Developmental Brain Injury 3805575726, 9783805575720

Citation preview

Developmental Brain Injury Proceedings and Abstracts of the Third Hershey Conference on Developmental Cerebral Blood Flow and Metabolism Hershey, Pa., June 6–9, 2002

Guest Editors

Susan J. Vannucci, New York, N.Y. Robert C. Vannucci, Hershey, Pa. Steven W. Levison, Hershey, Pa.

44 figures, 5 in color, and 7 tables, 2002

All papers have undergone the Journal’s usual peer review

Basel 폷 Freiburg 폷 Paris 폷 London 폷 New York 폷 Bangalore 폷 Bangkok 폷 Singapore 폷 Tokyo 폷 Sydney

S. Karger Medical and Scientific Publishers Basel 폷 Freiburg 폷 Paris 폷 London New York 폷 Bangalore 폷 Bangkok Singapore 폷 Tokyo 폷 Sydney

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Drug Dosage The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.

All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center (see ‘General Information’). © Copyright 2002 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–7572–6

Vol. 24, No. 5, 2002

Contents

347 Preface Vannucci, S.J. (New York, N.Y.); Vannucci, R.C.; Levison, S.W. (Hershey, Pa.)

Commentaries 349 Timing Is Everything – Delaying Therapy for Delayed Cell Death Ferriero, D.M. (San Francisco, Calif.) 352 New Approaches to Brain Injury in Preterm Infants Edwards, D. (London)

Review 355 Under What Circumstances Can Seizures Produce Hippocampal Injury:

Evidence for Age-Specific Effects Galanopoulou, A.S.; Vidaurre, J.; Moshé, S.L. (Bronx, N.Y.)

Perspectives 364 Animal Models of Developmental Brain Injury: Relevance to Human

Disease. A Summary of the Panel Discussion from the Third Hershey Conference on Developmental Cerebral Blood Flow and Metabolism Hagberg, H. (Göteborg); Ichord, R. (Philadelphia, Pa.); Palmer, C. (Hershey, Pa.); Yager, J.Y. (Saskatoon); Vannucci, S.J. (New York, N.Y.)

Original Papers 367 Prolonged Neonatal Seizures Exacerbate Hypoxic-Ischemic Brain Damage:

Correlation with Cerebral Energy Metabolism and Excitatory Amino Acid Release Yager, J.Y.; Armstrong, E.A.; Miyashita, H. (Saskatoon); Wirrell, E.C. (Calgary) 382 Neuroprotection of Creatine Supplementation in Neonatal Rats with

Transient Cerebral Hypoxia-Ischemia Adcock, K.H.; Nedelcu, J.†; Loenneker, T.; Martin, E.; Wallimann, T. (Zurich); Wagner, B.P. (Berne) 389 Inhibition of nNOS and iNOS following Hypoxia-Ischaemia Improves

Long-Term Outcome but Does Not Influence the Inflammatory Response in the Neonatal Rat Brain van denTweel, E.R.W.; Peeters-Scholte, C.M.P.C.D.; van Bel, F.; Heijnen, C.J.; Groenendaal, F. (Utrecht)

© 2002 S. Karger AG, Basel Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/dne_issues

396 Effects of Selective Nitric Oxide Synthase Inhibition on IGF-1, Caspases and

Cytokines in a Newborn Piglet Model of Perinatal Hypoxia-Ischaemia Peeters-Scholte, C.; Koster, J.; van den Tweel, E. (Utrecht); Blomgren, K. (Göteborg); Hamers, N. (Utrecht); Zhu, C. (Göteborg); van Buul-Offers, S. (Utrecht); Hagberg, H. (Göteborg); van Bel, F.; Heijnen, C.; Groenendaal, F. (Utrecht) 405 Evidence that p38 Mitogen-Activated Protein Kinase Contributes to

Neonatal Hypoxic-Ischemic Brain Injury Hee Han, B. (Seoul/St. Louis, Mo.); Choi, J.; Holtzman, D.M. (St. Louis, Mo.) 411 Hypoxic Preconditioning Increases Brain Glycogen and Delays Energy

Depletion from Hypoxia-Ischemia in the Immature Rat Brucklacher, R.M.; Vannucci, R.C. (Hershey, Pa.); Vannucci, S.J. (New York, N.Y.) 418 Early Appearance of Functional Deficits after Neonatal Excitotoxic and

Hypoxic-Ischemic Injury: Fragile Recovery after Development and Role of the NMDA Receptor Felt, B.T. (Ann Arbor, Mich.); Schallert, T. (Ann Arbor, Mich./Austin, Tex.); Shao, J.; Liu, Y.; Li, X.; Barks, J.D.E. (Ann Arbor, Mich.) 426 Damage to the Choroid Plexus, Ependyma and Subependyma as a

Consequence of Perinatal Hypoxia/Ischemia Rothstein, R.P.; Levison, S.W. (Hershey, Pa.) 437 IGF-I and NT-3 Signaling Pathways in Developing Oligodendrocytes:

Differential Regulation and Activation of Receptors and the Downstream Effector Akt Ness, J.K.; Mitchell, N.E.; Wood, T.L. (Hershey, Pa.)

