. Place the test object under the microscope (printed text is a good test sample).

104

Adam Santorelli et al.

2. Use the NI Device Monitor (in system tray of the connected workstation) to launch Test Panels for the DAQ. 3. Select the Analog Output tab, set the Channel Name to Dev1/ ao0 (the channel that is connected to the RF driver), adjust the Output Value to ~0.05 V, and click Update to apply the change. This voltage can be adjusted as necessary between 0 and 5 V to increase or decrease the modulation of the laser light with the AOM. 4. Launch the image collection software to start a live view of the camera. Ensure that the laser light power into the AOM has been previously optimized (see Note 3) and that the illumination optics are correctly functioning. 5. Adjust the height of the microscope until the object is in focus. 6. Center the laser beam in the camera field of view by adjusting the knobs on the final steering mirror (the mirror that is used to illuminate the sample). 7. Close the image collection software and Test Panels programs. 3.3

MESI Calibration

The objective of the MESI calibration stage is to ensure that there is an equal amount of light at each exposure time. This is accomplished by controlling the amplitude of the analog voltage signal applied to the RF driver of the AOM and modulating this voltage based on the pixel intensity at each exposure time. There are numerous methods to solve this iterative process. The method implemented by our lab (see Note 4 for complete details of the process) is described below. 1. Supply an initial guess voltage to the AOM. This guess should fall within the range of values that the DAQ can supply. 2. Record the average light intensity (pixel values) from this initial guess at each MESI exposure time (N total values, where N is the number of exposure times). 3. Iteratively, adjust the voltages at each exposure time to bring the intensities in line with each other. It is helpful to define an acceptable error tolerance. Additionally, it is useful to limit the maximum number of saturated pixels. 4. Save the voltages for each exposure time. These calibration voltages will be applied for any measurements that will occur with this sample. 5. Close the software.

3.4

MESI Acquisition

1. Open the image collection software to begin the MESI acquisition. 2. Set the number of MESI sequences to acquire and the desired output filename.

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow

105

3. Apply the calibration voltages to each exposure time. 4. Record a sequence of MESI images. The image collection software should automatically cycle through the total number of sequences defined and save the speckle images. 5. Close the running software. 3.5

Shutdown

1. Set the laser current to 0 mA, deactivate the laser, and turn off the laser diode controller. 2. Deactivate and turn off the temperature controller. 3. Turn off the AOM power supply and reference DC voltage to the DAQ. 4. Remove the USB and I/O plug from the camera (prevents possible burn out).

3.6

Data Processing

3.6.1 Calculating Speckle Contrast

Once the raw intensity images are obtained, it is necessary to extract the flow dynamics from the images. This section covers the primary steps in obtaining the ICT maps that are indicative of flow; additionally Fig. 3 provides a visualization of each step. 1. Obtain speckle contrast: Convert the collected raw speckle data into the speckle contrast image (using Eq. 1). This can be done by iteratively solving Eq. 1 at every pixel in the collected images. 2. Set the speckle contrast window size, typically N ¼ 7 (see Note 6). 3. Solve Eq. 1 and generate the speckle contrast images.

3.6.2 Imaging and Extracting Quantitative Flow Information

1. Use the computed speckle contrast, at each exposure time, to fit the measured data to the MESI equation (Eq. 3), and solve for the four unknown fitting parameters, including τc for a specific region of interest (see Note 7). 2. If a pixel-by-pixel map of τc is desired, then the fitting should be repeated at all pixels across the entire field of view from the measured speckle contrast vs exposure time. 3. Create ICT images to provide quantitative flow images that define the flow dynamics of the imaging scenario.

4

Notes 1. The LD785-SEV300 laser diode has a non-Gaussian beam profile. Optical correction of the abnormal beam shape is applied by coupling the light into an optical fiber. The laser was coupled into a single mode patch cable (125 μm cladding, P3-780A-FC-2, Thorlabs, Inc.) to obtain a Gaussian beam.

106

Adam Santorelli et al.

A

50µs

75µs

100µs

250µs

500µs

750µs

1ms

2.5ms

5ms

7.5ms

10ms

25ms

40ms

50ms

80ms

0.0

0.1

0.2

0.3

0.4

Speckle Contrast

B

Speckle Visibility Curve

C 0.14

Vessel Parenchyma

Speckle Variance, K2

0.12 0.10 0.08 0.06

c

= 3.6 x 10-4 s

0.04 c

= 4.7 x 10-5 s

0.02 0.00 10-5 3.0

3.4

3.8

4.2

4.6

10-4

10-3

10-2

10-1

Exposure Time (s)

log10(1/ c)

Fig. 3 (a) Example of the speckle contrast images of a mouse cortex (a cranial window is necessary so that the laser light reaches the brain surface) at all 15 MESI exposure times, (b) the resulting ICT image (log scale) for this set of MESI images after solving for τc at each pixel of the image, (c) plots of the MESI fits in each ROI from (b) with the calculated τc in each region. Note that the ICT values are much higher in the vessel, where blood flow is significantly higher, as opposed to the parenchyma, which is a region of comparatively low blood flow. (Image reproduced from [19])

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow

107

Fig. 4 Example of the light path to optimize to maximize power delivered to the system. The smaller red arrows indicate the path the illumination light follows, whereas the white arrows indicate the steering mirrors and the translational stages that can be used as control parameters to optimize the light delivered to the sample under test

The fiber output was re-collimated (F230APC-780, Thorlabs, Inc.) before being relayed to the free space AOM. 2. Maximize the power output of the system by ensuring that light is properly coupled into the fiber-optic cable. The following steps and Fig. 4 help this process: (a) Place a post-mounted optical power sensor in the holder downstream of the collimated fiber output. Center the laser beam on the sensor surface. Only use power sensors designed for your laser diode choice. Set the power meter to a wavelength of 785 nm (or if not using a 785 nm, match the wavelength of your optical source). (b) Using the power meter as reference, adjust the steering mirrors and translation stage for the fiber coupling lens to maximize the fiber output power. If necessary, the translation stage for the laser collimating optic can be adjusted as well. Warning: Be careful working around the free-space

108

Adam Santorelli et al.

laser beam. Consider reducing the current on the laser diode controller during this step. (c) Once the fiber output power has been maximized, remove the optical power sensor. (d) See Fig. 4 for an example of a system layout and the various steering mirrors and the translational stage that can be modified to optimize power. 3. It is highly suggested to create an image collection software that will allow for interaction with the suggested Basler cameras, the workstation, and the DAQ to acquire and save the MESI images. The pylon Camera Software Suite contains easy to use APIs, a viewer for live camera evaluation, and the drivers to interact with the Basler cameras. It is possible to develop dedicated software that can allow new users to easily acquire MESI images. Some suggestions for the development for this software are as follows. 4. Our lab has implemented a MATLAB (MathWorks, Inc.) script that uses the ANSI C NI-DAQmx library (National Instruments Corp.) to operate the DAQ to produce both the camera exposure trigger signals and AOM modulation voltages necessary for the MESI calibration stage (Fig. 5). While this functionality can be implemented directly in the acquisition software itself without relying on MATLAB, the concepts and methods for implementing the calibration defined here can be used. The camera exposure trigger signal and the AOM modulation voltage waveforms are both generated at 1 MHz with identical pulse durations that match the chosen camera exposure times. A slight temporal offset, +25 μs delay for the AOM 50µs

80ms

Camera Exposures

1V

AOM Modulation 10mV

Fig. 5 An example of the timing diagram for the MESI control signals. In this example we can see the 15 pulsed camera exposure trigger signals that make up a MESI frame and the corresponding modulation voltages that are applied to the RF driver for each specific exposure time. (Image reproduced from [19])

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow

109

signal, is used to guarantee that the actual camera exposures and the laser pulses were synchronized in time. The AOM modulation voltages for each exposure are determined using the calibration procedure outlined below: (a) An initial guess, between 0 and 1 V, for the modulation voltages is generated using a power law function. (b) These voltages are used to acquire a complete MESI frame containing 15 raw intensity images from different exposures. (c) The average intensity and total number of saturated pixels within a user-defined region is then calculated for each image. If the overall coefficient of variation and the number of saturated pixels is less than the defined thresholds, then the intensities are equalized and the calibration is complete. The thresholds define the tolerance of dissimilarity between the average pixel intensity at each exposure time. (d) If the average intensity variation is too high or if there are too many saturated pixels, then each of the modulation voltages are adjusted accordingly using the shortest exposure time as the target intensity. (e) This process repeats recursively until the stop conditions are achieved. (f) A typical calibration will take between 30 and 50 iterations and complete within less than a minute. 5. The exact choice, and number of exposure times, remains a topic of research. While the majority of the MESI studies from our lab have focused on using 15 exposure times spanning three decades of exposure times from 50 μs to 80 ms, the crucial aspect is to fully capture the MESI curve. The longer exposure times help us resolve the MESI curve in scenarios with low flow and slow changing dynamics, and the shortest exposure times are needed for higher flow and quickly changing dynamics (the lowest exposure time will be hardware limited by the camera and AOM minimum gating times). The range and number of exposure times can be further optimized for specific applications; it is possible to optimize the exposure times if a priori information about the sample or test case is known. This can help speed up MESI acquisition and move it closer to being a live, real-time imaging tool. 6. Computing speckle contrast: To compute the full speckle contrast image, as defined by Eq. 1, we must first define the NxN sliding window. Statistical analysis [1] and past precedent [18] have established N ¼ 7 as an optimal window size to ensure both image resolution and speckle contrast sampling. A

110

Adam Santorelli et al.

properly sampled speckle pattern will produce speckle contrast values ranging between 0 and 1 [6]. Iteratively go through each of the collected raw images and compute the speckle contrast images for each exposure time. These speckle contrast images are required to solve the ICT maps. 7. Levenberg-Marquardt nonlinear least squares optimization is used to fit the unknown MESI parameters (ρ, τc, ν, β) to the computed speckle contrast data collected from the experiment. This runs on a pixel-by-pixel iterative basis. Loop through the entire image to get the complete τc map (see Note 8 about initial guesses and the choice of β). 8. The β parameter is an instrumentation factor that theoretically should remain constant so long as the illuminating light and imaging optics are unchanged. However, if β is fitted for at every time point, it could vary significantly over time, impacting the stability of ICT. Microfluidics tests have shown that holding β fixed results in a more robust reproduction of stepped flow profiles. Therefore, when processing timeresolved MESI data, it is advantageous to perform a preliminary fit to estimate the value of β. Solve for beta over pre-defined region (e.g., the first 100 frames), and use this for subsequent analysis. Additionally, an initial guess for ρ, τc, ν is typically included in the nonlinear least squares optimization, and the initial guess can impact the speed at which the fit reaches a solution. An alternative to this fitting method is something like using a genetic algorithm to solve for the MESI parameters. References 1. Richards LM, Kazmi SS, Olin KE, Waldron JS, Fox DJ Jr, Dunn AK (2017) Intraoperative multi-exposure speckle imaging of cerebral blood flow. J Cereb Blood Flow Metab 37(9): 3097–3109 2. Parthasarathy AB, Weber EL, Richards LM, Fox DJ, Dunn AK (2010) Laser speckle contrast imaging of cerebral blood flow in humans during neurosurgery: a pilot clinical study. J Biomed Opt 15(6):066030 3. Raabe A, Beck J, Gerlach R, Zimmermann M, Seifert V (2003) Near-infrared indocyanine green video angiography: a new method for intraoperative assessment of vascular flow. Neurosurgery 52(1):132–139 4. Gruber A, Dorfer C, Standhardt H, Bavinzski G, Knosp E (2011) Prospective comparison of intraoperative vascular monitoring technologies during cerebral aneurysm surgery. Neurosurgery 68(3):657–673

˜ uela F, Hieshima G, 5. Martin NA, Bentson J, Vin Reicher M, Black K et al (1990) Intraoperative digital subtraction angiography and the surgical treatment of intracranial aneurysms and vascular malformations. J Neurosurg 73(4): 526–533 6. Boas DA, Dunn AK (2010) Laser speckle contrast imaging in biomedical optics. J Biomed Opt 15(1):011109 7. Dunn AK (2012) Laser speckle contrast imaging of cerebral blood flow. Ann Biomed Eng 40(2):367–377 8. Dunn AK, Bolay H, Moskowitz MA, Boas DA (2001) Dynamic imaging of cerebral blood flow using laser speckle. J Cereb Blood Flow Metab 21(3):195–201 9. He F, Sullender CT, Zhu H, Williamson MR, Li X, Zhao Z et al (2020) Multimodal mapping of neural activity and cerebral blood flow reveals long-lasting neurovascular dissociations

Multi-Exposure Speckle Imaging to Assess Cortical Blood Flow after small-scale strokes. Sci Adv 6(21): eaba1933 10. Kazmi SMS, Parthasarthy AB, Song NE, Jones TA, Dunn AK (2013) Chronic imaging of cortical blood flow using multi-exposure speckle imaging. J Cereb Blood Flow Metab 33(6): 798–808 11. Schrandt CJ, Kazmi SS, Jones TA, Dunn AK (2015) Chronic monitoring of vascular progression after ischemic stroke using multiexposure speckle imaging and two-photon fluorescence microscopy. J Cereb Blood Flow Metab 35(6):933–942 12. Sullender CT, Richards LM, He F, Luan L, Dunn AK (2022) Dynamics of isofluraneinduced vasodilation and blood flow of cerebral vasculature revealed by multi-exposure speckle imaging. J Neurosci Methods 366:109434 13. Soleimanzad H, Gurden H, Pain F (2018) Wide-field speckle imaging and two-photon microscopy for the investigation of cerebral blood flow in vivo in mice models of obesity. In: Biophotonics: photonic solutions for better health care VI, vol 10685. International Society for Optics and Photonics, p 1068508

111

14. Miller DR, Ashour R, Sullender CT, Dunn A (2021) Laser speckle contrast imaging for visualizing blood flow during cerebral aneurysm surgery: a comparison with indocyanine green angiography. medRxiv 15. Kazmi SS, Faraji E, Davis MA, Huang YY, Zhang XJ, Dunn AK (2015) Flux or speed? Examining speckle contrast imaging of vascular flows. Biomed Opt Express 6(7):2588–2608 16. Parthasarathy AB, Tom WJ, Gopal A, Zhang X, Dunn AK (2008) Robust flow measurement with multi-exposure speckle imaging. Opt Express 16(3):1975–1989 17. Kazmi SS, Balial S, Dunn AK (2014) Optimization of camera exposure durations for multiexposure speckle imaging of the microcirculation. Biomed Opt Express 5(7):2157–2171 18. Duncan DD, Kirkpatrick SJ, Wang RK (2008) Statistics of local speckle contrast. JOSA A 25(1):9–15 19. Sullender CT (2018) Quantitative optical imaging platform for studying neurovascular hemodynamics during ischemic stroke. Dissertation, University of Texas

Chapter 11 Wide-Field Optical Imaging in Mouse Models of Ischemic Stroke Jonah A. Padawer-Curry, Ryan M. Bowen, Anmol Jarang, Xiaodan Wang, Jin-Moo Lee, and Adam Q. Bauer Abstract Functional neuroimaging is a powerful tool for evaluating how local and global brain circuits evolve after focal ischemia and how these changes relate to functional recovery. For example, acutely after stroke, changes in functional brain organization relate to initial deficit and are predictive of recovery potential. During recovery, the reemergence and restoration of connections lost due to stroke correlate with recovery of function. Thus, information gleaned from functional neuroimaging can be used as a proxy for behavior and inform on the efficacy of interventional strategies designed to affect plasticity mechanisms after injury. And because these findings are consistently observed across species, bridge measurements can be made in animal models to enrich findings in human stroke populations. In mice, genetic engineering techniques have provided several new opportunities for extending optical neuroimaging methods to more direct measures of neuronal activity. These developments are especially useful in the context of stroke where neurovascular coupling can be altered, potentially limiting imaging measures based on hemodynamic activity alone. This chapter is designed to give an overview of functional wide-field optical imaging (WFOI) for applications in rodent models of stroke, primarily in the mouse. The goal is to provide a protocol for laboratories that want to incorporate an affordable functional neuroimaging assay into their current research thrusts, but perhaps lack the background knowledge or equipment for developing a new arm of research in their lab. Within, we offer a comprehensive guide developing and applying WFOI technology with the hope of facilitating accessibility of neuroimaging technology to other researchers in the stroke field. Key words WFOI, Wide-field optical imaging, OISI, Optical intrinsic optical imaging, GECI, Genetically encoded calcium indicator, BOLD, blood oxygen level dependent, RS-FC, Resting-state functional connectivity, ISA, Infra-slow activity

1

Introduction Functional neuroimaging has proven to be a powerful tool for evaluating functional brain organization in healthy subjects and its evolution following stroke [1–7]. For example, in the months

Authors Jonah A. Padawer-Curry and Ryan M. Bowen have contributed equally to this chapter. Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_11, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

113

114

Jonah A. Padawer-Curry et al.

following stroke, functional magnetic resonance imaging (fMRI) studies have shown that local brain circuits lost to infarction remap to peri-infarct cortex [8] and are more spatially focused in patients exhibiting more complete recovery. Patients with poorer recovery exhibit diffuse activation patterns that can involve both hemispheres [9–11]. Functionally, peri-infarct regions appear to undergo large-scale changes in neuronal response properties. For example, human and animal studies have shown that peri-infarct regions become more responsive to stimulation of somatomotor regions with which they are not typically associated, suggesting that peri-infarct cortex might take over the function of brain regions lost to stroke [12–16]. In rodents, remodeling of local circuitry in periinfarct cortex (“remapping”) correlates temporally with behavioral recovery [12, 13, 17]. Studies examining distributed patterns of intrinsic, resting-state activity throughout the brain reveal that global patterns of resting-state functional connectivity (RS-FC) [18, 19] are also altered after focal stroke. Acute disruption of interhemispheric, homotopic RS-FC predicts poor motor and attentional recovery [20, 21]. In rats, restoration of homotopic RS-FC correlates with improved behavioral recovery after stroke [22]. We [1, 23–25] and others [12, 26–28] have recapitulated many of these findings in mouse models of focal ischemia. Thus, bridge measurements can be made across animal models to enrich findings in human stroke populations. This chapter is designed to give an overview of functional optical neuroimaging methods for applications in rodent models of stroke, primarily in the mouse. The goal is to provide a protocol for researchers who might not have the background knowledge or equipment for developing a new arm of research in their lab, but who might want to incorporate a functional optical neuroimaging assay into their ongoing stroke studies. It should be noted that the techniques covered in this chapter date back nearly four decades from the time of this writing, beginning with the seminal work of Amir Grinvald and colleagues who noted that changes in diffuse reflectance collected from the exposed cat cortex were detectable during evoked activity of the visual system [29]. Because these signals occur from endogenous contrast (in this case, optical absorption due to hemoglobin), this mapping method is referred to as “optical intrinsic signal imaging” (OISI) or variations of that phrase. Later work by the same group [30], and then others, would perform spectral analysis to confer that changes in light reflectance were due to local changes in blood volume and oxygenation [30, 31]. Thus, through neurovascular coupling [32–35], OISI allows for indirect mapping of neuronal activity in a manner similar to blood oxygenation level-dependent functional magnetic resonance imaging (BOLD-fMRI). For the interested reader, several excellent reviews are available for further information on

Wide-Field Optical Imaging of Ischemic Stroke

115

neurovascular coupling [36], how neural activity relates to the BOLD signal [37–40], and functional optical neuroimaging [41]. Despite the power of blood-based imaging, compared to neuronal activity, hemodynamic measures of brain activity (e.g., OISI, FMRI) are slow and can be confounded by diseases affecting neurovascular coupling. Following stroke, altered neurovascular coupling can compromise blood-based measures of brain activity, emphasizing the need for more direct measures of neuronal activity following injury. Genetic engineering techniques in mice have provided several new opportunities for extending wide-field optical imaging methods to more direct measures of neural activity. Fluctuations in calcium or the local field potential can be imaged and visualized using fluorescent, genetically encoded calcium [42–46] or voltage indicators [47–51] (GECI or GEVI, respectively) [52– 57]. Over the past several years, OISI technology has rapidly expanded to include simultaneous measures of cortical calcium dynamics [23, 25, 41, 58–64]. Since 1975, over 1500 articles have been published on optical intrinsic signal imaging, and since 2014, over 100 articles have been written about wide-field calcium imaging. Throughout the rest of this chapter, we will collectively refer to OISI and fluorescence calcium imaging as wide-field optical imaging (WFOI). Below, we describe a series of protocols to assist other researchers implement WFOI in their mouse models of focal ischemia. 1.1 Animals and Imaging Contrasts, Housing, and Animal Preparation

The choice of mouse strain will depend on the imaging goals of the experiment. OISI can be applied in any strain of mouse (e.g., C57Bl6, Swiss Webster, etc.), while mapping of calcium dynamics requires mice expressing a GECI [44, 54, 65–67] (e.g., GCaMP, jRGECO1a, etc.), either through transgenic approaches or viral vector targeting [46, 48, 68–72]. When selecting mice for a stroke experiment, both sex and age of the mice used should be considered. Estrogen is a neuroprotective hormone that can cause sex-dependent discrepancies in outcomes after ischemic stroke [11], and it might be tempting to exclude female mice from stroke studies. However, failing to include both sexes in stroke research makes for an incomplete and uncomprehensive narrative, and both sexes should be included to maximize rigor of the experiment. Unless the goals of the experiments dictate otherwise, mice should have free access to food and water, and a 12-h light–dark cycle. Cages should be changed weekly. Mice are typically housed in standard caging from birth, but enriched housing provides a larger environment, multisensory stimulation (e.g., ladders, toys, huts, running wheels, etc.), greater social interaction, and the opportunity for exploration. Following stroke, exposure to enriched environments accelerates recovery [23, 73–75] and increases the dynamic range observable in behavioral and functional neuroimaging assays. Enriched environments should therefore be considered

116

Jonah A. Padawer-Curry et al.

as part of the study design. Mice should be placed in enriched housing 2 weeks prior to the collection of any baseline measures (e.g., behavioral performance, neuroimaging) to avoid confounds from plastic changes in brain function following exposure to the novel environment [76]. WFOI requires minimally invasive surgery to expose the skull prior to imaging. Scalp retraction is amenable for acute imaging and has also been successfully applied in longitudinal studies [77]. However, repeated scalp retractions can lead to the buildup of scar tissue on the scalp and granulation tissue on the skull, all of which can affect image quality. Longitudinal mapping of brain function (e.g., during the weeks following stroke) requires reliable long-term optical access to the cortex. Separately, movement artifacts will contaminate measures of brain activity in a similar manner as observed in fMRI studies [78]. In mice, particularly during awake imaging, movement is eliminated through head fixation. A variety of longitudinal cranial window surgeries and head fixing apparatuses have been used in wide-field optical imaging [1, 25, 58, 63, 64, 79–81]. Our solution involves securing a small Plexiglas window having pre-tapped holes onto the intact skull. Prior to window installation, the scalp is cleaned and retracted, followed by removal of any connective tissue on the skull. Once cleaned, the window is adhered to the skull with dental cement, and the mouse is allowed to recover before further intervention, imaging, or behavioral testing. 1.2 Wide-Field Optical Imaging System: Overview

While some commercial WFOI systems are available [82, 83], most are built in-house. Due to the custom nature of these systems, hardware components and system design for WFOI can vary widely across labs. Conceivably, one could map brain function with an imaging system comprised solely of a smart phone [84]; systems can also be tens of thousands of dollars and made up of multiple cameras, LEDs, drivers, data acquisition cards, computers, etc. Many research labs incorporate 1–4 LEDs for spectral analysis of hemoglobin absorption and fluorescence excitation of endogenous fluorophores (e.g., flavins) or GECIs (e.g., GCaMP). Light collection can be achieved using an off-the-shelf camera lens, detected by a back illuminated electron-multiplying charge coupled device (EMCCD) camera or a scientific complementary metal-oxide-semiconductor (sCMOS) camera. Image quality and noise characteristics of sCMOS cameras have dramatically improved over the past few years and exhibit performance characteristics that rival and are even superior to EMCCD cameras at approximately 25% of the cost.

Wide-Field Optical Imaging of Ischemic Stroke

117

1.3 Imaging Protocol Considerations During Anesthesia

WFOI under anesthesia allows for mapping brain activity for extended periods with benefit of reduced motion artifacts compared to awake imaging (provided mice are properly secured), and there is no need for acclimation as the mice are imaged while unconscious. However, the type and dose of anesthesia can dramatically affect cortical excitability and hemodynamic activity as measured by WFOI. One clear example observed under several forms of injectable anesthetics (e.g., ketamine/xylazine, dexmedetomidine) is the presence of large neuronal delta wave fluctuations that can obscure underlying spontaneous activity associated with functional brain network organization [59]. Additionally, titrating isoflurane anesthesia is notoriously difficult in the mouse and often results in either an awake-like state or dramatic changes in excitability and/or cortical silencing over a very small range of dosages [59, 85].

1.4 Imaging Protocols: Acclimation for Awake Imaging

Awake WFOI enables imaging cortical activity in behaving mice. To perform WFOI in awake animals, mice must be acclimated to the imaging environment. Specifically, the mice need to become accustomed to head fixation and novel sensory input (e.g., from strobing LEDs) while in the imaging system. Acclimation reduces stress and excessive movement that leads to imaging artifacts in neuroimaging data. Because awake WFOI avoids some of the confounds associated with anesthetized imaging, there is a trend in the neuroimaging community toward imaging under awake conditions when possible [63, 64, 67, 76, 80, 81, 86–100] (since 1980, approximately 2000 articles have been written on awake functional neuroimaging). However, because the mice can whisk, groom, walk, and/or sleep, care must be taken to monitor the state and behavior of the animal while performing awake imaging.

1.5 Imaging Protocols: Evoked Responses

In the months following stroke, cortical representations of function (i.e., local circuits) lost to infarction remap (apparent migration of function from damaged tissue to healthy tissue) to peri-infarct cortex [8]. Imaging this evolving process allows for examination of how manipulations designed to affect functional recovery relate to cortical remodeling. Below we outline procedures for measuring evoked activity in forelimb and whisker representations in somatosensory cortex. While whisker stimuli can be delivered in either awake or anesthetized conditions, functional neuroimaging of forelimb responses requires anesthesia due to potential pain from the preparation and stimuli delivered.

1.6 Imaging Protocols: RestingState Imaging

Performance deficits after stroke are best understood in terms of examining distributed functional systems [101]. These functional systems are efficiently studied using “resting state” neuroimaging methods (e.g., WFOI, BOLD-fMRI), i.e., functional neuroimaging acquired without imposed tasks [102, 103]. The topographies

118

Jonah A. Padawer-Curry et al.

revealed by such analyses are equivalently known as either resting state networks (RSNs) [104] or intrinsic connectivity networks [105]. The basis of resting state functional connectivity (RS-FC) is that spontaneous fluctuations of ensemble neuronal activity and blood oxygenation, measured with either WFOI [106] or fMRI [107], are temporally correlated within RSNs. RSNs are commonly mapped by extracting signals from regions-of-interest (see below) or in data-driven ways using, for example, independent component analysis [104]. Both methods yield reliable and broadly similar results provided that the data are relatively uncorrupted by artifact and acquired over a sufficiently long time [108]. The functional significance of RSNs derives from the observation that they topographically correspond to known sensory, motor, and “cognitive” functional systems [109, 110]. 1.7 WFOI Data Processing

Image sequences of diffuse reflectance at different illumination wavelengths and/or fluorescence emission require processing prior to any subsequent analysis. Standardizing analyses will allow for proper comparison across instructions and labs. General steps for processing raw image sequences into changes in hemodynamic and/or calcium concentration changes include spatiotemporal filtering, spatial normalization (to facilitate comparisons across animals), spectroscopic inversion (in the case of diffuse reflectance), and fluorescence correction (e.g., due to hemoglobin absorption). Once processed, several analysis strategies can be implemented for evaluating stimulus-evoked and/or spontaneous activity.

1.8 WFOI Data Analysis

Focal ischemia can cause remote dysfunction in distant, structurally intact brain regions, and small infarcts can result in widespread depression of circuit communication [111]. Thus, behavioral performance across multiple domains can be affected following focal injury. Functional recovery after stroke is associated with brain networks returning toward normal activation patterns and intrinsic network organization [12, 13, 17]. Below we outline several measures proven to be sensitive biomarkers of repair processes. Remapping metrics include evaluating evoked response time courses and response magnitude and area [1, 12, 24, 26–28] of evoked activity. Band-specific changes in spontaneous activity can be quantified through estimates of regional and global power spectral density [1, 112–114], how spectral content changes over time (e.g., via analysis of spectrograms), and mapping cortical power over the cortex (e.g., power topographies). Measures of RS-FC within and across brain regions strongly correlate with task performance [20– 22, 115–117], with homotopic RS-FC being a particularly sensitive biomarker [22]. Further, simultaneous imaging of calcium fluorescence and hemodynamic activity allow for mapping changes in neurovascular coupling [62, 93, 118–125] following cortical injury.

Wide-Field Optical Imaging of Ischemic Stroke

2

119

Materials

2.1 Animal Preparation for Serial Wide-Field Optical Imaging

Installing optical windows pre-tapped with screw holes allows for longitudinal, head-fixed WFOI. When mice are in their home cages, a second window screwed to the top of the first serves as a cover to prevent scratching. Sterile methods should be used for surgery. Materials for windowing surgery consist of: 1. Isoflurane anesthetic setup (bite bar and nose cone). 2. Plexiglas window (see CAD files). 3. 2 mm window screws. 4. 70% ethanol. 5. 1000 μL pipette and pipette tips. 6. Razor blade ×2. 7. Parkell MetabondⓇ L-Powder, Quick Base, Catalyst for C & B Metabond, adjustable precision applicator. 8. Ceramic mixing dish. 9. Surgical microscope. 10. Feedback temperature-controlled heating pad. 11. Eye lubricant. 12. Sterile Q-tips. 13. Isopropyl alcohol. 14. Betadine. 15. Injectable lidocaine hydrochloride (2%). 16. Surgical probe. 17. Large surgical scissors. 18. Small surgical scissors. 19. Surgical forceps ×2. 20. 0.30 mL syringe ×2. 21. Buprenorphine solution: 9 mL of sterile saline, 1 mL of 0.3 mg/mL buprenorphine HCl. 22. Alcohol prep pads.

2.2 Wide-Field Optical Imaging System

2.2.1

LEDS and Filters

General WFOI components include LEDs and drivers (see Note 1), a camera (see Note 2), lenses, associated filters (fluorescence excitation/emission), mounting hardware, and a computer with hardware control (e.g., via data acquisition card) for synchronizing the LEDs with camera acquisition. Essential materials include the following: 1. High-powered 470 nm blue LED, driver, and power supply. 2. 530 nm and 625 nm LEDs, drivers, and power supplies.

120

Jonah A. Padawer-Curry et al.

3. Narrow-band filters (~10 nm widths). 4. 1” Plano-convex lenses for uniform illumination. 2.2.2

Imaging System

1. Isolation table. 2. Scientific CMOS or EMCCD. 3. 75 mm f/1.4 focal length camera lens. 4. Excitation blocking, 515-nm-long pass filter. 5. Polarizer 72 mm. 6. Light-tight filter wheel for size Ø1” filters. 7. Mounting hardware. 8. Light-blocking black curtains.

2.2.3

Control Hardware

1. 16 bit, 8 channels, Ms/s analog output device. 2. Computer (I/O device). 3. Noise-rejecting, shielded BNC connector block. 4. BNC cables.

2.2.4 Head-Fixing Apparatus

1. Provided CAD Plexiglas head mount (one per mouse). 2. Provided CAD table post mount. 3. Posts for mounting ×4. 4. Fabric hammock for mouse. 5. Light-tight curtains. 6. Light box. 7. Small lab jack to adjust height. 8. Screwdriver for window screws. 9. Screwdriver for non-reflective screws. 10. Jack set for height micro-adjustments. 11. Non-reflective screws (three per mouse). 12. Alcohol prep pads.

2.3 Imaging Protocols: Anesthesia

A ketamine/xylazine cocktail can be prepared to anesthetize animals during WFOI. The materials to create the cocktail are shown below. 1. Ketamine-xylazine cocktail: 8.5 mL sterile saline, 1.0 mL ketamine HCl (100 mg/mL equivalent), 0.5 mL xylazine (20 mg/ mL); mix all solutions in a sterile 10 mL glass vial, vent with an empty needle while adding saline. 2. Insulin syringes.

2.4 Imaging Protocols: Acclimation for Awake

Prior to imaging under awake conditions, mice need to be acclimated to having their heads restrained while in the WFOI system. Note that for each mouse, there will be one window complete with three screws. The same head mount, table post mount, felt hammock, screwdrivers, and three non-reflective screws to secure a

Wide-Field Optical Imaging of Ischemic Stroke

121

mouse in the imaging head plate can be used for all mice being imaged. To acclimate a mouse, the following materials are needed: 1. Screwdriver for window screws. 2. Screwdriver for non-reflective screws. 3. Jack set for height micro-adjustments. 4. Non-reflective screws (three per mouse). 5. Alcohol prep pads. 2.5 Imaging Protocols: ForepawEvoked Response Imaging

Forepaw stimulation is painful and requires imaging mice under anesthesia. Patterned electrical stimuli can be easily configured using a stimulus isolation unit, or stimulus box, and triggered under computer control to time-locked electrical stimuli with WFOI. Components are as follows: 1. Ketamine-xylazine cocktail. 2. Banana plugs to test leads. 3. BNC cables. 4. Stimulus box. 5. Microvascular clips.

2.6 Imaging Protocols: WhiskerEvoked Response Imaging

Peripheral stimulation of the whiskers using air puffs can be performed under awake or anesthetized conditions. For whiskerevoked response imaging, obtain the following equipment: 1. Stimulus box. 2. Precision air puffer. 3. Function generator. 4. Coaxial cable. 5. Puffing tube. 6. Puffing nozzle. 7. Screw to hold and position nozzle.

2.7 Imaging Protocols: Resting State

Resting state imaging—whether awake or anesthetized—requires the mice to be head fixed and mounted. As such, all materials needed for resting state are the same materials seen in Subheading 2.4.

2.8 WFOI Data Processing and Analysis

All data can be processed using MATLAB and a computer with enough RAM to load raw imaging data (ideally ≥32 GB RAM). Raw imaging data for a 5 min session consists of 6000 images (300 s * 20 Hz acquisition frame rate) of size 128 × 128 pixels and a bit depth of 14. Increased computational power decreases computational time but is not required. Minimum requirements are an Intel Core i5 or AMD Ryzen 5, 1 TB of storage for processed data output, and Windows 10 or 11.

122

3

Jonah A. Padawer-Curry et al.

Methods

3.1 Animal Preparation for Serial Wide-Field Optical Imaging

Animals should be prepared for WFOI using the windowing procedure described as follows: 1. Lay out the tools for mixing the dental cement on a sterile towel drape beside the surgical setup. The tools for Metabond creation include Plexiglass windows, 70% ethanol, 1000 μL pipette, 1000 μL pipette tips, a razor blade, Parkell MetabondⓇ ingredients (Quick Base, Catalyst, powder), and Parkell Adjustable Precision Applicators. 2. Lay the other sterile towel drape over the heating pad and underneath the surgical microscope. 3. Place the eye lubricant, Betadine, isopropyl alcohol, lidocaine, 0.3 mL insulin syringes, and buprenorphine on the sterile towel drape on the surgical setup. 4. Once the surgical tools are sterilized, lay them out on the sterile towel drape on the surgical setup as well. 5. Turn on the feedback temperature-controlled heating pad on which the mouse will lay during the procedure, and set it to 37.0 °C; failure to regulate the mouse’s body temperature can worsen post-surgical outcomes [126]. 6. Sterilize the Plexiglas window by allowing it to soak in 70% ethanol before attachment. 7. Place the mouse in an induction chamber and induce anesthesia with 3–5% isoflurane, at 3.5 Liters per minute for 5 min. 8. Transfer the mouse into a stereotaxic frame equipped with a bite bar and nose cone that allow for continuous isoflurane delivery during surgery. 9. Use the surgical probe to move the mouse’s tongue out of the way and insert the top incisors into the hole in the bite bar. 10. Advance the nose cone over the mouse’s nose and fasten. 11. Maintain anesthesia around 1.25–1.75% isoflurane at 1.2 Liters per minute and ensure no response to toe pinch. 12. Apply eye lubricant to both eyes to prevent eye drying and corneal scratches. 13. Secure the ear bars of the stereotaxic frame. Gently touch the mouse’s skull at the anterior and ventral corner of each ear opening with the ear bars. Hard resistance from the skull should be felt when the ear bars are positioned properly. Using gentle force with the pinky fingers to keep the ear bars in place, tighten the ear bars down with the knobs on the stereotaxic frame. See Note 3.

Wide-Field Optical Imaging of Ischemic Stroke

123

14. Rub a pea sized amount of chemical depilatory into the scalp ranging from behind the ears to between the eyes and leave it on for 3–5 min. Avoid getting depilatory in the eyes. See Note 4. 15. Alternatively, the fur on the scalp can be plucked the day before the windowing procedure is performed. (a) Using 3–5% isoflurane at 3.5 Liters per minute, induce the mice until breathing has slowed to a rate of approximately 1 breath per second and is not responsive to painful stimuli (i.e., toe pinch). (b) Wearing sterile gloves, remove the mouse from the induction chamber, and pull the skin taut under the mouse’s neck, being cautious not to asphyxiate the mouse by pinching its trachea. (c) Remove hair by pinching tufts with thumb and forefinger and pulling. (d) Place the mouse back in the induction chamber after 30–45 s as it begins to awaken. (e) Repeat this procedure until fur on the entire scalp is removed, including the fur between the eyes, between the eyes and ears, and at the posterior edge of the ears. (f) Once the fur has been removed from the scalp, transfer the mouse to a temperature-controlled incubator to recover. 16. Once the mouse is appropriately set in the stereotaxic frame and a consistent anesthetic plane is achieved, apply Betadine to the exposed scalp using sterile cotton applicators. 17. Apply isopropyl alcohol to the exposed scalp using cotton applicators. 18. Repeat application of Betadine and isopropyl alcohol two more times. 19. Inject 100-150 microliters of lidocaine underneath the scalp along the midline of the skull, from approximately between the eyes to between the ears (along the path of the incision to be made). 20. Make a midline incision between the anterior edge of the ears up to the posterior edge of the eyes. 21. Using blunt dissection techniques with sterile cotton applicators and surgical forceps, clear away fascia and membranous tissue, and retract the scalp to produce an elliptical, bowl-like cavity. The goal of scalp retraction is to maximize the amount of neocortex visible for optical imaging. Thus, scalp retraction should expose cortical area from the occipital lobe up to the olfactory bulb on an anterior-posterior axis, as well as sensory

124

Jonah A. Padawer-Curry et al.

Fig. 1 Placement of a cranial window. Note the visibility of the anterior suture cranial landmark, as well as bregma and lambda

cortices in both hemispheres on a medial-lateral axis (see Fig. 1). 22. Secure a 1000 μL pipette tip onto the 1000 μL pipette, and set the pipette volume to 200 μL. Lay the pipette on its side and use one of the razor blades to slice the pipette tip at a 45 degree angle beginning at a location where the pipette tip is approximately 6–7 mm wide and slicing toward the end. See Note 5. 23. Next, the skull is prepared for application of the clear dental cement, Metabond. As the dental cement is quick-drying, the next few steps should be performed in a timely fashion. (a) Mix eight drops of Quick Base and two drops of Catalyst in the mixing dish. (b) Using the adjustable precision applicators, apply a thin film of this solution on top of the cortex. 24. See Notes 6–9 before performing this step. In a new well of the mixing dish, place four scoops of L-powder and eight drops of Quick Base, and mix well with an adjustable precision applicator. 25. Quickly check to ensure the scalp is retracted properly before advancing to the next step. 26. Add two drops of Catalyst into the powder and base solution in the mixing dish and mix well again.

Wide-Field Optical Imaging of Ischemic Stroke

125

27. Immediately begin extracting the Metabond from the mixing dish, but extract it as slowly as possible. 28. Slowly dispense the Metabond onto the skull, first applying it to the anterior point of the elliptical bowl-like shape and working around the inner edges of the cavity until the edges are filled. Then, fill the middle of the cavity, being careful not to overflow the cavity with Metabond. It is important to heed the advice of Notes 8 and 9 during this step. 29. Immediately after dispensing the Metabond onto the head, place the window onto the skull, with the narrower end of the window facing the mouse’s nose. The window should be centered on midline, with the posterior edge of the front screw flush with the anterior edge of the mouse’s eyes. 30. Once properly aligned, press gently for 15–30 s to ensure the window is well-adhered to the skull. 31. Using an adjustable precision applicator, apply the remainder of the Metabond in the mixing dish to the junctions of the window to the skull/scalp, allowing capillary action to pull the liquid cement into the crevices. 32. If any cement has been extruded onto the cheeks or ear bars while implanting the window, wait for the cement to dry slightly, but not too long, before attempting to remove it. See Note 10. 33. While the Metabond dries, clean out the ceramic mixing dish with an alcohol prep pad. 34. Loosen the screws of the window and use the other razor blade to gently remove the top leaf of the window (window cover), carefully slicing through any Metabond that is adhering the two leaves of the window together. 35. Label the underside of the top leaf of the window as desired with a Sharpie marker. 36. While continuing to wait for the Metabond to cure, administer 100 μL of buprenorphine solution subcutaneously in the mouse’s back. 37. Administer 200 μL of sterile saline to the mouse via IP injection to replace any fluids lost during the procedure. 38. Screw the top leaf of the window back onto the mouse. 39. Remove the ear bars, retract the nose cone, and remove the mouse from the bite bar. 40. Place the mouse in a temperature-controlled incubator to recover after the procedure is completed (approximately 30–45 min).