446 Abstracts 465 Author Index 466 Subject Index

346

Contents

Dev Neurosci 2002;24:347–348 DOI: 10.1159/000069042

Preface Susan J. Vannucci a Robert C. Vannucci b Steven W. Levison b a Department

of Pediatric Critical Care Medicine, Columbia University, New York, N.Y., and of Pediatrics and Neuroscience and Anatomy, Pennsylvania State University College of Medicine, Hershey, Pa., USA b Departments

This special issue of Developmental Neuroscience represents the proceedings of the Third Hershey Conference on Developmental Cerebral Blood Flow and Metabolism, which was held at the Hershey Hotel, Hershey, Pa., USA, June 6–9, 2002. Drs. Susan and Robert Vannucci had organized this conference following two previously successful conferences in June of 1997 and 2000. As in previous years, this year’s conference was an international meeting (with approximately 65 participants) that was highly interactive. Twenty-seven participants from 8 countries, including the USA, presented plenary talks, with 17 additional poster presentations. The conference brought together clinicians, basic scientists, fellows and graduate students in a relaxed setting to share their discoveries and thoughts with the goal of understanding the mechanisms that lead to perinatal brain injury and progress in identifying new treatments. The conference was funded largely by the National Institute of Neurological Disorders and Stroke (1R13 NS 43136) with an additional contribution from the Department of Pediatrics at Hershey. Scientific sessions began with a keynote address by Dr. Richard Traystman, of Johns Hopkins University School of Medicine. Dr. Traystman presented an enlightening, entertaining, and provocative seminar on the cerebrovascular effects of cocaine in the fetus, newborn and adult. The next talks focused on the mechanisms of cerebral

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0347$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

maturation and included talks by Mary McKenna, University of Maryland; William J. Pearce, Loma Linda University; Steven W. Levison, Penn State University; Raymond Koehler, Johns Hopkins University, and Elly Rocha, University of Western Ontario. The afternoon session of the first day focused on signal transduction pathways as targets for cell preservation after hypoxia/ischemia. Talks were presented by Donna Ferriero, University of California, San Francisco; Timothy Schallert, University of Texas, Austin; David Holtzman, Washington University School of Medicine; Henrik Hagberg, Sahlgrenska University Hospital, Goteborg, Sweden; Tony Cheung, University of Hong Kong, Hong Kong, SAR China, and Christine Fox, University of California, San Francisco. The morning session of the second day was divided into two sessions: the first focused on seizures, and the second session focused on white matter damage. Seizures are a frequent accompaniment to perinatal brain damage, and so, they were a topic of several presentations. Talks were presented by Harry Chugani, Wayne State University; Solomon Moshé, Albert Einstein College of Medicine, and Jerome Yager, University of Saskatchewan, Canada. White matter damage is frequently observed in survivors of perinatal hypoxic/ischemic brain damage and hence was the topic of several presentations. Talks on this subject were presented by Terri Wood, Penn State College of

Susan J. Vannucci Research Director, Pediatric Critical Care Medicine Morgan Stanley Children’s Hospital of New York/Columbia University 3959 Broadway, BHN 10–24, New York, NY 10032 (USA) Tel. +1 212 342 0275, Fax +1 212 342 2293, E-Mail [email protected]

Medicine; Stephen A. Back, Oregon Health Sciences University, and Stephen Miller, University of California, San Francisco. The afternoon session of the second day focused again on the mechanisms of hypoxic/ischemic brain damage. Talks were presented by David Edwards, Imperial College of Medicine, Hammersmith Hospital, London, UK; Sidhartha Tan, Northwestern University; Cacha PeetersScholte, University Medical Center, Wilhelmina Children’s Hospital, Utrecht, The Netherlands; Evelyn R. W. van den Tweel, University Medical Center, Wilhelmina Children’s Hospital, Utrecht, The Netherlands; Stéphane Sizonenko, University of Geneva, Geneva, Switzerland; Courtney Robertson, University of Maryland School of Medicine; Ben Wagner, Unversity Children’s Hospital, Berne Switzerland, and Andreas W. Loepke, University of Pennsylvania School of Medicine.

348

Dev Neurosci 2002;24:347–348

The final day of the meeting included a special talk by Dr. Marvin Cornblath, who presented a historical perspective on newborn hypoglycemia. Following his talk, the attendees participated in a panel discussion on animal models of developmental brain injury. A synopsis of this discussion is provided in this special issue. At the end of the conference, there was complete agreement that this had been yet another extremely exciting and enlightening meeting, and it was unanimously decided to schedule a fourth Conference on Developmental Cerebral Blood Flow and Metabolism for June 2004. It was further agreed that, although the conference will continue to be called the Hershey Conference, the location will now alternate to include the West Coast, East Coast, and Europe. Donna Ferriero offered to host the next meeting, which will likely be in the Bay area of California.