126

Jonah A. Padawer-Curry et al.

3.2 Wide-Field Optical Imaging System

3.2.1 To Build the System of LEDs

Construction of the imaging system consists of (1) assembling the LED components used to illuminate the skull, (2) assembling components for light collection, and (3) configuring a computer for synchronizing illumination and acquisition hardware. 1. Connect polarizers to each LED. 2. Mount the LEDs to the isolation table in such a way that all LEDs illuminate the focal plane. 3. Attach power and driver to the LEDs. 4. Connect the BNC cables to the LED drivers to allow the LEDs to receive timing signals. At this point, all LED drivers should have connected BNC cables that have the other ends unconnected. These will be used to receive timing signals and can be set aside until assembling control hardware. Then, to assemble the camera system: 5. Mount the camera on the isolation table approximately 12″ above the surface of the table. This will allow for the ample room for lenses and for the mouse stage. 6. Attach a polarizer to the end of the lens. 7. Place the filter into the filter wheel and attach the filter wheel to the C-mount side of the lens. Attach this to the C-mount end of the camera. 8. Assemble light-blocking black curtains to block out ambient light. 9. The data acquisition card and control board allow for synchronizing LED illumination sequences with the camera exposure. See Fig. 2 for an example of how these connections are made and a depiction of how each LED is strobed in synchrony with the camera’s acquisition. The length of time each LED is on (i.e., duty cycle) will depend on the native brightness of the LED, current settings of each LED driver, the numerical aperture of the lens, and the camera sensitivity. Procedures for setting light levels are described below in Steps 16–19.

3.2.2 Head Mounting Setup of the Imaging System (a Picture of Which Is Shown in Fig. 3)

1. Place a breadboard on top of the jack and secure it appropriately. The jack is used to bring the mouse‘s skull into the focal plane of the WFOI system. 2. Secure the two posts at the anterior corners of the breadboard. 3. Screw the large head fixing stage in tightly. 4. On the posterior side, secure a post of the same height as the anterior posts in the center of the board. 5. Slide the T-shaped post holder onto the post. Put another post through the other hole of the T-shaped post holder, thus

Wide-Field Optical Imaging of Ischemic Stroke

127

Fig. 2 Cartoon of WFOI system and example of control sequences sent to data acquisition card

creating a cross, one post oriented up and down and another attached to it that is oriented medially.

6. At both ends of the cross, secure another T-shaped post holder and place through, oriented in the anterior/posterior direction. 7. Hang the hammock on these two anterior/posterior posts. 8. The LED duty cycle should be adjusted to approximately 75% of the maximum number of electron counts on the sensor when the mouse is in the imaging system. For example, for a sensor with a 14 bit analog-to-digital converter, the maximum

128

Jonah A. Padawer-Curry et al.

Fig. 3 Image of head-fixing materials and whisker-stimulation materials. (a) Construction of head-fixing stage and whisker puffing system with mouse setup. (b) CAD files of the materials needed for head fixing to a stage and for whisker-puffing

counts are given by 214 = 16,384 counts, 75% of which is ~12,000. 9. Place the mouse in the imaging system using Steps 1–9 in Subheading 3.4.

Wide-Field Optical Imaging of Ischemic Stroke

129

10. Set the power of the LEDs to approximately 75% of its maximum power. 11. Initialize a block design that has equal LED and camera exposure times. 12. Adjust the LEDs and camera exposure times to achieve ~12,000 counts at the brightest location on the mouse’s skull. 3.3 Imaging Protocols: Anesthesia

1. Weigh the mouse in grams. 2. To deliver 100 mg/kg ketamine and 10 mg/kg xylazine, withdraw 10 μL of cocktail per gram of bodyweight with an insulin syringe and administer via IP injection (e.g., a 30 g mouse should receive 300 μL of cocktail). See Note 11. 3. Wait 10–20 min for anesthesia to take effect. The mouse is ready for anesthetized imaging when it is unresponsive to a firm pinch on its hind paw using a gloved hand.

3.4 Imaging Protocols: Acclimation for Awake Imaging

1. Remove the window cover by scruffing the mouse, then holding it by its window with one hand, and loosening the screws of its window with the other hand. 2. To place the head bracket onto the window, align the screws on the head bracket and cranial window. To achieve this, place your thumb and forefinger on the sides of the cranial window to secure the head to place the head-fixer. Align the screw taps on the cranial window and the head-fixer and screw in each of the screws. See Note 12. 3. Place the mouse in the head-fixing stage, securing the head bracket into position with the appropriate screws. 4. When the mouse has been mounted, wrap the mouse in the felt hammock. 5. Wipe the surface of the window with an alcohol prep pad to clear oils and debris which may obstruct imaging. 6. The first imaging run will require the stage height and x-y position to be adjusted so that the cortex is in focus and in the field of view of the camera. 7. Check that the polarizers are blocking all specular reflections by rotating the polarizer on the camera lens and running an imaging sequence. This step is achieved once there are no “hot spots” in image. 8. Once the stage is in the proper place and the window has been wiped clean, stop the video. 9. Before beginning image acquisition, turn off all lights in the imaging area and close the light-blocking curtains to avoid light contamination.

130

Jonah A. Padawer-Curry et al.

10. From here, to acclimate the mouse for awake imaging, repeat the following steps 3 days in a row before baseline imaging. 11. Run the LED sequence for 15–20 min. 12. Remove the mouse, replace the window cover, and return it to its cage. 3.5 Imaging Protocols: ForepawEvoked Response Imaging

1. Mount the mouse according to Steps 1–5 in Subheading 3.4 if not already done. 2. Set the stimulus box to deliver 0.3 ms, 1.0 mA electrical pulses at 3 Hz for 5 s. Stimulation blocks should be 30 s and designed as follows: 5 s rest, 5 s of stimulation, and 20 s of rest. Repeat this stimulation block 30 times for a total of 15 min. See Note 13. 3. Place a feedback temperature-controlled heating pad under the mouse after anesthesia is administered. See Note 14. 4. Attach microvascular clips on either side of the desired wrist in locations where skin can be stretched. See Note 15. 5. Once the microvascular clips are attached and the mouse is unresponsive to toe pinch, begin evoked response imaging. 6. When imaging is finished, remove the mouse from the imaging rig, replace the top leaf of its window, and place it in a temperature-controlled incubator to recover.

3.6 Imaging Protocols: WhiskerEvoked Response Imaging

1. Mount the mouse according to Steps 1–5 in Subheading 3.4 if not already done. 2. Set the Picospritzer to deliver 100 ms pulses of air at 1 Hz and 40–60 PSI for 5 s. See Note 13. 3. Set the stimulus box that is providing input to the Picospritzer to the 10 V output range (constant voltage, not constant current used for forepaw-evoked response imaging). 4. Connect the output of the Picospritzer to the puffing nozzle. 5. Attach the puffing nozzle to the underside of the mouse’s headplate, in front of the mouse’s nose and facing the mouse. See Note 16. 6. Begin the imaging sequence. 7. When imaging is completed, remove the mouse from the imaging stage and replace the window cover. 8. If anesthesia was used, place the mouse in a temperaturecontrolled incubator to recover until ambulatory before returning it to its cage. See Note 17.

Wide-Field Optical Imaging of Ischemic Stroke

3.7 Imaging Protocols: RestingState Imaging

131

1. Mount the mouse according to Steps 1–5 in Subheading 3.4 if not already done. 2. Anesthetize the mouse if so desired and wait until a proper plane of anesthesia has been established. Use a feedback temperature-controlled heating pad if using anesthesia. 3. Run the imaging software for the desired amount of time. 4. Remove the mouse from the imaging rig and return it to its home cage. If anesthesia was used, allow the mouse to recover in a temperature-controlled incubator before returning it to its cage.

3.8

WFOI Processing

1. The analysis process is visually summarized in Fig. 4 and can be performed in MATLAB using the Github link (https://github. com/BauerLabCodebase/WFOI-Textbook-Chapter). Each imaging session is comprised of multiple runs. For each run, raw data consist of a sequence of images in which each frame corresponds to diffuse reflectance or fluorescence emission from each LED channel. 2. Before data processing, raw data are reshaped to be images (Pixel rows × Pixel columns) × LED channel × Time; all data are spatially normalized to a common space (e.g., to the Paxinos atlas or Allen mouse brain atlas, Fig. 5) to facilitate comparisons across mice, and then a brain mask is created to identify cortical pixels from non-cortical pixels. 3. Load in an acquired frame and visually identify at least three anatomical landmarks used for co-registration. Two convenient landmarks are (1) the junction of the rostral rhinal sinus and the sagittal sinus (i.e., a T-shaped structure that separates the cortex from the olfactory bulb) and (2) the intersection of the lambdoid suture with the sagittal suture (i.e., lambda). See Note 18.

Fig. 4 Flow chart of WFOI processing stream

132

Jonah A. Padawer-Curry et al.

Fig. 5 Raw image sequences of the exposed cortex. For each channel and associated LED, a series of images is collected

4. Once landmarks are determined in scanner space, the same will be required for your reference space so that a transform matrix can be defined and applied to the image sequence. 5. Using the anterior and posterior landmarks in both scanner and reference space, create the affine matrix operator. 6. Multiply all images and masks by calculated affine matrix operator, thus aligning all images to a common reference space. 7. Identify non-brain pixels by drawing the outer cortical boundary on the skull (e.g., using roipoly, MATLAB, Fig. 6). This output of the operation is a binary image where all pixels within the mask are 1 (brain) and all pixels outside the mask are 0 (non-brain). See Note 19. 8. As an initial data quality check, calculate the average light level over the cortex of each channel over the duration of the imaging run. Plotting these time courses allows for visualization of large deviations in light level, often an indicator of head movement, large changes in cerebral blood volume, etc. Similarly, visualizing the power spectral density of the average raw light level over time can reveal the dynamic range in each channel, as well as physiologic signaling due to pulse and respiration. 9. For each channel, calculate the average light level within the brain mask and plot as a function of time. Then plot the standard deviation of each trace in Step 6. See Note 20. 10. Plot the power spectral density of each trace in Step 6. See Note 21. 11. Runs with poor data quality should be removed from further analysis. 12. Calculate the time average of the dark-frame image sequence, and subtract the resulting image from each frame in the raw data.

Wide-Field Optical Imaging of Ischemic Stroke

133

Fig. 6 White-light image of dorsal mouse skull. Region of interest is drawn around the cortex to distinguish cortex from skin and fur

13. Slow drifts (0.03 Hz are significantly attenuated

Fig. 10 Whole cortex functional connectivity matrix in healthy mice organized by hemisphere, then network assignment

9. For each pixel within the brain mask, regress the constructed global signal, ggsr(t). That is, for each pixel’s time-series, Si, assume that it can be approximated as Si = ggsr(t)βi + xi(t), such that ggsr(t) is a regressor, βi is the regression coefficient, and xi, is the signal within the ith pixel after regression. See Note 25. Whole cortex functional connectivity matrices (Fig. 10) can be created by following these short steps: 10. Filter the data into the frequency band of interest (e.g., ISA, Delta). See Note 26.

138

Jonah A. Padawer-Curry et al.

11. Normalize each pixel’s time trace within the brain mask to unit variance and zero mean (i.e., z-scored). 12. Calculate the Pearson R correlation coefficient (i.e., r = P ð x i - x Þ ðy i - y Þ P 2 ) between all pixel pairs within the brain √ ðx - x Þ2 ðy i - y Þ mask. An efficient strategy for this step is to reshape the image sequence to be pixels by time and calculate the inner product. The resulting “RSFC matrix” is square and symmetric about the main diagonal and has the same number of rows (and columns) as the number of pixels in the brain mask. 13. Organize the matrix by network assignments to visualize network relationships. 14. Note that this matrix contains all the RSFC information in the WFOI field-of-view. Each row in this matrix corresponds to a particular pixel and represents a map of functional connectivity for that pixel. Alternatively, one might be interested in only calculating RSFC for a defined set of cortical locations (e.g., pixel coordinates, functional region, etc., Fig. 11). This information can be extracted from the whole cortex RS-FC matrix or can be calculated separately using each ROI as a “seed region”: 15. Construct each ROI’s time trace by averaging the time traces within the region. 16. Correlate the ROI time trace with all other time traces within the bran mask. These steps produce a set of RSFC maps for each ROI. 17. To calculate an RSFC matrix for these ROIs, compute the Pearson r correlation coefficient between all ROI time traces, as done in Step 12. To calculate bilateral functional connectivity maps (Fig. 12), a few additional steps are required: 18. Symmetrize the brain mask about midline by multiplying the left side of the brain mask by the right side flipped vertically (e.g., MATLAB’s “fliplr” command) so that only shared pixels across hemispheres are included. 19. Choose a hemisphere as a reference; compute the Pearson R correlation coefficient between each pixel in that hemisphere and its contralateral counterpart (mirrored pixels about midline), creating a map of bilateral functional connectivity. This map is mathematically symmetric.

Wide-Field Optical Imaging of Ischemic Stroke

139

Fig. 11 Seed-based functional connectivity following 30–60 min of TMCAO. Mice were separated into three groups based on infarct size and location

Fig. 12 Bilateral functional connectivity maps of oxygenated hemoglobin for the same groups in Fig. 10

Estimates of neurovascular coupling can be calculated in several ways. Two strategies we have employed are described as “lagged correlation analysis” (Fig. 13a) and “multivariate gamma fitting”

140

Jonah A. Padawer-Curry et al.

Fig. 13 Neurovascular coupling in healthy mice, calculated using two methods. (a) lag analysis, in which spontaneous calcium activity is cross-correlated with activity of total hemoglobin in each pixel. Maps of lag time associated with peak correlation are plotted. (b, c) Iterative gamma variate fitting. Three parameters are used to characterize a gamma distribution, amplitude, time-to-peak, and width. This gamma function is convolved with measured calcium to obtain an estimate of subsequent hemodynamic activity. Iterations within this parameter space result in an estimate of the impulse response function between spontaneous calcium and hemodynamic activity. Parameters resulting in the highest correlation between predicted and measured hemoglobin at each pixel are plotted. (c) The average impulse response function across the cortex

(Fig. 13b). Lagged correlation analysis is performed using measured calcium fluctuations and hemodynamic activity: 20. Filter both the hemodynamic and calcium image sequences to eliminate physiological noise and instrumental drift (e.g., 0.01 Hz – 3 Hz). This range will depend on your system and experiment. 21. Cross correlate calcium and hemodynamic activity (e.g., MATLAB’s “xcorr” function) at each pixel. 22. Find the time shift (“lag”) associated with maximum correlation. Depending on the order of the inputs, this number will be positive (indicating hemoglobin lags calcium) or negative (indicated calcium leads hemoglobin). 23. Mapping lag times and lagged correlation values produces images reporting estimates of neurovascular coupling. Gamma variate fitting involves convolving measured calcium activity with a (parameterized) estimate of the hemodynamic response function and iterating on each parameter until the predicted hemodynamics match those measured. For example, three

Wide-Field Optical Imaging of Ischemic Stroke

141

parameters can characterize a gamma distribution: time to peak (T), width (W), and an amplitude (A). The process for solving for the best set of parameters is posed as an optimization problem, in which the cost function is the mean squared error (MSE) between the measured and predicted hemodynamics. 24. Filter both the hemodynamic and calcium image sequences to eliminate physiological noise and instrumental drift (e.g., 0.01 Hz – 3 Hz). This range will depend on your system and experiment. 25. Choose an initial estimate of T, W, and A. 26. Construct a gamma distribution parameters, T, W, and A.

with

the

current

27. Convolve the gamma distribution with the measured calcium signal. 28. Compute the MSE between the convolved neural signal and measured hemodynamic signal. 29. Update T, W, and A, using the appropriate update rule, and repeat Steps 26–28 until the MSE has been minimized.

4

Notes 1. Choice in LED Wavelengths: A single LED centered at an isosbestic wavelength (e.g.,~530 nm) is sufficient for mapping brain function [131] and provides absorption contrast that is insensitive to oxygenation [132]. If unmixing changes in oxyor deoxyhemoglobin is desired, a minimum of two wavelengths are needed. The choice of LED wavelengths can significantly affect oximetry measurements [41, 60, 132–134]. For imaging GCaMP fluorescence, the fluorescent yield is highest around 490 nm, and the largest portion of emitted photons occurs at an approximately 530 nm [41, 64], a wavelength that is also isosbestic. A common choice of LEDs for calcium imaging are LEDs centered at 470 nm, 530 nm, and 625 nm [64, 132, 135]. To avoid hemodynamic cross talk in absorption measurements, it is recommended to ensure LED spectra are relatively narrow-band [132, 135], which can be achieved by using filters [41, 132, 136]. As price is often a considerable factor, alternatives to expensive, narrow beam LEDs have been developed [137, 138]. For example, our group [135] and other’s [137] have incorporated low-cost LED rings capable of delivering spatially homogenous light [137]. Stable power supplies are also recommended as drifts in illumination will be perceived as changes in diffuse reflectance, which can become mapped to changes in hemodynamic activity [137].

142

Jonah A. Padawer-Curry et al.

2. Camera Selection: Stimulus-evoked and resting-state intrinsic optical signals present themselves as small (~0.1–1%) changes in light reflectance [139, 140] and require cameras with low readout noise, high dynamic range, and high sensitivity [135, 141]. Additionally, cameras with high frame rates and sensor pixel count are advised. To adequately sample (i.e., no aliasing) respiration and heart rate (up to ~10 Hz), a full frame rate of 20 Hz or higher is suggested. Thus, for a camera with four LEDs, camera acquisition must be performed at 80 Hz. Costs for common camera models found in literature range between $10,000 and $40,000 such as those from Andor (Xyla and iXon series), Teledyne Dalsa M60s or M30s, and Retiga R1. Alternatively, cheaper options include the Basler MED ace 2.3 MP 41 mono, Blackfly U3-23S6M, Basler Ace series (i.e., acA1440-22um, pro a2A2448-75 um Pro, pro a2A1920160um Pro), or Sony XC-ST70. 3. Ear Bar Placement: If Kopf brand ear bars are being used, the bars should be tightened at 4–4.5 mm on each side. 4. Chemical Depilatory Application: Leaving the depilatory on for more than 3–5 min can cause burning and long-term irritation. It is critical to avoid placement of chemical depilatory in the eyes or whiskers, as both can cause permanent damage and subsequent functional changes. 5. Pipette Tip Slicing: The purpose of slicing the pipette tip is to facilitate extraction of the viscous liquid cement from the mixing dish. 6. Metabond Creation: The dental cement creation and window application are time-sensitive and require a few steps of preparation to ensure optimal timing. First, the ceramic mixing dish should be awaiting use in a freezer to slow down the solidifying process before the surgery starts (the warmer the mixing bowl, the faster the cement will harden). It is easiest to have multiple mixing dishes if several windowing procedures are to be performed in series. 7. Metabond Extraction: One of the potential pitfalls of the windowing procedure is the development of air bubbles in the Metabond cement. Bubbles can be avoided by paying attention to a few key steps. First, the Metabond should be extracted from the mixing dish as soon as the Metabond is done being mixed. This ensures that the Metabond does not begin to harden sooner than is desired. However, the Metabond should be extracted from the mixing dish as slowly as possible to ensure that air does not leak into the pipette tip and form bubbles. The extraction process should take anywhere from 7 to 15 s. Second, the Metabond should be dispensed onto the skull similarly to how it was extracted: quickly after

Wide-Field Optical Imaging of Ischemic Stroke

143

extraction, but at as slow of a rate as possible. Dispensing as soon as possible will again prevent curing of the Metabond, while slowly dispensing from the pipette will ensure that no air bubbles are introduced into the layer of cement that covers the cranium. 8. Metabond Application: Several things can go wrong when applying Metabond to the skull. If the scalp bowl is constructed poorly, it can cause draining of the cement off the sides. A medium must be struck between stretching the scalp as much as possible and creating a cavity that can feasibly be filled with Metabond without overflowing. It may be beneficial to wait several seconds before dispensing the Metabond onto the skull, as this may prevent runny Metabond from dripping over the scalp wall and onto the mouse’s cheeks or ear bars. 9. Window Longevity: As the purpose of the windowing procedure is to allow for serial WFOI, the longevity of these windows is paramount to their success. To ensure that windows stay on throughout the time course of an experiment, two steps should be performed meticulously. First, when dispensing Metabond onto the skull, as much Metabond as the scalp can possibly hold should be used. The more Metabond between the window and the skull, the better the adherence between the two will be. Second, filling in the remaining Metabond in the junctions where the window meets the skull/scalp is very important. Filling these gaps in the Metabond and ridding the window of exposed edges as much as possible will not only prevent infection that could compromise the health of the animal, but also prevent free edges of the window from being exposed. This will help prevent excessive force from being applied to the window, decreasing the likelihood that it comes off when preparing the mouse for imaging. 10. Excess Metabond Removal: If Metabond does drip over the scalp wall, wait 20–40 s before trying to peel it off. This will give the Metabond time to cure slightly, making it easier to remove. Use the small surgical scissors and forceps to remove unwanted drips of Metabond cement from the cheeks, ears, ear bars, or any other place the Metabond may have gone. 11. Injection Placement: The location of the injection of ketamine/ xylazine cocktail is extremely important. Injections should be performed intraperitoneally (not subcutaneously) to ensure a proper anesthetic plane. However, injecting too deep risks puncturing organs. Aim for the location 4–5 mm posterior to the mouse’s hip approximately 1/3 the way from the ventral midline to the hip. 12. Securing the Head Bracket: Securing the head bracket onto the cranial window properly will prevent motion artifact during

144

Jonah A. Padawer-Curry et al.

awake imaging, which can ruin a run. Tighten the head bracket onto the cranial window tightly, so that the mouse is unable to move its head, but not too tight such that the screws cause stress fractures in the Plexiglas of the window. 13. Block Design: The block design outlined has been empirically proven to produce consistent and robust evoked responses. When choosing a block design and stimulus frequency, some titration in the beginning is prudent. Be sure to keep the block design consistent across experiments in which you would like to compare data. 14. Heating Pad Use: This will prevent hypothermia in the mouse, which is common under ketamine anesthesia and can result in death. 15. Microvascular Clip Placement: Having a firm connection to the wrists is of prime importance for evoked response imaging efficacy, as a weak connection of the microvascular clips to the wrists may cause the clips to fall off during an imaging run, resulting in poor quality data. 16. Puffing Nozzle Placement: The puffing nozzle should be able to pivot to puff air on the whiskers of both sides of the mouse’s snout. 17. Anesthetization: As whisker puffing is not a painful imaging paradigm, whisker-evoked response imaging can be done either while awake or while anesthetized. 18. Atlas Registration: It is important to note considerable effects of inter- and intra-rater variability in selection of landmarks and its effect on downstream analysis [142, 143], which in functional connectivity data can introduce significantly large changes in correlations, particularly further away from midline. To mitigate the effect of inter-rater variability, a single rater should be assigned. Additionally, an un-binned, high-resolution, single “scout” image can be used to help visualize landmarks more clearly. More recently, the use of automated machine learning-based registration and segmentation has been used but requires hundreds of images to adequately train the network and may not offer advantages over traditional methods [143]. 19. Masking: Differences in mouse-to-mouse position under the camera causes masks to vary. Across mice, window placement also guarantees differences in masks. Inherently, this causes data dropout, which can be compensated for by using rigorous statistical methods commonly seen in fMRI literature [144]. 20. Channel Averages and Standard Deviations: For each channel, average time traces should be smooth. If there are large jumps or artifacts, this may indicate that there was mouse movement, large shifts in LED illumination brightness, or other systematic

Wide-Field Optical Imaging of Ischemic Stroke

145

errors. When exploring the standard deviation of this time trace, often, raw light level traces having a standard deviation >1% from the mean tend to be associated with poorer data quality. 21. Spectral Shape: Functionally relevant signaling in the brain exhibits power spectral density having a “1/f-like fall off” (i.e., linear with negative slope on a log-log scale). Fluorescence emission from calcium signaling will extend to ~5 Hz (for GCAmP6f), while diffuse reflectance from the OIS channels will begin to taper and become flat (i.e., “whiten”) around 1–2 Hz. 22. General Note on Spectroscopy: Differential path lengths can be calculated analytically using the diffusion approximation to the radiative transport equation, assuming a semi-infinite geometry and assuming initial total hemoglobin concentrations and oxygen saturations [135, 145, 146]. Alternatively, differential path lengths can be estimated from Monte Carlo simulations [41] and are provided for select wavelengths in Table 1 for reference. For each LED, find the extinction coefficient of the central wavelength. Extinction coefficients for each species of hemoglobin are provided by Scott Prahl [147]. 23. Note on Evoked Responses: evoked responses at early time points post stroke can be difficult to detect. One strategy for defining whether a response was measured or not is through evaluation of contrast to noise pre- and post-stimulus onset. For example, pixels having a change in signal >2 standard deviations of pre-stimulus values are considered a response. 24. Power Spectral Density Estimates. Power spectral density (PSD) estimates can be performed several ways. Welch’s method splits the signal into several windows, taking the Fourier transform of each window and averaging the resultant spectrum. Compared to the fast Fourier method (FFT), noise will be reduced at higher frequencies, but lower frequency information will be lost due to the splitting of the signal. 25. Multiple Signal Regression. Hemodynamics occurring in infarct vs non-infarct tissue can exhibit markedly different spectral content and phase delays, resulting in artificial correlations where none exist [112]. Instead, we advocate using a multiple signal regression approach where two regressors are simultaneously removed from all cortical activity prior to RS-FC calculation. In this case, the regressors are calculated as the average signals within the infarct and non-infarcted areas, respectively. 26. Frequency bands for functional connectivity. While many studies examine RS-FC over infraslow (i.e., 0.01–0.08 Hz) and/or Delta-band (0.5–4.0 Hz), functional network activity exhibits coherence over a wide range of frequencies [148].

146

Jonah A. Padawer-Curry et al.

Acknowledgments We would like to thank past and present contributors to these projects (in alphabetical order): Annie Bice, Joseph Culver, Ernesto Gonzales, Andrew Kraft, Ron Perez, Zachary Rosenthal, Karen Smith, Avi Snyder, Brian White, Patrick Wright, and Ping Yan. This work was supported by National Institute of Health grants: R01-NS102870 (AQB), R37NS110699 (JML), R01NS084028 (JML), F31 NS122499-01A1 (RMB), and Imaging Science Pathway T32 Fellowship (JAP-C). References 1. Bauer AQ, Kraft AW, Wright PW et al (2014) Optical imaging of disrupted functional connectivity following ischemic stroke in mice. NeuroImage 99:388–401 2. Carter AR, Astafiev SV, Lang CE et al (2010) Resting interhemispheric functional magnetic resonance imaging connectivity predicts performance after stroke. Ann Neurol 67:365– 375 3. Friston KJ (1994) Functional and effective connectivity in neuroimaging: a synthesis. Hum Brain Mapp 2:56–78 4. Grefkes C, Fink GR (2011) Reorganization of cerebral networks after stroke: new insights from neuroimaging with connectivity approaches. Brain 134:1264–1276 5. Raichle ME, Mintun MA (2006) Brain work and brain imaging. Annu Rev Neurosci 29: 449–476 6. Silasi G, Murphy TH (2014) Removing the brakes on post-stroke plasticity drives recovery from the intact hemisphere and spinal cord. Brain 137:648–650 7. Silasi G, Murphy TH (2014) Stroke and the connectome: how connectivity guides therapeutic intervention. Neuron 83:1354–1368 8. Cramer SC, Riley JD (2008) Neuroplasticity and brain repair after stroke. Curr Opin Neurol 21:76–82 9. Calautti C, Leroy F, Guincestre JY et al (2001) Sequential activation brain mapping after subcortical stroke: changes in hemispheric balance and recovery. Neuroreport 12:3883–3886 10. Traversa R, Cicinelli P, Bassi A et al (1997) Mapping of motor cortical reorganization after stroke. A brain stimulation study with focal magnetic pulses. Stroke 28:110–117 11. Ward NS, Brown MM, Thompson AJ et al (2003) Neural correlates of motor recovery

after stroke: a longitudinal fMRI study. Brain 126:2476–2496 12. Brown CE, Aminoltejari K, Erb H et al (2009) In vivo voltage-sensitive dye imaging in adult mice reveals that somatosensory maps lost to stroke are replaced over weeks by new structural and functional circuits with prolonged modes of activation within both the peri-infarct zone and distant sites. J Neurosci 29:1719–1734 13. Dijkhuizen RM, Ren J, Mandeville JB et al (2001) Functional magnetic resonance imaging of reorganization in rat brain after stroke. Proc Natl Acad Sci U S A 98:12766–12771 14. Jaillard A, Martin CD, Garambois K et al (2005) Vicarious function within the human primary motor cortex? A longitudinal fMRI stroke study. Brain J Neurol 128:1122–1138 15. Nudo RJ, Wise BM, Sifuentes F et al (1996) Neural substrates for the effects of rehabilitative training on motor recovery after ischemic infarct. Science 272:1791–1794 16. Wei L, Erinjeri JP, Rovainen CM et al (2001) Collateral growth and angiogenesis around cortical stroke. Stroke 32:2179–2184 17. Gonzalez CL, Kolb B (2003) A comparison of different models of stroke on behaviour and brain morphology. Eur J Neurosci 18:1950– 1962 18. Felleman DJ, Van Essen DC (1991) Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex 1:1–47 19. Posner MI, Petersen SE, Fox PT et al (1988) Localization of cognitive operations in the human brain. Science 240:1627–1631 20. Carter AR, Patel KR, Astafiev SV et al (2012) Upstream dysfunction of somatomotor functional connectivity after corticospinal damage in stroke. Neurorehabil Neural Repair 26:7– 19

Wide-Field Optical Imaging of Ischemic Stroke 21. He BJ, Snyder AZ, Vincent JL et al (2007) Breakdown of functional connectivity in frontoparietal networks underlies behavioral deficits in spatial neglect. Neuron 53:905–918 22. Van Meer MPA, Van Der Marel K, Wang K et al (2010) Recovery of sensorimotor function after experimental stroke correlates with restoration of resting-state interhemispheric functional connectivity. J Neurosci 30:3964– 3972 23. Hakon J, Quattromani MJ, Sjolund C et al (2018) Multisensory stimulation improves functional recovery and resting-state functional connectivity in the mouse brain after stroke. Neuroimage Clin 17:717–730 24. Kraft AW, Bauer AQ, Smith KP et al (2015) Sensory deprivation following cortical focal ischemia facilitates remapping and accelerates behavioral recovery. In: BRAIN: international symposium on cerebral blood flow, metabolism and function, Vancouver 25. Quattromani MJ, Hakon J, Rauch U et al (2018) Changes in resting-state functional connectivity after stroke in a mouse brain lacking extracellular matrix components. Neurobiol Dis 112:91–105 26. Dijkhuizen Rm SA, Mandeville JB, Wu O, Halpern EF, Finklestein SP, Rosen BR, Lo E (2003) Correlation between brain reorganization, ischemic damage, and neurologic status after transient focal cerebral ischemia in rats: a functional magnetic resonance imaging study. J Neurosci 23:510–517 27. Harrison TC, Silasi G, Boyd JD et al (2013) Displacement of sensory maps and disorganization of motor cortex after targeted stroke in mice. Stroke 44:2300–2306 28. Winship IR, Murphy TH (2008) In vivo calcium imaging reveals functional rewiring of single somatosensory neurons after stroke. J Neurosci 28:6592–6606 29. Grinvald A, Lieke E, Frostig RD et al (1986) Functional architecture of cortex revealed by optical imaging of intrinsic signals. Nature 324:361–364 30. Malonek D, Grinvald A (1996) Interactions between electrical activity and cortical microcirculation revealed by imaging spectroscopy: implications for functional brain mapping. Science 272:551–554 31. Frostig RD, Lieke EE, Ts’o DY et al (1990) Cortical functional architecture and local coupling between neuronal activity and the microcirculation revealed by in vivo highresolution optical imaging of intrinsic signals. Proc Natl Acad Sci 87:6082–6086 32. Berwick J, Johnston D, Jones M et al (2005) Neurovascular coupling investigated with

147

two-dimensional optical imaging spectroscopy in rat whisker barrel cortex. Eur J Neurosci 22:1655–1666 33. Devor A, Dunn AK, Andermann ML et al (2003) Coupling of total hemoglobin concentration, oxygenation, and neural activity in rat somatosensory cortex. Neuron 39: 353–359 34. Jones M, Berwick J, Johnston D et al (2001) Concurrent optical imaging spectroscopy and laser-Doppler flowmetry: the relationship between blood flow, oxygenation, and volume in rodent barrel cortex. NeuroImage 13: 1002–1015 35. Sheth SA, Nemoto M, Guiou M et al (2004) Columnar specificity of microvascular oxygenation and volume responses: implications for functional brain mapping. J Neurosci 24:634– 641 36. Iadecola C (2017) The neurovascular unit coming of age: a journey through neurovascular coupling in health and disease. Neuron 96:17–42 37. Ekstrom A (2010) How and when the fMRI BOLD signal relates to underlying neural activity: the danger in dissociation. Brain Res Rev 62:233–244 38. Hillman EM (2014) Coupling mechanism and significance of the BOLD signal: a status report. Annu Rev Neurosci 37:161–181 39. Logothetis NK, Wandell BA (2004) Interpreting the BOLD signal. Annu Rev Physiol 66:735–769 40. Nair DG (2005) About being BOLD. Brain Res Rev 50:229–243 41. Ma Y, Shaik MA, Kim SH et al (2016) Widefield optical mapping of neural activity and brain haemodynamics: considerations and novel approaches. Philos Trans R Soc Lond Ser B Biol Sci 371:20150360 42. Grienberger C, Konnerth A (2012) Imaging calcium in neurons. Neuron 73:862–885 43. Hendel T, Mank M, Schnell B et al (2008) Fluorescence changes of genetic calcium indicators and OGB-1 correlated with neural activity and calcium in vivo and in vitro. J Neurosci 28:7399–7411 44. Oh J, Lee C, Kaang BK (2019) Imaging and analysis of genetically encoded calcium indicators linking neural circuits and behaviors. Korean J Physiol Pharmacol 23:237–249 45. Tian L, Hires SA, Looger LL (2012) Imaging neuronal activity with genetically encoded calcium indicators. Cold Spring Harb Protoc 2012:647–656 46. Tian L, Hires SA, Mao T et al (2009) Imaging neural activity in worms, flies and mice with

148

Jonah A. Padawer-Curry et al.

improved GCaMP calcium indicators. Nat Methods 6:875–881 47. Akemann W, Mutoh H, Perron A et al (2012) Imaging neural circuit dynamics with a voltage-sensitive fluorescent protein. J Neurophysiol 108:2323–2337 48. Broussard GJ, Liang R, Tian L (2014) Monitoring activity in neural circuits with genetically encoded indicators. Front Mol Neurosci 7:97 49. Inagaki S, Nagai T (2016) Current progress in genetically encoded voltage indicators for neural activity recording. Curr Opin Chem Biol 33:95–100 50. Momose-Sato Y, Sato K, Arai Y et al (1999) Evaluation of voltage-sensitive dyes for longterm recording of neural activity in the hippocampus. J Membr Biol 172:145–157 51. Yang HH, St-Pierre F (2016) Genetically encoded voltage indicators: opportunities and challenges. J Neurosci 36:9977–9989 52. Dana H, Chen TW, Hu A et al (2014) Thy1GCaMP6 transgenic mice for neuronal population imaging in vivo. PLoS One 9:e108697 53. Dana H, Mohar B, Sun Y et al (2016) Sensitive red protein calcium indicators for imaging neural activity. elife 5:e12727 54. Dana H, Novak O, Guardado-Montesino M et al (2018) Thy1 transgenic mice expressing the red fluorescent calcium indicator jRGECO1a for neuronal population imaging in vivo. PLoS One 13:e0205444 55. Dana H, Sun Y, Mohar B et al (2019) Highperformance calcium sensors for imaging activity in neuronal populations and microcompartments. Nat Methods 16:649–657 56. Fosque BF, Sun Y, Dana H et al (2015) Neural circuits. Labeling of active neural circuits in vivo with designed calcium integrators. Science 347:755–760 57. Shen Y, Dana H, Abdelfattah AS et al (2018) A genetically encoded Ca(2+) indicator based on circularly permutated sea anemone red fluorescent protein eqFP578. BMC Biol 16:9 58. Bauer AQ, Kraft AW, Baxter GA et al (2018) Effective connectivity measured using Optogenetically evoked hemodynamic signals exhibits topography distinct from resting state functional connectivity in the mouse. Cerebral Cortex (New York, NY : 1991) 28:370– 386 59. Brier LM, Landsness EC, Snyder AZ et al (2019) Separability of calcium slow waves and functional connectivity during wake, sleep, and anesthesia. Neurophotonics 6: 035002

60. Corlu A, Durduran T, Choe R et al (2003) Uniqueness and wavelength optimization in continuous-wave multispectral diffuse optical tomography. Opt Lett 28:2339–2341 61. Kraft AW, Bauer AQ, Culver JP et al (2018) Sensory deprivation after focal ischemia in mice accelerates brain remapping and improves functional recovery through Arc-dependent synaptic plasticity. Sci Transl Med 10:eaag1328 62. Ma Y, Shaik MA, Kozberg MG et al (2016) Resting-state hemodynamics are spatiotemporally coupled to synchronized and symmetric neural activity in excitatory neurons. Proc Natl Acad Sci U S A 113:E8463–E8471 63. Silasi G, Xiao D, Vanni MP et al (2016) Intact skull chronic windows for mesoscopic widefield imaging in awake mice. J Neurosci Methods 267:141–149 64. Wright PW, Brier LM, Bauer AQ et al (2017) Functional connectivity structure of cortical calcium dynamics in anesthetized and awake mice. PLoS One 12:e0185759 65. Macknik SL, Alexander RG, Caballero O et al (2019) Advanced circuit and cellular imaging methods in nonhuman primates. J Neurosci 39:8267–8274 66. Park K, Liyanage AC, Koretsky AP et al (2021) Optical imaging of stimulationevoked cortical activity using GCaMP6f and jRGECO1a. Quant Imaging Med Surg 11: 998 67. Ye L, Haroon MA, Salinas A et al (2017) Comparison of GCaMP3 and GCaMP6f for studying astrocyte Ca2+ dynamics in the awake mouse brain. PLoS One 12:e0181113 68. Akerboom J, Carreras Caldero´n N, Tian L et al (2013) Genetically encoded calcium indicators for multi-color neural activity imaging and combination with optogenetics. Front Mol Neurosci 6:2 69. Hires SA, Tian L, Looger LL (2008) Reporting neural activity with genetically encoded calcium indicators. Brain Cell Biol 36:69–86 70. Mank M, Santos AF, Direnberger S et al (2008) A genetically encoded calcium indicator for chronic in vivo two-photon imaging. Nat Methods 5:805–811 71. Mao T, O’connor DH, Scheuss V et al (2008) Characterization and subcellular targeting of GCaMP-type genetically-encoded calcium indicators. PLoS One 3:e1796 72. Tian L, Hires SA, Looger LL (2012) Imaging neuronal activity with genetically encoded calcium indicators. Cold Spring Harbor Protocols 2012:pdb top069609

Wide-Field Optical Imaging of Ischemic Stroke 73. De Ryck M, Van Reempts J, Duytschaever H et al (1992) Neocortical localization of tactile/proprioceptive limb placing reactions in the rat. Brain Res 573:44–60 74. Johansson BB, Ohlsson A-L (1996) Environment, social interaction, and physical activity as determinants of functional outcome after cerebral infarction in the rat. Exp Neurol 139: 322–327 75. Madinier A, Quattromani MJ, Sjo¨lund C et al (2014) Enriched housing enhances recovery of limb placement ability and reduces aggrecan-containing perineuronal nets in the rat somatosensory cortex after experimental stroke. PLoS One 9:e93121 76. Lay CC, Frostig RD (2014) Complete protection from impending stroke following permanent middle cerebral artery occlusion in awake, behaving rats. Eur J Neurosci 40: 3413–3421 77. Bero AW, Bauer AQ, Stewart FR et al (2012) Bidirectional relationship between functional connectivity and amyloid-beta deposition in mouse brain. J Neurosci 32:4334–4340 78. Power JD, Schlaggar BL, Petersen SE (2015) Recent progress and outstanding issues in motion correction in resting state fMRI. NeuroImage 105:536–551 79. Bergonzi KM, Bauer AQ, Wright PW et al (2015) Mapping functional connectivity using cerebral blood flow in the mouse brain. J Cereb Blood Flow Metab 35:367– 370 80. Goldey GJ, Roumis DK, Glickfeld LL et al (2014) Removable cranial windows for longterm imaging in awake mice. Nat Protoc 9: 2515–2538 81. Ren C, Komiyama T (2021) Wide-field calcium imaging of cortex-wide activity in awake, head-fixed mice. STAR Protoc 2:100973 82. Mightex In:Mightex, p Commercial Optical Imaging Hardware. 83. Scimedia In:SciMedia pCommercial Optical Imaging Hardware. 84. Stopczynski A, Stahlhut C, Larsen JE et al (2014) The smartphone brain scanner: a portable real-time neuroimaging system. PLoS One 9:e86733 85. Zhang X, Landsness EC, Chen W et al (2022) Automated sleep state classification of widefield calcium imaging data via multiplex visibility graphs and deep learning. J Neurosci Methods 366:109421 86. Desjardins M, Kılıc¸ K, Thunemann M et al (2019) Awake mouse imaging: from two-photon microscopy to blood oxygen level–dependent functional magnetic

149

resonance imaging. Biol Psychiatry Cogn Neurosci Neuroimag 4:533–542 87. Dinh TNA, Jung WB, Shim H-J et al (2021) Characteristics of fMRI responses to visual stimulation in anesthetized vs. awake mice. NeuroImage 226:117542 88. Drew PJ, Shih AY, Driscoll JD et al (2010) Chronic optical access through a polished and reinforced thinned skull. Nat Methods 7: 981–984 89. Gao Y-R, Drew PJ (2016) Effects of voluntary locomotion and calcitonin gene-related peptide on the dynamics of single dural vessels in awake mice. J Neurosci 36:2503–2516 90. Henckens MJ, Van Der Marel K, Van Der Toorn A et al (2015) Stress-induced alterations in large-scale functional networks of the rodent brain. NeuroImage 105:312–322 91. Krawchuk MB, Ruff CF, Yang X et al (2020) Optogenetic assessment of VIP, PV, SOM and NOS inhibitory neuron activity and cerebral blood flow regulation in mouse somatosensory cortex. J Cereb Blood Flow Metab 40:1427–1440 92. Mandino F, Cerri DH, Garin CM et al (2019) Animal functional magnetic resonance imaging: trends and path toward standardization. Front Neuroinform 13:78 93. Martin C, Martindale J, Berwick J et al (2006) Investigating neural-hemodynamic coupling and the hemodynamic response function in the awake rat. NeuroImage 32:33–48 94. Michelson NJ, Vazquez AL, Eles JR et al (2018) Multi-scale, multi-modal analysis uncovers complex relationship at the brain tissue-implant neural interface: new emphasis on the biological interface. J Neural Eng 15: 033001 95. Russo G, Helluy X, Behroozi M et al (2021) Gradual restraint habituation for awake functional magnetic resonance imaging combined with a sparse imaging paradigm reduces motion artifacts and stress levels in rodents. Front Neurosci-Switz 15:805679 96. Sencan ¸ I˙, Esipova T, Kılıc¸ K et al (2020) Optical measurement of microvascular oxygenation and blood flow responses in awake mouse cortex during functional activation. J Cereb Blood Flow Metab. 0271678X20928011 97. Shih AY, Drew PJ, Kleinfeld D (2014) Imaging vasodynamics in the awake mouse brain with two-photon microscopy. In: Neurovascular coupling methods. Springer, pp 55–73 98. Sunil S, Erdener SE, Lee BS et al (2020) Awake chronic mouse model of targeted pial

150

Jonah A. Padawer-Curry et al.

vessel occlusion via photothrombosis. Neurophotonics 7:015005 99. Turner KL, Gheres KW, Proctor EA et al (2020) Neurovascular coupling and bilateral connectivity during NREM and REM sleep. elife 9:e62071 100. Yang Q, Vazquez AL, Cui XT (2021) Longterm in vivo two-photon imaging of the neuroinflammatory response to intracortical implants and micro-vessel disruptions in awake mice. Biomaterials 276:121060 101. Siegel JS, Ramsey LE, Snyder AZ et al (2016) Disruptions of network connectivity predict impairment in multiple behavioral domains after stroke. Proc Natl Acad Sci U S A 113: E4367–E4376 102. Biswal B, Yetkin FZ, Haughton VM et al (1995) Functional connectivity in the motor cortex of resting human brain using echoplanar MRI. Magn Reson Med 34:537–541 103. Raichle ME (2015) The restless brain: how intrinsic activity organizes brain function. Philos Trans R Soc Lond Ser B Biol Sci 370: 20140172 104. Beckmann CF, Deluca M, Devlin JT et al (2005) Investigations into resting-state connectivity using independent component analysis. Phil Trans R Soc London Series B Biol Sci 360:1001–1013 105. Seeley WW, Menon V, Schatzberg AF et al (2007) Dissociable intrinsic connectivity networks for salience processing and executive control. J Neurosci 27:2349–2356 106. Eggebrecht AT, Ferradal SL, RobichauxViehoever A et al (2014) Mapping distributed brain function and networks with diffuse optical tomography. Nat Photonics 8:448–454 107. Shehzad Z, Kelly AM, Reiss PT et al (2009) The resting brain: unconstrained yet reliable. Cereb Cortex 19:2209–2229 108. Laumann TO, Gordon EM, Adeyemo B et al (2015) Functional system and areal organization of a highly sampled individual human brain. Neuron 87:657–670 109. Power JD, Cohen AL, Nelson SM et al (2011) Functional network organization of the human brain. Neuron 72:665–678 110. Smith SM, Fox PT, Miller KL et al (2009) Correspondence of the brain’s functional architecture during activation and rest. Proc Natl Acad Sci 106:13040–13045 111. Lim DH, Ledue JM, Mohajerani MH et al (2014) Optogenetic mapping after stroke reveals network-wide scaling of functional connections and heterogeneous recovery of the peri-infarct. J Neurosci 34:16455–16466

112. Assenza G, Zappasodi F, Pasqualetti P et al (2013) A contralesional EEG power increase mediated by interhemispheric disconnection provides negative prognosis in acute stroke. Restor Neurol Neurosci 31:177–188 113. Fanciullacci C, Bertolucci F, Lamola G et al (2017) Delta power is higher and more symmetrical in ischemic stroke patients with cortical involvement. Front Hum Neurosci 11: 385 114. Ramanathan DS, Guo L, Gulati T et al (2018) Low-frequency cortical activity is a neuromodulatory target that tracks recovery after stroke. Nat Med 24:1257–1267 115. Biswal B, Zerrin Yetkin F, Haughton VM et al (1995) Functional connectivity in the motor cortex of resting human brain using echoplanar MRI. Magn Reson Med 34:537–541 116. Fox MD, Raichle ME (2007) Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging. Nat Rev Neurosci 8:700–711 117. White BR, Snyder AZ, Cohen AL et al (2009) Resting-state functional connectivity in the human brain revealed with diffuse optical tomography. NeuroImage 47:148–156 118. Buxton RB, Uludag K, Dubowitz DJ et al (2004) Modeling the hemodynamic response to brain activation. NeuroImage 23(Suppl 1): S220–S233 119. Ciuciu P, Poline JB, Marrelec G et al (2003) Unsupervised robust nonparametric estimation of the hemodynamic response function for any fMRI experiment. IEEE Trans Med Imaging 22:1235–1251 120. Gossl C, Fahrmeir L, Auer DP (2001) Bayesian modeling of the hemodynamic response function in BOLD fMRI. NeuroImage 14: 140–148 121. Hossein-Zadeh GA, Ardekani BA, SoltanianZadeh H (2003) A signal subspace approach for modeling the hemodynamic response function in fMRI. Magn Reson Imaging 21: 835–843 122. Kennerley AJ, Berwick J, Martindale J et al (2005) Concurrent fMRI and optical measures for the investigation of the hemodynamic response function. Magn Reson Med 54:354–365 123. Lindquist MA, Wager TD (2007) Validity and power in hemodynamic response modeling: a comparison study and a new approach. Hum Brain Mapp 28:764–784 124. Rajapakse JC, Kruggel F, Maisog JM et al (1998) Modeling hemodynamic response for analysis of functional MRI time-series. Hum Brain Mapp 6:283–300

Wide-Field Optical Imaging of Ischemic Stroke 125. Uga M, Dan I, Sano T et al (2014) Optimizing the general linear model for functional near-infrared spectroscopy: an adaptive hemodynamic response function approach. Neurophotonics 1:015004 126. Marschner JA, Schafer H, Holderied A et al (2016) Optimizing mouse surgery with online rectal temperature monitoring and preoperative heat supply. Effects on postischemic acute kidney injury. PLoS One 11: e0149489 127. Valley MT, Moore MG, Zhuang J et al (2020) Separation of hemodynamic signals from GCaMP fluorescence measured with widefield imaging. J Neurophysiol 123:356–366 128. Ma Y, Shaik MA, Kozberg MG et al (2016) Resting-state hemodynamics are spatiotemporally coupled to synchronized and symmetric neural activity in excitatory neurons. Proc Natl Acad Sci 113:E8463–E8471 129. Fox MD, Snyder AZ, Vincent JL et al (2005) The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci U S A 102:9673– 9678 130. Fox MD, Zhang D, Snyder AZ et al (2009) The global signal and observed anticorrelated resting state brain networks. J Neurophysiol 101:3270–3283 131. Kura S, Xie HY, Fu BY et al (2018) Intrinsic optical signal imaging of the blood volume changes is sufficient for mapping the resting state functional connectivity in the rodent cortex. J Neural Eng 15:035003 132. White BR, Padawer-Curry JA, Ko T et al (2021) Wavelength censoring for spectroscopy in optical functional neuroimaging. Phys Med Biol 66:065026 133. Brendel B, Nielsen T (2009) Selection of optimal wavelengths for spectral reconstruction in diffuse optical tomography. J Biomed Opt 14:034041 134. Pouratian N, Cannestra AF, Martin NA et al (2002) Intraoperative optical intrinsic signal imaging: a clinical tool for functional brain mapping. Neurosurg Focus 13:e1 135. White BR, Bauer AQ, Snyder AZ et al (2011) Imaging of functional connectivity in the mouse brain. PLoS One 6:e16322 136. Cramer JV, Gesierich B, Roth S et al (2019) In vivo widefield calcium imaging of the mouse cortex for analysis of network connectivity in health and brain disease. NeuroImage 199:570–584

151

137. Harrison TC, Sigler A, Murphy TH (2009) Simple and cost-effective hardware and software for functional brain mapping using intrinsic optical signal imaging. J Neurosci Methods 182:211–218 138. Mcvey BFP, Prabakar S, Gooding JJ et al (2017) Solution synthesis, surface passivation, optical properties, biomedical applications, and cytotoxicity of silicon and germanium nanocrystals. ChemPlusChem 82:60–73 139. Grinvald A, Frostig RD, Lieke E et al (1988) Optical imaging of neuronal activity. Physiol Rev 68:1285–1366 140. Morone KA, Neimat JS, Roe AW et al (2017) Review of functional and clinical relevance of intrinsic signal optical imaging in human brain mapping. Neurophotonics 4:031220 141. Juavinett AL, Nauhaus I, Garrett ME et al (2017) Automated identification of mouse visual areas with intrinsic signal imaging. Nat Protoc 12:32–43 142. Padawer-Curry JA, Jahnavi J, Breimann JS et al (2021) Variability in atlas registration of optical intrinsic signal imaging and its effect on functional connectivity analysis. J Opt Soc Am A 38:245–252 143. Xiao D, Forys BJ, Vanni MP et al (2021) MesoNet allows automated scaling and segmentation of mouse mesoscale cortical maps using machine learning. Nat Commun 12:1– 13 144. White BR, Padawer-Curry JA, Cohen AS et al (2019) Brain segmentation, spatial censoring, and averaging techniques for optical functional connectivity imaging in mice. Biomed Opt Express 10(11):5952–5973 145. Boas D, Strangman G, Culver J et al (2003) Can the cerebral metabolic rate of oxygen be estimated with near-infrared spectroscopy? Phys Med Biol 48:2405 146. Schweiger M, Arridge SR, Delpy DT (1993) Application of the finite-element method for the forward and inverse models in optical tomography. J Math Imag Vision 3:263–283 147. Prahl S (1999) Tabulated molar extinction coefficient for hemoglobin in water. http:// omlc.ogi.edu/spectra/hemoglobin/sum mary.html 148. Vanni MP, Chan AW, Balbi M et al (2017) Mesoscale mapping of mouse cortex reveals frequency-dependent cycling between distinct macroscale functional modules. J Neurosci 37:7513–7533

Chapter 12 Post-mortem Magnetic Resonance Imaging of Degenerating and Reorganizing White Matter in Post-stroke Rodent Brain Vera H. Wielenga, Rick M. Dijkhuizen, and Annette Van der Toorn Abstract Magnetic resonance imaging (MRI) allows noninvasive and non-destructive imaging of brain tissue. More specifically, the status of white matter fibers can be measured with diffusion-weighted MRI, enabling assessment of structural degeneration or remodeling of white matter tracts in diseased brain. Here, we describe the preparation of post-stroke rodent brain samples for post-mortem high-resolution 3D diffusion-weighted MR imaging, accompanied with guidelines for acquiring and processing the images. Key words Diffusion-weighted MRI, Rodent brain, Neuroimaging, Microstructure

1

Introduction Magnetic resonance imaging (MRI) has evolved tremendously since its first application by Mansfield [1] and Lauterbur [2] and presently also enables the non-destructive imaging of tissues at high spatial resolution. Post-mortem high-resolution MRI can generate tissue images at spatial resolutions as low as 21.5 μm and has therefore also been termed MR histology [3, 4]. Diffusionweighted MR imaging (DWI) is based on the measurement of the velocity by which protons diffuse in a gradient field [5] and can be used to assess the diffusion of water in biological tissues [6–8]. In biological tissue, water diffusion is often hindered or restricted by the local micro-environment. For example, in brain tissue, myelin sheets surrounding the nerve fibers, or axons, in white matter form a (directional) barrier, causing water to diffuse more freely along the axonal direction compared to its perpendicular directions. Based on these characteristics, quantitative diffusion-weighted imaging (qDWI) allows tracking of white matter fibers and estimation of their integrity [8, 9]. Many brain disorders, including stroke, are accompanied by direct or indirect axonal degeneration, which may be followed by

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_12, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

153

154

Vera H. Wielenga et al.

structural remodeling. Post-mortem high-resolution DWI of the rodent brain after experimental stroke can reveal microstructural alterations reflecting enduring white matter reorganization [10, 11]. While in vivo DWI is typically limited by scan time, post-mortem DWI can be executed over days as long as the sample temperature is kept constant. Longer scanning times allow the acquisition of higher-quality data and thus more accurate and specific analysis of small structures. This chapter describes the procedures to prepare a rodent brain for post-mortem high-resolution MRI. A major advantage of the described method is that the brain tissue remains intact and can be used afterward for additional histological analyses. Guidelines for the acquisition of diffusion-weighted MR images are presented. These guidelines can be adjusted based on the research question at hand. Some typical considerations are described in the Notes. Additionally, a general post-processing pipeline for the acquired data is described that can be adapted to specific research requirements.

2

Materials

2.1 Transcardial Perfusion–Fixation

1. Peristaltic pump with two input lines (one for each solution; can be fitted with a tap) and an output line with an attached needle (rat, blunt, 19G; mouse, sharp, 25G) (manual control, variable speed, medium flow (minimal range, 5–50 ml/min), e.g., VWR product number 70730-062). 2. 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (see Notes 1 and 2). 3. Phosphate-buffered saline (PBS) with 0.05% sodium azide. 4. 0.9% saline solution (NaCl) (see Note 3). 5. Two Kocher clamps. 6. Dissecting scissors (straight, blunt tip). 7. Cardiovascular scissors (curved, blunt tip). 8. Thumb forceps (rat tooth tip). 9. Isoflurane.

2.2 Preparation of Sample for Scanning

1. PBS with 0.05% sodium azide. 2. Perfluoropolyether (PFPE) fluid (e.g., Galden® D 05 (Solvay) or Fomblin® (Solvay Solexis)). 3. Sample container: id 19.8 mm, od 21 mm, height 47.3 mm (including lid of 5 mm thickness). 4. Vacuum desiccator.

Post-Mortem MRI of Rodent Brain

155

Fig. 1 Mouse brains (inside the skull) prepared for MRI (a); in the lower brain a region is darker because blood remained in stroke-damaged tissue. Three mouse brains inserted in the container (b–e). Rat brain (inside the skull) prepared for MRI (f) and inserted in the container (g–i), note the tubing to fix the brain during scanning

Fig. 2 RF coil used for post-mortem imaging of rodent brains. The coil is fixed in a cradle which narrowly fits the bore of the gradient. A sliding wedge can be turned up or down to fix the cradle in the gradient. The rods coming out at the right side are used for fixing the wedge, tuning and matching the RF coil, and inserting the cradle 2.3

MR Scanning

1. MR system. 2. High-strength gradient system (e.g., going up to 1000 mT/m). 3. Container with sample (Fig. 1). 4. RF coil, for example, vertically oriented three-turn solenoid id 26 mm (Fig. 2).

3

Methods

3.1 Transcardial Perfusion–Fixation of Rat and Mouse Brains

The perfusion–fixation procedure is similar for mice and rats. The following protocol applies to both species and specifies differences where applicable: 1. Prepare a bottle with 4% PFA (in phosphate buffer) and a bottle with 0.9% NaCl. Both solutions should be ice-cold. Flush and fill the lines to and from the peristaltic pump. Check if large air

156

Vera H. Wielenga et al.

bubbles have been flushed from the lines, so that no air is pumped into the vascular system (see Note 4). 2. Anesthetize the animal with an overdose of isoflurane. 3. When the animal is heavily anesthetized, but still alive (breathing, heart beating), make an incision in the abdomen. Cut open the diaphragm, without damaging any organs (specifically lungs, liver and heart), to gain access to the heart. 4. Put a Kocher clamp on the sternum, and cut the rib cage on both sides, so the ribcage can be folded open. Cut the ribcage far enough so that the weight of the clamp is enough to keep it open, and both hands are free. Now the heart and aorta are accessible. If the aorta is not visible, move any obstructing tissue out of the way. 5. Put the needle (connected to the output line of the peristaltic pump) in the apex of the heart and insert it until it becomes visible through the vessel wall in the bow of the aorta just above the heart. For mice: Pull back a little because the vessels are fragile and can rupture. For rats: Fix the heart with a Kocher clamp at the level of the ventricles before insertion of the needle. When the needle is correctly positioned in the aorta, clamp the heart to fix the needle in place. 6. Switch the pump on to start perfusion, and cut the right atrium with scissors to allow the blood to leave the system. Start with 0.9% NaCl to clear the blood from the system (mice, 15 ml, 5 ml/min; rats, 100 ml, 40 ml/min) (see Note 5). 7. After flushing the vessels with saline, turn off the pump and switch the input to ice-cold 4% PFA. Switch the pump on again to resume perfusion (mice, 30 ml, 5 ml/min; rats, 250 ml, 35 ml/min) (see Note 6). 8. After 30 ml (for mice) or 250 ml (for rats), switch the pump off. Remove the needle from the heart. The animal should be completely fixed and stiff. 9. For post-mortem MRI of the brain, remove the head from the body via cervical decapitation. Clean the skull by removing skin, muscle, eyes, and the lower jaw (see Note 7). 10. Trim the bone from the nose until the olfactory bulb is reached. 11. Leave the brain in 4% PFA for post-fixation. 12. After 7 days, move the brain to PBS with 0.05% sodium azide at 4 °C for further storage. After a minimum of 5 days in PBS with sodium azide, post-mortem MRI can be performed (see Note 8).

Post-Mortem MRI of Rodent Brain

3.2 Preparation of Sample for Scanning

157

1. Place the brain (in the skull) in the container it will be scanned in (see Note 9). In the case of mouse brains, three samples fit into the container, and the container is filled at the bottom with some Delrin® poly-acetal discs to position the brains at the center of the coil. Plastic sheets with markings at the top divide the mouse brains, and the markings at the top are used to keep track of the brain positions. Only one rat brain can fit into the container and can be fixed in place with a piece of tubing on the superior part of the skull (Fig. 1). 2. Fill the container with PBS with 0.05% sodium azide. Make sure the brain is completely immersed in PBS. Add some marker (e.g., a piece of tubing with remaining PBS that remains visible on the MR images) to distinguish the right and left sides of the sample. A marker is also inserted with the mouse brains to keep track of the positions of the mouse brains in the scanner. 3. Place the container in a vacuum desiccator. Turn it on and leave the brains at a pressure of -0.9 bar for 10 min. You will see bubbles appearing on the skull and bubbling to the surface as air is being pulled from the sample. 4. Turn the vacuum desiccator off and slowly let the pressure return to normal. 5. Let the air bubbles dissolve in the PBS overnight at 4 °C. 6. On the day of MRI, replace the PBS with PFPE fluid by injecting it into the container with a syringe. As PFPE fluid is denser than PBS, it will naturally accumulate at the bottom and push PBS out at the top. Add PFPE fluid until the skull is completely immersed with it. Make sure the skull is not exposed to air during the process (see Note 10). 7. Remove any residual PBS at the top with a tissue or syringe. Also check the sample for residual air bubbles, and if necessary, try to remove them with a syringe. 8. Place the lid on the container (see Note 11).

3.3

MR Scanning

1. Insert the sample in the RF coil (Fig. 2) and position both in the center of the magnet. The field of view (FOV) will be very restricted so a few millimeters off-center can have a significant effect (see Note 12). 2. Determine and set the resonance frequency (see Note 13). 3. Obtain an initial scout or survey image to confirm the correct position of the brain. 4. Optimize pulse power (see Note 13). 5. Shim the sample. Scanner vendors usually offer multiple shim methods both for localized shimming and for whole volume

158

Vera H. Wielenga et al.

(global) shimming. In the case of post-mortem whole brain scanning, global shimming is sufficient. In cases where only a small region is considered, local shimming may provide a better homogeneity in that region and improve the resulting images. At 9.4 T the rat brains are usually shimmed to a linewidth at half height of about 40–55 Hz for the full sample. Three mouse brains together in one container can be shimmed to a linewidth at half height of 55–80 Hz. 6. Diffusion-weighted imaging can be executed with different sequences, for example, a 3D diffusion-weighted spin-echo EPI sequence with half-sine diffusion encoding gradient pulses around the 180° RF pulse (see Note 14). As the number of diffusion-weighted directions, number of shells (b-values), and height of the b-value(s) dictate the type and reliability of the analysis [6, 12, 13], a highly versatile protocol that allows for diffusion tensor imaging (DTI), diffusion kurtosis imaging (DKI), and constrained spherical deconvolution (CSD) analysis is described here (see Note 15), where diffusion-weighted directions were increasingly divided over the four b-values (10 × 2500, 20 × 5000, 20 × 7500, 60 × 10,000) and acquired in random order with a b = 0 image after every ten images (see Note 16). Other sequence parameters and options for rat and mouse brains are shown in Table 1. Some typical imaging results are shown in Fig. 3 for rats and Fig. 4 for mice. 7. When using an EPI sequence, which is sensitive to susceptibility-induced geometric distortions, it is useful to include reversed phase-encoding image(s) in the data to be able to correct these distortions during pre-processing [14, 15]. For example, one extra b = 0 image with reversed phase-encoding can be added at the end of the acquisition (for 1 b = 0 pair) (see Note 17). 8. Images should be comparable to the images shown in Fig. 3a–d for rats and to Fig. 4a–d for mice. 3.4 Analysis of Diffusion MRI Data 3.4.1

Pre-processing

As diffusion MRI is inherently sensitive to various types of imaging artifacts, which can affect further analysis, certain pre-processing steps are necessary. The DESIGNER pipeline is a standardized pipeline, which contains the most important pre-processing steps for diffusion MRI. Below we will briefly summarize each step. For further details, consult the paper by Ades-Aron et al. [16], which also contains a link to the DESIGNER Github page (see Note 18). 1. First, all diffusion-weighted images should be exported for further processing. All MR vendors allow export of the data as DICOM files. These files contain usually all needed information on the scanning parameters (see Note 19).

Post-Mortem MRI of Rodent Brain

159

Table 1 Relevant scan parameters for diffusion weighted image protocols regularly used at 9.4 T used imaging sequences Scan

DWI

DWI

DWI

Animal

Rat

Rat

Mouse

Sequence

DW SEEPI

DW SEEPI

DW SEEPI

FOV (mm)

33 × 19.2 × 16.5

35.1 × 19.2 × 16.5

24 × 21.6 × 20

Matrix size

220 × 128 × 108

234 × 128 × 110

160 × 144 × 140

Resolution (mm)

150 × 150 × 150

150 × 150 × 150

150 × 150 × 150

TR (ms)

500

500

500

TE (ms)

32.4

33.5

33

Number of shots

8

8

8

Averages

1

1

1

Gdiff,max (mT/m)

77

90

78.2

δ/Δ (ms)

4/32.4

5.5/15.5

5.5/8.83

5

14

6

3842/60

2500/10

1000/30

5000/20

2500/30

7500/20

4000/30

Number of b0 values -3

b values (10e

2

s/mm )/ nr of directions

10,000/60 Total time (h)

7.8

15.2

14.9

Abbreviations: DWI diffusion-weighted imaging, DW SEEPI diffusion-weighted spin-echo echo planar imaging, FOV field of view, TR repetition time, TE echo time, Gdiff,max maximum diffusion gradient strength, δ length of diffusion gradient, Δ time between diffusion gradient pair

2. Then, preferably, the data should be sorted by b-value and stored in one big data file containing all b-values and all diffusion-gradient directions. For example, the software package MRtrix offers the data type .mif, in which the DWI data, bvalues, and diffusion-gradient directions can be combined in one file. However, it is also possible to work with a separate NifTI-file for the DWI data and two text files that contain a list of the b-values and b-vector directions. 3. Processing by tool of choice, processing pipelines should contain the following steps: (a) Denoising using a MP-PCA-based algorithm (see Note 20). (b) Gibbs artifact correction (see Note 21). (c) Rician bias correction (see Note 22). (d) Motion and eddy current correction (see Note 23).

160

Vera H. Wielenga et al.

Fig. 3 Axial images of a rat brain after stroke by middle cerebral artery occlusion at the right side. Diffusionweighted images with b-values of 2500, 5000, 7500, and 10,000 s/mm2 (a–d) and without b-weighting (b0, e). Maps of mean diffusivity (f) and fractional anisotropy (g) and an RGB display of the main eigenvector (h)

(e) B1 bias field correction (see Note 24). (f) Outlier detection (see Note 25) in parameter maps. (g) Model selection and fitting. To give an impression of what processed data may look like, Figs. 3 and 4 show example images of the classical DTI model. The most well-known of these are the mean diffusivity map (MD, Figs. 3f and 4e), the fractional anisotropy map (FA, a measure of the directionality of the diffusion, Figs. 3g and 4f), and maps of the principal eigenvalues and eigenvectors that can be used as input for fiber tracking algorithms. The principal eigenvector can be combined with the FA-map to create RGB-color-coded maps where the intensity of the color indicates the main orientation of the anisotropic diffusion in each voxel (Fig. 3h). 3.4.2 Registration to Rodent Brain Atlas

Whatever analysis strategy is chosen, a brain template is probably required [17]. Regions of interest (ROIs) can be manually drawn, if necessary. However, brain atlases are freely available to register commonly used ROIs to the calculated maps. For rat brain, the

Post-Mortem MRI of Rodent Brain

161

Fig. 4 Axial images of a mouse brain after stroke by a middle cerebral artery occlusion (right side), diffusionweighted denoised images with b-values of 0 (a), 1000 (b), 2500 (c), and 4000 mm2/s (d). Also maps of mean diffusivity (e, in mm2/s) and fractional anisotropy (f) are shown

Sprague Dawley brain atlas from the NeuroImaging Tools and Resources Collaboratory (NITRC) can be recommended (https://www.nitrc.org/projects/whs-sd-atlas/). For mouse brain, Janke et al. [18] published an atlas on https://imaging.org. au/AMBMC/Model. A good overview of rat brain atlases is described in Johnson et al. [19]. 3.4.3 Analysis of Diffusion Parameters

There is a wide range of possibilities for analysis of diffusion data, which are all highly specific and need optimalization dependent on the research question [6]. Single voxel or ROI analysis of the generated maps is possible, eventually combined with morphometric analyses. Also fiber tracts can be generated and analyzed for specific changes within the tracts or used to assess connectivity within the brain. A single analysis pipeline would be too specific to be applicable for most aims, and reviewing possible analysis strategies is beyond the scope of this chapter. Review papers by Martinez-Heras et al. [6] and Novikov et al. [7] provide a good starting point.

162

4

Vera H. Wielenga et al.

Notes 1. Dissolve 10.9 g of Na2HPO4 in 800 ml demi-water in an Erlenmeyer. After everything is dissolved, the pH should be between 8.5 and 9.5. Place the Erlenmeyer on a heating plate set to 60 °C and add 40 g of PFA. If all is dissolved, add 3.1 g of NaH2PO4. Set the pH to 7.4 and add demi-water until total volume is 1 L. PFA should be as fresh as possible for perfusionfixation, and not older than 7 days if stored in the fridge at 4 ° C. 2. Gd-containing compounds (Gadovist® or ProHance®) may be added to the fixative or the PBS solution. It has been used to enhance contrast and reduce T1 relaxation, allowing faster MR acquisition [4, 19–24]. In that case, repetition times (TR) can be significantly reduced, and the diffusion-weighted acquisition can speed up. 3. Researchers inexperienced with performing transcardial perfusions may benefit from adding heparin (1 ml/L) to the 0.9% NaCl solution. As the diaphragm is cut and the lungs collapse, the animal effectively starts dying. An inexperienced researcher may not be fast enough to start the perfusion before clots start forming in the blood vessels. To facilitate successful perfusion of all vessels, heparin helps to dissolve any clots. 4. During perfusion fixation, it is absolutely necessary that all blood is flushed away and no air bubbles are introduced in the brain vessels. Both cause local distortions of the B0-field during the MR experiments and will lead to artifacts, usually in the form of locally decreased or disappeared signal. In the case of air bubbles, these effects can have a larger range than the air bubbles itself (blooming effect). 5. When perfusion proceeds correctly, the liver (and all other organs) should change color as they are cleared of blood. If the liver shows a blotched pattern of brown and purple, this is an indication of incomplete blood clearance in all organs. 6. The muscles will start twitching and limbs will move as the PFA reaches every part of the vascular system. This will stop as soon as the muscles become fixed. 7. Removal of tissue outside the skull reduces inhomogeneities during post-mortem MRI. 8. Switching the brains to PBS in sodium azide causes the T1 and T2 relaxation times to increase after fixation with paraformaldehyde, which shortens these times. The relaxation time will more closely resemble those of in vivo tissue. 9. Brains remain in their skulls during preparation and scanning. The main advantage of this is that the brain is less prone to

Post-Mortem MRI of Rodent Brain

163

damage during the preparation and does not deform when fixed in the container. The brains can also be stabilized in agarose gel. The disadvantage is that a large amount of signal remains present outside the tissue of interest making it necessary to increase the FOV or to add compounds to the outside fluid to reduce its T1 relaxation times. If necessary, fix the brain in the container with non-magnetic and non-MRI visible materials (e.g., pieces of plastic or tubing) so that the brain will not be able to move during MRI, because fast gradient switching may cause vibrations. 10. PFPE fluid replaces the outer volume in the samples. These are fluorinated compounds that are comparable to oil. Because these compounds are heavier than water, water will be replaced by the compound, unless the space between the container and the PBS is very small (e.g., when discs are inserted in the container to increase the sample height). In that case it is useful to fill the bottom space with a small amount of PFPE fluid before filling with PBS. No air bubbles will form by the exchange of water with PFPE fluid. Air bubbles cause local B0 inhomogeneities in the sample that will affect the quality of the scan. Both PFPE fluids have electrical properties that prevent local B0 inhomogeneities to occur at the transition area between tissue and the outer fluid, leading to improved image quality. To our knowledge, immersion in PFPE fluid does not affect later histology. It can be easily removed after the experiment. 11. It is useful to have a hole in the top of your container (that can be closed) to be able to replace residual air bubbles at the top of the sample after placing the lid with PFPE fluid via syringe. 12. The sample and RF coil allow insertion into the magnet of both in a top-down orientation. However, either the shape or diameter of the RF coil or the bore size may make it necessary to orient the sample in the direction of the main magnetic field. In that case the container may leak PFPE fluid very easily because it is a liquid used for greasing. In addition, because diffusion is temperature dependent, it is recommended to arrange temperature regulation to keep the sample temperature constant. Air blowing in at a constant temperature may be sufficient. 13. Methods to determine the resonance frequency and pulse powers vary between magnet vendors and MR laboratories. Often these parameters are set automatically. However, it is good practice to check these values, and they should remain similar between experiments. 14. Here a multi-shot spin-echo EPI acquisition is used to reduce the acquisition time per volume [10, 11, 25, 26]. However, in other studies [4, 19–21, 23, 24, 27], a regular spin-echo

164

Vera H. Wielenga et al.

diffusion-weighted sequence with or without compressed sensing is used. In these cases, a Gd chelate was added as a T1 shortening agent allowing the repetition time to be reduced to less than 100 ms and thus shortening the acquisition time per volume. Gd chelates also shorten T2*, increasing the amount of EPI induced artifacts if EPI acquisition is used in combination with Gd-induced relaxation shortening. Another way to decrease the total acquisition time per volume is by using a fast spin echo (FSE) sequence with a diffusion-weighted preparation. However, the power deposition of multiple 180° RF pulses may cause heating of the sample, making this a less likely solution for long post-mortem acquisitions. An important parameter that affects the signal-to-noise ratio and is easy to adjust is the voxel volume. It is recommended for fiber tracking algorithms to keep the voxel dimensions isotropic. However, small steps in voxel dimension result in volume changes to the third power. Signal-to-noise increases only with the square root of the volume. Therefore, doubling the pixel dimension in all three directions will lead to an eightfold increase pffiffiffi of the signal from the voxel volume, but results in only a 2 2-fold increase of the SNR. However, increasing the voxel volume will require less phase-encoding steps, which speeds up the total acquisition time. Similarly, decreasing the voxel volume causes large increases in acquisition time in a comparable way. These considerations are at the root of the choice to acquire post-mortem data with an isotropic spatial resolution of 150 μm. Therefore, optimal acquisition schemes and voxel volumes are very dependent on the equipment used in each laboratory. 15. The current consensus is that a minimum of 60 diffusionweighted directions is needed for CSD analysis and a b-value of 3000 s/mm2 in vivo [28]. However, the apparent diffusion coefficient changes in post-mortem (fixed) tissue compared to in vivo tissue need to be considered [29, 30]. Therefore, the bvalue needs to increase by a factor of 3–4 compared to in vivo values to obtain a similar angular contrast. These considerations require acquisition with b = 10,000 as the highest b-value (comparable to b = 2500–3300 in vivo). However, for other analyses, such high b-values are not always necessary, and bvalues of 2500 and 5000 may suffice (respectively, 600–800 and 1200–1700 in vivo) for DTI analysis. Table 1 also shows acquisitions that perform well with lower b-values and less angular directions. 16. The diffusion gradients, especially at high b-values, may increase the sample temperature. The measured diffusion is temperature dependent. Therefore, measures should be taken to keep the sample temperature constant. Randomizing b-

Post-Mortem MRI of Rodent Brain

165

values allows the system to cool down regularly (making temperature regulation easier) and reduces long-term effects on the measurement caused by external circumstances (such as the day–night rhythm). Additionally, interleaving b0-images can also be used to detect (and possibly correct) signal drift during long experiments. 17. To correct for susceptibility-induced geometric distortions, one could opt to acquire all images with normal and reversed phase encoding. This doubles the total acquisition time, unless diffusion-weighted directions are halved to maintain the same total acquisition time. Another (more time-conservative) option is to simply acquire one or more b = 0 pairs, where half of the volumes have the same phase encoding as the other diffusion-weighted images, and the other half has the opposite phase-encoding direction. If no reversed phase-encoding images are acquired, the geometric distortions can be (partially) resolved by registration to an undistorted image. 18. There are multiple software packages for analysis of diffusion MRI data (e.g., ExploreDTI [31] (http://exploredti.com/); FSL (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/); DSI Studio (https://dsi-studio.labsolver.org/), especially in case of tractography-based analysis; and MRtrix3 [32] (https://www. mrtrix.org/), which particularly specializes in constrained spherical deconvolution (CSD)-based tractography and fixelbased analysis). All contain a wide range of necessary functions such as motion and eddy current correction, calculating DTI metrics, and creating functional connectomes. A hitchhiker’s guide to DTI that describes many analysis stages is published by Soares et al. [33]. 19. Many differences in data format are present between vendors [34]. A versatile data format is the NifTI format (Neuroimaging Informatics Technology, https://nifti.nimh.nih.gov/), which is used in many pipelines as input format. Dicom format can be converted with relative ease to NifTI format if necessary. Multiple tools are available, look, e.g., at https:// opensourcelibs.com/lib/dcm2niix. 20. The use of a denoising algorithm improves the image quality and the accuracy of diffusion parameter estimation. This should always be performed as the first step, as interpolation or smoothing in other processing steps will change noise characteristics, thus violating the assumptions of denoising algorithms, rendering them ineffective. This also includes zeropadding the data before Fourier transform, which therefore should be avoided. Marchenko–Pastur Principal Component Analysis (MP-PCA)-based denoising is a fast and robust method for denoising of DWI data [35].

166

Vera H. Wielenga et al.

21. The Gibbs artifact (also known as “Gibbs ringing” or “truncation artifact”) occurs as a consequence of Fourier transforming finitely sampled signals in the MR image. The Gibbs artifact is especially pronounced in regions with high-contrast boundaries (e.g., between cerebrospinal fluid and gray matter) and significantly affects diffusion parameters [36]. 22. Noise in low SNR magnitude data (especially SNR < 2), such as DWI data, follows a Rician distribution rather than Gaussian, which creates a positive bias. This may affect fitting of diffusion parameters in images with high b-values, which typically have the lowest SNR [37]. 23. Diffusion MR images are acquired with spin-echo EPI sequences, which make use of a strong and rapidly switching diffusion-gradient field. These sequences cause geometric distortions (scaling and shearing) and eddy currents [38]. 24. Radiofrequency (RF) field inhomogeneities lead to a spatial bias in MR signal intensities (i.e., the B1 bias field). If left uncorrected, this bias can vary the tensor fitting within tissues of the same type and propagate major artifacts to diffusion model parameters [16]. 25. Detection of outliers in the diffusion data during the fitting process increases the accuracy of the model [16].