Vannucci/Vannucci/Levison

Commentary Dev Neurosci 2002;24:349–351 DOI: 10.1159/000069048

Received: August 22, 2002 Accepted: October 13, 2002

Timing Is Everything – Delaying Therapy for Delayed Cell Death Donna M. Ferriero Departments of Neurology and Pediatrics, University of California San Francisco, San Francisco, Calif., USA

Key Words Cell death W Therapy W Neonate W Brain

Damage after a hypoxic-ischemic (HI) insult is both region and cell population specific [Johnston, 1998]. This observation is critically evident in, and important to, the pathogenesis of the insult in the immature central nervous system. In the term human neonate, damage to the deep grey nuclei is seen using magnetic resonance imaging (MRI) modalities such as diffusion tensor imaging, MR spectroscopy, and structural MRI [Barkovich et al., 1999, 2001]. The loss of T2-weighted signal in the posterior limb of the internal capsule has become the gold standard in predicting neurological injury with MRI [Rutherford et al., 1995]. Early spectroscopic changes of a variety of metabolites like lactate, N-acetyl aspartate, choline and creatine using single-voxel techniques, coupled with the presence of clinical neonatal seizures can aid in the accurate prediction of neurodevelopmental outcome after HI in the term newborn [Miller et al., 2002]. Neonatal physiological profiles in the first 24 h of life also contribute to a better prediction of outcome in the mildly and severely affected newborn [Newton et al., 2001]. Despite these advances in the recognition of injury and prediction of neurodevelopmental outcome after HI, therapeutic advances are essentially nonexistent for the term

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0349$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

newborn HI encephalopathy and stroke. This is in part due to misguided assumptions that animal studies in the mature nervous system extrapolate to those of the neonatal brain, and further assumptions that those data can be translated to clinical trials. In fact, that approach has failed to produce a significant reduction of morbidity and mortality from stroke in the adult, and certainly no therapies exist for HI or stroke in the newborn [Stroke Progress Review Group, 2002], even though a plethora of basic science contributions regarding the cell biology of the pathological processes exist [Dirnagl et al., 1999; Vexler and Ferriero, 2001]. A potential reason for this failure is that the therapeutic time window for treatment is thought to be limited to a few hours after the event. In point of fact, the tissue plasminogen activator trial for adult stroke excluded enrollment after 6 h, and the recent ‘head cooling’ trial for perinatal asphyxia did likewise [Gunn et al., 1998; Thoresen and Whitelaw, 2000]. However, if one critically examines the existing cell biological data, especially for the developing nervous system, it is clear that neuropathological changes evolve over weeks [Towfighi and Mauger, 1998], and that both the severity [Geddes et al., 2001] and age at the time of insult [Towfighi et al., 1997] dictate regional vulnerability of brain structures. In both cortex and striatum, there are at least biphasic stages of neurodegeneration after an HI insult in the neonatal rodent: early (1.5– 3 h) and late (6 days!) [Northington et al., 2001].

Donna M. Ferriero, MD Departments of Neurology and Pediatrics, University of California San Francisco 521 Parnassus Avenue C215 San Francisco, CA 94143-0114 (USA) Tel. +1 415 502 5820, Fax +1 415 502 5821, E-Mail [email protected]

Fig. 1. Gene expression after ischemia. This cascade is documented for the mature rodent after focal ischemia, but is

only presumed for the neonatal brain. Adapted from Ellison et al. [1999], copyright 1999, New York Academy of Sciences, USA.

Careful anatomic studies demonstrate injury evolution and ongoing cell death through 168 h [Nakajima et al., 2000] in rat cortex and striatum, as well as in mouse hippocampus [Sheldon et al., 2001]. The nature of this cell death is complex, and has recently been termed the ‘apoptotic-necrotic continuum’ by Martin et al. [Martin et al., 1998]. Even in pure ischemia models in the neonatal rodent, where blood flow is only transiently perturbed, injury evolves with a distinct time course and morphology [Manabat et al., 2000] in the core and penumbral regions of the infarct. One proposal for future investigations is to treat early, and treat again, using therapies (note plurality) to address the cascade and time course of toxic events as they occur in the young brain (fig. 1). When the existing literature from studies mapping gene and protein expression after neonatal HI is compiled, there are clear profiles that emerge for a variety of cytotoxic and repair processes [Vexler and Ferriero, 2001; Ferriero, 2001]. Stress (HI)

350

Dev Neurosci 2002;24:349–351

activates the intermediate early genes (c-fos, c-jun, HSP) [Herdegen and Leah, 1998]. Simultaneously glutamate receptors are activated by energy depletion and release of presynaptic glutamate, triggering neuronal NOS activity in vulnerable regions [Ferriero and Ashwal, 2002]. These injurious signals lead to intracellular oxidation of lipids, proteins and nucleotides which results in structural damage to mitochondria, endoplasmic reticulum, golgi, and plasma membranes. Selective cell death results in the availability of resistant killer (nNOS-containing) neurons partnered with invading microglia to stimulate inflammatory processes and release more neurotoxins (OONO-, cytokines, chemokines) that must be silenced in order for the vicious cycle to end [Beckman and Koppenol, 1996]. Repair can then occur when growth factors, adhesion molecules and structural proteins and constituents, and nurturing transcription factors like HIF1· are expressed in regions of cell demise [Bergeron et al., 2000].