Acknowledgments This work was supported by the Graduate Program (022.006.001) of the Dutch Research Council (NWO). We thank Dr. Martijn Froeling of the University Medical Center Utrecht for his assistance and expert input. References 1. Mansfield P (1977) Multi-planar image formation using NMR spin-echoes. J Phys C Solid State Phys 10:L55–L58 2. Lauterbur PC (1973) Image formation by induced local interactions: examples employing nuclear magnetic resonance. Nature 242:190 3. Wang N, White LE, Qi Y et al (2020) Cytoarchitecture of the mouse brain by high resolution diffusion magnetic resonance imaging. NeuroImage 216:1–29. https://doi.org/ 10.1016/j.neuroimage.2020.116876 4. Johnson GA, Ali-Sharief A, Badea A et al (2007) High-throughput morphologic phenotyping of the mouse brain with magnetic resonance histology. NeuroImage 37:82–89.

https://doi.org/10.1016/j.neuroimage. 2007.05.013 5. Stejskal EO, Tanner JE (1965) Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J Chem Phys 42:288–292. https://doi.org/10.1063/1. 1695690 6. Martinez-Heras E, Grussu F, Prados F et al (2021) Diffusion-weighted imaging: recent advances and applications. Semin Ultrasound CT MR 42:490–506. https://doi.org/10. 1053/j.sult.2021.07.006 7. Novikov DS, Fieremans E, Jespersen SN, Kiselev VG (2019) Quantifying brain microstructure with diffusion MRI: theory and parameter

Post-Mortem MRI of Rodent Brain estimation. NMR Biomed 32:1–53. https:// doi.org/10.1002/nbm.3998 8. Johansen-Berg H, Behrens TEJ (2014) Diffusion MRI. Elsevier, San Diego 9. Zhang F, Daducci A, He Y et al (2022) Quantitative mapping of the brain’s structural connectivity using diffusion MRI tractography: a review. NeuroImage 249:118870. https://doi. org/10.1016/j.neuroimage.2021.118870 10. Umesh Rudrapatna S, Bakker CJG, Viergever MA et al (2017) Improved estimation of MR relaxation parameters using complex-valued data. Magn Reson Med 77:385–397. https:// doi.org/10.1002/mrm.26088 11. Sinke MR, Otte WM, van Meer MP et al (2018) Modified structural network backbone in the contralesional hemisphere chronically after stroke in rat brain. J Cereb Blood Flow Metab 38:1642–1653. https://doi.org/10. 1177/0271678X17713901 12. Jeurissen B, Tournier JD, Dhollander T et al (2014) Multi-tissue constrained spherical deconvolution for improved analysis of multishell diffusion MRI data. NeuroImage 103: 411–426. https://doi.org/10.1016/j. neuroimage.2014.07.061 13. Dhollander T, Clemente A, Singh M et al (2021) Fixel-based analysis of diffusion MRI: methods, applications, challenges and opportunities. NeuroImage 241:118417 14. Andersson JLR, Skare S, Ashburner J (2003) How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. NeuroImage 20: 870–888. https://doi.org/10.1016/S10538119(03)00336-7 15. Smith SM, Jenkinson M, Woolrich MW et al (2004) Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage 23:208–219. https://doi.org/ 10.1016/j.neuroimage.2004.07.051 16. Ades-Aron B, Veraart J, Kochunov P et al (2018) Evaluation of the accuracy and precision of the diffusion parameter EStImation with Gibbs and NoisE removal pipeline. NeuroImage 183:532–543. https://doi.org/10. 1016/j.neuroimage.2018.07.066 17. Hess A, Hinz R, Keliris GA, Boehm-Sturm P (2018) On the usage of brain atlases in neuroimaging research. Mol Imaging Biol 20:742– 749 18. Janke AL, Ullmann JFP (2015) Robust methods to create ex vivo minimum deformation atlases for brain mapping. Methods 73:18–26. https://doi.org/10.1016/j.ymeth.2015. 01.005

167

19. Johnson GA, Laoprasert R, Anderson RJ et al (2021) A multicontrast MR atlas of the Wistar rat brain. NeuroImage 242:118470. https:// doi.org/10.1016/j.neuroimage.2021. 118470 20. Papp EA, Leergaard TB, Calabrese E et al (2014) Waxholm Space atlas of the Sprague Dawley rat brain. NeuroImage 97:374–386. https://doi.org/10.1016/j.neuroimage. 2014.04.001 21. Johnson GA, Badea A, Brandenburg J et al (2010) Waxholm Space: an image-based reference for coordinating mouse brain research. NeuroImage 53:365–372. https://doi.org/ 10.1016/j.neuroimage.2010.06.067 22. Johnson GA, Cofer GP, Gewalt SL, Hedlund LW (2002) Special report with MR microscopy : the visible mouse 1. Radiology 222: 789–793 23. Johnson GA, Calabrese E, Badea A et al (2012) A multidimensional magnetic resonance histology atlas of the Wistar rat brain. NeuroImage 62:1848–1856. https://doi.org/10.1016/j. neuroimage.2012.05.041 24. Calabrese E, Badea A, Watson C, Johnson GA (2013) A quantitative magnetic resonance histology atlas of postnatal rat brain development with regional estimates of growth and variability. NeuroImage 71:196–206. https://doi. org/10.1016/j.neuroimage.2013.01.017 25. van Tilborg E, Achterberg EJM, van Kammen CM et al (2018) Combined fetal inflammation and postnatal hypoxia causes myelin deficits and autism-like behavior in a rat model of diffuse white matter injury. Glia 66:78–93. https://doi.org/10.1002/glia.23216 26. Vaes JEG, van Kammen CM, Trayford C et al (2021) Intranasal mesenchymal stem cell therapy to boost myelination after encephalopathy of prematurity. Glia 69:655–680. https://doi. org/10.1002/glia.23919 27. Veraart J, Leergaard TB, Antonsen BT et al (2011) Population-averaged diffusion tensor imaging atlas of the Sprague Dawley rat brain. NeuroImage 58:975–983. https://doi.org/ 10.1016/j.neuroimage.2011.06.063 28. Dell’Acqua F, Tournier J-D (2019) Modelling white matter with spherical deconvolution: how and why? NMR Biomed 32:e3945. https://doi.org/10.1002/nbm.3945 29. Bourne R, Bongers A, Charles N et al (2013) Effect of formalin fixation on biexponential modeling of diffusion decay in prostate tissue. Magn Reson Med 70:1160–1166. https://doi. org/10.1002/mrm.24549 30. D’Arceuil H, de Crespigny A (2007) The effects of brain tissue decomposition on

168

Vera H. Wielenga et al.

diffusion tensor imaging and tractography. NeuroImage 36:64–68. https://doi.org/10. 1016/j.neuroimage.2007.02.039 31. Leemans A, Jeurissen B, Sijbers J, Jones DK (2009) ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion MR data. In: Proceedings of 17th Annual Meeting of Intl Soc Mag Reson Med, p. 3527, Hawaii, USA 32. Tournier JD, Smith R, Raffelt D et al (2019) MRtrix3: a fast, flexible and open software framework for medical image processing and visualisation. NeuroImage 202:116137 33. Soares JM, Marques P, Alves V, Sousa N (2013) A hitchhiker’s guide to diffusion tensor imaging. Front Neurosci 7:1–14. https://doi. org/10.3389/fnins.2013.00031 34. Li X, Morgan PS, Ashburner J et al (2016) The first step for neuroimaging data analysis: DICOM to NIfTI conversion. J Neurosci

Methods 264:47–56. https://doi.org/10. 1016/j.jneumeth.2016.03.001 35. Veraart J, Novikov DS, Christiaens D et al (2016) Denoising of diffusion MRI using random matrix theory. NeuroImage 142:394– 406. https://doi.org/10.1016/j.neuroimage. 2016.08.016 36. Veraart J, Fieremans E, Jelescu IO et al (2016) Gibbs ringing in diffusion MRI. Magn Reson Med 76:301–314. https://doi.org/10.1002/ mrm.25866 37. Gudbjartsson H, Patz S (1995) The rician distribution of noisy MRI data. Magn Reson Med 34:910–914. https://doi.org/10.1002/mrm. 1910340618 38. Andersson JLR, Sotiropoulos SN (2016) An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. NeuroImage 125: 1063–1078. https://doi.org/10.1016/j. neuroimage.2015.10.019

Part III Methods to Identify Molecular and Immune Mechanisms Supporting Recovery

Chapter 13 Quantitative Spatial Mapping of Axons Across Cortical Regions to Assess Axonal Sprouting After Stroke Mary T. Joy, Samuel P. Bridges, and S. Thomas Carmichael Abstract Neurological disease such as a stroke causes death of brain tissue and loss of connectivity. Paradoxically, the stroke itself induces growth of new axonal collaterals, a phenomenon that is restrained in the normal adult brain. Enhancements in sprouting of axons have been linked with enhancements in motor function. Here, we describe a method developed in-house using standard reagents to map and quantitatively assess differential sprouting responses in stroke and following treatment with candidate molecular or pharmacological targets. This method allows for measurements of axonal growth responses that act as structural correlates for neural repair processes in the brain that aid in stroke recovery. Key words BDA-labeling, Stroke recovery, Axonal mapping technique, Quantitative assessment, Axonal sprouting

1

Introduction Stroke causes permanent loss of brain tissue resulting in loss of structural and functional connectivity at local and distant sites. In the days to weeks following a stroke, a molecular growth program [1, 2] is initiated that supports growth of new axonal collaterals in brain regions adjoining the stroke site. This structural growth response termed “axonal sprouting” diminishes with age [1] and can be amplified by modulating genes that underlie pathways that support synaptic plasticity [3], learning [3, 4], axon guidance [5], and growth signaling [6, 7]. While a direct causal link between newly sprouted axons and changes in behavioral function is yet to be established, a substantial body of evidence shows that enhancements in motor function are associated with differential sprouting responses across cortical areas in densities or spatial locations unique to treatments that enhance function when compared to stroke alone [1–7]. Hence, measuring these growth changes allow us to determine how potential candidate genes or drug targets alter

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_13, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

171

172

Mary T. Joy et al.

Fig. 1 Overview of methods

the inherent molecular program to initiate repair process at the structural level in the brain that support restoration of function. Here, we describe a step-by-step protocol to label axons using readily available chemical reagents, processing of tissue for histology and immunohistochemistry, and tracing of axons using a commercially available software and spatial mapping using a custom software that can be easily deployed with minimal computer hardware requirements. 1.1 Overview of Methods

2 2.1

Animals are induced with a stroke and cortical sprouting resulting from a stroke or stroke with treatment are assessed 1–3 months post stroke. A typical workflow (Fig. 1) includes stroke induction followed by biotinylated dextran amine (BDA) injection at 2 months after stroke. The brains are harvested 7 days after BDA injection. The cortex is dissected, flattened, fixed with fixative, sectioned, and histochemically stained for BDA. Sections are then imaged. Labeled axons are traced semi-automatically, and data is plotted in Cartesian coordinates with statistical significance computed.

Materials Axonal Tracer

2.2 Surgical Reagents

1. Biotinylated dextran amine (BDA)—10,000 MW (Molecular Biosciences). 1. Animals. 2. Anesthesia (isoflurane). 3. Antiseptics (70% ethanol and Betadine). 4. Analgesics (Marcaine). 5. 0.9% saline. 6. Sterile cotton swabs. 7. Eye lubricant. 8. Vetbond (3 M). 9. Sterile drapes and gloves. 10. Rose bengal, 10 mg/ml (Fisher Scientific).

Quantitative Axonal Mapping Across the Mouse Cerebral Cortex

2.3 Surgical Equipment

173

1. Stereotaxic frame connected to anesthesia dispenser (Kopf/ Sterling). 2. Microinjector (Drummond). 3. Glass needles (WPI). 4. Pipette puller (Sutter). 5. Sterilized surgical tools (Fine Sciences). 6. Light source for stroke induction (Thorlabs).

2.4 Histology Reagents

1. 4% paraformaldehyde. 2. 1× phosphate buffer saline. 3. Glass slides (FisherBrand). 4. Stainless steel hex nuts (Amazon, 18-8 Stainless Steel Plain Finish, #2-56 thread size, 5/32″ width across flats, 1/16″ thick). 5. Bone wax (Ethicon). 6. Blocking solution (20 ml 1× PBS, 0.5 ml donkey serum, 10ul Tween, 2 ul Triton-X, 0.2 mg Bovine Serum Albumin). 7. Biotin binding protein (Streptavidin) conjugated to a fluorophore (Thermo Fisher).

2.5 Microscopy and Analysis

1. Epifluorescent microscope (Nikon). 2. Neurolucida 360 (MBF). 3. R package. 4. HeatMap (custom package).

3 3.1

Methods Animal Surgery

Stroke Induction All surgeries are to be conducted using aseptic techniques and in compliance with the institution’s animal care committee. 1. Anesthetize animal with 5% isoflurane in combination with 100% oxygen. Maintain anesthesia at 1.2–2%. 2. Fix the animal’s head onto a stereotaxic frame and maintain body temperature at 37 ± 0.5 °C with a heating pad monitored through a rectal probe throughout the course of surgery. 3. Prep the animal for surgery by clipping hair from the head and sterilize skin with alternate swabs of Betadine and 70% alcohol. Repeat swabbing twice. 4. Using sterile forceps and scissors, make an incision through the midline to expose the skull.

174

Mary T. Joy et al.

5. Mark bregma and position the light source for stroke induction at 1.5 mm from bregma. Inject 10 mg/ml of filter sterilized rose bengal solution at 10 mg/kg intraperitoneally and wait for 5 min. Turn on the light source for 15 min. The resulting photochemical reaction from light exposure during this time frame induces a stroke (see Note 1). 6. Following stroke induction, apply Marcaine around the incised skin, and close the wound using tissue glue Vetbond. 7. Place the animal in a new cage, and allow the animal to recover from anesthesia. Provide antimicrobials (TMS) in drinking water for 7 days. Axonal Labeling Following 4, 8, or 12 weeks after a stroke, prep animals for injections with axonal tracers (see Note 2). 8. Repeat steps 1–4. Using bregma as reference, drill a hole into the skull that correspond to coordinates for injection of the tracer (see Note 3). 9. Back fill a pulled glass needle with mineral oil and lodge the needle into a microinjector in accordance with manufacturer’s instructions. 10. Draw a desirable volume of tracer, and inject 0.4 μl of tracer at 0.1 μl/s into the brain at a depth of 0.75 mm (for cortical injections). Following injection, wait for 5 min and slowly retract the needle (see Note 4). Repeat steps 6 and 7. 3.2

Histology

At day 7 after injection of the tracer, animals are prepped for histology. 1. Subject animal to terminal anesthesia and transcardially perfuse with 4% paraformaldehyde using standard perfusion technique. 2. Dissect out the brain (see Note 5). Keep the brain moist with PBS during subsequent steps of dissection. Using a scalpel blade, cut through the midline dividing the brain into its two hemispheres. Using a small spatula and curved forceps, lift the cortex by gaining access through its posterior end close to the cerebellum and gently lift away. Sever with forceps in the areas that are joined to its subcortical structures. Microdissect out the corpus callosum and hippocampus if dissected out with the cortex. 3. Take 4 hex nuts that are 1 mm in width, and coat these in bone wax. Place the coated hex nuts on four edges of a glass slide. The bone wax acts as an adhesive. Add a drop of PBS or 4% paraformaldehyde in the center of the slide and place the cortex. Using another glass slide, press down gently on the

Quantitative Axonal Mapping Across the Mouse Cerebral Cortex

175

cortex such that the cortex is sandwiched between two glass slides adhered by bone wax. The width of the hex nuts limits the extend of tissue compression such that the cortex is flattened to ~1 mm of thickness. 4. Place the prep in a petri dish with 4% PFA, and post fix tissue overnight at 4 °C without agitation. Following post-fixation, cryoprotect tissue by incubating in 30% sucrose for 24 h at 4 ° C. 5. Embed tissue on a cutting block with OCT (optimal cutting temperature) media and section tissue using a cryostat by slicing through tangentially at 40 μm thickness. Ensure that the tissue is perpendicular to the blade to avoid sectioning at angles. Collect one section per well of a 24-well plate in PBS or anti-freeze solution for long-term storage. Collection of one section per well allows estimation of the cortical depth from which the section has been sampled allowing for the correspondence of cortical section to cortical lamina. 6. Wash sections twice with 1× PBS for 5 min with gentle shaking. Following washes, treat sections with blocking solution for 30 min. Aspirate blocking solution and add Streptavidin conjugated to a fluorophore (e.g., 547) diluted at 1:1000 in blocking solution. Incubate at room temperature for 2 h with gentle shaking. Following incubation, wash sections 3 times with PBS and gentle shaking for 5 min for each wash. 7. Mount sections onto a glass slide (superfrost) pre-coated with gelatin. Allow sections to dry. Treat slides with increasing gradients of ethanol—50, 70, and 95% followed by dehydration with xylene. Cover slip slides with DPX as mounting media. Allow slides to dry overnight. 3.3 Microscopy and Semi-automated Axonal Tracing

Images of sections are captured with an epifluorescent microscope using StereoInvestigator from MicroBrightField (MBF). BDA-labeled axons in the acquired images are semi-automatically traced with Neurolucida 360 (MBF). 1. Images of sections are acquired using the 2D slide scanning module (MBF). This module takes multiple images under higher magnification that span the entire section, and the images are stitched to form a large image with higher resolution of the entire flattened cortical section. For detailed instructions, please see manufacturer’s guidelines (https:// www.mbfbioscience.com/help/stereo_investigator/Default. htm#Ribbons/Acquire/SlideScanWF.htm). Briefly (Fig. 2), visualize section by clicking on “live image” with low magnification (4×). Draw a contour that borders the section. Enable the scan slide module by clicking on “scan slide.” The trim and blend options are set to 1 μm and 40 pixels/section,

176

Mary T. Joy et al.

Fig. 2 Steps in slide scanning. (a) Setting trim and blend values. (b) Window showing options to display all scan sites (grids). (c) Focus map generated from selected grids by focusing with 20× objective. Once focus map has been generated, start scanning slide

Fig. 3 Semi-automated tracing of axons. (a) A field of view with labeled axons. (b) Seed placement on detected axons for tracing with Neurolucida 360

respectively. After setting trim and blend options, all scan sites are shown that appear as a grid placed over the contoured area. Select a subset of grids that will share the same focal plane at 20× magnification as its neighboring grids. Select 20–40 grids. Focus with 20× objective as the slide automatically moves to the regions that correspond to the selected grids, thus creating a focus map. Initiate scan.

2. Saved images are further processed to trace axons with Neurolucida 360 (MBF) (Fig. 3). I. Open image in Neurolucida 360. II. Set a new reference point in the center of the tracer injection. (a) Important: failure to do this will result in the reference point being in the top right corner of the image, causing all coordinates to be inaccurate.

Quantitative Axonal Mapping Across the Mouse Cerebral Cortex

177

III. Optional: change the pseudo coloring to grayscale. (a) Having the image in grayscale will make the automated detection much simpler. IV. Select contour and outline the injection site (and stroke if present). Save these measurements for further analysis. V. Zoom in to a region with axons present. VI. Select a marker type and start the “Mark Objects” workflow.

VII. Select “Use Full Image.” VIII. Set sensitivity to 75.3%. IX. Exclude objects smaller than 0.1 μm and larger than 4 μm. X. Set object color sufficient to label spots of fluorescence within the preview window. XI. Select “Separate Objects by Average Size,” and set average size to 3 μm. XII. Once all settings are confirmed, select “Mark Objects.” XIII. Note: settings need to be optimized to tracer used, but should be kept the same for all sections analyzed in an experiment. XIV. Once all markers have been placed, remove incorrectly placed markers by selecting “Select Objects.” (a) Markers are often placed within the stroke core and edges of the tissue section due to autofluorescence. (b) Important: once all incorrectly placed markers have been removed, deselect “Select Objects.” Failure to do so prior to exporting marker coordinates will result in wrong marker coordinates being included in the exported coordinates. XV. Under File, select export -> export marker coordinates, to get coordinates in a text file. XVI. Select “New File,” save the data file generated, and proceed with the next image starting at step I. 3.4

Analysis

Text files generated from Neurolucida tracings contain x and y coordinates for every traced axon with reference to the injection site. The files are opened with a custom written program—HeatMap for plotting and statistical analysis (see Note 6). 1. Save the program HeatMap (folders: HeatMap_Stats and Stats) in C drive of a PC. This program is only compatible with Windows operating system. Additionally, download the

178

Mary T. Joy et al.

statistical package R. Note: The version of R most robustly tested is R-2.15.0. 2. HeatMap expects data files in a specific format. The data files generated from Neurolucida are copied into an excel sheet and saved as a comma separated value (csv) file. The first row of the .csv file should have the following entries in each column:

. For example: dataset 2A2-3L1 treated c/d x and y coordinates are then entered in two separate columns. Data from multiple sections and animals corresponding to one treatment type can be entered into a single excel sheet where data from each section is separated by information as entered in the first column of the excel sheet. Alternatively, data can be stored in separate .csv files. 3. Open R and the executable file HeatMap Variable Injection. exe. This opens a graphical user interface where data can be loaded by clicking on “open data file” on the left panel. Subsequent data files can be opened with “append data file” (Fig. 4).

Fig. 4 HeatMap graphical user interface (GUI). (a) Layout of GUI. (b) Example dataset loaded and visualized with a three-color scatter plot. The left panel shows file information. (c) Alternate visualization of data using a polar plot. (d) Statistics tab opens binned data for computation of statistical significance

Quantitative Axonal Mapping Across the Mouse Cerebral Cortex

179

4. Data can be plotted and saved as a three-color or two-color scatter (Fig. 4). A three-color scatter is preferred to visualize projections unique to treatment vs control as well as regions of overlap. Additionally, data can be visualized in polar coordinates that captures the magnitude change in radial projections from treated vs control conditions. 5. Statistical significance is computed with Hoteling’s T squared test. To compute significance, click on “Statistics.” This displays binned data. Default values for radial and polar bins are 24 and 10, respectively. If using different values, ensure that the data is similarly binned across all conditions being assessed. Click on “copy to clipboard” and then open r_go.bat located in the Stats folder. Wait for a few minutes. Statistical values computed are stored in “results.csv” in the Stats folder.

4

Notes 1. Here we describe a protocol to assess sprouting in a photothrombotic stroke model. However, the above method is adaptable to any stroke model to assess cortical sprouting. 2. The investigator can choose from a range of axonal tracers. BDA has been robustly tested in our laboratory to anterogradely label axons. Notably, BDA can also label to a lesser extent, axons in a retrograde fashion. Other commonly used tracers include cholera toxin subunit B (CTb) for retrograde tracing. With developments in viral design, more precise directional labeling can be achieved with adeno-associated viruses. For a general overview of viral methods for beginners to trace axons, we recommend reading this blog from Addgene: https://blog. addgene.org/using-aav-for-neuronal-tracing. 3. Injection coordinates for the tracer should be determined by the investigator and the needs of the study. We routinely asses sprouting of axons from the rostral forelimb area of the cortex adjacent to the stroke to the remainder of the cortex and hence inject at a single site at 1.5–1.8 mm anterior from bregma. 4. To fairly assess differential sprouting responses, care should be taken to avoid variability in injection volume or location across animals. The mediolateral and anteroposterior coordinates of the injection site and the injection volume should be assessed across animals by measurements in tissue sections. Animals with high variability should be excluded from the study as changes in injection location or volume can produce a different pattern of sprouting, occluding the overall conclusions of the study.

180

Mary T. Joy et al.

5. We recommend practicing dissections prior to starting an experiment and watching the dissection section of this video prior to starting (https://www.jove.com/v/56992/visualiza tion-of-cortical-modules-in-flattened-mammalian-cortices). 6. A major limitation of HeatMap is its non-compatibility with other operating systems outside of Windows. Alternatively, the functions outlined in HeatMap can be scripted in Matlab or Python using standard plotting functions and the same statistical tests.

Acknowledgments This work was funded by the Miriam and Sheldon G. Adelson Foundation for Medical Research. References 1. Li S, Overman JJ, Katsman D, Kozlov SV, Donnelly CJ, Twiss JL, Giger RJ, Coppola G, Geschwind DH, Carmichael ST (2010) An age-related sprouting transcriptome provides molecular control of axonal sprouting after stroke. Nat Neurosci 13:1496–1504 2. Carmichael ST, Kathirvelu B, Schweppe CA, Nie EH (2017) Molecular, cellular and functional events in axonal sprouting after stroke. Exp Neurol 287(Pt 3):384–394 3. Caracciolo L, Marosi M, Mazzitelli J, Latifi S, Sano Y, Galvan L, Kawaguchi R, Holley S, Levine MS, Coppola G et al (2018) CREB controls cortical circuit plasticity and functional recovery after stroke. Nat Commun 9:2250 4. Joy MT et al (2019) CCR5 is a therapeutic target for recovery after stroke and traumatic brain injury. Cell 176:1143–1157.e13

5. Overman JJ, Clarkson AN, Wanner IB, Overman WT, Eckstein I, Maguire JL, Dinov ID, Toga AW, Carmichael ST (2012) A role for ephrin-A5 in axonal sprouting, recovery, and activity-dependent plasticity after stroke. Proc Natl Acad Sci U S A 109:E2230–E2239 6. Li S, Nie EH, Yin Y, Benowitz LI, Tung S, Vinters HV, Bahjat FR, Stenzel-Poore MP, Kawaguchi R, Coppola G, Carmichael ST (2015) GDF10 is a signal for axonal sprouting and functional recovery after stroke. Nat Neurosci 18:1737–1745 7. Wahl AS et al (2014) Neuronal repair. Asynchronous therapy restores motor control by rewiring of the rat corticospinal tract after stroke. Science 344:1250–1255

Chapter 14 Quantitative Evaluation of Cerebral Microhemorrhages in the Mouse Brain Rudy Chang and Rachita K. Sumbria Abstract Cerebral microhemorrhages are microscopic bleeds in the brain parenchyma and are the pathological substrates of cerebral microbleeds. Clinically and in mouse models, detection of cerebral microhemorrhages involves the use of magnetic resonance imaging and/or postmortem neuropathology techniques including hematoxylin and eosin (H & E) staining to detect extravasated red blood cells and fresh/acute microhemorrhages and Prussian blue staining to detect iron released from extravasated red blood cells and subacute/old microhemorrhages. Here we describe the step-by-step procedure for mouse brain processing and H & E and Prussian blue staining and quantification of acute (H & E-positive) and subacute (Prussian blue-positive) cerebral microhemorrhages in mouse brain tissues. Key words Cerebral microhemorrhages, Hematoxylin and eosin, Prussian blue, Mouse brain

1

Introduction Cerebral microhemorrhages are microscopic bleeds (clinically >Live Cells (c) >>>CD45+ Leukocytes 2. Click “Add Keyword” in the activity ribbon (Fig. 3a) to create keyword parameters that denote all experimental variables of interest. Examples include *Animal, *Group, *Tissue, and *Timepoint (see Note 7). 3. You will see a new column for each keyword added in FlowJo™ Software. Under each keyword column, add a numerical value for each sample to identify individual samples or create sample sets. (a) For example, if the keyword *Group is made to denote groups and there are two groups in the experiment, then all samples in Group 1 receive a keyword value of 1 in the *Group column. Samples in Group 2 receive a keyword value of 2 in the *Group column. Continue assigning keyword values to samples in this fashion for each keyword column (Fig. 3b). (b) Create keyword value series. The process of adding keyword values can be automated by creating a keyword value series. To do this, first add a keyword with the “Add

236

Thomas A. Ujas et al.

Fig. 3 (a) Where to add a keyword. (b) Two groups and how the numerical values are added to each sample. (c) How to select equivalent nodes

Keyword” button, as described above. Click “Create keyword value series. . .” in the keywords band. Select the newly created keyword at the top of the dialog window. Use the Numerical Keys tab to set up the pattern of integers to be used as keyword values for the samples

Unbiased Flow Cytometry Analysis

237

Fig. 4 (a) Where to select export/concatenate populations. (b) The selections in the control window for how to concatenate populations

based on the order in which they are displayed in the workspace. 4. Right-click the target population node, and choose “Select Equivalent Nodes” (Fig. 3c). 5. While those are highlighted, right-click and select Export/ Concatenate Populations (Fig. 4a). 6. In the dialog window of “Populations: Export or Concatenate” (Fig. 4b), select the following: (a) Populations: Concatenate (b) Format: FCS3 (c) Destination: location where the concatenated fcs file will be exported (d) Include events > Include no more than > Chose a value (Input a number of events to be taken from each sample into the concatenated file: this number can vary but is an opportunity to normalize the event number across samples and downsample the data (see Note 8)). (e) Parameters: All uncompensated parameters if fcs files share a single compensation matrix. Otherwise, select all compensated parameters. (f) Advanced Options > Additional Parameters > Choose. . . Select the added *Keywords (see Note 9). 7. Select Concatenate at the bottom. After the export is complete, the option is to either load files into “New Workspace” or “Existing Workspace.” Either option works, for this example, select Existing Workspace. This results in a default file name of

238

Thomas A. Ujas et al.

Fig. 5 (a) The histogram of the keyword “*group” and how to label the peaks on that histogram. (b) The FlowSOM control window and the settings. (c) The initial output of FlowSOM and the FlowSOM populations. Highlighted in gray is the red FlowSOM operation node

“concat_1.fcs” file unless the name was manually changed in step 6 (see Note 10). 8. Select the new concatenated file by double-clicking it. A graph window will pop up. On the y-axis parameter dropdown, select “Histogram.” On the x-axis parameter dropdown, select one of the created keywords. Click the T button beside the x-axis and set this to linear axis. There should be peaks corresponding to each unique keyword value in numerical order. 9. Next, in the top left, select the range “I—I” icon, and drag it over the first peak. A small “Subpopulation Identification” window pops up. Here, label the population to denote its identity based on its keyword value on the x-axis, assigned on step 3. Notice that these steps create nodes under the concatenated file. Continue manually labeling each peak until all are labeled. See Fig. 5a for an example with two groups “stimulated” and “unstimulated.” After creating the keyword of *Group, stimulated was assigned the number 1 and unstimulated the number 2 in step 3. (a) As a shortcut for this step, right-click the plot, and select “Create Gate on Peaks.” This will automatically draw and name a range gate on each peak; the gated populations can

Unbiased Flow Cytometry Analysis

239

then be renamed in the workspace by right-clicking the population node and selecting “Rename” (see Note 11). 10. Repeat steps 8 and 9 for additional keywords. 3.3

Using FlowSOM

1. Select the concatenated file, navigate to “Plugins” in the activity ribbon, and select “FlowSOM” [11] (see Note 12). 2. Highlight the compensated parameters, selecting only those that offer discriminatory power (i.e., are not homogenous in expression level across cells in the concatenated file). For example, if the cells that went into the concatenated file were pre-gated as live CD45+ cells, then CD45 and the viability dye should be excluded from the algorithm because they will not contribute to identifying unique populations (see Fig. 5b for the FlowSOM window). 3. In “Number of meta clusters,” select the number of populations for the program to uniquely identify (see Note 13). 4. Leave the rest of the options on default and select “OK” to run FlowSOM. 5. FlowSOM will output subpopulations labeled as Pop0-PopN (see Fig. 5c for how this looks for 15 meta clusters). Drag and drop the red FlowSOM operation node onto Layout Editor to see the meta-clustering result as a minimum spanning tree visualization. FlowSOM will also create PDFs of the tree maps in a new FlowSOM output folder as a subfolder in the one containing the current .wsp file. 6. Be sure to select the FlowSOM operation node and delete it after running the plugin, only keeping the derived parameter node (“blue a/b”) and the Pop0-PopN nodes (see Fig. 5c, to see the Pop nodes, “blue a/b” node, and the operation FlowSOM node, highlighted in gray in the figure, that needs to be deleted after it is finished calculating) (see Note 14).

3.4 Visualization in Two-Dimensional Space Using tSNE or UMAP

1. After clustering has been performed in high-dimensional space, the data can be visualized in two-dimensional space using tSNE or UMAP plots. Running both is an option, to see which visualization may best suit your data set (see Note 15). 2. Select tSNE in the activity ribbon, choosing the compensated parameters of interest. Just like FlowSOM, for both UMAP and tSNE, be sure to only map the parameters that add new information. If the live cells and prior populations were already gated out manually, do not select those markers for mapping. For this walk-through, we recommend selecting “Auto (opt-SNE)” as a learning configuration and keeping the other default settings, then select “Run.” See below for more information on the settings in the tSNE window (see Fig. 6a to see how these parameters would look in the tSNE window).

240

Thomas A. Ujas et al.

Fig. 6 (a) The tSNE control window and parameters. (b) Depiction of how the UMAP control window looks like. (c) The control window of ClusterExplorer and how to select your parameters

(a) Learning configuration: Auto (opt-SNE) is optimized to finish faster and can go beyond one million events. Iteration number and learning rate are optimized on Kullback– Leibler divergence (KLD). For more detail on opt-SNE, please refer to Belkina et al. [2]. (b) KNN (k-nearest neighbor) algorithm: Exact (vantage point tree) takes longer to run but may reveal more detail in larger data sets. ANNOY calculates “near” instead of “nearest” neighbors, allowing faster computation. However, we found that for tSNE plots of five million events or less, the results are very similar for exact or approximate. (c) Gradient algorithm: Barnes–Hut is the original algorithm that has been used on mass cytometry data since Amir et al. [1] first reported its use. Fit-SNE uses Fourier transform for much faster computation of repulsive forces [4]. (d) Perplexity: This value reflects the level of attraction between similar events and can be adjusted to improve resolution. Increasing the perplexity creates more separation between islands but compresses cells within an island, while lowering perplexity creates the opposite effect. 3. Select UMAP [5] under the “Plugins” section. Select your compensated parameters, keeping all default settings, and press “OK” (see Fig. 6b for how this would look like for a default UMAP run). For details on settings, see below:

Unbiased Flow Cytometry Analysis

241

(a) Distance metric: Options for calculating distances. We recommend staying with the Euclidean algorithm. (b) Nearest Neighbors: Sets the number of neighbors to use to approximate geodesic distance. Larger numbers will focus more on the global structure (distance between islands), whereas smaller values will focus on local structure (distance between cells in an island). We recommend starting at 15. (c) Minimum distance: Determines how tight the data is clustered. Smaller values will result in more clumping of the data, while larger values will disperse the data more. 0.5 is a good number to start with. (d) Number of components: How many UMAP parameters to be created. We recommend staying with two components to be able to view them in a two-dimensional format. Three components will allow a three-dimensional view with other plugins in FlowJo™ Software (e.g. iCellR). 4. After the plot is generated, select the concatenated file, and then under “Plugins” choose ClusterExplorer. 5. FlowSOM should be highlighted in blue under “Derived Cluster Columns.” Under “Relevant Measurements and Derived Parameters,” select the parameters that were considered in dimensionality reduction and clustering (see Fig. 6c for how this will look). 6. Under “Select Derived Map Parameters,” choose tSNE and/or UMAP. 7. Make sure that “Ignore Nodes belonging to unselected Cluster sets” is selected, and press “OK”. 8. ClusterExplorer will run and produce various charts and map your FlowSOM populations onto your two-dimensional plot (tSNE or UMAP). The following interactive windows will pop up: (a) Control panel: contains various options for displaying data in the plugin, such as plotting relative values as medians or means. The display of clusters can also be animated. (See Fig. 7a–e, for an example of the 15 populations identified by FlowSOM and how they are mapped on the tSNE plot). (b) Bar chart: frequency (%) of clusters in the concatenated file. Each cluster has a unique color used throughout the plugin interface (Fig. 7b). (c) 2D plot: defaults to tSNE/UMAP but can be changed to visualize other markers (Fig. 7c).

242

Thomas A. Ujas et al.

Fig. 7 The outputs of ClusterExplorer. (a) The control panel. (b) The bar chart displaying the frequency (%) of each cluster. (c) The 2D plot of your tSNE or UMAP plot. (d) The cord plot with relative expression of each parameter. (e) The heat-map with relative fluorescent intensity of the parameters

(d) Cord plot: relative expression level of parameters (Fig. 7d). (e) Heat-map: relative fluorescence intensity of parameters (Fig. 7e). 9. Based on the expression pattern of markers, the clusters can be annotated. For example, in Fig. 8, we looked at pop 6. This population has high relative expression for CD8, Granzyme B, and CD107a. Granzyme B is a serine protease released by

Unbiased Flow Cytometry Analysis

243

Fig. 8 An example on how to use ClusterExplorer. (a) The cord plot that shows expression levels of each parameter for population 6. (b) An example of how a gated subpopulation would look like with a FlowSOM population. (c) A comparison of tSNE plots in layout editor of both the FlowSOM population and the Granzyme B high expressing CD8 T cell population

cytotoxic T cells, and CD107a is a degranulation marker. Thus, we could phenotype this subset of CD8 T cells as releasing Granzyme B. In Fig. 8a the chord plot was used to annotate this population. 10. To confirm annotations, clusters can be manually gated in the concatenated file based on their marker expression profile shown in Cluster Explorer. TheFlowSOM clusters could then be dragged onto the manually drawn gates to see if most of the events in the candidate cluster fall within the manually gated population. The manually drawn population can also be overlaid onto the tSNE or UMAP plot to see if it is in the same place as the candidate cluster. In Fig. 8b, high levels of Granzyme B expression were gated within population 6 so that in Fig. 8c, we can further look at that subset of CD8 T cells with the highest Granzyme B expression. This is a good way to check the FlowSOM populations on your tSNE or UMAP plots.

244

Thomas A. Ujas et al.

Fig. 9 (a) A comparison of tSNE plots between the stimulated group and the unstimulated group. (b) Depiction of how the FlowSOM nodes can be applied to an individual sample or group

11. Continue going through the populations, and annotate clusters of interest. 12. In “Layout Editor,” the entire UMAP or tSNE plot can be dragged and dropped onto the layout editor page (double-click your “concat_1.fcs” file to modify the axes for UMAP or tSNE parameters), with clusters overlaid onto the plot. In addition, UMAP or tSNE plots only displaying events from an individual sample or group can be dragged and dropped to compare trends visually. See Fig. 9 for a comparison plot for the stimulated versus unstimulated group. From Fig. 9a, we can see which populations are missing or only present in one group. For example, Population 6, the CD8 Granzyme BHi T cells are present in the stimulated plot but not in the unstimulated plot. 13. To export the frequency of clusters within each individual sample, select and highlight the populations labeled “FlowSOM.Pop0-N” (by FlowSOM in Subheading 3.3) in the main FlowJo workspace then drag and drop that analysis onto each individual sample, which exists as a subpopulation gated from the concatenated file (Fig. 9b shows the FlowSOM populations being applied to Sample 1). FlowSOM clusters can also be dragged onto groups instead of individual samples.

Unbiased Flow Cytometry Analysis

245

14. To export these numbers into excel, select “Save As. . .” and then select “Export to Excel (XLS).” 15. Choose your file destination. 16. Statistics or graphs can be generated in Excel or other programs.

4

Notes 1. At time of publication, these are the most recent software versions, but please be cognizant of updated versions of all software and plugins. 2. These protocols are based on the analysis of data from a published 30-parameter flow cytometry panel publicly available [3], the data is publicly available on FlowRepository (FR-FCM-ZYRX) [10]. For lab data sets, it is recommended to work off files saved directly onto your computer’s hard drive, as working off a network drive or cloud storage may cause performance or connection issues in FlowJo™ Software. We also advise labeling parameters with their corresponding fluorochromes and markers during acquisition on the cytometer; this will facilitate interpreting the analysis. If this was not done with your data, stain names can be modified in FlowJo™ Software by adding or editing the keyword value associated with each stain name’s respective keyword. 3. FlowAI is an automated tool that allows users to clean up their flow cytometry data by removing acquisition-associated anomalies. Without proper sample clean-up, it is possible to incur erroneous events that can lead to false discoveries. The three quality control checks FlowAI employs are: flow rate, signal stability, and dynamic range. The flow rate check removes large deviations that trend away from the median flow rate. The signal acquisition check removes regions where statistics are shifted from the most stable region. The dynamic range check removes outlier events in the lower limit and margin events in the upper limit. See Monaco et al. [6], for further details on the FlowAI algorithm. 4. Completion of steps 6–9 will export your good events into a new workspace to be used for downstream analysis. Alternatively, downstream analysis can continue in the same workspace on the good events subpopulations. We recommend excluded events in the “FlowAIBadEvents” subpopulation be inspected to evaluate over- or underexclusion of anomalous events. 5. FlowAI strictness: FlowAI may be too strict for some data sets. To ease restriction, the changepoint penalty can be increased from 200 to roughly 500; the higher the value, the less strict is

246

Thomas A. Ujas et al.

the detection of anomalies. The box that removes outliers prior to the signal stability check can also be selected to try and prevent large sections of events from being excluded. 6. Focusing downstream analysis on populations further down the hierarchy increases resolution. For example, you can manually gate to live CD45+ populations, apply the high dimensional analysis, and achieve a more global view of your data. Alternatively, rather than including all mononuclear cells, NK cells could be the input for the analysis, therefore making it easier to resolve rare NK populations. 7. Why are keywords important? These keyword designations facilitate separating and comparing individual samples or experimental variables after files are merged during concatenation (as described in Subheading 3.2, steps 4–7). Keyword values need to be integers in order to use gating in the graph window to separate samples or variables. A tip is to add an asterisk [*] before your keyword (e.g., *Group or *TreatmentDay). This will group the user-created keywords together, making it easier to find them during step 6f, in Subheading 3.2, when concatenating and selecting additional parameters. 8. This step provides an opportunity to normalize and downsample events. Normalizing event number ensures all samples have the same weight when performing an unsupervised analysis. This is especially crucial if interrogating rare populations, which may appear at a higher frequency in samples with significantly more events. Selecting an event number to normalize samples to is partially influenced by the sample containing the least number of events. In one example, say one sample has 1000 CD45+ leukocytes but the other samples contain 20,000 CD45+ leukocytes. In this case, incorporating the 1000 event sample limits concatenation to 1000 cells from each sample. Thus, if one sample only contains a really low number of CD45+ cells, then it is not advised to include that sample in your data set. It is possible to exclude samples from the computational analysis and manually gate the samples with low events to confirm findings from the other samples. Additionally, it is important to be aware of computational overhead of having many events. In cases where devices are limited in memory, downsampling may be necessary. When considering the number of cells to include from each sample, take care not to exclude rare events (350 g rat, tubes that are a 10 cm inner diameter × 30 cm inner length are not restrictive for the rats to burrow in. The diameter of the tubes selected, however, will need to be increased for larger rats. The tubes should be closed at one end. 2. Wood chips (Envigo 7090 Sani-chips). 3. Scale. 4. Supply food and water as usual.