Ferriero

Therapy should be instituted ideally before the insult if we could be aware of impending doom (unfortunately babies do not have classical transient ischemic attacks). However, there is a wide therapeutic window of opportunity in the developing nervous system that would allow for rescue of dying cells (both neurons and oligodendrocytes), even after the injury is complete. Stem cell biologi-

cal advances can lead to promising areas of rescue in the promiscuous developing brain. Broadening the therapeutic window, and designing age-appropriate therapies that address both cellular and regional selective vulnerability will eventually make perinatal HI encephalopathy, like other neurological disorders, a treatable condition.

References Barkovich AJ, Baranski K, Vigneron D, Partridge JC, Hallam DK, Hajnal BL, Ferriero DM (1999): Proton MR spectroscopy in the evaluation of brain injury in asphyxiated, term neonates. AJNR Am J Neuroradiol 20:1399– 1405. Barkovich AJ, Westmark KD, Bedi HS, Partridge JC, Ferriero DM, Vigneron DB (2001): Proton spectroscopy and diffusion imaging on the first day of life after perinatal asphyxia: Preliminary report. AJNR Am J Neuroradiol 22:1786– 1794. Beckman JS, Koppenol WH (1996): Nitric oxide, superoxide, and peroxynitrite: The good, the bad, and ugly. Am J Physiol 271(5 Pt 1): C1424–C1437. Bergeron M, Gidday JM, Yu AY, Semenza GL, Ferriero DM, Sharp FR (2000): Role of hypoxia-inducible factor-1 (HIF-1) in hypoxia-induced ischemic tolerance in neonatal rat brain. Ann Neurol 48:285–296. Dirnagl U, Iadecola C, Moskowitz MA (1999): Pathobiology of ischaemic stroke: An integrated view. Trends Neurosci 22:391–397. Ellison JA, Barone FC, Feuerstein GZ (1999): Matrix remodeling after stroke: De novo express of matrix proteins and integrin receptors. Ann NY Acad Sci 890:204–222 (fig. 1, p 209). Ferriero DM (2001): Oxidant mechanisms in neonatal hypoxia-ischemia. Dev Neurosci 23:198– 202. Ferriero DM, Ashwal S (2002): Effects of nitric oxide on neuronal and cerebrovascular function; in Donn SM, Sinha SK, Chiswick ML (eds): Birth Asphyxia and the Brain: Basic Science and Clinical Implications. Armonk, Futura, pp 153–187.

Delaying Therapy for Delayed Cell Death

Geddes R, Vannucci RC, Vannucci SJ (2001): Delayed cerebral atrophy following moderate hypoxia-ischemia in the immature rat. Dev Neurosci 23:180–185. Gunn AJ, Gluckman PD, Gunn TR (1998): Selective head cooling in newborn infants after perinatal asphyxia: A safety study. Pediatrics 102(4 Pt 1):885–892. Herdegen T, Leah JD (1998): Inducible and constitutive transcription factors in the mammalian nervous system: Control of gene expression by Jun, Fos and Krox, and CREB/ATF proteins. Brain Res Brain Res Rev 28:370–490. Johnston MV (1998): Selective vulnerability in the neonatal brain. Ann Neurol 44:155–156. Manabat C, Derugin N, Muramatsu K, Wendland M, Gregory G, Ferriero D, Vexler Z (2000): Oxidative injury after neonatal transient middle cerebral artery occlusion in the rat. Stroke 31:344. Martin LJ, Al-Abdulla NA, Brambrink AM, Kirsch JR, Sieber FE, Portera-Cailliau C (1998): Neurodegeneration in excitotoxicity, global cerebral ischemia, and target deprivation: A perspective on the contributions of apoptosis and necrosis. Brain Res Bull 46:281–309. Miller SP, Weiss J, Barnwell A, Ferriero DM, Latal-Hajnal B, Ferrer-Rogers A, Newton N, Partridge JC, Glidden DV, Vigneron DB, Barkovich AJ (2002): Seizure-associated brain injury in term newborns with perinatal asphyxia. Neurology 58:542–548. Nakajima W, Ishida A, Lange MS, Gabrielson KL, Wilson MA, Martin LJ, Blue ME, Johnston MV (2000): Apoptosis has a prolonged role in the neurodegeneration after hypoxic ischemia in the newborn Rat. J Neurosci 20:7994–8004. Newton NR, Miller SP, Ferriero DM, Kaufman SA, Barkovich AJ, Partridge JC (2001): SNAPPE as a predictor of neurodevelopmental outcome at 1 and 2.5 years of age following perinatal depression. Pediatr Res 49:310A.