Assessing Depression and Cognitive Impairment

3

267

Methods

3.1 Barnes Maze Test

1. Position maze in the center of the room (see Note 5 for other physical considerations). 2. Position lights on opposite sides of the maze, about 2 feet away from the table and pointing downward. Lights must be angled so that there are no shadows on the table. 3. Attach large black geometric cues (about 8 × 10 inches) to walls close to the maze, approximately 10 inches above the surface of the table. 4. Create a hiding area for the experimenter and the recording computer, behind a solid panel which can also serve as a landmark in the room.

3.1.1

Acclimation

1. Handle the rats 1 week prior to the experiment for 5 min daily, in the experiment room. 2. Two days prior to the learning trials, place the rat inside the escape box (Fig. 1), in the center of the Barnes maze. 3. Cover the box with a lid, so that the animal is unable to escape. 4. After 2 min, take the rat out of the escape box, and gently place them into the rectangular start box, also in the center of the Barnes. 5. Attach the escape box to the Barnes maze table, in a predetermined position that will not be used during the testing trials. 6. Turn on the flood lights on each side of the table, and remove the start box to allow the rat to move freely in the maze. 7. If the rat does not find and enter the escape box within 5 min, gently guide the rat into the box, and then turn off the lights. 8. Clean the entire apparatus with a disinfectant thoroughly between animals.

3.1.2 Testing Spatial Learning and Memory

1. Place the escape box in a predetermined location under the Barnes maze table. 2. Place the start box in the center of the table, and place the rat inside the start box. 3. After 30 s, remove the start box and turn the flood lights on. 4. Allow the rat to move freely around the Barnes maze for 2 min. If the rat is unable to find the escape box within 2 min, gently guide the rat into the escape box (without removing the box from under the table). 5. Once the rat is in the box, turn off the flood lights and allow them to sit in the box for 30 s before returning them to their holding cage.

268

Karienn A. de Souza et al.

6. Repeat steps 2–5 three times per day, with a 15-min intertrial interval. 7. Clean the entire apparatus with a disinfectant thoroughly between animals. 8. Rotate the table every day to minimize intramaze cues (odors), but always place the escape box in the same spatial location. 9. Each rat should receive three 2-min test trials per day for 4 consecutive days. 10. After 2 days of rest, memory can be tested with a probe trial. For probe trial, repeat steps 1–4, but remove the target box, so that all boxes under the hole are shallow and the rat has no available escape. 11. Allow rat to explore for 2 min under the light. 3.1.3

Data Analyses

The video detection software can acquire and analyze multiple variables, but the latency (seconds) to entrance into the escape box and distance traveled (centimeters) are generally used to show whether the animals have learned the location of the escape during learning trials. Speed (cm/s) influences the latency variable and can be used to demonstrate whether a group is struggling with motivation or motor performance. Data should be analyzed with repeated measures analysis of variance (ANOVA) across days. For models of injury and stroke, search strategies may be analyzed as indices of learning to reduce confounding effects of reduced locomotor ability with injury. Learning strategies can be separated into three broad categories: hippocampal-based (which includes “Direct,” “Corrected,” “Focused,” and “Long Correction,” seen in Fig. 2 inside the orange box), non-hippocampalbased (Serial), or unsuccessful strategies (Random or Fail) [17–19]. A Barnes score can also be developed based on search strategy [18–20]. A predetermined score is assigned to each of the search strategies (Direct = 1, Corrected = 0.75, Focused Search = 0.5, Long Corrected = 0.5, Serial = 0.25, Fail/Random = 0). Each rat’s scores for the trials/day are added to produce a summative score. The average of the daily summed scores for each group can then be compared across days with a repeated measure analysis of variance (ANOVA). For probe trials, a number of errors are widely used for group comparisons. Percent time or distance in the target quadrant can also quantify the animals’ insistence about the location of the target. A quadrant is defined by dividing the Barnes into four, such that each quadrant has the same number of holes and the escape is in the middle of the target quadrant (Fig. 3). Because rodents can habituate overtime, we have observed subjects that remember the location of the target (great “learners” as defined by their performance in the learning trials) will give up looking for

Assessing Depression and Cognitive Impairment

269

Fig. 2 Examples of search strategies. (Images were acquired using Ethovision Software by Noldus)

Fig. 3 For probe analysis, the arena is divided into quadrants (A) so that behavior in the area of interest can be analyzed

270

Karienn A. de Souza et al.

Fig. 4 Schematic of object familiarization and novel object recognition trials. (Created with BioRender.com)

it during probe trials. Thus, although the rodent stays on the Barnes maze for the full length of the trial, we focus the analysis of their behavior during the initial period of 30–60 s. Group means from the probe trial can be analyzed with one-way ANOVAs or predetermined t-tests. 3.2 Novel Object Recognition Test 3.2.1

Place the rat in the empty arena and allow it to freely explore for 10 min per day for 2 consecutive days.

Acclimation

3.2.2 Testing Novel Object Recognition

1. On Day 3 (after the 2 days of acclimation), place the rat in the center of the open-field chamber with two identical objects (A + A) positioned diagonally from each other (Fig. 4). 2. Allow the rat to explore the arena and the objects for 10 min. 3. Return the rat to its home cage for 1 h (retention interval). 4. Add two objects in the same location, one that was previously used in the familiarization (A) and the other that was novel (B). 5. Place the rat in the arena and record for 5 min. 6. Return the rat to the holding cage.

3.2.3

Data Analysis

Data can be analyzed with a video behavioral analysis software such as ANYmaze (Wood Dale, IL) or Ethovision. If the data collection is performed manually by the experimenter with video analyses, the amount of time spent exploring the novel object is determined from the recordings by an investigator blind to the experimental condition. Exploration of an object is defined as the rat sniffing or touching the object with its snout at a distance of 80% 30 trials within response omissions on 12.5 min on each of each of at least five five consecutive consecutive sessions sessions) w/reinstatement, dependent on reinstatement method [62]

Autoshaping (AUTO)

4–5 sessions

N/A

≥2 sessions

~14 sessions Dependent on reinstatement method

7–10 sessions

a

Adapted from Ref. [43] Pretraining is typically only required once. If multiple touchscreen tests are performed on the same cohort of mice, at the conclusion of one task the mice would continue into the Task-Specific Training c If multiple touchscreen tests are performed on the same cohort of mice, the Total can be reduced by the duration of the Pretraining. Seea d As with many parameters in behavioral tests, there are slight differences in protocols across laboratories. The original work on PAL stated the acquisition criteria for this task is the completion of all trials with an accuracy of 80% for two consecutive sessions. For our purposes, we use 30–45 sessions and do not hold mice to an accuracy percentage b

able to perform a range of cognitive tasks on the touchscreen platform [38, 39, 45, 74, 106, 107], which underscores the importance of looking beyond motor function in such injury models [86]. Readers are referred to these and many other excellent studies and reviews for broad guidance on how to perform rigorous, reproducible, and meaningful research using an operant touchscreen platform in models of CNS injury. As with noninjury studies as well, experiments must be carefully designed to not just detect changes in cognition but to anticipate and account for fundamental changes in motor function, vision, motivation, attention, stamina, etc. With performing such long training and testing protocols, rodents can easily become sated on the reward stimulus (e.g., milk-based liquid or sugar pellets), so some groups have moved to a 5-day on/2-day off training protocol to maintain motivation.

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

285

Notes are provided throughout the Methods to highlight key issues and papers of interest in this regard, but three comments are worth mentioning here. First, while touchscreen studies typically do not report data from Pretraining steps, the Pretraining data (including the proportion of mice from control and experimental groups that did not complete Pretraining) can provide valuable insight about the mice even prior to Task-Specific Training and Testing (e.g., age-related effects on learning). Second, as with any manipulation in behavioral neuroscience, the timing of the manipulation relative to training or testing is important (see Note 4). For example, if probing the impact of injury on memory, the CNS injury could occur after Pretraining, Task-Specific Training, and Baseline Task Testing have occurred. Third, the consistency and reproducibility of touchscreen testing are amenable to longitudinal testing of mice, something that is central to understanding CNS injury and recovery [10, 11, 27, 35, 49, 64, 72]. Therefore, readers may want to apply one test (we recommend LDR) to the same cohort at different points post-injury. Finally, if the injury impacts the ability of the mice to learn certain tests, researchers might consider changing aspects of the test to better suit the question at hand, just as some researchers have done with image selection to eliminate potential visual bias [7]. This chapter is our contribution to a very dynamic literature and growing community focused on using touchscreens for cognitive testing. The rapid spread of this technology worldwide, as well as the relatively high level of standardization, has been made possible largely due to the efforts of the pioneers of this technology, Lisa Saksida and Tim Bussey. Together with their colleagues, they have published protocols that have set standards for the field [37, 62, 76]. Along with others in the field, they have asked research questions that ultimately improve both the protocols and processes of running these experiments. To further aid in what has become a very dynamic and evolving field, Saksida, Bussey, and colleagues have established an online community: touchscreencognition.org [28]. There many standard operating procedures, protocols, methods, and modifications are shared. They host an annual conference focused on the use of touchscreens for cognitive testing; this has been held virtually in recent years, which allows researchers from all over the world to present and attend. Another major advance in the touchscreen community was spearheaded by Marco Prado and Boyer Winter: MouseBytes (mousebytes.ca), an open science data repository for touchscreen data [10]. MouseBytes allows scientists to upload their raw data for other groups to analyze and review, underscoring both the cutting-edge and the community-oriented nature of this field. We hope this methods chapter, along with these listed resources and the many other resources available in the literature and online (see Note 5), will help expand the ranks of CNS injury researchers that join the vibrant and supportive touchscreen community.

286

2

Katherine M. Cotter et al.

Materials 1. Mice (see Note 6). 2. Mouse identification tools: (a) Ear tag applicator and metal ear tags with numbers or ear punch (see Notes 7 and 8). (b) Permanent marker for tail marking (e.g., Sharpie®) (see Note 9). 3. Mouse Touch Screen Chamber (Easy Install Package; Lafayette Instrument, USA) (see Notes 10 and 11), which includes: (a) Reinforcer (or reward) magazine with infrared (IR) beam and light (LED, 75.2 lux; Lafayette Instrument, USA Cat. #80614-7) (Fig. 5) (see Note 12). (b) Microliter accuracy liquid feed pump (Lafayette Instrument, USA Cat. #80204M) (Fig. 5) (see Note 13). (c) House LED light (3 W). (d) Tone and click generator. (e) Isolation chamber (sound attenuation level of ~35 DB) with ventilation fan. (f) (Optional) IR camera (Lafayette Instrument, USA Cat. #80051-1-NTSC); Quad Channel HDD Video Recorder (Cat. #80050A-3); SVGA cable; HDMI Monitor (see Note 14). 4. Touchscreen “masks” for associated tasks: (a) 2-Window mask for AUTO (Lafayette Instrument, USA, Cat. #80614-M5). (b) 3-Window mask for PAL and EXT (Lafayette Instrument, USA, Cat. #80614-M2). (c) 12-Window mask for LDR (Lafayette Instrument, USA, Cat. #80614-M3). 5. Easy-Install Kit (Lafayette Instruments, USA, Cat. #80614-20). 6. ABET II touch screen software (or “touchscreen software”; Lafayette Instruments, USA, Cat. #89505) and individual ABET II programs for specific touchscreen tasks (see Notes 15 and 16): (a) PAL is available for separate purchase (ABET II software, Cat. #89541). (b) LDR is available for separate purchase (ABET II software, Cat. #89546-6) (c) EXT is available for separate purchase (ABET II software, Cat. #89547). (d) AUTO is available for separate purchase (ABET II software, Cat. #89544).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

287

Fig. 5 Reward magazine positioned opposite the touchscreen wall. In this photograph, taken under the red light that illuminates our touchscreen behavioral room, the reward magazine and touchscreen can be seen positioned on opposite sides of the chamber. This image also emphasizes the “funneling” of the side walls flanking the reward magazine, thus guiding the mouse toward the reward magazine. Also highlighted are the prongs in the reward magazine, from where the reward dispenses

7. Control PC for ABET CORE (Lafayette Instruments, USA, Cat. #88530-W10) or a PC computer with the following minimum specs [2]: (a) Microsoft® Windows® 10 or 11 (recommended). (b) Minimum of 3.2 GHz Quad Core Processor (i5 or higher). (c) Two PCI Express Card Slots (full size) for: (i) ABET II interface. (ii) Quad Video/Graphics card (chambers 1–4).

288

Katherine M. Cotter et al.

(d) One PCI Express Slot for 4-Port RS232 card (chambers 1–4). (e) On-Board VGA or DVI for PC operation. (f) 8 GB of RAM or higher recommended. (g) Hard Drive: (i) 150 MB of available hard-disk space for installation, additional free space required during installation (cannot install on removable flash storage devices). (ii) Additional disk space required for data files and other working files (500 GB recommended). (h) Microsoft-compatible keyboard and mouse. (i) Microsoft.NET Framework 4.0 support required. (j) Whisker multi-media single station license for each connected chamber (Lafayette Instruments, USA, Cat. #80698-1). (k) UPDD, Universal Pointer Device Driver single license. 8. Reinforcer, such as a milk-based liquid like Ensure® Original Strawberry Nutrition Shake (here referred to as “milkshake”) (see Notes 17, 18, and 19): (a) Glass jars to hold and dispense liquid reward (Fig. 6) (see Note 20). (b) Shallow (petri-like) dishes (if using a liquid reward system) for placement of food reward in cages. 9. Large rack with shelves on which cages can be placed for habituation to the environment room (see Note 21). 10. Waist-high transportation cart (see Note 22). 11. Scale to weigh the mice. 12. Weigh bucket or other holder to place mouse in for weighing. 13. Room red light if you prefer to perform your studies in red light. 14. Paper copy of Daily Run Sheet with animal IDs, chamber assignments, and experimenter information (see Notes 23 and 24). 15. Personal protective equipment (gloves, gown, shoe covers) as required by the animal facility or your protocol. 16. Cleaning and maintenance: (a) Hot water for flushing reward delivery lines (also referred to as tubing). Deionized or tap water is acceptable. (b) 50% ethanol (EtOH). (c) Paper towels. (d) KimWipes (or dry laboratory wipes suitable for delicate surfaces).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

289

Fig. 6 Reward delivery system. In this photograph, taken under the red light that illuminates our touchscreen behavioral room, the glass jars (white outline) that hold our reward (milkshake) are stabilized in an upright position by placement inside the circular holders in the metal attachment. Note the height of the jar is low enough to access the reward pump toggle (teal outline) but tall enough to hold enough reward for the daily sessions. The diameter of each jar is ~3″. The reward pumpline (green arrow) delivers the milkshake from the jar to the Microliter Accuracy Liquid Feed Pump (yellow outline). It then is delivered into the reward magazine (see Figs. 5 and 7)

(e) Sani-Cloth wipes (or similar germicidal disposable wipes). (f) Hemostats for grasping tubes and to aid with cleaning. (g) 30 gauge needles and 26 gauge needles for unclogging reward delivery tubing and reward magazine prongs.

290

3

Katherine M. Cotter et al.

Methods

3.1 Pre-experiment Preparation

3.1.1

Mouse Preparation

The steps in this section must occur prior to beginning the actual touchscreen experiment and encompass mouse preparation and acclimation to a reward-based paradigm (see Note 25). 1. Acclimate mice to the animal facility for a minimum of 4–7 days if received from an outside source. 2. Set up a mouse ID system, such as ear tagging or the Universal Ear Punch Mouse Numbering System [3, 42] (see Notes 7, 8, and 9).

3.1.2

Food Restriction

If using a reward that requires food restriction, follow the steps below to begin food restriction BEFORE introducing the reward. 1. Weigh the animals for 3 consecutive days with ad libitum food and water. Calculate the mean free-feeding weight of each animal. 2. At the conclusion of the acclimation period and mean free-feed weight calculation, slowly begin food restriction, gradually reducing the weight of the animals to the goal weight (~85% of the mean free-feeding weight of each animal) (see Note 26). 3. Make sure each mouse can be easily and reliably identified (see Notes 7, 8, and 9). 4. Weigh the mice daily after the start of the food restriction process. Avoid allowing the mice to lose >5% of their freefeeding weight per day and for food-restricted weight to be no less than 85% of their free-feeding weight.

3.1.3 Habituation to Reward

During or after the acclimation period (if doing food restriction), introduce the mice to the reward being used during training/ testing in the home cages. 1. Place a small amount of liquid reward in a petri dish and place it on the floor of each cage. 2. Provide fresh liquid reward in the petri dish every day for 3–5 days.

3.2 Daily Session Protocol

3.2.1 Mouse Habituation to Touchscreen Testing Environment

The steps below are done for each experiment regardless of stage of testing (Pretraining, Task-Specific Training, or Task-Specific Testing). A general day-to-day touchscreen experimental schedule is provided in Table 2. 1. Using a cart, bring cages of mice to be tested for the day into the testing room or nearby (see Notes 21, 22, and 27). 2. Cross-check the mice on the cart with the mice on the run sheet and check other information (experimenter, etc.) (see Notes 23 and 24).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

291

Table 2 Sample schedule of daily touchscreen (TS) tasks Time

Task

Subheading

8:00 AM

Begin habituation (to testing environment) for first cage(s)

3.2.1, see Note 27

8:05 AM

Review Daily Runsheet

See Note 23

8:10 AM

Load reward

3.2.2

8:20 AM

Test chambers

3.2.3

8:25 AM

Load session schedules and place mice in chamber

3.2.4

8:40 AM

Run TS experiment

3.2, 3.3, 3.4, 3.5, and 3.6

8:43 AM

Begin habituation (to testing environment) for remaining subjects

3.2.1, see Note 27

9:10 AM

Record end of session schedule variables

3.2.5

9:11 AM

Return subjects to home cage and prepare chambers for the next group

3.2.5

9:15 AM

Load session schedules and place mice in chamber

3.2.4 and 3.2.5

9:20 AM

Run TS experiment

3.2, 3.3, 3.4, 3.5, and 3.6

9:20 AM–12:40 PM

Continue experiment for remaining cages

12:40 PM

Run TS experiment (final cage(s))

3.2, 3.3, 3.4, 3.5, and 3.6

12:45 PM

Begin returning cages to animal room, provide food hopper

See Note 26

1:10 PM

Record end of session schedule variables

3.2.5

1:11 PM

Return final cages to holding room and provide food hopper

See Note 26

1:15 PM

Clean chambers

3.2.6

1:30 PM

Clean testing room

3.2.6

1:35 PM

Prepare next day schedule files

See Note 28

1:50 PM

Weigh animals

3.1.2

(Weekly)1:50 PM

Refresh tail marks

See Note 9

5:00 PM

Remove food hopper from cages

3.1.2

(Optional) 5:30 PM

Check all schedule files for next day

See Note 28

292

Katherine M. Cotter et al.

3. Allow mice to habituate to the lighting and sounds in the testing room for ~1 h. (a) Turn on the room’s red light and white light lamp, if applicable. 3.2.2 Chamber Setup: Loading Reward

This is done only once a day, typically as one of the first setup steps. 1. Assemble chambers (e.g., insert mask) based on the touchscreen step or task being run. 2. Follow this order for powering on equipment: Turn on the power strip for chambers, then Quad Channel HDD video recorder, and then computer (see Note 29). 3. Empty EtOH from the previous day’s experiment from the pumplines by switching the toggle on the pump upwards. (a) After seeing the EtOH dispense from each reward magazine, place a tightly folded paper towel under the reward magazine prongs from where the liquid reward directly dispenses (Fig. 5). (i) Be careful not to damage or bend the prongs. (ii) Place the reward magazine face down with the paper towel in place, allowing the prongs and paper towel to orient toward the chamber floor. (b) If no EtOH visibly dispenses, check for clogging (see Note 30). 4. Fill four glass jars with hot water and place the pumpline end into the jar opening (Fig. 6) (see Note 20). 5. Flush the pumplines with hot water for 20 s. 6. Empty the pumplines of all water: (a) Remove the paper towel and visually check no liquid is dispensed from the reward magazine prongs. (b) If bubbles are visible coming from the reward magazine prongs, continue emptying the pumplines. 7. Load the reward magazine with milkshake (see Notes 17, 18, and 19): (a) Shake the bottle thoroughly before filling milkshake glass jars halfway with the liquid. (b) Place the pumpline into the jar opening (Fig. 6). (c) Turn the pumpline off by switching the pump toggle down once you can see the milkshake is dispensed from the reward magazine prongs. (d) Wipe the area around the reward magazine prongs with KimWipe to clean excess milkshake. 8. Attach the reward magazine to the chamber walls (Fig. 7). 9. Repeat steps 6–9 for all chambers.

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

293

Fig. 7 View of opened touchscreen chamber. Photographed are the reward magazine (green outline, Fig. 5), pumpline (teal outline, Fig. 6), Microliter Accuracy Liquid Feed Pump (purple outline, Fig. 6), chamber lid (yellow outline), far-inside chamber wall (orange), IR beam, and IR beam toggle (white outline) 3.2.3 Chamber Setup: Testing Touchscreen, Reward Magazine, and IR Beam

1. Open Whisker Server first then ABET II Touch program from screen. If Whisker Server is not opened before the ABET II program, chambers won’t work properly. 2. At the start of each day of testing, check that each chamber’s IR beam, light stimuli, and reward magazine are functioning reliably: (a) The ABET II software includes programs to test the functioning of these equipment pieces [37] (see Note 31). (b) Under the Execution Manager tab in ABET II, open and run the “Mouse Test line” schedule appropriate for the mask being used in the training (e.g., “6 × 2” reflecting the mask response layout for the LDR task). (c) Clean the reward magazine by wiping the area around the prongs with KimWipe after testing. (d) Stop the “Mouse Test line” program after testing the chamber setup.

294

Katherine M. Cotter et al.

3. Configure ABET II software with appropriate training tasks for each chamber to begin training for the day. This includes schedule and database selection, max trial and time criterion, and suggest animal ID and user entries. 3.2.4 Chamber Setup: Load Session Schedules and Load Mice into Chambers

1. Load Schedules onto ABET for respective cages (see Note 28): (a) Select “Load Schedule” symbol to choose the appropriate files labeled by cage ID and date. (b) Check schedule name, schedule variables, chamber assignments, animal IDs, and database. See Table 3 for sample Pretraining ABET Schedules and Session Variables. All tests follow a similar formatting to these examples (see Note 32). 2. Load mice into chambers: (a) Place the cage(s) on the transport cart. (b) Check the cage cards and animal IDs against the Runsheet while in testing room. (c) Open all touchscreen chambers, slide out the apparatus shelf, and remove chamber lids (Fig. 7). Chamber lids can be placed on top of the chamber shelf while loading mice. (d) Open one cage at a time and place mice in their corresponding chamber (see Notes 33 and 34). (e) Once a mouse is placed in the chamber, gently put the lid back on the chamber and slide the shelf back into place. Carefully close the chamber door. 3. Once all mice are loaded, press the play symbol to start the program (see Note 35). 4. Record session start time and experimenter initials on Daily Runsheet.

3.2.5 Chamber Setup: Between Subjects

1. Once a session ends, record the end of Session Variables on the Daily Runsheet to prepare for the next day’s schedule setup (see Note 36). 2. Unload the completed schedules by selecting the red square with an X. 3. Load schedules as detailed in step 1 of Subheading 3.2.4 for the next cages. 4. Remove mice from chambers and return to home cages (see Note 37). 5. Clean chambers: (a) Between subjects, briefly wipe the complete inside of chambers (walls, mask, floor, reward magazine wall, and visible screen) with a paper towel sprayed with EtOH (see Note 38).

Session Variables

Variables

Value

0

10 min Max Time &0s

Session Variables

Variables

Max Trials

Max Time

_Trial_Counter (Int.)

0

0

False

0

0

0

_Trial_Counter (Int.)

reward_beam (Int.)

Reward_First (Bool.)

reward_to_screen (Int.)

Screen_Beam (Int.)

Screen_bottom (Int.)

Screen_First (Int.) False

Notes

Notes

Max Trials

Variables

Session Variables

10.000

280

Feeder_Pulse _Time (Int.)

0

0

1

0

0

Correct_Grid _Position (Int.)

Correct _Counter (Int.)

Blank_Touch _Counter (Int.)

A3Sound3 (Int.)

A2Sound2 (Int.)

A1Sound1 (Int.)

_Trial_Counter (Int.)

Notes

User

Max Trials

Variables

Session Variables

0

0

0

Correct_Grid 0 _Position (Int.)

Correct_Counter (Int.)

Blank_Touch 0 _Counter (Int.)

0

0

1

0

0

A3Sound3 (Int.)

A2Sound2 (Int.)

A1Sound1 (Int.)

_Trial_Counter (Int.)

0

1

0

0

Notes

User

Max Trials

Variables

Session Variables

Mouse LDR Must Initiate Training v2

Correct_Grid _Position (Int.)

Correct _Counter (Int.)

Blank_Touch _Counter (Int.)

A3Sound3 (Int.)

A2Sound2 (Int.)

A1Sound1 (Int.)

Max Trials

Variables

Session Variables

0

0

0

0

1

0

Correct_Percent (Dec.)

Correct _Grid_Position (Int.)

Correct _Counter (Int.)

A3Sound3 (Int.)

A2Sound2 (Int.)

A1Sound1 (Int.)

_Trial _Counter (Int.)

Notes

User

(continued)

0

0

0

0

1

0

0

30 min &0s

30

Value

Mouse LDR Punish Incorrect Training v2

30 min Max Time &0s

30

Value

_Trial_Counter (Int.) 0

Notes

User

30 min Max Time &0s

30

Value

Mouse LDR Must Touch Training v2

30 min Max Time &0s

30

Value

Mouse LDR Initial Touch Training v2

20 min Max Time &0s

30

Value

Delay_Time (Dec.)

Acclimatisation _time (Dec.)

A3Sound3 (Int.)

A2Sound2 (Int.)

A1Sound1 (Int.)

User

User

Max Trials

Mouse LDR HAB2

Mouse LDR HAB1

Table 3 Pretraining ABET Schedules and Session Variables

Touchscreen Cognitive Tests in Mouse Models of Brain Injury 295

0

0

Screen_top (Int.)

SessionMinCnt (Int.) True

1000

Pulse_Tone0 (Bool.)

Tone_Duration (Int.)

6000

0

Screen_to_reward (Int.)

Value

Prime_Feed _Time (Int.)

Variables

Value

Variables False

Session Variables

Session Variables

Houselight _Normally _On (Bool.)

Mouse LDR HAB2

Mouse LDR HAB1

Table 3 (continued)

1000

False

False

Houselight_Normally _On (Bool.)

True

True

280

False

0

False

0

Value

Increment_this _trial (Bool)

First_Trial (Bool)

Tone_Duration (Int.)

Tone_Duration (Int.) 1000

0

Right_Touches _During_ITI (Int.)

1000

First_Analysis (Bool)

0 Right_Touches _During_ITI (Int.) 0

Right_Touches _During_ITI (Int.)

0

Left_Touches _During _ITI (Int.) Tone_Duration (Int.)

Feeder_Pulse_Time (Int.) 0

Left_Touches _During_ITI (Int.)

0

Left_Touches _During_ITI (Int.)

Correction_Trials_Set (Bool)

10

ITI (Int.)

10

10

ITI (Int.)

Correction_Trial _Correct_Counter (Int.)

Correction_Trial (Bool)

Correction _Counter (Int.)

Variables

Session Variables

Mouse LDR Punish Incorrect Training v2

ITI (Int.)

30

Houselight_Normally False _On (Bool.)

False

Houselight _Normally_On (Bool.)

False

True

280

Value

First_Trial (Bool)

Feeder_Puls e_Time (Int.)

Variables

True

280

Value

Session Variables

Mouse LDR Must Initiate Training v2

First_Trial (Bool)

Feeder_Pulse _Time (Int.)

Variables

Session Variables

Mouse LDR Must Touch Training v2

True

280

Value

Image Time (Int)

Houselight _Normally _On (Bool.)

First_Trial (Bool)

Feeder_Pulse _Time (Int.)

Variables

Session Variables

Mouse LDR Initial Touch Training v2

296 Katherine M. Cotter et al.

10 0

1

ITI (Int.) Left_Touches _During_ITI (Int.) No_Incorrects _before_Correction _Trial (Int.)

0

5 1000

0

Sequential_Correction _Trials (Int.) Time_Out (Int.) Tone_Duration (Int.) Total_No_Correction _Trials (Int.)

Right_Touches_During 0 _ITI (Int.)

False

Increment_this _trial (Bool)

Touchscreen Cognitive Tests in Mouse Models of Brain Injury 297

298

Katherine M. Cotter et al.

(b) Do not disassemble the chambers during a testing session to wipe. 6. Load mice and begin the next session as detailed in Subheading 3.2.4. 3.2.6 Chamber and Room Cleaning: At End of Day

Following use (after all daily sessions are complete), clean as follows: 1. Remove the reward magazine from the chamber by sliding it out of the notches. 2. Remove reward magazine, walls, trays, and mask, placing on a rack for later cleaning. 3. Dispense any remaining reward from the pumpline. Refer to Subheading 3.2.2 on how to dispense liquid from the reward magazine. 4. Replace the glass milkshake reward jar with a clean glass jar filled with hot water. Allow the reward lines to run with hot water for 1 min, or soak through the paper towel. 5. Remove the reward line from the water jar and allow the lines empty of water for 20 s. Remove the paper towel from the reward magazines and check that all water has been emptied. 6. Place the reward line into a glass jar with EtOH and load the reward pump with EtOH. Turn off the reward pump once EtOH is dispensed from the reward magazine prongs. 7. Wipe the area around the reward magazine prongs with a KimWipe. 8. Wipe down the masks, touchscreens, and IR beams with a paper towel sprayed with EtOH. 9. Wipe down the chamber frames, walls, trays, and floors with Sani-Cloth wipes. Use a hemostat to help hold onto the wipe when cleaning the under-floor area. 10. Clean the behavior room, including sweeping the floor.

3.3

Pretraining

This is the standard Pretraining (referred to as “General Touchscreen Training” in some publications) used for three of the four touchscreen tasks covered in this chapter. During Pretraining, mice gradually obtain the general rule of touchscreen-associated learning. Some tasks require alterations or additions to the Pretraining procedure (see Note 39). The touchscreen mask used for Pretraining should be the same as the mask required for the task the animal is first being trained/tested on (e.g., use a 3-window mask for PAL, 12-window mask for LDR, etc.).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury 3.3.1 Preparing for Pretraining

299

1. If using a reward that requires food restriction, continue food restriction according to the following steps during the duration of all training and testing. Some rewards may not require food restriction to enhance motivation/expedite learning (see Note 19): (a) Remove food from each cage at ~17:00 h (5 pm) the day prior to training or testing. (b) Give food ad libitum in each cage for 3–4 h immediately following daily touchscreen training/testing and following completion of training/testing on Fridays until 5 pm on Sunday (see Notes 40 and 41). 2. Prepare an experiment schedule the day before testing. See Table 3 for schedule variables for each Pretraining Stage (see Notes 36 and 42).

3.3.2 Running Pretraining

1. Habituation (Hab)/Phase 1 (a) Place the mouse into their designated touchscreen chamber (maximum of 10 min per session, 1 session per day). See Subheading 3.2.4 for instruction on loading mice to chambers. (b) Start (play) Habituation I program (see Note 43). (c) Habituation 1 is a 10 acclimation session in which nothing happens in the chamber (see Note 44). 2. Habituation (Hab)/Phase 2 (a) Place the mouse into their designated touchscreen chamber (maximum of 30 min per session, 1 session per day) (see Notes 34 and 42). (b) Play Habituation 2 program: (i) An initial reward (150 μL of milkshake) is loaded into the chamber reward magazine in conjunction with a reward tone (70 decibel [dB] at 3 kHz, 1000 ms). (ii) When the mouse has removed its head from the tray, the tray light will turn off and a 10 s delay will begin. At the end of the delay, the tray light will illuminate along with simultaneous reward tone, and a standard liquid reward (7 μL of milkshake) is dispensed. (iii) If the mouse has not removed its head from the reward magazine or has replaced its head into the tray at the end of the 10 s delay, an additional 1 s delay is added (see Note 45). (c) The criterion for mice to move on to Initial Touch (IT)/ Phase 2 is the initiation and collection of 25 reward within the 30 min session time (see Notes 42, 43, 46, and 47).

300

Katherine M. Cotter et al.

Remove mice from the touchscreen chamber and return to their home cage immediately after they receive their 25th reward in order to reduce extinction learning (see Note 37). 1. Initial Touch (IT)/Phase 2 (a) Place the mouse into their designated touchscreen chamber (maximum of 30 min per session, 1 session per day) (see Note 34). (b) For each trial, a stimulus is shown on the touchscreen. (i) For PAL and LDR only: The stimulus shown is a single white square in a random location. (ii) For mice that only go through EXT: The stimulus shown is one of various black-and-white shapes in a random location. (c) Upon stimulus appearance, the mouse has 30 s to touch the stimulus location, which they usually do with their nose. There are two possible outcomes for each trial: the mouse does not touch the stimulus location within 30 s (i), or the mouse does touch the stimulus location (ii): (i) If the mouse does not touch the response location within 30 s of the stimulus being displayed, the image is removed, the reward magazine is lit and a reward is dispensed, and an audible tone is played. If the mouse touches a blank response area, nothing happens, and the stimulus remains on the screen until it is touched or the 30 s trial has completed. (ii) If the mouse touches the stimulus location, the stimulus immediately disappears, and the mouse receives a reward that is 3 times the standard reward volume (21 μL of milkshake) into the lit reward magazine, and the reward tone is played. (d) Following reward retrieval, a 20 s intertrial interval (ITI) begins, followed by the automatic initiation of the next trial (see Note 48). (e) Mice advance to Must Touch (MT)/Phase 3 after they complete 25 total trials within the 30 min session time (see Notes 45 and 47), regardless of trial outcome. (f) Remove the mouse from the touchscreen chamber and return them to their home cage immediately after receiving their 25th reward in order to reduce extinction learning (see Note 37).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

301

2. Must Touch (MT)/Phase 3 (a) Place the mouse into their designated touchscreen chamber (maximum of 30 min per session, 1 session per day) (see Note 34): (i) For each trial, a stimulus is shown on the touchscreen and remains on the screen until touched. (ii) Upon touch, the image will disappear, and the reward magazine light will illuminate and a standard reward is dispensed, along with the reward tone being played. If the mouse touches a blank location (the image is not displayed there), there is no reward; the image remains lit in the location it was at until the mouse touches the screen. (b) Following reward retrieval, a 20 s ITI begins, followed by the automatic initiation of the next trial (see Note 48). (c) Mice advance to the Must Initiate (MI)/Phase 4 after they receive 25 rewards within the 30 min session time (see Notes 43 and 45). (d) Mice should be removed from the touchscreen chamber and returned to their home cage immediately after receiving their 25th reward in order to reduce extinction learning (see Note 37). 3. Must Initiate (MI)/Phase 4 (a) Place the mouse into their designated touchscreen chamber (maximum of 30 min per session, 1 session per day) (see Note 34): (i) This is similar to MT/Phase 3, except the mouse is required to initiate the training by placing its head into the already lit reward magazine. A stimulus is displayed on the screen, and the mouse is required to touch the image to receive a reward (reward magazine lit, standard reward dispensed, reward tone played). Following reward retrieval, the mouse must remove its head from the reward magazine and then reinsert it after the ITI in order to initiate the next trial (see Note 48). (ii) At the beginning of the session, a free standard reward can be given in a lit reward magazine to encourage the initiation of the first trial. (b) Mice advance to Punish Incorrect (PI)/Phase 5 after they receive 25 rewards within the 30 min session time (see Notes 43, 45, and 47). (c) Mice should be removed from the touchscreen chamber and returned to their home cage immediately after receiving their 25th reward in order to reduce extinction learning (see Note 37).

302

Katherine M. Cotter et al.

4. Punish Incorrect (PI)/Phase 5 PI is only for PAL and LDR mice; it is not used for AUTO or EXT mice: (a) Place the mouse into their designated touchscreen chamber (maximum of 30 min per session, 1 session per day) (see Note 34). (b) Building on MI/Phase 4, this phase adds the inclusion of a “punishment” if the blank location is touched. This punishment is defined as the removal of the image and the illumination of the house light for a 5 s time-out period. After 5 s, the house light turns off, and, following the ITI, the mouse has the ability to initiate a correction trial (CT). A correction trial is automatically initiated (the same way as a regular trial) following an incorrect response (blank location or incorrect image, depending on the task), where the same image will appear in the same location on the screen as the previous trial. CTs are repeated until the mouse correctly responds and receives the reward. Following incorrect responses to a CT, the mouse undergoes the punishment as described above and will initiate another CT. Correct responses to CTs are NOT counted toward the final percent criteria (see Note 48). (c) The criteria for mice to move on to Task-Specific Training is to complete 25–30 trials within 30 min at ≥80–85% (≥20–22 correct) for 2 consecutive days. Mice should be removed from the touchscreen chamber and returned to their home cage immediately after their 25th trial in order to reduce extinction learning (see Note 37). When all mice have reached the criteria of Phase 5 at least once, run a baseline of all subjects so the group can advance to the next step of their task (Task-Specific Training for PAL, LDR, or EXT or for testing in AUTO) together. If subjects are to receive an experimental manipulation, this period of time (following Pretraining) is one in which these can be given (see Notes 42 and 49). 3.3.3 Pretraining Data Collection

1. Record the number of sessions it takes for each subject to achieve each Pretraining stage. 2. Record the total number of sessions required to complete all Pretraining stages (see Notes 36, 42, and 43). 3. Record additional training performance notes during the Punish Incorrect Training stage (i.e., total number of trials, total session length to reach the criteria and % correct, etc.).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

3.4 Paired Associates Learning (PAL)

3.4.1 PAL Task-Specific Training

303

The PAL task is used to evaluate object-location associations learning [40]. Adapted from the CANTAB PAL task for humans and nonhuman primates, the rodent PAL task uses three objectlocation associations, in which each of the three distinct objects (Fig. 1) or three distinct line patterns has a correct location (left, middle, or right) on the screen (see Note 50). There are two variations on the PAL task: the sPAL (same PAL), where the same image is shown at two locations, with one location having the correct object-location association (the stimulus that when selected results in reward delivery, termed S+), and the dPAL (different PAL), where two different images are shown on the screen simultaneously, with one image having the correct location association (S +) and the other image does not have the correct association (the stimulus that when selected does not result in reward delivery, termed S-) [98]. 1. Ensure all mice have completed all five stages of Pretraining. 2. Set up touchscreen chamber(s) with 3-window mask(s) and set up ABET II PAL Task software. 3. Place mice in their designated touchscreen chamber (see Notes 34). 4. Begin session with a free reward (7 μL of milkshake). Once the mouse inserts its head into the reward magazine, the first trial will be initiated, and two stimuli are presented (the same stimulus presented for sPAL or two different stimuli for dPAL), with one image shown in the correct location (S+). One of the three potential response locations will be blank. There are two possible outcomes for each trial: (a) The mouse touches the stimulus in the correct location (S +), upon which the stimuli are removed from the screen, the reward magazine is lit and a reward is dispensed, and a reward tone is played. After reward collection, the tray light turns off, and a 20 s ITI begins. Following the ITI, the tray light is illuminated, and the mouse is required to insert its head to initiate a new trial. (b) The mouse touches a stimulus in an incorrect location (S), the stimuli are immediately removed from the screen, the house light is illuminated for a 5 s timeout period, and no reward is dispensed. Following the timeout period, a 20 s ITI begins. Following the ITI, the tray light is illuminated for initiation of the next trial. After an incorrect response, the following trial is a correction trial (CT), where the stimuli shown are in the same locations as the previous trial. These CTs will continue until the correct response occurs, where the reward signals described above

304

Katherine M. Cotter et al.

will occur. Correct responses on CTs are not included in accuracy measures, and they do not contribute to the number of trials within a session. 5. The first sessions are likely to have a worse response accuracy relative to later sessions, resulting in multiple CTs. For these initial sessions, set session limits to 18 trials completed within 60 min. Use these parameters until the mice can complete 18 trials within 30 min, at which time increase the session parameters to complete 36 trials within 60 min. 6. Continue training on PAL tasks for 5–7 days/week until all subjects have reached task criteria. Criteria have been suggested to be >80% (excluding CTs) accuracy from all 36 trials in two consecutive sessions; we say that mice have to complete 30–45 sessions with no accuracy requirement (Table 1). 3.4.2

PAL Test

1. Upon reaching task criteria, mice can continue on the task as described above for the Task-Specific Training. 2. We suggest mice continue for a minimum of 2 additional sessions (Table 1). However, this is also a time when post-training experimental manipulations can be implemented (see Note 49). 3. At the conclusion of each session, promptly return the animal to its home cage to reduce extinction learning (see Note 37).

3.4.3 [37]

PAL Data Collection

1. Record the length of each session (time and number of trials). 2. Record the percent of correct responses (excluding CTs) of each session. 3. Record the number of trials/errors required to reach criterion for each subject. 4. Calculate the average latency to reward collection, measured as the latency of the mouse to enter the magazine to collect reward following a correct response, across all sessions. 5. Calculate the average reaction time of correct vs. incorrect responses, measured as latency to response after the stimulus appears on the screen, across all sessions. 6. Calculate the percentage of bias for each subject. 7. Perform trial type analysis by calculating the accuracy percentage for each of the 6 types separately for each subject.

3.5 Location Discrimination Reversal (LDR)

The LDR task was first developed for rats as a way to analyze the role of the hippocampus in the functions of memory, specifically mnemonic discrimination, also called behavioral pattern separation [65] (see Note 51). This paradigm has also been adapted for mice and used by many groups [32, 76, 85, 103], although slight differences among published protocols exist (see Note 52). In

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

305

general, using a twelve-window mask, two windows are illuminated and one of the locations—either the left or right lit square—is designated as correct (S+). If it is touched while still on the screen, a reward is dispensed. LDR includes a training stage (LDR Train) where the windows are always presented with an intermediate spacing (two lit windows in a horizontal line separated by two black windows). After LDR Train criteria are reached, the mouse advances to the LDR Test where the windows are placed with a Large Separation (two lit windows in a horizontal line separated by four black windows) or a Small Separation (two lit windows are adjacent to each other in a horizontal line) (Fig. 2). This is the concept of “load”; it is “easier” (less of a load on behavioral pattern separation ability) when faced with a Large Separation vs. a Small Separation (where it is more challenging to tell the left from the right position). The LDR Train and Test also employ reversal learning, so that when a correct response is given in 7/8 consecutive trials, the formerly correct location becomes incorrect, while the incorrect window becomes the correct window. This aspect of the LDR task probes cognitive flexibility (see Note 53). 3.5.1

LDR Train

An additional training step (LDR Train) is required prior to LDR Test, where two identical white squares with intermediate spacing are lit. 1. Ensure all mice have completed all five stages of Pretraining and have undergone reminder sessions as needed to maintain performance at criteria. 2. Set up touchscreen chamber(s) with 12-window mask(s) and set up ABET II LDR task software. 3. Split mice into two groups, counterbalanced with the number of sessions to complete Pretraining. Difficulty should be consistent across mice, but assign groups to different positive stimuli (i.e., correct side is on the left, or correct side is on the right). 4. Place mice in their designated touchscreen chamber (see Note 34). 5. Run the LDR Train schedule and record final schedule variables at the end of each session (see Note 36): (a) A touch to the correct position results in a reward, while a touch to the incorrect position results in a time out period described in Pretraining. (b) Acquisition is achieved when mice correctly respond in 7 out of 8 consecutive trials. (c) Once acquisition is met, within the same session, the correct lit window (S+) will automatically become S-, and S- will become S+, termed re-acquisition or reversal.

306

Katherine M. Cotter et al.

(d) The last window that was correct in one session will serve as the correct window at the beginning of the next session. (e) Mice can move on to the LDR Test phase once they reach the criteria (at least +1 reversal) in 2 out of 3 consecutive sessions. 3.5.2

LDR Test

1. Run mice through a counterbalanced design (based on final LDR Train day) where they are presented with two lit windows that are spaced by either a Large or Small Separation. The rewarded position (left or right) is also counterbalanced among the mice. Mice only receive Large or Small Separation on a given day. 2. Run the LDR in “blocks,” where one block is four consecutive days consisting of 2 days with Large Separation and then 2 days with Small Separation (or vice versa). The number of blocks varies, but it is typical to run the test for 6 blocks (24 days) total [94]. At this stage, there is no trial and reversal limit, and the session ends when 30 min have elapsed. 3. At the conclusion of each session, promptly return the animal to its home cage to reduce extinction learning.

3.5.3 LDR Data Collection [76]

1. Record the number of sessions and trials per session it takes for each subject to complete the intermediate training stage. 2. Calculate the average number of trials it takes each subject to achieve acquisition criterion (initial 7 correct out of 8 consecutive trials) of the Large Separation and Small Separation conditions separately (see Note 52). 3. Calculate the average reaction time of correct vs. incorrect responses, measured as latency to response after the stimulus appears on the screen, across all probe sessions. 4. Calculate the average magazine latency, measured as the latency of the mouse to enter the magazine to collect reward following a correct response, overall probe sessions. 5. Record the number of times each subject touches a window during ITIs and timeouts.

3.6 Autoshaping (AUTO)

In AUTO, the mouse learns to create an association between a visual stimulus (lit window) in a specific location with receipt of a reward [37, 71] (Fig. 3). The reward magazine is positioned to be directly in front of the touchscreen such that it divides the screen. In this location, the reward magazine also divides the front ambulatory IR beam into two, which provides means to record the number of approaches the mouse makes to the screen (left vs. right) [37]. Due to the introduction of position bias,

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

307

AUTO is typically not run prior to any other tasks. However, some researchers report success in performing AUTO reversal as well, showing that the initial AUTO-induced side bias is malleable [101]. 3.6.1 AUTO-Specific Training

AUTO-specific Training is needed due to the repositioning of the reward magazine: 1. Set up the touchscreen chamber so that the reward magazine is centered directly in front of the screen (Fig. 3), and use a fitted cover to seal the hole in the back of the chamber where the reward magazine was previously. 2. Set up the ABET II software to run the AUTO task with the desired parameters. 3. Beginning all subjects on the same day, run the first session as a 20 min habituation session. 4. Provide a single delivery of a reward (7 μL of milkshake) at the start of the session. 5. Perform a second habituation session on day 2, utilizing a variable interval (VI, 10–40 s); a reward was delivered along with simultaneous illumination of the magazine and a reward tone being played: (a) The mice are required to insert their nose into the magazine to trigger the next VI trial. (b) Mice advance to task acquisition after they complete 40 trials within 60 min.

3.6.2 AUTO-Specific Testing

1. Divide the subjects into two groups (and account for experimental groups as needed): Half has conditioned stimulus (S+) on the left side, while the rest has S+ on the right side. Counterbalance the groups according to the number of sessions it took for the subjects to complete AUTO Pretraining. 2. Place each mouse in its designated chamber (see Note 34) and begin each session with a free reward to trigger the first trial: (a) When mice break the beam near the rear chamber, new trials are started. (b) A session is finished when the subject has completed 40 trials (20 presentations each of S+ on a given side in a pseudorandom order) or 60 min have elapsed, whichever occurs first. 3. At the conclusion of each session, promptly return the animal to its home cage to reduce extinction learning (see Note 37). 4. Test sessions occur once per day, for 5–7 days per week, with each having a minimum of two sessions.

308

Katherine M. Cotter et al.

5. Continue running a set number of sessions or until the control group exhibits the learned behavior, with discriminated approach to S+ (left or right, depending on the group the subjects are in). 6. Additional post-acquisition manipulations may be run, depending on the experiment design. 3.6.3 AUTO Data Collection [37]

1. Record the number of sessions it takes each subject to complete AUTO-specific Pretraining. 2. Record the number of approaches to each side made when a stimulus is displayed. 3. Record the latency to approach after stimulus presentation. 4. Record the latency to touch after stimulus presentation and the number of touches to each side. 5. Record the latency to reward collection following reward delivery.

3.7

Extinction (EXT)

3.7.1 EXT-Specific Training (Acquisition of Stimulus-Response)

The EXT task is used to evaluate response inhibition [71]. Mice are first trained on Acquisition of a simple stimulus-response task: touch a white square on the screen and then receive a reward (Fig. 4). A daily acquisition session is complete once the mouse touches the S+ window 30 times or when 30 min has passed. Once the mouse completes 30 trials within 12.5 min on each of five consecutive sessions (criteria for Acquisition), the mouse advances to EXT. 1. Set up the touchscreen chamber, with the 3-window mask, and the ABET II software to run the EXT task with the desired parameters. The experiment design is initially the same as Pretraining Phase 5. 2. Place each mouse in their designated chamber (see Note 34). 3. Begin each session with a free reward to trigger the first trial. 4. The session concludes when the mouse has completed 30 trials, or 30 min have elapsed. 5. Perform once-daily acquisition sessions 5–7 days per week until acquisition criteria is reached, which is completing all 30 trials within 12.5 min over five consecutive sessions.

3.7.2 EXT-Specific Testing

1. Advance mice to EXT testing either individually upon achieving criteria or as a group by resting subjects (and giving reminder sessions), while the remainder of the group completes training. 2. Reduce EXT session times to 10 min, and no reward (food, tone, magazine light illumination) or punishment (house light illumination, timeout period) is given for any stimulus response.

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

309

(a) The stimulus is presented for 10 s and is removed upon a response (without reward), or the 10 s has elapsed. (b) Mice are not required to initiate each trial. EXT continues until the subject has an omission percentage of 77% (23 out of 30 trials) for two consecutive sessions. 3.7.3 [62]

EXT Data Collection

1. Record the number of responses and omissions for each session extinction session. 2. Calculate the response rate per unit of time (minute). 3. Record the time taken to complete each session. 4. Record the number of trials and/or sessions it took each subject to reach criterion. 5. Record response latency (time taken to response following stimulus presentation). 6. Record the latency to reward collection (time taken for mouse to retrieve the reward after dispensed) (acquisition phase) or the time taken to check the food magazine (extinction). 7. Record, if applicable, how quickly the animals relearn to reacquire the task following completion of extinction (calculate the difference between responding during reinstatement of reward and responding during the last extinction session). 8. In addition to looking at how long it takes for the subjects to “unlearn” the task, how quickly they “relapse” or relearn the task after undergoing extinction learning when some or all of the conditioned rewards (food, light, tone) are reintroduced can also be analyzed [12, 13] (see Notes 54 and 55).

3.8

4

Troubleshooting

A number of common issues may arise during touchscreen experiments. Many problems can be dealt with quickly by having spare parts available, including touchscreen connector cables, IR beam assembly pieces, touchscreens, and tubing/syringes/needles for liquid rewards. Table 4, adapted from Refs. [37, 62, 76], summarizes some obstacles, why they may arise, and possible solutions, although step-specific troubleshooting issues appear throughout the protocol above.

Notes 1. The touchscreen platform offered by Lafayette Instrument (Camden Instruments in the UK) is the platform with which we have had experience and which is the focus of this chapter. For this platform, there are currently ten verified paradigms available for rodent testing, nine of which are specifically verified for mice. Aside from Lafayette, there are other sources for

310

Katherine M. Cotter et al.

Table 4 Overview of Troubleshooting Tipsa Problem

Possible reason

Solution

Incomplete reward consumption

Insufficient food restriction (if using reward requiring food restriction)

Monitor cage for evidence of food grinding (see Note 41), increase food restriction (within regulations) Provide reward in home cage for additional days

Animal is insufficiently habituated to reward Unstable or poor performance

Inconsistent motivation

Aversion to mask or touchscreen

Excessive fighting in home cage

Stressors in the housing room (noise, light, etc.) Poor learning ability Abrupt decline in performance and/or trial completion

Touchscreen error (nonresponsive, not Check physical connections; clean; displaying images) run test program; recalibrate; reboot the system (see Note 35) Inconsistent or stopped reward Check for reward line blockage (see delivery Note 46) or disconnection; check for interface error; replace (line, cords, or whole reward dispenser) Initiation not detected Clean magazine photobeam; check physical connections and replace if necessary Controlling system error (software or Check physical connections; reboot hardware) system; change hardware if necessary

Animal appears to make Infrared beam failure unusually low/high number of beam crosses ABET II failure to load next run or save data

Increase attention to weight control; consider temporary feeding separation according to rate of response Increase exploration of mask and screen by applying a food reward directly to the mask or touchscreen Monitor home cage and general health, separate animals if necessary Frequently observe rooms and cages, move animals if needed Exclusion from study may be necessary (see Note 42)

Clean infrared beam pathway; check position of infrared switch; replace faulty beams if necessary

Database number of saved files reached If the next trial can still be loaded and data collected, the database can be changed after data collection by selecting “yes.” If the program doesn’t load the next trial, the error message “You will need to create a new database. Exit out of the ABET II TOUCH program, and restart it” will appear (continued)

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

311

Table 4 (continued) Problem

Possible reason

Solution

Incorrect variables within schedule files of an environment template

Incorrectly entered while creating the Click on the box that must be updated and type in a new session experiment file; once parameters are variable value. To ensure the finalized, save the file changes are applied, click outside of the box or select enter on the keyboard. Check that the updates were saved by selecting a different environment ID, then reselecting the environment to which the changes occurred

Whisker Server or other The Whisker Server window is open on Do NOT exit out of Whisker Server; minimize the window on the windows appear on the computer monitor computer monitor and it should the touchscreen [20] The “Topmost” setting in Whisker – disappear from the touchscreen Display Settings is unchecked. This display for immediate resolution would usually only occur if settings Long term solution should be to were reset as result of a windows check Whisker Server—Display update Devices—Topmost setting a

Adapted from Refs. [37, 62, 76]

touchscreen platforms [81], and instructions are also available for do-it-yourself touchscreen setups for rodents and other lab animals [73, 99]. If the cost and effort of setting up for your individual lab is still out of reach, another cost-effective approach is to work toward having an institutional core facility with a dedicated staff for running touchscreen experiments. With these alternatives, we hope the perceived cost of setting up touchscreen testing for mice will not dissuade interested readers. There are also additions and modifications that can be made to existing touchscreen platforms to further automate experimentation and thus aim to decrease stress via handling by researchers [82, 83]. 2. Recent studies have also shown that performance on LDR is also impaired following stroke to the prefrontal cortex, highlighting that these tests rely on a spatial component that requires intact hippocampal-prefrontal circuitry to perform. 3. Of the four paradigms mentioned here, PAL, LDR, and EXT require the standard Pretraining (sometimes referred to in papers as “General Touchscreen Training”) prior to TaskSpecific Training. The training times listed under each task are the duration of Task-Specific Training after Pretraining has been completed. AUTO does not require the Pretraining procedure, only Task-Specific Training. See Table 1 and Subheading 3 for more details.

312

Katherine M. Cotter et al.

4. If the goal of a study is to evaluate how the experimental manipulation influences overall learning, experimental manipulations can take place after Pretraining but prior to the TaskSpecific Training. However, if the goal of the study is to evaluate how the experimental manipulation influences retention of a previously learned task, then the experimental manipulation can take place after the Task-Specific Training or even after baseline performance in the Test phase of the Task has been collected [33, 57, 67]. 5. Readers may find watching the Lafayette Instrumentsponsored 2016 webinar “Using Touchscreen Operant Systems to Study Cognitive Behaviors in Rodents” [43] as a good starting point in understanding the background and use of the touchscreen systems prior to reading existing literature. In addition to the webinar by Inside Scientific [43], Lafayette Instrument has short videos on their website of the tasks being run [1] as well as lists of citations which use each task. 6. Researchers should consider strain, sex, and age when planning their studies. Some mouse strains exhibit deficits in some touchscreen tasks. Examples: The R6/2 strain shows deficits in PD [66], the 3xTgAD strain shows deficits in the 5-CSRT task [84], and the DBA/2 J strain shows deficits in EXT [62]. Other mouse strains may also have phenotypes that lead to deficits in touchscreen task performance, particularly if the strain is susceptible to visual impairments. Researchers are advised to carefully review the literature and to perform pilot studies with their strain to ensure that the range of output measures is as expected. Also, if both female and mice are to be tested, researchers should familiarize themselves with best practices for working with both sexes, as well as persistent myths that prevent many researchers from considering sex as a biological variable in their studies [8, 21, 63, 80, 89– 91]. Finally, with regard to age, touchscreens have been widely used to assess rodent cognitive changes that occur with aging (up to 21 months of age) [17] and in early life as well (as young as 5 weeks of age) [81], and some studies report both young (3-month-old) vs. older (12- or 15-month-old) mice on touchscreen tasks in the same publication [34, 102]. 7. Ear tagging is an effective system to identify mice during weighing and mouse management activities. During ear tag application, we found it is also helpful to apply even numbered tags to the right ear of the animal and odd-numbered tags to the left ear as an additional way to distinguish mice in a cage. However, confirmation of tag numbers requires scruffing, which can introduce an additional stress factor to experimental mice. If opting for this method of identification, it is important that the experimenter is able to scruff effectively prior to

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

313

working with experimental mice. It is also important to limit the number of events in which mice must be scruffed. Furthermore, it is difficult to read the small metal ear tags when in the red light during the experiment; therefore, the secondary form of identification (e.g., tail marks) is necessary. 8. We found ear punches especially helpful when verifying the tail marks during testing and weighing, since the mouse can be placed on the arm of the experimenter to view the punches and quickly identify the animal. Ear punches are an easy way for experimenters with little prior rodent handling to minimize scruffing the animals, which is helpful if creating rotation of the experimenter who performs the daily touch screen experiment. However, if ear punches are not done consistently, it can cause error when identifying the animals. Ensure the experimenter creating the ear punches has a secure scruff. Furthermore, since the ear tissue is thin, the ear punches can occasionally become deformed. It is important to record any changes of the ear punch appearance as soon as they are noticed during experimentation. 9. Multiple methods of permanent mouse identification are acceptable. However, we find a second form of identification is helpful specifically for identifying mice in the red light of a behavior room. The red light can make reading ear tags or ear punches difficult, and constant scruffing before the TS experiment can stress the mice. Therefore, it is best to avoid relying solely on the ear identification system during experimentation. Tail marking is an effective secondary measure to identify animals [16] and can be done with a permanent marker (e.g., Sharpie®) near the base of the tail. We suggest tail marks are re-applied weekly, before or after weighing the animals. Certain colors and brands fade at different rates; however, in our experience, dark purple or dark blue Sharpies work best when red light is in use. If using red light, colors such as red, pink, and orange are very difficult to distinguish. When assigning tail marks to animals, keep markings consistent with chamber assignment to decrease experimental error, especially if using multiple chambers. For example, the mouse with no mark is assigned to Chamber 1, the mouse with one mark is assigned to Chamber 2, the mouse with two marks is assigned to Chamber 3, and the mouse with three marks is assigned to Chamber 4. 10. This chapter is written with regard to the Bussey-Saksida Mouse Touch Screen Chamber by Lafayette Instrument (Cat. #80614, Lafayette, USA) and the Easy-Install System for Mouse Touch Screen Systems (Cat. #80614-20, Lafayette, USA) for dispensing liquid reward. This system contains four pre-cabled Bussey-Saksida Mouse Touch Screen Chambers in separated sound-attenuated cubicles. Each chamber comes

314

Katherine M. Cotter et al.

with glass jars which are used to deliver liquid reward and the tubing as well; replacements numbers are Lafayette #80204-4 10 glass jars, 30 mL Watson-Marlow #913.A005.016 0.5 mm ID/1.6 mm OD silicone tubing. Other suppliers may be used, but the hardware and interface of equipment setup and use may differ from the methods listed here. In addition, your cleaning and maintenance will depend on the specific parts in your setup. For example, if you have a different pump (perhaps to deliver a thicker liquid reward), you may have different tubing and then would use a different needle to clean the tubing. Also, here our reward magazine (the point in the reward hopper where the liquid is precisely dispensed) is quite delicate, and therefore we use KimWipes or any other fine lab wipes to clean this area. If you opt to use a pellet, your cleaning wipe may be distinct from this. The Easy-Install system is a one-piece, pre-cabled system with all the stations connected to the control interface. This piece requires only one AC power plug, which is electrically safe and reduces the number of cables. The computer and interface software are not included in the Easy-Install system but can be ordered from Lafayette Instrument and will be incorporated into the system station. Additionally, IR-sensitive cameras and equipment for video monitoring of the chamber interiors can be procured and added to the EasyInstall system. 11. These touchscreens are really a grid of infrared (IR) beams, not a pressure-sensitive screen. Therefore, they do not require a physical “touch” or pressure to be exerted by the mice, just the breaking of the IR beam. Also, there is a distinction here between touching the stimulus location vs. touching the screen (blank touch). Touching the stimulus location will result in a reward, whereas touching the screen in any other location (blank touch) will not. 12. The terms “reinforcer” or “reward” are often used throughout the field of behavioral neuroscience to indicate something that reinforces a behavior. We will use these terms interchangeably. 13. The equipment that delivers the reinforcement and appropriate tubing are available for liquid or pellets (for liquid, a pump; for pellets, a dispenser). 14. Utilizing IR-sensitive camera equipment with computer monitors is helpful to visualize the mice when they are actually in the touchscreen chambers to ensure the equipment is working appropriately during test sessions. 15. Touchscreen software is usually offered by the touchscreen chamber supplier. The system used here is ABET II Touch (Cat. #89505, Lafayette Instrument, USA).

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

315

16. The general ABET II Touch software includes functions only for initial base training (which is applicable to multiple tasks) and EXT. Additional software is purchased for touchscreen tasks, including the AUTO, LDR, and PAL discussed in this chapter. Other touchscreen tasks not included in this chapter are available from Lafayette Instrument and are listed below; readers are encouraged to consider what their specific needs are for testing prior to purchase: (a) Pairwise Discrimination (PD)/Reversal Testing (ABET II software, Cat. #89540) (b) Visuomotor Conditional Learning (VMCL; ABET II software, Cat. #89542) (c) 5-Choice Serial Reaction Time Task (5-CSRT; ABET II software, Cat. #89543) (d) Trial-Unique Nonmatching-to-Location Task (TUNL; ABET II software, Cat. #89545-1) (e) Progressive Ratio (PR) and Effort-Related Choice Task (ERC; ABET II software, Cat. #89549) (f) Rodent Continuous Performance Task (rCTP; ABET II software, Cat. #89551) 17. As with many operant behaviors employed in the field of behavioral neuroscience, there are a variety of reinforcers (or rewards) that can be used with the touchscreen platform. A common one is a milk-based liquid reward [53]. This current protocol is written with the milk-based nutritional shake Ensure® Original Strawberry Nutrition Shake (“milkshake”) as the reinforcer. For other operant behaviors, groups use food or sucrose pellets, peanut oil, juice, or even water. Whether a given reward requires food (or water), restriction should be empirically assessed in each laboratory and likely for each strain, sex, and age of mouse tested. 18. An important consideration in comparing data across laboratories is the nutritional and motivational content of the reinforcer used, particularly in light of the worldwide use of touchscreens and the lack of a universal source or manufacturer for reinforcers that are used. Studies on this question conclude that regardless of reinforcer used, caloric content should be matched across laboratories in an effort to overcome noncaloric aspects of the reinforcers [53, 78]. 19. If using a milk-based liquid reward as a reinforcer, as we do, you may want to follow protocols for food restriction to increase motivation [103, 104]. However, food restriction protocols themselves come with caveats and therefore some groups opt to not restrict food intake [28, 44, 55, 61, 92]. When selecting a reward, make sure to follow the

316

Katherine M. Cotter et al.

supplier’s instructions to preserve quality of the food. For example, with milk-based rewards, products can be kept at room temperature until the product is opened; however, upon opening, the remaining product must be kept in a refrigerator. Record the dates opened on the container and throw away the container after using all of the product or after the expiring date has passed. When planning a touch screen experiment, it is important to have enough reward on hand to ensure execution of the experiment is possible; however, do not stock more than you can use before the reward expiration date. 20. Each chamber comes with a set of two jars and 1 m of reward tubing standard. The jars used for cleaning (water and EtOH) should be kept separate from jars used for dispensing the reward (milkshake). All jars should be labeled to avoid confusion between liquids. This will prevent contamination of the reward jars with EtOH and thus reduce cleaning time. Jars should be thoroughly rinsed daily with hot water if using a milk-based reward to prevent residue and mold. 21. This rack is needed to allow the mice to habituate to a quiet environment, ideally the touchscreen room (or an environment very similar to it). A large/tall rack allows all the cages to share a similar environment rather than spread over a larger space. If the large rack can be kept next to the chambers or at least nearby the testing room, this will minimize disturbance (e.g., mice won’t have to be transported on a cart or carried a long distance immediately before testing). Keep a record of any disturbances that may occur in the habituation environment during the habituation period. Note that this “habituation to environment” is distinct from the “Habituation” stage of touchscreen Pretraining discussed in a later section. 22. Use the cart to transport cages to and from the testing room. If the cart is waist-high, it doubles as a surface to open cages and handle mice while in the testing room. Mouse handling should not be done on the habituation rack (see Note 22). If a waisthigh cart is unavailable, a work table with wheels can be used. The cart or table can be moved outside of the testing room when cleaning occurs if space is needed. 23. The Daily Runsheet is helpful to keep track of chamber assignments while in the testing room. Information on the Runsheet may include: session start time, cage number, animal number, tail mark number, and experiment details. For example, when running the software “LDR1 Choice Reversal,” it is important to keep track of the correct side of the positive stimuli assigned and the difficulty assigned to each animal throughout testing. A sample Runsheet is provided here for the reader’s convenience: https://bit.ly/3ABPQP7.

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

317

24. Throughout a given experiment, ideally the identity of the experimenters performing specific daily tasks is kept constant (e.g., the same experimenter(s) should give weekly tail marks). Rodents are sensitive to variables such as odor and pheromones, and variance in handling techniques can create extraneous stressors. Ensure experimenters are trained in a standardized fashion to allow for similar testing conditions for all subjects. Keep track of the experimenters and each task they perform for record keeping purposes. 25. Prior to your experiments, ensure your chambers are assembled and software is installed and updated. After setting up all devices and hardware, manually disconnect the computer from the Internet. If the computer is left connected to the Internet, software updates may be downloaded without notification. Software updates may interfere with the testing schedule. Reconnecting to the Internet may be necessary for specific troubleshooting; however, after troubleshooting, ensure the Internet is turned off again. 26. Our food restriction protocol steps all occur in the animal housing room (not in the behavior testing room). Mice receive food ad libitum in their food hopper in their home cage for 3–4 h immediately after the daily touchscreen session is complete. If mice only run on touchscreen testing Monday through Friday, mice also receive food ad libitum from after testing is complete on Friday until Sunday at 5 pm (when food is taken away). Food is provided in the cage hopper; however, to minimize the dust and crumbs falling into the cage (which would defeat the purpose of food restriction), we suggest removing the food hopper from the cage, filling it with whole pellets (not partial or crumbling pellets), firmly tapping the hopper over the food bin or trash can to remove dust or crumbs, and then place the hopper into the cage. As cages are staggered in the completion of their daily touchscreen sessions, providing food for the 3–4 h period can also be staggered, as long as all cages receive their food for the same period of time. Document the time food is provided to the last cage; the animals should have access to food for 3–4 h a day. When removing food in the evening, remove food from the cages in the same order it was provided so that all cages get equal amount of time with food. The food restriction weight threshold may differ between institutions; follow your institution’s animal care and use guidelines. A common guideline is that mice who reach a weight below 80% of their mean free-feeding weight should be removed from the study. In our experience, we have not had animals reach or even approach this threshold, but CNS injuries have broad effects so weight and weight gain must be monitored carefully.

318

Katherine M. Cotter et al.

27. At the start of each experimental day (8:00 AM), retrieve the cages for just the day’s first round of touchscreen behavior from the animal holding room; if you have 8 chambers, these might be two cages of 4 mice each. Transport these via cart to the testing room and begin habituation to the room while preparing the chamber setup. See Subheading 3.2 for setup tasks, and keep noise levels minimal while the first mice are habituating. As soon as the experimental session begins for the first two cages, move the rest of the subjects into the test room on a larger habituation rack. Continue to be mindful about the noise because the mice in the chambers are extra sensitive to sound and vibrations. Record habituation times on the Daily Runsheet. 28. The day before each experimental day, it is important to prepare the schedule files that will be used during the experiment. See Table 3 for sample ABET II Schedules and Session Variables. It is recommended to prepare the experimental file immediately after cleaning the testing room (prior to food deprivation). Before running the program, the experimenter should quickly verify each schedule is properly assigned to the animals, and the variables are correct. It is optional, but sometimes necessary, to have a third check of the experimental files before they are loaded. The third check can be done after food deprivation. Each file name should include the cage number, test type (PAL, LDR, EXT, AUTO), and date (e.g., Cage3_LD04082022). Each file should include animal ID, group ID/cage number, schedule, database assignment, and environment (chamber assignment). 29. We recommend turning on the instruments in this order to avoid technological issues during setup. If the devices are turned on out of order, the video monitoring system may not work. Furthermore, allow a 30 s pause before turning on the next device in this sequence. A beep should sound signaling the device is operating. Overloading the system all at once may lead to the need for a reboot before testing. 30. If no EtOH is visible coming out of the reward magazine, flush with hot water for 1 min. If there is still no visible liquid, this indicates clogging. Detach the reward pumpline and probe the tubing with a 30 gauge needle. Probe the reward dispensary prongs in the reward magazine with a 26 gauge needle. If a clump is visible within the tube that is out of reach of the needle, you may also try massaging the tube with a hemostat or fingers to loosen the material. Follow these steps by rinsing again with hot water. 31. When testing the functionality of the IR beams using the Mouse Test Lines schedule, place a hand between the beam and the chamber. A sound should go off when the right beam is

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

319

broken and the house light should turn on when the left beam is broken. If no response from the chamber is received, wipe down the IR beam with a paper towel sprayed with EtOH. Check that the switch on the beam is toggled to the left side. If the switch is toggled to the right, the beam will incorrectly detect mouse movement during trials. To ensure the touchscreen works prior to experimentation, “become the mouse.” Tap each window with a gloved hand and visually check the screen illuminates in that window. Tap each window again to turn off the screen. Place a gloved finger inside the reward magazine to check proper dispensing of reward. Reward should dispense upon finger insertion, and a 3 Hz tone should sound. 32. Loading/configuring the ABET II software ahead of time reduces the time the mice spend in the chamber not attending to a task. 33. When working with multiple cages during one session, ensure only one cage is opened at a time to keep lids paired with their respective cages. 34. To reduce variability in performance, place each mouse into the same touchscreen chamber for each session and each stage (Pretraining, Task-Specific Training, and Task-Specific Testing). 35. Before starting (playing) the program, verify only one rodent is in each chamber via the video monitors. After playing the program, watch the video monitors to ensure the reward magazine and touchscreen are working properly. If errors in the experiment are noticed in the first 5 min of the session, stop the session, fix the errors, and restart the session. If errors are noticed after 5 min, let the session run to completion to keep the experimental conditions for the mice consistent. 36. It is important to record the end of Session Variables for each animal at the end of every session to aid in the preparation of the next day’s experimental file and Daily Runsheet creation. These variables can be found under the “Monitor Tab.” Some sessions may not have a value if the mouse did not reach acquisition within the session time limit. In analysis, include experimental conditions and separation. 37. If a mouse finishes a session before the session time elapses, quietly remove it from the chamber and place it back into its home cage to minimize extinction learning. This must be done quietly, as the performance of other mice still testing can be influenced by sounds and vibrations. If a mouse finishes after 25 min, do not remove it early; take it out when mice in all the other chambers have completed their sessions.

320

Katherine M. Cotter et al.

38. Do not spray EtOH directly into the chamber or on the screen. Too much EtOH applied directly to the chamber interior may be harmful to the chamber materials and to the mice. Excessive EtOH may result in a lingering, aversive odor which may influence behavior. 39. Of the tasks covered in this chapter, AUTO [37] and EXT [62] require the use of a modified Pretraining procedure; AUTO only requires habituation, whereas EXT requires Habituation, Initial touch, Must touch, and Must initiation training but not Punish Incorrect. Tasks not covered in this chapter that also require a modified Pretraining procedure include 5-CSRT [62], TUNL [76], and VMCL [37]. 40. If using a reward that does not require food restriction, then continue to Habituation. 41. Note if mice are grinding their food in their home cage; a thick dusting and/or crumbs will be evident near the food hopper, on top of the bedding, and even on the bottom of the cage/ under the bedding. Such “grinders” may need to have their cages replaced more frequently, as the food restriction may not be effective in increasing motivation. In our facility, we have special cage cards that indicate when signs of grinding have been observed by animal caretakers or experimenters to ease monitoring of these mice. 42. If a mouse does not meet the criteria for the Pretraining stage during a daily session, have the mouse repeat the Pretraining stage until completion. If the mouse is unable to complete all Pretraining and LDR Train stages before beginning the LDR Test stage, have the mouse complete the LDR Test stage as if normal. Keep the animal with its littermates to reduce potential stress to the mice that are still running. Exclude data for this animal from analysis. This is a generally applicable statement to all tasks; it is not LDR-specific. 43. Individual mice will vary in how quickly they advance through the Pretraining/Task-Specific Training stages. However, it is not recommended to keep running mice that have reached training criteria just so other mice can “catch up”; overtraining can have undesired effects on later performance. One solution to the conundrum of the varying speed of mice in these training steps is to “rest” mice that have completed Pretraining, running them only 1–2× per week (reminder sessions). If a mouse has completed training and then falls below the criteria for PI/Stage 5 or the Task-Specific Training, resume regular training until the criteria are reached again. After the whole cohort has achieved criteria at least once, run a baseline (all animals trained daily), and then proceed to the next step. This allows the group to be synchronized on the same step while

Touchscreen Cognitive Tests in Mouse Models of Brain Injury

321

also reducing variation in performance across the group and minimizing overtraining. Alternatives to this approach are also described in the literature [76]. 44. Habituation 1 will look slightly different during execution than all other Pretraining stages. Habituation 1 is an acclimation period for the mice to become adjusted to the chamber. There are no criteria for the subjects to complete; therefore, all animals move onto Habituation 2 the next day and no notes regarding the schedule variables or criteria reaching performance should be recorded on the Daily Runsheet. Also, because there is no chamber activity, do not be alarmed when the house light, reward chamber, and touch screen do not activate while the mouse is in the chamber. Continue to check that all rodents are in the chamber via the touchscreen monitor before starting the program. 45. At these and other steps, check the magazine to make sure all reward has been consumed. 46. If the reward pumpline becomes clogged, stop the program, and use the “manual feed/prime” switch on the pump and then check if the reward is dispensed. If it does not dispense properly, there may be an air bubble in the line. 47. Some studies use 30 responses in a 60 min session for Pretraining [62, 76], and some omit Pretraining altogether [97]. 48. A 20 s intertrial interval (ITI) is the standard ITI duration, but it can be lengthened or shortened. EXT can use a 5 or 10 s ITI, and AUTO often uses a 10–40 s variable ITI. 49. Sample experimental manipulations are drug treatments, stroke, TBI, exercise, etc. If the goal of the study is to evaluate how the experimental manipulation influences overall learning, experimental manipulations can take place after Pretraining but prior to the Task-Specific Training (LDR, PAL) or Acquisition (EXT). However, if the goal of the study is to evaluate how the experimental manipulation influences retention of a previously learned task, then the experimental manipulation will take place after the Task-Specific Training or Acquisition or even after baseline performance in the Test phase of the Task has been collected [37, 62, 76]. 50. In the first paper to describe PAL for the rodent touchscreen platform [98], the three images used were a flower, airplane, and spider, known as the “flower-plane-spider” stimulus combination. An alternate presentation of stimuli uses distinct line patterns and has been proposed to have less variability [37, 85]. 51. Behavioral pattern separation is thought to be reflective of the computational process, pattern separation; these are distinct

322

Katherine M. Cotter et al.

terms. Interested readers are referred to excellent resources elsewhere for more on this difference and the importance of this difference [24, 31, 41, 50, 51, 56, 95]. 52. Some studies do not include a reversal component, and some expose rodents to both Large and Small Separation on the same day. Most studies expose rodents to only one kind of separation per day (Large or Small). 53. Differences of opinions and protocols exist in regard to cognitive flexibility. For example, some studies consider reversal number as relevant (or equivalent) to behavioral pattern separation. Our perspective is that performance up until the first reversal trials each session reflects behavioral pattern separation ability, while the performance after the first reversal each session reflects cognitive flexibility [97]. 54. If doing relapse analysis following extinction learning, various methods have been used due to the dependency on the experiment. As summarized elsewhere [62], these include contingent partial re-exposure, noncontingent partial re-exposure, and variations of when these are reintroduced and to how many sessions they apply. 55. Python codes that the Eisch Lab used for EXT analysis and other behavioral tests are available at https://github.com/ raymon-shi/ts-data-analysis-app

Acknowledgments NIH (NS 088555-07A1, PI: AMS), NASA (80NSSC21K0814, PI: SY), University of Pennsylvania (Dept Radiation Oncology Pilot Grant, co-PI: AJE; Undergraduate Research Foundation, PI: SY; Trainees: GLB, HAH), the Brain and Behavior Research Foundation (2019 NARSAD Young Investigator Grant, PI: SY), Children Hospital of Philadelphia (2022 Forderer Grant, co-PI: SY). References 1. Anonymous Bussey-Saksida mouse touch screen chamber package 2. Anonymous Specifications: Minimum Computer System Requirement 3. Anonymous Universal Earpunch Mouse Numbering System 4. Arulsamy A, Corrigan F, Collins-Praino LE (2019) Age, but not severity of injury, mediates decline in executive function: validation of the rodent touchscreen paradigm for preclinical models of traumatic brain injury. Behav Brain Res 368:111912 5. Bartko SJ, Vendrell I, Saksida LM et al (2011) A computer-automated touchscreen paired-

associates learning (PAL) task for mice: impairments following administration of scopolamine or dicyclomine and improvements following donepezil. Psychopharmacology 214:537–548 6. Basso DM (2004) Behavioral testing after spinal cord injury: congruities, complexities, and controversies. J Neurotrauma 21:395–404 7. Belarde JA, Chen CW, Rafikian E et al (2021) Optimizing touchscreen measures of rodent cognition by eliminating image bias. bioRxiv 8. Beltz AM, Beery AK, Becker JB (2019) Analysis of sex differences in pre-clinical and

Touchscreen Cognitive Tests in Mouse Models of Brain Injury clinical data sets. Neuropsychopharmacology 44:2155–2158 9. Benevento M, Oomen CA, Horner AE et al (2017) Haploinsufficiency of EHMT1 improves pattern separation and increases hippocampal cell proliferation. Sci Rep 7:40284 10. Beraldo FH, Palmer D, Memar S et al (2019) MouseBytes, an open-access high-throughput pipeline and database for rodent touchscreenbased cognitive assessment. Elife 8:e49630 11. Bharmal AV, Kent BA, Bussey TJ et al (2015) Performance of transgenic TgTau-P301L mice in a 5-choice serial reaction time task (5-CSRTT) as a model of Alzheimer’s disease. Psychiatr Danub 27(Suppl 1):S515–S525 12. Bouton ME (2002) Context, ambiguity, and unlearning: sources of relapse after behavioral extinction. Biol Psychiatry 52:976–986 13. Bouton ME, Winterbauer NE, Todd TP (2012) Relapse processes after the extinction of instrumental learning: renewal, resurgence, and reacquisition. Behav Process 90:130–141 14. Brigman JL, Feyder M, Saksida LM et al (2008) Impaired discrimination learning in mice lacking the NMDA receptor NR2A subunit. Learn Mem 15:50–54 15. Broschard MB, Kim J, Love BC et al (2021) Category learning in rodents using touchscreen-based tasks. Genes Brain Behav 20:e12665 16. Burn CC, Mazlan NHB, Chancellor N et al (2021) The pen is milder than the blade: identification marking mice using ink on the tail appears more humane than ear-punching even with local anaesthetic. Animals (Basel) 11:1664 17. Buscher N, Van Dorsselaer P, Steckler T et al (2017) Evaluating aged mice in three touchscreen tests that differ in visual demands: impaired cognitive function and impaired visual abilities. Behav Brain Res 333:142–149 18. Bussey TJ, Holmes A, Lyon L et al (2012) New translational assays for preclinical modelling of cognition in schizophrenia: the touchscreen testing method for mice and rats. Neuropharmacology 62:1191–1203 19. Bussey TJ, Muir JL, Robbins TW (1994) A novel automated touchscreen procedure for assessing learning in the rat using computer graphic stimuli. Neurosci Res Commun 15: 103–110 20. Cardinal RN, Aitken MRF (2010) Whisker: a client-server high-performance multimedia research control system. Behav Res Methods 42:1059–1071 21. Carroll ME, Lynch WJ (2016) How to study sex differences in addiction using animal models. Addict Biol 21:1007–1029