Northington FJ, Ferriero DM, Graham E, Traystman RJ, Martin LJ (2001): Early neurodegeneration after hypoxia-ischemia in neonatal rat is necrosis while delayed neuronal death is apoptosis. Neurobiol Dis 8:207–219. Report of the Stroke Progress Review Group: NIH Public No. 02.5117, April 2002. Rutherford MA, Pennock JM, Schwieso JE, Cowan FM, Dubowitz LM (1995): Hypoxic ischaemic encephalopathy: Early magnetic resonance imaging findings and their evolution. Neuropediatrics 26:183–191. Sheldon RA, Hall JJ, Noble LJ, Ferriero DM (2001): Delayed cell death in neonatal mouse hippocampus from hypoxia-ischemia is neither apoptotic nor necrotic. Neurosci Lett 304:165– 168. Thoresen M, Whitelaw A (2000): Cardiovascular changes during mild therapeutic hypothermia and rewarming in infants with hypoxic-ischemic encephalopathy. Pediatrics 106(1 Pt 1): 92–99. Towfighi J, Mauger D (1998): Temporal evolution of neuronal changes in cerebral hypoxia-ischemia in developing rats: A quantitative light microscopic study. Brain Res Dev Brain Res 109:169–177. Towfighi J, Mauger D, Vannucci RC, Vannucci SJ (1997): Influence of age on the cerebral lesions in an immature rat model of cerebral hypoxiaischemia: A light microscopic study. Brain Res Dev Brain Res 100:149–160. Vexler ZS, Ferriero DM (2001): Molecular and biochemical mechanisms of perinatal brain injury. Semin Neonatol 6:99–108.

Dev Neurosci 2002;24:349–351

351

Commentary Dev Neurosci 2002;24:352–354 DOI: 10.1159/000069054

Received: June 6, 2002 Accepted: June 20, 2002

New Approaches to Brain Injury in Preterm Infants David Edwards Weston Laboratory, Department of Pediatrics, Imperial College Faculty of Medicine, and the Medical Research Council Clinical Sciences Centre, Hammersmith Hospital, London, UK

Key Words Intrauterine infection W Hypoxia W Ischemia W Cytokines W Neonatal encephalopathy

Preterm birth is strongly associated with neurodevelopmental impairment, but the causes of cerebral injury are not completely understood. It is often assumed that damage is due to cerebral hypoxia-ischemia secondary to the extreme stresses induced by prolonged intensive care. However, the primary role of reduced cerebral blood flow is questioned by the findings of several studies that normal cerebral blood flow can be as low as 5–10 ml/ 100 g –1/min –1 in infants who have normal neurological outcome [1, 2]. Indeed, some commonly used drugs, particularly indomethacin, but also aminophylline, cause significant falls in cerebral blood flow without any clear evidence that they lead to cerebral injury [3, 4]. Recent evidence suggests that a considerable number of extremely preterm infants might suffer some element of cerebral damage in utero. Previous results using cerebral ultrasonography suggested that perhaps one in twenty infants were delivered with definable cerebral abnormality [5]. However, in a recent study in which a consecutive cohort of preterm infants born before 30 weeks gestation

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0352$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

were studied using magnetic resonance (MR) imaging on the first days after delivery about half of the infants showed long-standing abnormalities [6]. Several factors have led researchers to consider the possible role of intrauterine infection or inflammation in cerebral damage. First, paediatricians are very familiar with the neurological sequelae of meningitis and septicaemia; second, there is increasing epidemiological evidence of a link between maternofetal infection and neonatal encephalopathy in term infants [7]; third, the strong relation between preterm delivery before 30 weeks gestation and infection offers the possibility of unifying hypothesis connecting both delivery and brain injury; fourth, intrauterine inflammation has been shown to be associated with lung damage and chronic lung disease in preterm infants [8, 9]. The evidence that preterm birth is strongly associated with intrauterine infection and inflammation is strong. Certainly, local uterine inflammation has been demonstrated in association with preterm delivery, and both chorioamnionitis and high amniotic fluid concentrations of the proinflammatory cytokines are found among infants born before 30 weeks of gestation [10]. Indeed, amniotic fluid proinflammatory cytokines can be increased some weeks before preterm birth [11]. There is also evidence of a systemic maternal immune response:

A. David Edwards, MD Weston Laboratory, Division of Paediatrics, Obstetrics and Gynaecology Imperial College School of Science, Technology and Medicine Hammersmith Hospital, Du Cane Rd, London W12 0NN (UK) Tel. +44 208 383 3326, E-Mail [email protected]

maternal plasma IgM concentrations were found to be increased in mid-pregnancy in women who went on to deliver preterm [12]. Although the effects of a chronically inflamed intrauterine environment on the human fetus are poorly understood, the developing brain seems to be specifically vulnerable to inflammatory injury. In immature rat pups, inflammatory stimuli cause increased blood-brain barrier breakdown with enhanced phagocyte diapedesis mediated by CXC chemokines, leading to more severe brain injury than in mature rats [13]. Immunohistochemical studies of human infants suffering periventricular leucomalacia have demonstrated increased tumor necrosis factor-· and interleukin-6 in brain tissue [14]. There are a certain amount of data from studies in animals that suggest that inflammation and infection might directly cause damage in the developing brain. In a groundbreaking study in 1976, Gilles and his colleagues showed that injection of endotoxin into kittens produces white matter lesions similar to the telencephalic leucoencephalopathy they described in preterm human infants [15]. Experimental intrauterine infection with E. coli cause lesions in the cerebral white matter in fetal rabbits [14]. Nevertheless, it remains to be shown definitively that this effect is not due to systemic hypotension and changes in cerebral perfusion. Presently, the data to support a role for intrauterine infection in human cerebral damage is largely circumstantial and to some degree inconsistent. Clinical studies, particularly from Romero’s group, have related evidence of inflammation to cerebral injury. Interleukin-6 was found in high concentrations in the amniotic fluid and cord blood of infants who went on to develop ultrasound evidence of cerebral damage or neurodevelopmental impairment [16, 17]. The studies referenced here suffer from the use of ultrasonography or neurological examination as measures of brain injury. Ultrasonography has particular problems, being both a sensitive and specific test for haemorrhagic, but an insensitive although specific test of hypoxic-ischemic lesions [18]. Its value for detecting diffuse white matter damage has not been adequately assessed, but is generally believed to be poor. In addition to this problem, both ultrasound and neurological examination can only be applied weeks or months after delivery, making it impossible to distinguish intrauterine from postnatal cerebral damage. We have used magnetic resonance (MR) imaging to improve the definition of cerebral damage in preterm infants. We have installed within our neonatal intensive care unit a dedicated 1.0 T MR system with full

intensive care facilities, so that we are now able to obtain images of the brains of even the smallest and sickest infants very soon after delivery [19]. In a consecutive cohort of over 40 infants born in our hospital, we have found a strong association between chorioamnionitis and brain injury detected by MR imaging within the first 2 days after delivery. Pro-inflammatory cytokines were higher in umbilical cord blood of infants with white matter disease. Brain abnormalities were also associated with higher fractions of CD45RO positive T cells; these are memory cells which signify exposure to antigen at least 7–10 days earlier [20]. Other mechanisms might also be involved in white matter injury. Thyroid hormone deficiency might play a role in brain injury in preterm infants [21]. It has been suggested that treatment with magnesium sulphate reduces neurodevelopmental impairment by an unknown mechanism [22], although this proposal has been disputed [23]. The genetic background of the individual might be relevant in an analogous fashion to the role played by thrombophilic disorders in term infants suffering perinatal brain lesions. One particular complexity is the fact that the neural response to hypoxia-ischaemia often involves inflammatory pathways [24], while it is clear that inflammation can interrupt haemodynamic stability. The interrelation between hypoxia-ischaemia and inflammation may be profound; for example, it has recently been shown that hypoxia-ischaemia induces neural cells to express Fas, a classical receptor for lymphocyte-mediated killing, and that cross-linking of the receptor with antibodies is sufficient to cause neural apoptosis [25]. The picture is complex and placental inflammation might only be one piece in a subtle mechanism of damage and repair within the developing brain.

New Approaches to Brain Injury in Preterm Infants

Dev Neurosci 2002;24:352–354

353

References 1 Edwards AD, Wyatt JS, Richardson C, Delpy DT, Cope M, Reynolds EOR: Cotside measurement of cerebral blood flow in ill newborn infants by near infrared spectroscopy. Lancet 1988;2:770–771. 2 Greisen G, Pryds O: Low CBF, discontinuous EEG activity, and periventricular brain injury in ill, preterm neonates. Brain Dev 1989;11: 164–168. 3 Edwards AD, Wyatt JS, Richardson C, Potter A, Cope M, Delpy DT, Reynolds EOR: Effects of indomethacin on cerebral haemodynamics in very preterm infants. Lancet 1991;335: 1491–1495. 4 McDonnell M, Ives NK, Hope PL: Intravenous aminophylline and cerebral blood flow in preterm infants. Arch Dis Child 1992;67:416– 418. 5 Sinha SK, D’Souza SW, Rivlin E, Chiswick ML: Ischaemic brain lesions at birth in preterm infants: Clinical events and developmental outcome. Arch Dis Child 1990;65:1017–1020. 6 Maalouf E, Duggan PJ, Rutherford MA, Counsell S, Fletcher AM, Battin M, Cowan FM, Edwards AD: Magnetic resonance imaging of the brain in a cohort of extremely preterm infants. J Pediatr 1999;135:351–357. 7 Badawi N, Kurinczuk JJ, Keogh JM, Alessandri LM, O’Sullivan F, Burton PR, Pemberton PJ, Stanley FJ: Intrapartum risk factor for newborn encephalopathy: The Western Australian case-control study. BMJ 1998;317:1554–1558. 8 Yoon BH, Romero R, Jun JK, Park KH, Park JD, Ghezzi F, Kim BI: Amniotic fluid cytokines (interleukin-6, tumor necrosis factor-alpha, interleukin-1 beta, and interleukin-8), and the risk for the development of bronchopulmonary dysplasia. Am J Obstet Gynecol 1997b;177:825–830.