323

22. Chow WZ, Ong LK, Kluge MG et al (2020) Similar cognitive deficits in mice and humans in the chronic phase post-stroke identified using the touchscreen-based paired-associate learning task. Sci Rep 10:1–8 23. Chow WZ, Ong LK, Kluge MG et al (2020) Similar cognitive deficits in mice and humans in the chronic phase post-stroke identified using the touchscreen-based paired-associate learning task. Sci Rep 10:19545 24. Christian KM, Ming G-L, Song H (2020) Adult neurogenesis and the dentate gyrus: predicting function from form. Behav Brain Res 379:112346 25. Chudasama Y, Robbins TW (2003) Dissociable contributions of the orbitofrontal and infralimbic cortex to pavlovian autoshaping and discrimination reversal learning: further evidence for the functional heterogeneity of the rodent frontal cortex. J Neurosci 23: 8771–8780 26. Delotterie D, Mathis C, Cassel J-C et al (2014) Optimization of touchscreen-based behavioral paradigms in mice: implications for building a battery of tasks taxing learning and memory functions. PLoS One 9:e100817 27. Doostdar N, Airey J, Radulescu CI et al (2021) Multi-scale network imaging in a mouse model of amyloidosis. Cell Calcium 95:102365 28. Dumont JR, Salewski R, Beraldo F (2020) Critical mass: the rise of a touchscreen technology community for rodent cognitive testing. Genes Brain Behav 20:e12650 29. Fouad K, Ng C, Basso DM (2020) Behavioral testing in animal models of spinal cord injury. Exp Neurol 333:113410 30. Fujimoto ST, Longhi L, Saatman KE et al (2004) Motor and cognitive function evaluation following experimental traumatic brain injury. Neurosci Biobehav Rev 28:365–378 31. Gilbert PE, Holden HM, Sheppard DP et al (2016) Pattern separation: a key processing deficit associated with aging? In: Jackson PA, Chiba AA, Berman RF, Ragozzino ME (eds) The neurobiological basis of memory: a system, attribute, and process analysis. Springer International Publishing, Cham, pp 115–135 32. Graf R, Longo JL, Hughes ZA (2018) The location discrimination reversal task in mice is sensitive to deficits in performance caused by aging, pharmacological and other challenges. J Psychopharmacol 32:1027–1036 33. Graybeal C, Bachu M, Mozhui K et al (2014) Strains and stressors: an analysis of touchscreen learning in genetically diverse mouse strains. PLoS One 9:e87745 34. Harb MR, Sousa N, Zihl J et al (2014) Reward components of feeding behavior are

324

Katherine M. Cotter et al.

preserved during mouse aging. Front Aging Neurosci 6:242 35. Heath CJ, O’callaghan C, Mason SL et al (2019) A touchscreen motivation assessment evaluated in Huntington’s disease patients and R6/1 model mice. Front Neurol 10:858 36. Heath CJ, Phillips BU, Bussey TJ et al (2016) Measuring motivation and reward-related decision making in the rodent operant touchscreen system. Curr Protoc Neurosci 74:8–34 37. Horner AE, Heath CJ, Hvoslef-Eide M et al (2013) The touchscreen operant platform for testing learning and memory in rats and mice. Nat Protoc 8:1961–1984 38. Houlton J, Barwick D, Clarkson AN (2021) Frontal cortex stroke-induced impairment in spatial working memory on the trial-unique nonmatching-to-location task in mice. Neurobiolof Learn Mem 177:107355 39. Houlton J, Zhou LYY, Barwick D et al (2019) Stroke induces a BDNF-dependent improvement in cognitive flexibility in aged mice. Neural Plast 2019:1460890 40. Hvoslef-Eide M, Nilsson SRO, Saksida LM et al (2016) Cognitive translation using the rodent touchscreen testing approach. In: Robbins TW, Sahakian BJ (eds) Translational neuropsychopharmacology. Springer International Publishing, Cham, pp 423–447 41. Hvoslef-Eide M, Oomen CA (2016) Adult neurogenesis and pattern separation in rodents: a critical evaluation of data, tasks and interpretation. Front Biol 11:168–181 42. Inglis JK (1980) Introduction to laboratory animal science and technology. Pergamon Press 43. Insidescientific (2016) Using touchscreen operant systems to study cognitive behaviors in rodents. In:Youtube, p 3741 sec 44. Janickova H, Kljakic O, Robbins TW et al (2021) Evaluating sequential response learning in the rodent operant touchscreen system. Curr Protoc 1:e268 45. Jobson DD, Hase Y, Clarkson AN et al (2021) The role of the medial prefrontal cortex in cognition, ageing and dementia. Brain Commun 3:fcab125 46. Jones NC, Cardamone L, Williams JP et al (2008) Experimental traumatic brain injury induces a pervasive hyperanxious phenotype in rats. J Neurotrauma 25:1367–1374 47. Kangas BD, Bergman J (2017) Touchscreen technology in the study of cognition-related behavior. Behav Pharmacol 28:623–629 48. Kazanis I (2005) CNS injury research; reviewing the last decade: methodological errors and

a proposal for a new strategy. Brain Res Brain Res Rev 50:377–386 49. Kent BA, Heath CJ, Kim CH et al (2018) Longitudinal evaluation of Tau-P301L transgenic mice reveals no cognitive impairments at 17 months of age. Brain Behav 8:e00896 50. Kesner RP, Lee I, Gilbert P (2004) A behavioral assessment of hippocampal function based on a subregional analysis. Rev Neurosci 15:333–351 51. Kesner RP, Rolls ET (2015) A computational theory of hippocampal function, and tests of the theory: new developments. Neurosci Biobehav Rev 48:92–147 52. Kim CH, Heath CJ, Kent BA et al (2015) The role of the dorsal hippocampus in two versions of the touchscreen automated paired associates learning (PAL) task for mice. Psychopharmacology 232:3899–3910 53. Kim EW, Phillips BU, Heath CJ et al (2017) Optimizing reproducibility of operant testing through reinforcer standardization: identification of key nutritional constituents determining reward strength in touchscreens. Mol Brain 10:31 54. Kim M, Kwak C, Yu N-K et al (2016) Optimization of the touchscreen paired-associate learning (PAL) task for mice and its dorsal hippocampal dependency. Anim Cells Syst 20:229–236 55. Krakenberg V, Wewer M, Palme R et al (2021) Regular touchscreen training affects faecal corticosterone metabolites and anxiety-like behaviour in mice. Behav Brain Res 401:113080 56. Lacar B, Parylak SL, Vadodaria KC et al (2014) Increasing the resolution of the adult neurogenesis picture. F1000Prime Rep 6:8 57. Lafayette Instrument Company I (2022) Bussey-Saksida Touch Screen – Standard Tasks and Bibliography 58. Levine DA, Galecki AT, Langa KM et al (2015) Trajectory of cognitive decline after incident stroke. JAMA 314:41 59. Lim J, Kim E, Noh HJ et al (2019) Assessment of mGluR5 KO mice under conditions of low stress using a rodent touchscreen apparatus reveals impaired behavioural flexibility driven by perseverative responses. Mol Brain 12:37 60. Lopez-Cruz L, Bussey TJ, Saksida LM et al (2021) Using touchscreen-delivered cognitive assessments to address the principles of the 3Rs in behavioral sciences. Lab Anim 50: 174–184 61. Mallien AS, Palme R, Richetto J et al (2016) Daily exposure to a touchscreen-paradigm

Touchscreen Cognitive Tests in Mouse Models of Brain Injury and associated food restriction evokes an increase in adrenocortical and neural activity in mice. Horm Behav 81:97–105 62. Mar AC, Horner AE, Nilsson SRO et al (2013) The touchscreen operant platform for assessing executive function in rats and mice. Nat Protoc 8:1985–2005 63. Mccarthy MM, Woolley CS, Arnold AP (2017) Incorporating sex as a biological variable in neuroscience: what do we gain. Nat Rev Neurosci 18:707 64. Mcquail JA, Dunn AR, Stern Y et al (2020) Cognitive reserve in model systems for mechanistic discovery: the importance of longitudinal studies. Front Aging Neurosci 12: 607685 65. Mctighe SM, Mar AC, Romberg C et al (2009) A new touchscreen test of pattern separation: effect of hippocampal lesions. Neuroreport 20:881–885 66. Morton AJ, Skillings E, Bussey TJ et al (2006) Measuring cognitive deficits in disabled mice using an automated interactive touchscreen system. Nat Methods 3:767–767 67. Muthukumar S, Mehrotra K, Fouda M et al (2022) Prenatal and postnatal insults differentially contribute to executive function and cognition: utilizing touchscreen technology for perinatal brain injury research. Exp Neurol 354:114104 68. Nagahara AH, Bernot T, Tuszynski MH (2010) Age-related cognitive deficits in rhesus monkeys mirror human deficits on an automated test battery. Neurobiol Aging 31: 1020–1031 69. Nichols JN, Hagan KL, Floyd CL (2017) Evaluation of touchscreen chambers to assess cognition in adult mice: effect of training and mild traumatic brain injury. J Neurotrauma 34:2481–2494 70. Nilsson SRO, Celada P, Fejgin K et al (2016) A mouse model of the 15q13.3 microdeletion syndrome shows prefrontal neurophysiological dysfunctions and attentional impairment. Psychopharmacology 233:2151–2163 71. Nithianantharajah J, Komiyama NH, Mckechanie A et al (2013) Synaptic scaffold evolution generated components of vertebrate cognitive complexity. Nat Neurosci 16:16–24 72. O’brien KR, Brait VH, Jackman KA et al (2020) Capturing longitudinal impacts on cognition following stroke in rodent models using touchscreen testing. Alzheimer’s Dement 16:e044156 73. O’leary JD, O’leary OF, Cryan JF et al (2018) A low-cost touchscreen operant chamber using a Raspberry Pi™. Behav Res Methods 50:2523–2530

325

74. O’sullivan MJ, Oestreich LKL, Wright P et al (2022) Cholinergic and hippocampal systems facilitate cross-domain cognitive recovery after stroke. Brain 145:1698–1710 75. Ong LK, Chow WZ, Tebay C et al (2018) Growth hormone improves cognitive function after experimental stroke. Stroke 49: 1257–1266 76. Oomen CA, Hvoslef-Eide M, Heath CJ et al (2013) The touchscreen operant platform for testing working memory and pattern separation in rats and mice. Nat Protoc 8:2006– 2021 77. Palmer D, Dumont JR, Dexter TD et al (2021) Touchscreen cognitive testing: crossspecies translation and co-clinical trials in neurodegenerative and neuropsychiatric disease. Neurobiol Learn Mem 182:107443 78. Phillips BU, Heath CJ, Ossowska Z et al (2017) Optimisation of cognitive performance in rodent operant (touchscreen) testing: evaluation and effects of reinforcer strength. Learn Behav 45:252–262 79. Phillips BU, Lopez-Cruz L, Hailwood J et al (2018) Translational approaches to evaluating motivation in laboratory rodents: conventional and touchscreen-based procedures. Curr Opin Behav Sci 22:21–27 80. Prendergast BJ, Onishi KG, Zucker I (2014) Female mice liberated for inclusion in neuroscience and biomedical research. Neurosci Biobehav Rev 40:1–5 81. Remmelink E, Chau U, Smit AB et al (2017) A one-week 5-choice serial reaction time task to measure impulsivity and attention in adult and adolescent mice. Sci Rep 7:42519 82. Rivalan M, Munawar H, Fuchs A et al (2017) An automated, experimenter-free method for the standardised, operant cognitive testing of rats. PLoS One 12:e0169476 83. Rivalan M, Munawar H, Fuchs A et al (2017) Correction: an automated, experimenter-free method for the standardised, operant cognitive testing of rats. PLoS One 12:e0176807 84. Romberg C, Bussey TJ, Saksida LM (2013) Paying more attention to attention: towards more comprehensive cognitive translation using mouse models of Alzheimer’s disease. Brain Res Bull 92:49–55 85. Saifullah MB, Komine O, Dong Y et al (2020) Touchscreen-based location discrimination and paired associate learning tasks detect cognitive impairment at an early stage in an App knock-in mouse model of Alzheimer’s disease. Mol Brain 13:1–13 86. Sanchez-Bezanilla S, Hood RJ, CollinsPraino LE et al (2021) More than motor impairment: a spatiotemporal analysis of

326

Katherine M. Cotter et al.

cognitive impairment and associated neuropathological changes following cortical photothrombotic stroke. J Cereb Blood Flow Metab 41:2439–2455 87. Schallert T (2006) Behavioral tests for preclinical intervention assessment. NeuroRx 3: 497–504 88. Semple B, Noble-Haeusslein L (2016) Behavioral testing after experimental traumatic brain injury. In: Cox CS (ed) Cellular therapy for neurological injury. CRC Press, pp 167–191 89. Shansky RM (2019) Are hormones a “female problem” for animal research? Science 364: 825–826 90. Shansky RM, Murphy AZ (2021) Considering sex as a biological variable will require a global shift in science culture. Nat Neurosci 24:457–464 91. Shepherd A, Tyebji S, Hannan AJ et al (2016) Translational assays for assessment of cognition in rodent models of Alzheimer’s disease and dementia. J Mol Neurosci 60:371–382 92. Siegel A, Bianchi DW, Guedj F (2020) Translating the human CANTAB touchscreen based tasks to evaluate learning and memory in mouse models of Down syndrome. bioRxiv 93. Soares FC, De Oliveira TCG, De Macedo LDED et al (2015) CANTAB object recognition and language tests to detect aging cognitive decline: an exploratory comparative study. Clin Interv Aging 10:37–48 94. Soler I, Yun S, Reynolds RP et al (2021) Multi-domain touchscreen-based cognitive assessment of C57BL/6J female mice shows whole-body exposure to 56Fe particle space radiation in maturity improves discrimination learning yet impairs stimulus-response rulebased habit learning. Front Behav Neurosci 15:722780 95. Stark SM, Kirwan CB, Stark CEL (2019) Mnemonic similarity task: a tool for assessing hippocampal integrity. Trends Cogn Sci 23: 938–951 96. Sullivan JA, Dumont JR, Memar S et al (2021) New frontiers in translational research: touchscreens, open science, and the mouse translational research accelerator platform. Genes Brain Behav 20:e12705 97. Swan AA, Clutton JE, Chary PK et al (2014) Characterization of the role of adult neurogenesis in touch-screen discrimination learning. Hippocampus 24:1581–1591 98. Talpos JC, Winters BD, Dias R et al (2009) A novel touchscreen-automated paired-

associate learning (PAL) task sensitive to pharmacological manipulation of the hippocampus: a translational rodent model of cognitive impairments in neurodegenerative disease. Psychopharmacology 205:157–168 99. Toegel F, Toegel C, Perone M (2021) Design and evaluation of a touchscreen apparatus for operant research with pigeons. J Exp Anal Behav 116:249–264 100. Tucker LB, Burke JF, Fu AH et al (2017) Neuropsychiatric symptom modeling in male and female C57BL/6J mice after experimental traumatic brain injury. J Neurotrauma 34: 890–905 101. Van Den Boom BJG, Mooij AH, Misevicˇiu ¯ te˙ I et al (2019) Behavioral flexibility in a mouse model for obsessive-compulsive disorder: impaired Pavlovian reversal learning in SAPAP3 mutants. Genes Brain Behav 18: e12557 102. Van Den Broeck L, Sierksma A, Hansquine P et al (2021) Comparison between touchscreen operant chambers and water maze to detect early prefrontal dysfunction in mice. Genes Brain Behav 20:e12695 103. Whoolery CW, Yun S, Reynolds RP et al (2020) Multi-domain cognitive assessment of male mice shows space radiation is not harmful to high-level cognition and actually improves pattern separation. Sci Rep 10:1–15 104. Yang J-H, Presby RE, Jarvie AA et al (2020) Pharmacological studies of effort-related decision making using mouse touchscreen procedures: effects of dopamine antagonism do not resemble reinforcer devaluation by removal of food restriction. Psychopharmacology 237:33–43 105. Zeleznikow-Johnston AM, Renoir T, Churilov L et al (2018) Touchscreen testing reveals clinically relevant cognitive abnormalities in a mouse model of schizophrenia lacking metabotropic glutamate receptor 5. Sci Rep 8: 16412 106. Zhao Z, Hood RJ, Ong LK et al (2021) Exploring how low oxygen post conditioning improves stroke-induced cognitive impairment: a consideration of amyloid-beta loading and other mechanisms. Front Neurol 12:231 107. Zhou LYY, Wright TE, Clarkson AN (2016) Prefrontal cortex stroke induces delayed impairment in spatial memory. Behav Brain Res 296:373–378

Chapter 22 Using Operant Reach Chambers to Assess Mouse Skilled Forelimb Use After Stroke Dene Betz, April M. Becker, Katherine M. Cotter, Andrew M. Sloan, Ann M. Stowe, and Mark P. Goldberg Abstract Skilled forelimb reaching and grasping are important components of rodent motor performance. The isometric pull task can serve as a tool for quantifying forelimb function following stroke or other CNS injury as well as in forelimb rehabilitation. This task has been extensively developed for use in rats. Here, we describe methods of setup and training of an operant reach chamber for mice. Using a reward of peanut oil, mice are adaptively trained to pull a handle positioned slightly outside of an operant chamber, with automated recording of the number of attempts, force generated, success rate, and latency to maximal force. Key words Forelimb function, Operant reach chambers, Stroke, Motor Impairment, Isometric pull task

1

Introduction Motor impairment is one of the most common long-term effects of stroke [10]. With deficits often affecting skilled upper limb function, stroke is the leading cause of adult disability [9]. While rehabilitation can facilitate recovery, it is typically incomplete, establishing a need for interventions that enhance the current therapeutic ceiling. Rodent models of stroke provide critical insights toward these efforts, given their excellent translational features, particularly, the ability to model fine motor impairment and its recovery trajectory after stroke [5, 6]. With this in mind, it is important that mouse motor assays exhibit the sensitivity and accuracy needed to capture these relatively subtle features of post-stroke deficits. The skilled reaching task, which requires a rodent to reach through a slit in the chamber wall to grasp and retrieve a food

Dene Betz and April M. Becker contributed equally. Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_22, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

327

328

Dene Betz et al.

pellet, has proven an effective measure of forelimb function after CNS injury [3], but assays of forelimb function are often based on labor- and time-intensive video/visual scoring. An automated method for measuring forelimb function, the isometric pull task, provides computer-controlled training and rapid quantitative data on multiple parameters of a reach-and-grasp movement [4, 11]. Originally developed for rat studies, we adapted this behavioral assay to mouse models to take advantage of the wide availability of genetic and pharmacological tools in mice [2, 8]. We utilize a pre-assembled operant chamber modified for mice (Vulintus, Lafayette, CO), with software for force measurement, adaptive training, and automated data collection [2]. Mice are trained to reach through a slit in the chamber wall to grasp and pull a metal handle attached to an isometric force transducer. The location of the slit limits use to either the left or right forelimb (we use left cortical stroke and the right forelimb reach task). An infrared beam records each pass through the slit as an attempted pull. Once a minimum force threshold is exceeded (typically 20 g per pull after training), the system dispenses a small droplet of peanut oil. We have found that peanut oil is a strong reinforcer for mice, and prior food or water deprivation is not required which is particularly useful for the well-being of post-stroke animals. The system provides automated assessments of the several facets of the pull task, including force generation, success rate, overall pull attempts, and latency to maximal force [4]. In this chapter, we describe the setup and use of the mouse operant forelimb reach chamber apparatus [1, 2, 7, 8] and a method for adaptive training to the isometric pull task. This protocol is based on experiments with adult male and female C57/Bl mice exposed to unilateral stroke as described in [2, 8].

2

Materials 1. Behavioral room with optional reverse light cycle: Housing in the same room is preferred, particularly if a reverse light cycle setup is chosen, to minimize environmental sources of variance. The room should be isolated either with a door or interior walls to minimize noise and light changes. Red-light emitting headlamps (available online or at local hardware stores) should be used for dark-cycle behavioral sessions. 2. Table or shelving: Tables can be used for up to 4 reach box setups. Foam panels may be used to isolate adjacent boxes. 3. Mouse operant chambers with isometric pull assembly, MotoTrak controller (MotoTrak Mouse, Vulintus, Lafayette, CO), and liquid dispensing apparatus with requested modifications (Fig. 1):

Automated Skilled Forelimb Reach Task

329

Fig. 1 Overall view of the reach box apparatus, liquid dispenser, and controller. (a) The dispenser and controller are positioned to either side of the apparatus. Note that cords are not included in this schematic. Other parts depicted here include: handle assembly (A), feeding needle port (B), dividing wall (C), water bottle (D).

(a) The liquid dispensing needle port should positioned adjacent to the slit, 0.25 cm directly diagonal from the upper right corner of the slit (8 mm × 15 mm). (b) Chamber should be modified to hold a water bottle on one side. 4. MotoTrak software (Vulintus, Lafayette, CO) including Mototrak 2.0 Behavioral Program (Vulinus.com) and Mototrak Autoshaping Program (available from Vulintus upon request). 5. Weights for handle calibration: Mass hanger (50 g) and slotted (total 250 g) weights included, Vulintus, Lafayette, CO. 6. 0.5 inch, 18-gauge blunt-tipped needle for liquid dispensing. 7. Spare parts for operant chambers: (a) Solenoid pinch valve for liquid dispenser (available from Vulintus or Nresearch) (for troubleshooting, see Note 1) (b) 100 g mini load cell (available from Vulintus or online) (c) 6 Pin mini Din cable (d) 8 Pin mini Din cable (e) USB 2.0 A to B cable (f) Adaptive 15-20 V power supplies for each MotoTrak controller

330

Dene Betz et al.

8. 10 mL syringe without the plunger. 9. Computer or laptop with 4 GB RAM, running Microsoft Windows 7 or higher. Each device can drive 2–4 MotoTrak controllers using separate USB ports. 10. USB hubs: Each reach box apparatus will require 1 USB port and an additional port is required for each webcam (if applicable). 11. Motion-activated webcam (optional, for remote observation of mouse training and performance). If webcams are desired, please note this upon ordering the reach box apparatus as the reach box can be readily modified to allow for webcam installation. 12. Foam insulating panels for sound isolation (optional) (see Note 2). 13. 1 mL transfer pipettes (for shaping and peanut oil exposure). 14. Water bottles to fit water bottle apparatus. 15. Peanut oil: We use LouAna’s Peanut Oil (see Note 3).

3

Methods The reach task requires that mice are trained to consistently operate the isometric pull assembly to a specific force threshold, prior to stroke or brain injury. The training process for mouse skilled reach performance can be labor-intensive. To increase the throughput and efficacy of this behavioral task, we developed a protocol that uses a mixture of manual (i.e., shaping), group, and automated adaptive paradigms. For this protocol, each operant apparatus can be used to train and test up to 4 mice per experimental cohort per apparatus. Mice receive initial acclimation to the chamber and task (group training) before shifting to individualized adaptive training to complete a skilled-reaching task with a minimum 20 g force threshold. Task performance is automatically tracked and analyzed via the MotoTrak Behavioral Program, providing measures such as success rate, latency to reach success criterion, and the number of attempts per session. After a three-day baseline measurement, mice are subjected to CNS injury such as stroke, and recovery is evaluated weekly post-injury (see Note 4).

3.1

Hardware Setup

1. Assemble the chamber with the central partial divider wall to either side of the central slit opening. The active limb will be closest to the divider (shown as “C” in Fig. 2), and the opposite paw should not be able to reach through. Position the force transducer/handle at the central slit opening, initially as far as possible out from the cage interior (to protect the handle/load cell, while the door is being opened or closed).

Automated Skilled Forelimb Reach Task

331

Fig. 2 Views of the apparatus with the divider. (a) Side view of the reach box apparatus with the following labels: handle assembly (A) with mounted force transducer handle positioning motor (B), dividing wall (C), floor with tray (D), water bottle spout (E), and peanut oil dispenser (F). (b) In this image, the dividing wall is positioned just to the right of slit with force handle, so that only the right forelimb can be used

2. Place the liquid feeding needle in the port above the reaching slot, fully protruding into the chamber (shown as “F” in Fig. 2; see Note 5). 3. Place the computer-controlled liquid dispenser outside of the operant chamber, ensuring that it is above the feeding needle, as the liquid dispenser is gravity-driven. 4. Connect the MotoTrak controller to the computer via the USB cable. 5. Connect the MotoTrak controller to the force transducer, liquid dispenser, and infrared beam using supplied cables. (It is important to correctly match the cables and check the integrity of the pins on each cable. Replace any cable with bent pins.) Pictures of all connections are included in the user manual. 6. Attach the power supply to the MotoTrak controller. Please note that we do not connect the auto-positioner as described in the manual. This does not affect the function of the box and, in our experience, can reduce behavioral variation that may be introduced by its movements or sounds. 3.2 Calibration and Testing (Initial Setup and Weekly)

Remove any strong electromagnetic sources, such as cell phones, during these steps. 1. Use the “feed” setting within the MotoTrak behavioral program to verify that activation of the liquid dispenser causes the release of a peanut oil droplet. This should also be done at the beginning of every session. Additionally, press the blue button

332

Dene Betz et al.

Fig. 3 Liquid dispensing apparatus with labeled parts

on the liquid dispenser to confirm its function (shown in left panel in Fig. 3). If this is not working properly, check for air bubbles and/or clogs in the system. Replace tubing or pinch valve if necessary (see Note 6).

2. With all parts of the assembly in place, including the feeding needle inserted into the port on the door, check the size of the peanut oil droplet dispensed during a “feed.” Droplets should be approximately 2 μL and can be adjusted by extending the number of milliseconds that the pinch valve (shown in right panel in Fig. 3) remains open. This number is depicted in green on the digital screen located on the liquid dispenser (see Note 7). 3. Calibrate the handle force and load cells according to the instructions provided with the equipment. Weights for handle calibration are included with the operant chamber. Verify that the handle can distinguish pulls as soft as 0.1 g and as hard as ~40 g. Replace any load cells which show unstable baseline fluctuations or do not respond to light touch (see Note 8). 4. Verify that the infrared beam sensor is functioning properly. The MotoTrak program should be set to detect the sensor signal (red box shown in Fig. 4a), which changes the graph output to transduce the signal from the IR detector to pink peaks. By waving a finger or transfer pipette in front of the emitter and/or detector (Fig. 4b), an alteration of IR signal can be visualized in the graph of the MotoTrak program if the IR

Automated Skilled Forelimb Reach Task

333

Fig. 4 (a) Mototrak 2.0 behavioral program box set (via settings in red box) to display signal (pink peaks) from IR apparatus (b), located on reach box door

sensor is properly functioning. If steady offsets or other signal variations are observed with no object in the light path, carefully clean the emitter and detector using a KimWipe and 70% ethanol. 3.3 Training—Group Phase (5 Consecutive Days, 1 Week)

The purpose of this stage is to train mice to respond to the reinforcement apparatus (“magazine training”), to recognize peanut oil as a reinforcer, and to reach for the isometric pull assembly (handle) with the appropriate paw. We use a combination of noncontingent and contingent peanut oil delivery, and both group and individual training paradigms, to improve task rate, throughput, and increase the baseline success rate measurement: 1. The entire assembly should be cleaned prior to each group training session to reduce behavioral variation. Prior to cleaning, all parts of the apparatus should be labeled (including the handle, controller, and liquid dispenser apparatus) so that they are all kept together. For example, if one reach box apparatus is labeled “BLUE”, the handle, controller, and liquid dispenser should also be labeled “BLUE”. This can later be used as the mouse group identifier. The apparatus can accumulate excessive debris during group training sessions, so we recommend removing each reach box apparatus (not including liquid dispenser or controller) and taking them to a sink. 2. Remove the tray located underneath the floor and throw away debris from the previous session. Wash with soapy water, spray with 70% ethanol, wipe dry, and slide back into place. Thoroughly wipe down the metal floor and acrylic walls with 70%

334

Dene Betz et al.

ethanol. The handle should be very carefully wiped down with a KimWipe and 70% ethanol. Use extreme caution not to bump the handle as this can damage the load cell. Sweep up all surrounding debris from the floor, shelving, and surrounding tables and wipe down all surfaces with 70% ethanol. 3. Transfer fresh peanut oil into the dispenser (see Note 9), assuring that the feeding needle protrudes into the chamber and that the peanut oil flows freely without bubbles when the valve is open. Additionally, once the apparatus is plugged back into the computer and powered, press the blue button on the liquid dispenser to trigger a manual feed and ensure that a peanut oil droplet has been dispensed on the end of the needle. 4. Place a group of up to four same-sex cagemate mice in each reach chamber for up to 6 h per session (see Note 10). Time should be consistent each day. We provide access to water during this stage (see Note 11). 5. Open the MotoTrak Autoshaping software application and select the proper “BOOTH” via the setup instructions provided. Each reach box assembly requires a separate instance of the MotoTrak software, meaning that a new MotoTrak browser should be opened for each reach box assembly. Enter the mouse group identifier for “NAME.” Set the program parameters to deliver noncontingent reinforcement (peanut oil delivery) randomly every 0.5–2 min and, in parallel, contingent reinforcement of handle pulls with the force threshold set to the minimum value (typically 1-3 g, value must exceed the baseline electrical noise of the transducer as displayed in the MotoTrak software). Do not start the session until the handle is in place. 6. Position the handle ~0.25 cm from the inside edge of the chamber (Fig. 5a). For the initial session the handle should be placed just far enough to allow for licking or reaching behaviors and prevent the ability to trigger a reward through biting. 7. Let the session run for up to 6 h. Mice should occasionally be observed (at least 3 time per session, either directly or via webcam) to monitor the acquisition of reaching behavior (see Note 12). Once at least one mouse per group exhibits reaching behavior, position the handle further back, another ~0.5 cm or until licking can no longer trigger a reward. If a drastic decline in reaching behaviors is observed, the handle should be moved back to its original position to avoid extinction. 8. At least once per day, shaping is recommended. To shape, periodically place a drop of peanut oil on the handle to encourage reach attempts. Excessive need for shaping indicates a need to adjust handle position and/or sensitivity (lowering the force threshold or recalibrating).

Automated Skilled Forelimb Reach Task

335

Fig. 5 Handle assembly (black) positioned at initial (a) and final (b) positions

9. Follow group response rates via the MotoTrak software interface as training progresses. High pulling rates (200+ pulls in 30 min) by day 3 indicate good progress and the handle should be moved further back (~0.50 cm) to the final position. In the final position, the handle should not be within reach of the mouse tongue and require full extension of the mouse forelimb (Figure 5b). Low response rates may indicate the need for adjustment of equipment or training (handle position, sensitivity, oil deliveries, oil quality, more frequent oil baiting on the handle, or increased manual shaping) (see Note 13). 3.4 Training: Individual Phase (5 Sessions per Week, up to 3 Weeks)

The purpose of this phase is to train the animals to consistently pull the handle to the expected force threshold: 1. Begin individual training for mice with a consistently timed session daily. We suggest a 2-hour session. Mice should be trained in the same chamber throughout the entire experiment, including the group-training phase. This means that a group of four group-trained mice will later require four individual mouse training sessions in the same chamber per day. Differentiate individual mice within a group via tail-marked numbers (i.e. BLUE1, BLUE2, etc.) to avoid excessive handling needed for ear-tag based identification. We suggest naming the group after their reach box apparatus. 2. The handle should be positioned back to the initial position outside the reach box, ~5 mm from the inside edge of the chamber, allowing for mice to be able to lick the handle, but only trigger a reward with a paw touch. A small droplet of peanut oil should be placed on the handle prior to the session throughout the first week of training.

336

Dene Betz et al.

Fig. 6 Mototrak 2.0 Behavioral program set to depict signal from the force transducer (purple). A. Trace shows applied force (grams) in purple for two pull attempts, each exceeding the specified force threshold. X-axis shows time in milliseconds since initiation of the reach attempt

3. Open the MotoTrak 2.0 Behavioral program with a separate program for each chamber. Enter the mouse identifier in the section labeled “NAME.” From the drop-down menu, STAGE, select the ADAPTIVE 1 g training program, which sets the initial force threshold at 1 g and each subsequent threshold to the lower quartile of the previous 15 pulls. Start the session and deliver 2 manual feeds in quick succession (Fig. 6). 4. End the session by pressing “STOP.” Then, record the success rate (in the form of success/number of trials) final session force criterion for each animal. Once all programs have been stopped, replace the mouse with the next in the group. Please ensure that mice are being loaded and removed through the door furthest from the isometric pull assembly to avoid bumping the handle. 5. Close each instance of the Mototrak 2.0 Behavioral program and open a new instance for each reach box apparatus as before. While it is possible start a new behavioral session from the same Mototrak 2.0 Behavioral program, doing so causes the program to freeze mid-session, which can result in behavioral extinction. Enter the mouse identifier and start a new session.

Automated Skilled Forelimb Reach Task

337

6. For the first individual training session, a small droplet of peanut oil should be placed on the handle and mice should be shaped for around 15 minutes. Shaping can be accomplished by holding a peanut oil filled transfer pipette close to the handle. The mice should approach the slit, fail to lick the peanut oil, and then attempt reaching to retrieve it. We recommend alternating between the mice in separate reach boxes after 1-2 mins of shaping/mouse for the first 15 min of the training session 7. To determine the appropriate STAGE for the next training session we recommend the following: maintain the Adaptive 1 g stage and initial handle position for mice exhibiting sparse reaching behaviors (100 pulls in one hour), move the handle to the final position (described above). As soon as reaching behaviors are observed, proceed with daily stage determination in the followng manner: (a) If the prior day’s success rate exceeds 50%, set the Adaptive threshold to the highest value which is less than the prior two final criteria for each animal. For example, if a mouse has reached a final criterion of 12 g for two consecutive sessions, the animal should be started on Adaptive 10 g in the next session. (b) If the success rate was between 30% and 50%, maintain the same adaptive threshold. (c) If the success rate was lower than 20%, move to a lower Adaptive threshold stage. If the animal is on Adaptive Stage 1, consider introducing noncontingent peanut oil deliveries, adjusting the handle closer to the inside edge of the chamber (temporarily), manual shaping, or excluding the animal (see Note 14). 8. Once the animal on Adaptive 15 g reaches a final criterion of 20 g for two consecutive days, move to the Static 20 g program. Once this is complete, mice are advanced to the baseline measurement phase of the experimental protocol. Exclude mice that do not reach these criteria. 3.5 Individual Baseline Measurement (3 Consecutive Days)

The purpose of this phase is to obtain a measure of baseline performance for statistical comparisons in the experimental phase. 1. Continue at the Static 20 stage until 3 consecutive sessions show greater than 50 pulls in 30 min and variance of 1 indicate that the transporter facilitates permeability of the endothelial cells while values 1 indicate that the transporter restricts permeability of the endothelial cells while ER values 350 g. The steps to prepare the suture are shown as below:

Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models. . .

471

1. Cut the surgical nylon filament into 3 cm length. 2. Smooth one end of the filament by burning with the cauterizer tip to round up the end. Check the tip smoothness under the surgical microscope. 3. Mark the filament with silver sharpie marker at 1.8 cm from the smoothed end. In recent studies, sutures specifically made for the MCAO surgery (Doccol Corp) were used, and the size was chosen according to animal BW per manufacturer’s instructions. In our studies, the suture sizes 403745PK10 and 403945PK10 were used. 3.4 Pre-op Care and Preparation (Fig. 1, See Note 8)

1. Start treating the animals with chew toys for extra enrichment and diet supplement including peanut butter, sunflower seeds, and DietGel in the cage one week prior to the planned surgery date. Repeat the same process and supplement in the post-op care to help with the survival. 2. Pack all the surgical instruments in the sterilize bag to autoclave. Date the bag and open only right before the surgery. If the surgical instruments will be used for more than one surgery, the instruments will be cleaned with 70% isopropyl alcohol, sterilized in the hot bead sterilizer at 100 °C for 10 s between the surgeries, and cooled down in the sterilized water before touching the animal tissue.

Fig. 1 Presurgical preparation for middle cerebral artery occlusion

472

Weiguo Li et al.

3. Inject 0.3 mg/kg buprenorphine to the animals at 1 h before the surgery (see Notes 9 and 10). 4. Clean the tabletop with 10% bleach solution to disinfect the surface. 5. The surgeon should wash hands well before surgery and wear a clean lab coat, a mask, and a pair of sterile gloves. Change the gloves between the surgeries and every time the surgeon touches unsterilized area. Use the Glad Press’N Seal wrap as sterile drapes to cover the surgery table, the animal, and the designated area for the surgical instruments. 6. Anesthetize the animals with 5% isoflurane inhalation in the induction chamber. Confirm the successful anesthesia with toe pinch (see Note 10). 7. Shave the surgical area on the neck with electric clipper. The shaved area should be 3 times the size of surgical area. Sterilize the surgical area with Betadine and 70% isopropyl alcohol. Using a swab, cleansing is done in a circular motion beginning at the center of the shaved area and working outward to the edge. Repeat the procedure three times with different swabs to avoid contamination of the already scrubbed area. Apply the ophthalmic ointment to protect the eyes. 3.5 Operation Procedures (See Note 10)

1. Connect the animal to the nose cone of the anesthesia workstation and maintain the 2–3% isoflurane inhalation (see Note 11). Place the animal on temperature control warming pad to maintain the body temperature during the surgical process. Check the anesthesia status by toe pinch every 10–15 min. 2. Make a midline cervical incision on the skin with scalpel to expose the underneath tissue and neck muscle group. 3. Use Dumont forceps to separate tissue and muscles to expose common carotid artery (CCA) under the surgical microscope (see Note 12). 4. After separating CCA from surrounding tissue, tie a silk suture on the CCA to secure the vessel and block the blood flow. Hold the suture with a hemostat. 5. Follow the run of CCA and find the bifurcation of external carotid artery (ECA) and internal carotid artery (ICA), separate the surrounding tissue on ICA first, and tie a silk suture and hold it with a hemostat (see Note 13). 6. Separate the surrounding tissue on ECA, tie two double knots on ECA with silk suture, and use the tissue scissors to cut ECA in between the two knots. 7. Hold the ECA stump at the bifurcation with a hemostat, tie a loose knot with silk suture, and use Vanna’s spring scissors to cut a small incision on ECA stump.

Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models. . .

473

8. Embolic MCAO model: Insert the PTFE-160 catheter containing a 4 ± 0.5 cm blood clot and saline into the stump of ECA. Tighten the preplaced loose knot in Step 7 to close the incision on the ECA stump. Release the knot on ICA and proceed the catheter into the ICA. Confirm the catheter passing the bifurcation of the pterygopalatine artery to entering the skull without resistance (see Note 14 and 15). Stop the catheter at 1.8 cm mark, inject the clot with 100 μL saline into the vessel at a low speed, and wait 30 s after injection to slowly withdraw the catheter (see Note 16). 9. Suture MCAO model: Insert a nylon suture monofilament into the ECA stump and tighten the preplaced loose knot to close the incision on the ECA stump. Release the silk suture on ICA which is tied in Step 5. Advance the filament into the ICA and confirm it passing the bifurcation of the pterygopalatine artery. Stop the filament at the 1.8 cm mark (see Note 14 and 15). For permanent stroke model, leave the filament in the vessel without further operation until the designated time, and euthanatize the animals (see Note 17 and 18). For transient stroke model, anesthetize the animals again with isoflurane inhalation at designated time and withdraw the occlusion filament at a very low speed (see Note 16). 10. After the catheter or the suture filament reach to the 1.8 cm mark, clean the surgical area with cotton tip and confirm no active bleeding spot. Release the knot on CCA slowly to confirm the restoration the antegrade blood flow. 11. Pull back the muscle group to cover the vessel and close the incision on the skin with wound clips and apply the betadine solution. The clips are removed after 7 days if the animals survived, and any infection is monitored till the removal. 12. The surgery is typically completed in 15 to 25 min (see Note 19). 3.6 Post-op Care (Fig. 2)

Animals are monitored after surgery on warming pads until they regain consciousness and transferred into new cages. Till the incision healed, they are housed singly. Due to the stroke and surgical injuries, animals eat and drink less and may have 15–20% loss of BW within 24 h of the surgery if not supplemented with fluids. In some cases, the weight loss will continue until the fourth day after stroke, despite the animal showing signs of recovery (see Note 20). Thus, for long-term experiments (14 days or longer), we developed a monitoring post-operative plan as described below to allow the animals to remain on study beyond the recommended 25% weight loss. All of this information is documented in the surgery record (see Note 10):

474

Weiguo Li et al.

Fig. 2 Postsurgical care in middle cerebral artery occlusion

1. Body condition score (BCS) is evaluated daily and included for better assessment/monitoring of the animals [15]. Body condition score is evaluated as follow: 5 = obese; 4 = overconditioned; 3 = well-conditioned; 2 = under conditioned; 1 = emaciated. 2. Day of surgery: Animals receive Ringer’s solution immediately after MCAO (10 mL i.p. injection based on 8–11 mL/100 g of BW/day of normal consumption) and another 5 mL saline within 12 h. The amount of fluid will provide 50% of the normal daily water intake. Weigh all the animals daily for 5 days after surgery and then once a week till euthanasia. In addition, provide softened food (regular chow or HFD pellets softened in 5 mL water in a petri dish) for 5 days. Give peanut butter and recovery gel (DietGel and HydroGel) in the cage and replace daily. 3. Days 1 and 2 after surgery: Animals receive 5 mL i.p. injection of Ringer’s solution twice daily. Monitor the weight loss by weighing the animals daily. If animals lost 15% BW despite fluid supplementation by the end of Day 2, increase the fluid to 10 mL twice a day.

Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models. . .

475

4. Day 3 after surgery: Contact the Lab Animal Services if the animals lost 20% BW and follow the recommendations of the attending veterinarian. Euthanasia decision is based on the status of the animal. If animals are bright, alert, responsive, eating/drinking well, and BCS = 3, continue the monitoring. If BW loss continues by Day 5 or when BW loss reaches greater than 25% (whichever comes first), euthanize the animals. Additionally, if, by Day 3, animals are depressed, lie on side in cage, and, slow or not reacting, euthanize the animals. 3.7 Surgery Evaluation

Successful MCA occlusion can be confirmed with laser imaging system as described in previous studies [16–18]. However, there are some cases that the animal develops transient ischemia during the surgery and has fully recovery to normal blood flow after the surgery. Therefore, we developed an inclusion/exclusion criterion based on the motor behavioral outcome scores and weight loss measured 24 h after the surgery to ensure the successful ischemia and avoid bias. The criteria include the items listed as below: 1. Neurological behavior score: Bederson score for each animal is obtained by using parameters listed below: (A) spontaneous ipsilateral circling walk, graded as 2 (no circling), 1 (partial circling), and 0 (continuous circling); (B) resistance to push, the animal is slightly pushed from the back while walking, graded as 1 (resistance) and 0 (no resistance); and (C) contralateral hind limb and (D) forelimb retraction which measures the ability of the animal to replace the limb after it is displaced laterally by 2 to 3 cm, graded as 2 for immediate replacement, 1 for replacement after a moment, and 0 for no replacement. Maximum score of 7 is allotted to a normal rat without ischemia injury. 2. Adhesive removal test (ART): Adhesive paper dots are placed on the forepaws of the animals. Contact and removal latency of the adhesive dot are recorded on Day 1 and Day 3 after MCAO. For each day, the average is taken from 3 trials with a maximum removal latency of 180 s per trial. 3. Inclusion criteria: The animals with ART >35 s on Day 3 AND Bederson score ≤4 on Day 0 or >10% weight loss on Day 3 are considered successful MCAO surgery. The predictive value of these criteria of a successful MCAO surgery was confirmed by magnetic resonance imaging on Day 3 after MCAO.

4

Notes 1. We chose the HFD/STZ model because it mimics many aspects of DM. It also allows us to incorporate female animals into our studies. As in genetic models of diabetes, female rats

476

Weiguo Li et al.

hardly develop hyperglycemia at the same level and develop it much later in their lifespan. HFD alone or low dose STZ alone does not result in elevated blood glucose. 2. The result of diabetes induction may vary depending on the STZ source, the freshness of the solution and buffer, and the actual dose injected into the animals. Animals were not fasted before STZ injection in our studies. 3. Be aware of the toxicity of STZ to human; the personal protection equipment including mask, gloves, and gown are necessary to prepare and inject the STZ. All injections are done in a chemical hood. Laboratory Animal Services personnel are notified before the injection and special cards indicating STZ treatment are placed on the cages. It is important to follow these precautions which may be different at each institution. 4. STZ is a toxic chemical, and the direct effect on the brain is often questioned. Animals that do not develop hyperglycemia can be used as additional controls to address the issue of STZ toxicity in the brain. In our experience, at the doses we describe, we have not observed any pathology in the brain or the vasculature in these animals. 5. The animal’s age also needs to be considered for the diabetes model [19]. We have experienced 90% mortality 3 days after STZ injection in 48-week-old Wistar male rats, while we have never experienced that in younger animal groups. 6. The impact of BG level on cerebral infarct size has been well studied by many groups including us [7, 20]. In our previous studies with animal models having mild or moderately increased BG level, we have found increased infarct size comparing to the control group only in female but not in male animals [10, 16]. Interestingly, the infarct size was significantly increased in the animals with acutely increased BG level above 450 mg/dL in other group’s study [21]. 7. Animal BW has big impact on the stroke outcome as well. Animals survive better if the BW is higher than 300 g, and 300–400 g range gives the most reproducible stroke outcomes. When both sexes are included in the experimental design, the significant BW difference between male and female animals limits the age-matched studies as low BW increases mortality in the female group. 8. The surgery room is sectioned into pre-op, surgery, and postop areas. We have placed posters with the key elements of pre-op preparation, surgery, and post-op care in each area (Figs. 1 and 2). 9. Studies in the last several years have shown buprenorphine only provides adequate pain coverage in rodents up to 6–8 h. Thus,

Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models. . .

477

the sustained release form is used in our studies for a longer coverage and need to be given at least 1 h before the surgery. 10. A surgery record for each animal is highly recommended. The record can be a table to convey all the major information of the animal and the surgery, which may include surgery date and time, animal number, BW, BG, anesthesia information, pre-op and post-op care information, etc. 11. The anesthesia workstation supplies both gas anesthetic delivery and ventilatory support to the animal. The workstation needs to be connected to an isoflurane vaporizer, an oxygen gas cylinder (or any other gas mixture of the researcher’s choice), and a canister containing active charcoal to collect the waste anesthesia gas. The animal BW range for this equipment is from 150 g to 7 kg. The surgeon is also playing the role of anesthetist to pay attention to the breathing pace of the animal on the surgery table. If the animal loses spontaneous breath because of the deep anesthesia during the surgery, the ventilation can be turned on and the vaporizer can be turned off to allow the anesthesia gas to be ventilated out from the system of the animal. The ventilation can be turned off and the vaporizer can be turned back on to 2–3% after the animal regains the spontaneous breath. The settings in our studies are airway pressure high limit 25 cmH2O, oxygen flow rate 0.4 lpm, breath rate 60 bpm. 12. The carotid artery is underneath the neck muscle groups and can be found without cutting or breaking through the muscles. The access point to reach CCA is surrounded by muscle groups of sternomastoideus, omohyoideus and sternohyoideus, and digastricus. Retractors can be used to pull the muscles and expose the surgical area. The surrounding nerves and small vessels can be separated with blunt dissection. Be aware of and avoid over-touching the vagus nerve. 13. The silk suture is used instead of the artery clips to secure the vessels and block the blood flow of the CCA and ICA based on personal preference. The artery clips have a possibility to slip off and occupy more space on the arteries. However, one needs to be cautious when using the silk suture to avoid the damage to the muscular layer of the arteries. 14. The direction of the nylon filament or the catheter needs to be confirmed under the surgical microscope to pass the bifurcation of the pterygopalatine artery and ICA. It is in the wrong direction into the pterygopalatine artery if there is resistance before reaching the 1.8 cm marker. 15. One of the common complications of the MCAO surgery is the puncture of the MCA, which causes intracerebral hematoma and is one of the main reasons for the post-operation mortality. The outer diameter of the PTFE Sub-Lite 160 catheter used in

478

Weiguo Li et al.

the embolic MCAO model is 0.4 mm. The diameter of the nylon filament suture used in the suture MCAO model is 0.37 to 0.41 mm based on the animal BW. The length of the catheter or the suture filament inserted into the vessels is approximately 1.8 cm. 16. For both embolic and suture/reperfusion MCAO model, the withdrawal of catheter or the suture filament from ICA need to be slow to avoid the robust reperfusion into the cerebral ischemic area which may cause significant secondary injury known as hemorrhagic transformation and increase the mortality rate. 17. The successful occlusion of MCA can be confirmed by the laser imaging system [16, 18] which requires extra surgical procedures to open the scalp layers to expose the skull. In some cases, thinned skull or even cranial window are needed. 18. Before starting the project, a few trial surgeries with permanent MCAO can be done to confirm the successful surgery. The animals can be euthanatized instantly after the surgery without withdrawing the filament suture. The autopsy can show whether the filament suture entered the ICA and the Willis’ circle at the bottom of the brain. The occlusion is considered successful if the tip of the filament suture reached or passed the origin of the MCA. 19. The surgery time is variable depending on the skill level of the surgeon and the successfulness of the anesthesia. The longer surgery time results in greater exposure of the animals to the anesthesia and impacts the outcome of the surgery. 20. After successful MCAO surgery, the animals have very limited ability to access the feeder and the water bottle which are usually at the top of the cage. The nutrient and fluid supplement are necessary for the post-op care, especially for the diabetic animals. As described above, the Ringer’s solution is preferred because of the components of the solution. Saline can be used in general as well. HydroGel and DietGel (ClearH2O Inc.) are also commonly used as the nutrient and hydration supplement. Other supplements include peanut butter, sunflower seeds, or mixed unshelled seeds.

Acknowledgments This work was supported by Veterans Affairs (VA) Merit Review (BX000347), VA Senior Research Career Scientist Award (IK6 BX004471), National Institute of Health (NIH) RF1 NS083559 (formerly R01 NS083559), and R01 NS104573 (multi-PI, Susan C. Fagan as co-PI) to AE and South Carolina Clinical & Translational Research (SCTR) Institute (UL1TR001460) discovery grant (SCTR2201) to YA.

Acute Ischemic Stroke by Middle Cerebral Artery Occlusion in Rat Models. . .

479

References 1. Chen R, Ovbiagele B, Feng W (2016) Diabetes and stroke: epidemiology, pathophysiology, pharmaceuticals and outcomes. Am J Med Sci 351:380–386 2. Demaerschalk BM, Kleindorfer DO, Adeoye OM et al (2016) Scientific rationale for the inclusion and exclusion criteria for intravenous Alteplase in acute ischemic stroke: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 47:581–641 3. Ergul A, Kelly-Cobbs A, Abdalla M et al (2012) Cerebrovascular complications of diabetes: focus on stroke. Endocr Metab Immune Disord Drug Targets 12:148–158 4. Fagan SC, Lapchak PA, Liebeskind DS et al (2013) Recommendations for preclinical research in hemorrhagic transformation. Transl Stroke Res 4:322–327 5. Bruno A, Liebeskind D, Hao Q et al (2010) Diabetes mellitus, acute hyperglycemia, and ischemic stroke. Curr Treat Options Neurol 12:492–503 6. Demchuk AM, Morgenstern LB, Krieger DW et al (1999) Serum glucose level and diabetes predict tissue plasminogen activator-related intracerebral hemorrhage in acute ischemic stroke. Stroke 30:34–39 7. Hafez S, Abdelsaid M, Fagan SC et al (2018) Peroxynitrite-induced tyrosine nitration contributes to matrix Metalloprotease-3 activation: relevance to Hyperglycemic ischemic brain injury and tissue plasminogen activator. Neurochem Res 43:259–266 8. Abdelsaid M, Prakash R, Li W et al (2015) Metformin treatment in the period after stroke prevents nitrative stress and restores angiogenic signaling in the brain in diabetes. Diabetes 64: 1804–1817 9. Jackson L, Dong G, Althomali W et al (2020) Delayed Administration of Angiotensin II Type 2 receptor (AT2R) Agonist compound 21 prevents the development of post-stroke cognitive impairment in diabetes through the modulation of microglia polarization. Transl Stroke Res 11:762–775 10. Li W, Ward R, Valenzuela JP et al (2017) Diabetes worsens functional outcomes in young female rats: comparison of stroke models, tissue plasminogen activator effects, and sexes. Transl Stroke Res 8:429–439

11. Prakash R, Johnson M, Fagan SC et al (2013) Cerebral neovascularization and remodeling patterns in two different models of type 2 diabetes. PLoS One 8:e56264 12. Prakash R, Li W, Qu Z et al (2013) Vascularization pattern after ischemic stroke is different in control versus diabetic rats: relevance to stroke recovery. Stroke 44:2875–2882 13. Prakash R, Somanath PR, El-Remessy AB et al (2012) Enhanced cerebral but not peripheral angiogenesis in the Goto-Kakizaki model of type 2 diabetes involves VEGF and peroxynitrite signaling. Diabetes 61:1533–1542 14. Jackson-Cowan L, Eldahshan W, Dumanli S et al (2021) Delayed Administration of Angiotensin Receptor (AT2R) Agonist C21 improves survival and preserves sensorimotor outcomes in female diabetic rats post-stroke through modulation of microglial activation. Int J Mol Sci 22:1356 15. Ullman-Cullere MH, Foltz CJ (1999) Body condition scoring: a rapid and accurate method for assessing health status in mice. Lab Anim Sci 49:319–323 16. Li W, Prakash R, Kelly-Cobbs AI et al (2010) Adaptive cerebral neovascularization in a model of type 2 diabetes: relevance to focal cerebral ischemia. Diabetes 59:228–235 17. Li W, Qu Z, Prakash R et al (2013) Comparative analysis of the neurovascular injury and functional outcomes in experimental stroke models in diabetic Goto-Kakizaki rats. Brain Res 1541:106–114 18. Ward R, Valenzuela JP, Li W et al (2018) Poststroke cognitive impairment and hippocampal neurovascular remodeling: the impact of diabetes and sex. Am J Physiol Heart Circ Physiol 315:H1402–H1413 19. Wang-Fischer Y, Garyantes T (2018) Improving the reliability and utility of Streptozotocininduced rat diabetic model. J Diabetes Res 2018:8054073 20. Ergul A, Li W, Elgebaly MM et al (2009) Hyperglycemia, diabetes and stroke: focus on the cerebrovasculature. Vasc Pharmacol 51:44– 49 21. Kuroki T, Tanaka R, Shimada Y et al (2016) Exendin-4 inhibits matrix Metalloproteinase-9 activation and reduces infarct growth after focal cerebral ischemia in hyperglycemic mice. Stroke 47:1328–1335

Chapter 34 The DOCA-Salt Model of Hypertension for Studies of Cerebrovascular Function, Stroke, and Brain Health T. Michael De Silva and Frank M. Faraci Abstract The brain renin-angiotensin-aldosterone system (RAAS) regulates many physiological processes including fluid and electrolyte balance, vascular structure and function, blood pressure, cognition, and other aspects of brain function. Treatment with the mineralocorticoid deoxycorticosterone acetate and salt stimulates the local RAAS within the brain. In this chapter, we describe the surgical procedures used to induce activation of the brain RAAS with deoxycorticosterone acetate and salt. This technique can be used for studies of hypertension, cerebrovascular biology and dysfunction, and other diseases that impact brain health. Key words Cerebral artery, Deoxycorticosterone acetate, Renin-angiotensin-aldosterone system, Angiotensin II

1

Introduction Hypertension is a major cause of morbidity and mortality worldwide [1]. With regard to the brain, hypertension causes cerebrovascular dysfunction and is the major modifiable risk factor for stroke and multiple forms of dementia [2]. Despite the fact that a large number of stroke patients are hypertensive, preclinical studies very often do not examine the effect of ischemic or hemorrhagic stroke in the presence of hypertension. The lack of studying stroke or potential stroke therapeutics in the presence of comorbidities may be an important contributing factor to the failure of almost all preclinical stroke therapies in clinical trials. Essential hypertension is a complex disease with several underlying mechanisms that account for the presence of vascular disease and chronic increases in arterial pressure. These mechanisms include endothelial cell dysfunction; inward remodeling of resistance vessels; stiffening of the aorta and other large conduit arteries; activation of the sympathetic nervous system, the systemic or local renin-angiotensin-aldosterone systems (RAAS) (Fig. 1a),

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0_34, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

481

482

T. Michael De Silva and Frank M. Faraci

A)

B) RAAS

C) DOCA-Salt and the Brain RAAS

Plasma Renin

Angiotensinogen

Aldosterone

Renin Ang I ACE ACE2 Ang II

32% 12%

Deoxycorticosterone acetate C57Bl/6 mice (3 wks) Tap H2O or 0.15 mol/L NaCl No nephrectomy

Ang (1-7) Activation of the central RAAS

MR

AT1R

AT2R 56%

• vascular disease • vascular remodeling • vascular stiffening • increased SNS • thirst + salt appetite • increased vascular resistance • increased blood pressure • atherosclerosis

• Increased Ang II in CSF • Activation of AT1-receptors • Activation of sympathetic nerves

Normal High Low • Endothelial dysfunction • Increased arterial pressure • Suppression of the peripheral RAAS

Fig. 1 (a) Major components of the renin-angiotensin-aldosterone system (RAAS). Angiotensin II (Ang II) has classically been considered to be the main effector of the RAAS. Ang II is formed from angiotensin I (Ang I) by angiotensin-converting enzyme (ACE). Effects of Ang II occur as a result of activation of specific receptor subtypes. In relation to Ang II-driven signaling, two receptors are primarily involved (AT1 and AT2 receptors). In general, most of what are considered detrimental effects of Ang II are mediated by activation of the AT1 receptor (AT1R). In addition to activation of AT1R, Ang II can be degraded by ACE2 (forming Ang (1–7)) or can stimulate the production of aldosterone. Effects of aldosterone are mediated via mineralocorticoid receptors (MR). SNS sympathetic nervous system. (b) Distribution of plasma renin levels in patients with essential hypertension [3]. (c) Central effects of DOCA-salt treatment in C57BL/6 mice. These effects include activation of the brain RAAS, resulting in increased local production of Ang II, increased concentrations of Ang II in cerebrospinal fluid (CSF), activation of central AT1Rs, and sympathetic outflow. The consequences of these changes include increased arterial pressure and suppression of the peripheral RAAS. See text for additional details

and the immune system; and renal dysfunction [4, 5]. Interestingly, the underlying cause of hypertension can differ between populations or subgroups and with age. For example, hypertension with low circulating levels of renin occurs in approximately 30% of hypertensive individuals but is more common in blacks, the elderly, and in resistant hypertension (Fig. 1b) [3, 6]. These populations are also at increased risk of stroke. Although increases in circulating renin is the least common change seen in humans with essential hypertension (Fig. 1b), chronic systemic infusion of angiotensin II is the most common model of hypertension in rodents [7]. Induction of hypertension using deoxycorticosterone acetate (DOCA) and NaCl in the drinking water was first reported by Selye et al. in the 1940s [8]. Since then, many studies—including some that focused on the cerebral

DOCA-Salt and the Cerebral Circulation

483

circulation or stroke—have examined effects of DOCA-salt treatment on blood pressure and end-organ damage [9–13]. In the majority of studies, this model has included an additional modification (a uninephrectomy) to induce greater levels of hypertension and more severe injury. The DOCA-salt model of hypertension is relevant to human hypertension as it involves sodium retention, volume expansion, activation of the sympathetic nervous system, and vascular dysfunction (Fig. 1c). As a result of the increase in arterial pressure, there is simultaneous suppression of the peripheral RAAS [9, 11, 14]. Conversely, DOCA-salt treatment increases the activity and expression of the brain RAAS and increases cerebrospinal fluid levels of angiotensin II (Fig. 1c) [15, 16], which makes it a useful model for studying cerebrovascular dysfunction, stroke, and effects on neurological function and brain health. The DOCA-salt model is a “low renin” model of hypertension but one that produces two “hits” on the cerebral circulation. One hit is due to increased arterial pressure, while the second hit is due to increased local levels of angiotensin II in the brain [2, 9]. Here, we outline the methodology used to induce activation of the central RAAS using DOCA-salt (Fig. 1c).

2

Materials

2.1 Pellet Preparation (See Notes 1 and 2)

1. Deoxycorticosterone acetate.

2.2 Surgical Instruments and Setup (See Note 3)

1. Scissors—straight tip (e.g., Fine Science Tools Strabismus scissors 14574-11).

2. Parr Instruments pellet press with 0.25-inch punch and die set.

2. Forceps × 2 (e.g., Fine Science Tools standard pattern forceps 11000-16). 3. Anesthetic—inhalation (e.g., isoflurane) or injectable (e.g., ketamine [80–100 mg/kg] and xylazine [10–12.5 mg/kg]) (see Note 4). 4. Analgesic (e.g., carprofen 5 mg/kg) (see Note 4). 5. Hair clippers (see Note 5). 6. Sterile surgical gloves. 7. Surgical gown. 8. Heat mat. 9. Surgical drape. 10. Chlorhexidine scrub. 11. Cotton tip applicators. 12. Wound closure (see Note 6):

484

T. Michael De Silva and Frank M. Faraci

• Fine Science Tools Olsen-Hegar Needle Holders with Suture Cutters 12002-14 and non-absorbable suture (e.g., Ethilon 4-0 nylon suture with reverse cutting needle). • Wound clips (Fine Science Tools AutoCLIP system 1202000). 2.3

3

Drinking Water

1. 0.15 M NaCl in distilled H2O.

Methods All procedures should be completed under aseptic conditions to reduce the risk of infection in the animal. All surgical instruments and consumables should be sterilized using an autoclave.

3.1 Pellet Preparation (If Using Pre-purchased Pellets, Proceed to 3.2)

1. Weigh 50 mg of DOCA and compress into a pellet using the pellet press. Pellets should be stored under sterile conditions until used.

3.2 Subcutaneous Implantation of DOCA Pellet

1. Weigh the animal that is to be operated on and record the weight. 2. Turn on the heating pad and prepare anesthetic and analgesic according to the animal weight. 3. Anesthetize animal using the chosen method of anesthesia. 4. Inject animal with analgesic. Analgesics are more effective if administered prior to surgery. 5. Use hair clippers to shave the fur on the dorsal surface of the neck, from the scapulae to the base of the skull. 6. Apply chlorhexidine surgical scrub to sterilize the surgical area using a sterile cotton tip applicator. Chlorhexidine should be applied in a circular motion starting at the incision site and moving outward. 7. Investigator is to clean hands with surgical scrub and then wear surgical gloves. 8. Drape the animal to isolate the surgical site. 9. Use sterile scissors to make a midline 2 cm incision between the scapulae and base of the skull. 10. Use the sterile forceps to create a pocket between the skin and underlying muscle. The pocket should extend to the flank of the animal (Fig. 2). 11. Place pellet in the pocket. The pellet should be placed as far away from the incision site as possible. This reduces the chances of the pellet interfering with healing of the incision site.

DOCA-Salt and the Cerebral Circulation

485

Fig. 2 Location of subcutaneous pocket that is created with forceps for placement of the DOCA pellet. Created with Biorender.com

12. Close incision site with suture or wound clips. 13. Place animal in a heated recovery box until they regain consciousness. Sham-operated mice will undergo the same procedure except that no DOCA pellet will be placed into the pocket. 3.3 Post-Surgical Monitoring and Supply of 0.15 M NaCl in Drinking Water

1. Animals should be monitored for signs of pain, distress, and infection as per institutional guidelines. Administer 5 mg/kg/ day carprofen subcutaneously for 3 days post-surgery according to institutional guidelines. 2. Provide animals with both normal drinking water and water containing 0.15 M NaCl. The duration of DOCA-salt treatment can vary, although the use of a three-week protocol is common [9, 11]. Cerebrovascular changes in this DOCA-salt model occur in a time-dependent manner [9]. At the end of a given experimental period, mice may be euthanized for acute studies including assessment of cerebral blood flow or cerebrovascular function in vivo. Alternatively, cerebral arteries or arterioles may be removed and studied in vitro using pressure myography [9]. DOCA-salt treated mice may also be used in studies of stroke or cognitive function.

4

Notes 1. DOCA pellets can be purchased from Innovative Research of America (catalogue # M-121). 2. A pellet press with various size punch and die sets are available from Parr Instruments. 3. This is a suggestion for surgical instruments that may be used. Similar instruments may be suitable.

486

T. Michael De Silva and Frank M. Faraci

4. Choice of anesthesia and analgesia should be determined by the user in accordance with institutional and other regulatory guidelines. 5. Hair clippers are recommended for clearing hair from the surgical site. Chemical hair removal creams may induce skin burns. 6. Suture 3-0 or smaller is the recommended size for mice. 7. As we have shown previously [9, 11], the effectiveness of the DOCA-salt intervention can be confirmed in several ways. (1) DOCA-salt treatment will increase kidney weight compared with sham controls; (2) renal renin (Ren1) mRNA expression will be reduced following DOCA-salt treatment; or (3) mRNA expression of angiotensinogen (Agt) and angiotensin converting enzyme (Ace) is increased in the brain and cerebral vasculature. 8. The severity of the model can be increased by performing uninephrectomy and only providing 0.15 mol/L NaCl drinking water as has been performed previously [13].

Acknowledgments Studies that have been performed using this methodology have been supported by the National Institutes of Health (NS-096465, NS-108409). References 1. Mills KT, Stefanescu A, He J (2020) The global epidemiology of hypertension. Nat Rev Nephrol 16:223–237 2. Basting T, Lazartigues E (2017) DOCA-salt hypertension: an update. Curr Hypertens Rep 19:32 3. Alderman MH, Madhavan S, Ooi WL et al (1991) Association of the renin-sodium profile with the risk of myocardial infarction in patients with hypertension. N Engl J Med 324:1098–1104 4. Drummond GR, Vinh A, Guzik TJ et al (2019) Immune mechanisms of hypertension. Nat Rev Immunol 19:517–532 5. Oparil S, Acelajado MC, Bakris GL et al (2018) Hypertension. Nat Rev Dis Primers 4:18014 6. Carey RM, Calhoun DA, Bakris GL et al (2018) Resistant hypertension: detection, evaluation, and management: a scientific statement from the American Heart Association. Hypertension 72:e53–e90

7. Galis ZS, Thrasher T, Reid DM et al (2013) Investing in high blood pressure research: a national institutes of health perspective. Hypertension 61:757–761 8. Selye H, Hall CE, Rowley EM (1943) Malignant hypertension produced by treatment with Desoxycorticosterone acetate and sodium chloride. Can Med Assoc J 49:88–92 9. De Silva TM, Modrick ML, Grobe JL et al (2021) Activation of the central reninangiotensin system causes local cerebrovascular dysfunction. Stroke 52:2404–2413 10. Faraco G, Park L, Zhou P et al (2016) Hypertension enhances Abeta-induced neurovascular dysfunction, promotes beta-secretase activity, and leads to amyloidogenic processing of APP. J Cereb Blood Flow Metab 36:241–252 11. Grobe JL, Buehrer BA, Hilzendeger AM et al (2011) Angiotensinergic signaling in the brain mediates metabolic effects of deoxycorticosterone (DOCA)-salt in C57 mice. Hypertension 57:600–607

DOCA-Salt and the Cerebral Circulation 12. Matin N, Pires PW, Garver H et al (2016) DOCA-salt hypertension impairs artery function in rat middle cerebral artery and parenchymal arterioles. Microcirculation 23:571–579 13. Thomas JM, Ling YH, Huuskes B et al (2021) IL-18 (Interleukin-18) produced by renal tubular epithelial cells promotes renal inflammation and injury during Deoxycorticosterone/salt-induced hypertension in mice. Hypertension 78:1296–1309 14. Yemane H, Busauskas M, Burris SK et al (2010) Neurohumoral mechanisms in

487

deoxycorticosterone acetate (DOCA)-salt hypertension in rats. Exp Physiol 95:51–55 15. Itaya Y, Suzuki H, Matsukawa S et al (1986) Central renin-angiotensin system and the pathogenesis of DOCA-salt hypertension in rats. Am J Phys 251:H261–H268 16. Kubo T, Yamaguchi H, Tsujimura M et al (2000) Blockade of angiotensin receptors in the anterior hypothalamic preoptic area lowers blood pressure in DOCA-salt hypertensive rats. Hypertens Res 23:109–118

INDEX A Aging ..................................... 87, 88, 181, 264, 312, 454 Alternative .................................................. 26, 37, 53, 93, 110, 136, 260, 342, 416, 431 Angiotensin II ...................................................... 482, 483 Animal model ........................................... 4, 9, 29, 47, 49, 57, 114, 193, 213, 345, 376, 389, 392, 393, 430, 453–464, 467, 476 Appetitive behavior ....................................................... 279 Axonal mapping technique.................................. 171–180 Axonal sprouting................................... 70, 171–180, 372

B Barnes maze.........................................264, 265, 267–270 BDA-labeling........................................................ 175, 179 Behavior..................................................35, 38, 117, 269, 270, 275, 295, 308, 312–315, 317, 318, 320, 322, 334, 338, 341, 342, 346, 356, 359–365, 374, 442, 446, 447, 449, 474 Bilateral carotid artery stenosis (BCAS) ..................39–45 Blood-brain barrier (BBB).................................. 191–201, 380, 403–416, 419 Blood oxygen level dependent (BOLD)...................... 115 Brain........................................................4, 17, 21, 29, 39, 47, 55, 70, 83, 106, 113, 153, 171, 181, 191, 206, 213, 251, 264, 280, 330, 356, 370, 380, 392, 403, 419, 430, 444, 453, 467, 481 Brain drug disposition .................................................. 403 Brain injury...........................................22, 220, 264, 280, 330, 371, 372, 386, 430–432, 467 Burrowing............................................265, 266, 272–275

C Cell genesis .................................................................... 356 Cell proliferation ................................................ 13–20, 70 Central nervous system (CNS).........................9, 13, 191, 192, 214, 220, 251, 257, 263, 280–322, 328, 330, 379, 385, 403, 404, 413, 419 Cerebral artery ..................................................48, 55, 57, 60, 83–85, 394, 443, 453–464, 467, 485 Cerebral blood flow (CBF) ................. 39–45, 48, 52–54, 61, 83–95, 97, 98, 399, 422, 423, 446, 485 Cerebral hypoperfusion ............................................40, 47

Cerebral ischemia .............................................................. 8 Cerebral microhemorrhages ................................ 181–190 Cerebrovascular ............................................ 47, 391, 394, 396–398, 400, 442, 453, 481–483, 485 Chitosan/β- glycerophosphate............................ 381–384 Chronic cerebral hypoperfusion..................................... 49 Chronic stroke...................................................... 429–436 Cognitive function ....................... 39, 264, 280, 423, 485 Cognitive impairment ................................ 39, 47, 48, 69, 263–275, 280, 468 Cranial window7, 70, 71, 74–77, 79, 88, 106, 116, 124, 127, 143, 144, 478 Cylinder test ............................... 346, 348–351, 363, 374

D Deoxycorticosterone acetate (DOCA) ............... 482–485 Diabetes ...................................................... 376, 394, 424, 429, 430, 432, 435, 436, 453, 467–478 Diffusion-weighted MRI .............................................. 154 Distal MCAO (dMCAO) ............................................. 5, 6 Drug development .......................................380, 441–450 Drug permeability ......................................................... 419 Drugs promoting functional recovery ......................... 369 Drug transport .............................................................. 404

E EdU (2’-Deoxy-5-ethynyluridine)..............13–15, 17–20 Enriched environment (EE) ................................ 355–365 Executive function ........................................................ 279

F FITC-dextran ...............................................193–196, 200 Flow cytometry .......................................... 214, 217, 226, 227, 231, 232, 240, 395 Flow cytometry analysis ...............................214, 231–248 FlowJo ............................... 228, 232, 233, 240, 244, 247 Fluorescently activated cell sorting/staining (FACS) ................... 214, 217, 218, 221–225, 227 Forelimb function ................................................ 328, 346 Functional impairment .........................21, 192, 357, 376 Functional recovery ......................................3–9, 70, 117, 118, 214, 263–275, 345, 355–365, 370, 371, 375, 376

Vardan T. Karamyan and Ann M. Stowe (eds.), Neural Repair: Methods and Protocols, Methods in Molecular Biology, vol. 2616, https://doi.org/10.1007/978-1-0716-2926-0, © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

489

NEURAL REPAIR: METHODS AND PROTOCOLS

490 Index G

Genetically encoded calcium indicator (GECI) .......... 115 Glial scar ........................................................................ 355 Grid-walking.................................................345–352, 374

H Hematoxylin and eosin (H & E) ................181–187, 189 Hemorrhagic transformation .............................. 467, 478 Hippocampus ....................................... 4, 22, 29–38, 174, 207, 256, 259, 282, 304, 363 Hydrogel.............................................................. 380–387, 389, 468, 474, 478 Hyperglycemia..................................................... 429–432, 435–437, 467–478

I Immune cell co-culture ................................................ 258 Infra-slow activity.......................................................... 113 Internal carotid artery.................................. 8, 47, 51, 52, 84, 85, 92, 410, 446, 455, 472 Intraoperative imaging.................................................... 97 Inverted laser speckle contrast imaging ......................... 90 In vitro ................................................251–260, 381, 385, 386, 403–416, 421, 485 In vivo drug delivery............................................ 419, 421 Ischemia.................................................. 4–9, 49, 62, 114, 115, 118, 364, 397, 399, 421, 432, 454, 455, 460, 474 Isometric pull task................................................ 328, 342

L Laser speckle contrast imaging (LSCI) .................. 40, 45, 83–95, 97–99 Lymphocytes ............................................... 213, 216, 397

M Magnetic nanoparticle ......................................... 8, 55–65 Magnetized red blood cells (mRBCs) .....................55–65 Material...................................... 8, 14, 15, 22, 23, 30–32, 40, 41, 49, 50, 56–60, 73, 88–90, 94, 100–101, 119–121, 128, 154–155, 163, 172, 173, 182–184, 195, 196, 206, 207, 211, 214–216, 220, 232, 253, 254, 265, 266, 286, 288, 289, 318, 320, 328–330, 346, 347, 357–360, 364, 380–382, 387, 404–406, 414, 420, 421, 433, 443–444, 455–457, 463, 468–469, 483–484 Micro-coil .................................40, 41, 43, 45, 48, 50–53 Micro-dissection.......................................... 195, 206, 208 Microstructure .............................................................. 153

Middle cerebral artery occlusion (MCAO) ............4–6, 8, 55, 72, 83–95, 160, 161, 360, 396, 397, 415, 443, 447, 454, 455, 464, 467–478 Mixed cortical cultures ........................................ 251–260 Model..................................................3–9, 18, 21, 22, 25, 29–38, 40, 47–56, 59, 63, 72, 83–95, 97, 99, 103, 113–145, 160, 166, 178, 182, 189, 220, 222, 251, 263, 264, 268, 275, 280, 284, 327, 328, 345, 346, 352, 356, 357, 370, 371, 380, 381, 385, 389, 396–398, 411, 412, 415, 419, 421, 429–436, 442, 454, 455, 467–478, 482, 483, 485, 486 Morphometric ............................................. 161, 194, 199 Motor impairment ..........................................7, 264, 327, 346, 370, 374 Mouse ................................................... 4, 6, 7, 13–20, 22, 23, 25–27, 29–42, 44, 47–64, 70, 71, 74, 75, 77, 78, 83–95, 99, 106, 113–145, 154–156, 158, 182, 184, 186, 187, 189, 195–197, 215, 216, 218, 219, 227, 253, 254, 264, 266, 272–274, 280–322, 327–343, 345–352, 356–361, 363, 364, 370, 381, 386, 419, 421–423, 430–432, 434, 435, 437, 441–450, 454, 455, 457, 460–464, 482, 485, 486 Mouse brain................................................ 30, 37, 59, 70, 71, 73–75, 131, 155, 157, 158, 161, 181–190, 199, 200, 207, 209, 213–228 Mouse permanent occlusion stroke model..............21–27 Moyamoya .................................................................47–54 Multi-exposure speckle imaging (MESI)............. 99–101, 104–106, 108–110

N Neuroimaging .....................................113–117, 161, 165 Neuro-immune ............................................................. 253 Neurorecovery........................................ 9, 214, 263–275, 355, 356, 363–365 Neurosurgery ................................................................ 475 Neurotherapeutics............................................................. 7 N5-(1-iminoethyl)-L-ornithine HCL............................ 14 Non-linear dimensionality reduction ........................... 232 Novel object recognition test (NORT) ......................264, 266, 270, 271, 274

O Operant behavior .......................................................... 315 Operant reach chambers ...................................... 327–343 Optical imaging..............................................97, 113–145 Optical intrinsic optical imaging (OISI)............. 114, 115

NEURAL REPAIR: METHODS Outcomes ..................................................4–9, 25, 39–45, 53, 115, 122, 300, 303, 345, 356, 357, 360, 371, 373–375, 392, 394, 397–399, 420, 430, 441, 442, 454, 467, 468, 474, 476, 478

P Perinatal arterial ischemic stroke (PAIS) .................55, 56 Perinatal stroke................................................................ 55 PFA-fixation .................................................................. 205 Pharmacotherapy..........................................370–373, 375 Photothrombosis...............................................21, 22, 25, 27, 29, 30, 56, 72, 79, 136, 352, 360 Photothrombotic stroke ...................................21, 22, 70, 72, 74, 77, 178, 380, 386 Plasticity.............................................. v, 6, 9, 69–79, 171, 260, 356, 371–373, 375 Post-stroke recovery ...................... 25, 69, 369, 419–424 Preclinical stroke study .......................................... 97, 404 Protocol ................................................13, 18, 19, 30, 37, 48, 70, 88, 99, 100, 114, 115, 117–118, 120–121, 127–131, 155, 158, 172, 178, 184, 200, 205–207, 209, 211, 214, 218, 221–224, 227, 228, 232, 240, 254, 259, 264, 266, 272–274, 284, 285, 288, 290–295, 304, 309, 315, 317, 322, 328, 330, 337, 342, 343, 346, 356, 357, 359, 360, 385, 395, 400, 416, 442, 444, 450, 455, 485 Prussian blue ........................................................ 181–187 PVA-tyramine ......................................381, 382, 384–385

Q Quantitative...................................................97–110, 153, 194, 197, 328 Quantitative assessment ................................................ 193

AND

PROTOCOLS Index 491

Reperfusion ................................................ 4, 5, 8, 22, 56, 63, 83, 84, 88, 92, 93, 415, 421, 422, 446–448, 450, 455, 458, 461–463, 478 Resting-state functional connectivity (RS-FC) ..........114, 118, 138, 145 RNA extraction ............................................................. 206 RNA sequencing (RNA-seq)............................... 205–211 Rodent ....................................................4, 6, 8, 9, 21, 22, 31, 39, 55–65, 84, 97, 114, 264, 268, 270, 271, 274, 275, 280, 283, 284, 303, 309, 311–313, 315, 317, 319, 321, 322, 327, 356, 357, 361–364, 370–375, 381, 386, 397, 405, 420, 429–436, 449, 454, 455, 476, 482 Rodent brain ........................................................ 153–166 Rotating pole test................................................. 362, 363

S SIMPLE................................................................ 8, 56–63 SIMPLeR ................................................. 8, 56, 57, 59–63 Single-cell suspension ................................. 214, 221, 223 Small interfering RNA (siRNA) ..................373, 419–424 Sociability ...................................................................... 272 Spleen............................................................213–228, 423 Streptozotocin.....................................430, 433–435, 468 Stroke......................................................3, 13, 21, 29, 47, 55, 69, 83, 97, 113, 153, 171, 181, 192, 213, 232, 251, 263, 280, 327, 345, 355, 369, 379, 391, 403, 419, 430, 441, 453, 467, 481 Stroke mouse model .....................................v, 3–9, 21–27 Stroke pharmacotherapy .....................370, 372, 373, 375 Stroke recovery................................................6–9, 21, 26, 345–352, 364, 365, 369–376, 386 Sucrose preference .............................................. 264, 266, 271, 272, 274, 275, 423

R

T

Rat.........................................................7, 40, 56, 84, 154, 155, 157–161, 216, 264, 266–268, 270, 272–274, 328, 357, 358, 396, 397, 410, 419, 434, 435, 437, 467–478 Recovery ............................................... 4, 6–9, 22, 23, 25, 37, 44, 52, 69–71, 75–77, 114, 115, 213, 285, 327, 330, 338, 346, 356, 359, 360, 362–364, 370–376, 379, 386, 389, 423, 454, 460, 461, 463, 468, 473, 474, 485 Rehabilitation .................................... 6, 70, 71, 213, 327, 338, 342, 346, 356, 360, 370–373, 375 Remodeling ........................... 70, 98, 114, 117, 356, 481 Renin-angiotensin-aldosterone system (RAAS)...................................................... 481–483

Therapeutics ...........................................7, 9, 25, 83, 213, 327, 345, 370, 372, 375, 376, 379–389, 391–400, 403, 419, 420, 423, 424, 467, 481 3D printed imaging platform ..................... 89, 90, 92, 94 Tight junction protein .................................................. 191 Toxicity ......................................................... 57, 372, 386, 420, 423, 424, 430, 436, 462, 470, 476 Toxicology ..................................................................... 441 Transfection..................................................420–422, 424 Translational ............................................. 3, 87, 102, 107, 108, 214, 280, 327, 360, 369, 370, 380, 392, 393, 396–398, 403, 430 T-stochastic neighbor embedding (tSNE) .................231, 232, 236, 240–245, 247, 248

NEURAL REPAIR: METHODS AND PROTOCOLS

492 Index

Two-photon microscopy ................................................ 72 Type 1 diabetes .................................................... 394, 429

Vascular dementia ..................................... 9, 39, 220, 394 Vascular stenosis .............................................................. 47

U

W

Uniform manifold approximation projection (UMAP).......................................... 231, 232, 236, 240–245, 247, 248

White matter degeneration............................................. 39 White matter stroke (WMS) ................................ 9, 13–20 Wide-field optical imaging (WFOI)................... 115–129, 131–134, 138, 143

V Vasoconstriction ............................................................ 7, 9