354

9 Ghezzi F, Gomez R, Romero R, Yoon BH, Edwin SS, David C, Janisse J, Mazor M: Elevated interleukin-8 concentrations in amniotic fluid of mothers whose neonates subsequently develop bronchopulmonary dysplasia. Eur J Obstet Gynecol Reprod Biol 1998;78:5–10. 10 Berry SM, Romero R, Gomez R, Puder KS, Ghezzi F, Cotton DB, Bianchi DW: Premature parturition is characterized by in utero activation of the fetal immune system. Am J Obstet Gynecol 1995;173:1315–1320. 11 Wenstrom KD, Andrews WW, Hauth JC, Goldenberg RL, DuBard MB, Cliver SP: Elevated second-trimester amniotic fluid interleukin-6 levels predict preterm delivery. Am J Obstet Gynecol 1999;178:546–550. 12 Holzman C, Jetton J, Fisher R, Mohan M, Paneth N: Association of maternal IgM concentrations above the median at 15–19 weeks of gestation and early preterm delivery. Lancet 1999;345:1095. 13 Anthony D, Dempster R, Fearn S, Clements J, Wells G, Perry VH, Walker K: CXC cytokines generate age-related increases in neutrophilmediated brain inflammation and blood-brain barrier breakdown. Curr Biol 1998;923–926. 14 Yoon BH, Romero R, Kim CJ, Koo JN, Choe G, Syn HC, Chi JG: High expression of tumor necrosis factor-alpha and interleukin-6 in periventricular leukomalacia. Am J Obstet Gynecol 1997c;177:406–411. 15 Gilles FH, Leviton A, Kerr CS: Endotoxin leucoencephalopathy in the telencephalon of the newborn kitten. J Neurol Sci 1976;27:183– 191. 16 Yoon BH, Romero R, Yang SH, Jun JK, Kim IO, Choi JH, Syn HC: Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am J Obstet Gynecol 1996;174:1433–1440.

Dev Neurosci 2002;24:352–354

17 Yoon BH, Jun JK, Romero R, Park KH, Gomez R, Choi JH, Kim IO: Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1beta, and tumor necrosis factor-alpha), neonatal brain white matter lesions, and cerebral palsy. Am J Obstet Gynecol 1997a;177: 19–26. 18 Hope PL, Gould SJ, Howard S, Hamilton PA, Costello AM, Reynolds EO: Precision of ultrasound diagnosis of pathologically verified lesions in the brains of very preterm infants. Dev Med Child Neurol 1988;30:457–471. 19 Battin M, Maalouf E, Counsell S, Herlihy AH, Edwards AD: Magnetic resonance imaging of the brain of premature infants within a neonatal intensive care unit. Lancet 1997;349:1741. 20 Duggan PJ, Maalouf E, Watts TL, Sullivan MHF, Counsell S, Allsop JM, Al-Nakib L, Rutherford M, Battin M, Roberts I, Edwards AD: Intrauterine T cell activation and increased pro-inflammatory cytokine concentrations in preterm infants with cerebral lesions. Lancet 2001;358:1699–1700. 21 Reuss ML, Paneth N, Pinto Martin JA, Lorenz JM, Susser M: The relation of transient hypothyroxinemia in preterm infants to neurologic development at two years of age. N Engl J Med 1996;334:821–827. 22 Nelson KB, Grether JK: Can magnesium sulfate reduce the risk of cerebral palsy in very low birthweight infants? Pediatrics 1995;95:263– 269. 23 Bennet P, Edwards AD: Use of magnesium sulphate in obstetrics. Lancet 1997;350:1491. 24 Rothwell NJ, Strijbos PJ: Cytokines in neurodegeneration and repair. Int J Dev Neurosci 1995;13:179–185. 25 Felderhoff-Muser U, Taylor DL, Greenwood K, Joashi U, Kozma M, Stibenz D, Edwards AD, Mehmet H: Fas/Apol/CD95 can function as a death receptor for neuronal cells in vitro and in vivo and is upregulated following cerebral hypoxic-ischemic injury to the developing rat brain. Brain Research 2000, in press.

Edwards

Review Dev Neurosci 2002;24:355–363 DOI: 10.1159/000069047

Received: September 16, 2002 Accepted: October 13, 2002

Under What Circumstances Can Seizures Produce Hippocampal Injury: Evidence for Age-Specific Effects Aristea S. Galanopoulou a, d Jorge Vidaurre a, d Solomon L. Moshé a–d Departments of a Neurology, b Neuroscience, and c Pediatrics, and d Montefiore/Einstein Epilepsy Management Center, Albert Einstein College of Medicine, Bronx, N.Y., USA

Key Words Epilepsy W Hippocampus W Neuroprotection W Febrile seizure W Adult W Immature W Rat W Mesial temporal sclerosis

Abstract Mesial temporal sclerosis (MTS) is the characteristic hippocampal pathology of temporal lobe epilepsy in adults. Both clinical and experimental studies indicate that although the immature brain is highly susceptible to seizures, it is more resistant to the development of the seizure-induced hippocampal pathology akin to MTS, compared with the adult brain. However, seizures in the immature brain may produce age-specific effects on hippocampal morphology or function. The spectrum of these effects is still unknown. Factors such as the presence of prior neurological abnormalities, age, etiology of the seizures, repetitive seizures and genetic predisposition may affect the range and severity of hippocampal changes. The key point is to identify the significance of these changes and design age-appropriate preventative treatments. Copyright © 2002 S. Karger AG, Basel

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0355$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

Introduction

Current experimental and clinical studies support the notion that the development of mesial temporal lobe epilepsy (MTLE) and mesial temporal sclerosis (MTS) is an age-specific phenomenon, triggered or accelerated by genetic or environmental factors and even by recurrent severe seizures. In this paper, we will briefly review the hippocampal anatomy, the principal features of MTLE and factors linked with the genesis of this epileptic syndrome and its characteristic pathology. We will present data indicating that seizures produce age-specific hippocampal changes, with the immature brain showing a relative resistance to seizure-induced hippocampal injury compared with adults.

Hippocampal Anatomy

The mesial temporal region consists of the amygdala and the temporal segment of the prepiriform cortex rostrally, and the hippocampal allocortex. The latter comprises the perirhinal cortex, the parahippocampal gyrus (entorhinal cortex, parasubiculum, presubiculum, and parahippocampal proisocortex) and the hippocampal formation [1]. The hippocampal formation includes the

Aristea S. Galanopoulou, MD, PhD Albert Einstein College of Medicine 1410 Pelham Parkway South, Kennedy Center Room 311 Bronx, NY 10461 (USA) Tel. +1 718 430 2620, Fax +1 718 430 8899, E-Mail [email protected]

Fig. 1. Anatomy of the mesial temporal lobe structures and seizure-induced hippocampal injury in adult rats. a Nisslstained coronal sections of an adult rat brain illustrate the mesial temporal lobe structures. The rostral cut (3.60 mm posterior to bregma) depicts the anterior dorsal hippocampus, amygdala, perirhinal and periamygdaloid prepiriform cortex. The caudal cut (6.10 mm posterior to bregma) depicts the hippocampus, which includes the subiculum, as well as the entorhinal cortex. b Lithium-pilocarpine-induced SE induces a characteristic pattern of neuronal loss (indicated with black arrows) in the anterior dorsal hippocampus of adult rats, compared with vehicle-injected controls. Neuronal loss is prominent at the CA1 area. The end folium is thinner and less densely populated in rats subjected to SE compared with controls. The CA2 sector, located between the red arrows, is relatively resistant to SE-induced injury. The dentate gyrus granular neurons appear intact.

subiculum, the pyramidal cell layer and the dentate gyrus. The pyramidal cell layer of the hippocampus is divided into 3 fields: CA1, CA2, and CA3 [2]. The term CA4 has been used to indicate the polymorphic layer of the dentate gyrus. The CA4 region and the dentate hilus comprise the end folium [3]. Figure 1 depicts the anatomy of the mesial temporal structures.

356

Dev Neurosci 2002;24:355–363

The Syndrome of MTLE

MTLE accounts for 70% of intractable epilepsy patients [4]. It presents with partial seizures, which usually appear in the mid or late childhood [5, 6]. Seizures may remit for a period of time and then become medically intractable [7]. Interictal abnormalities, such as depression, may occur [7]. The syndrome is associated with characteristic hippocampal pathology (MTS or hippocampal sclerosis). In a recent workshop (2002) organized

Galanopoulou/Vidaurre/Moshé

by the International League against Epilepsy, MTS was defined by the following features: (1) CA1 and dentate hilar neuronal cell loss and gliosis with relative sparing of transitional cortex, although with detailed counting all hippocampal regions may show neuronal loss and gliosis to various degrees [8–10]; (2) short- and long-term structural and functional glial changes [8, 10]; (3) evidence of structural and functional reorganization including but not limited to mossy fibers [8, 11]; (4) granular cell dispersion [11]; (5) extrahippocampal lesions or atrophy [8], and (6) dysplastic or structurally altered neurons may also be seen [11, 12]. More detailed studies have also revealed the presence of an increased number of proliferating neuronal precursors in the subgranular matrix of the dentate gyrus [8, 13]. Furthermore, MTS-associated functional changes in glutamate or GABA receptors or the expression of growth factors have been observed [8]. Figure 1b demonstrates the characteristic distribution of neuronal loss in a rat hippocampus with MTS. Magnetic resonance imaging (MRI) has been helpful in diagnosing the sclerotic hippocampus. The radiologic diagnostic criteria of MTS are hippocampal atrophy on T1-weighted images, increased hippocampal signal on T2weighted images, and loss of gray /white differentiation in the hippocampus [14–16]. The excellent seizure outcome following anterior temporal lobectomy of MTLE patients with ipsilateral MTS has emphasized the importance of MTS as the pathologic substrate of seizures in this syndrome [17–19]. However, it should be pointed out that the presence of MTS is not sufficient to cause TLE in the absence of other predisposing factors. MTS-like pathology may be an incidental finding in mental retardation, dementia, schizophrenia, alcoholism or other nonepileptic disorders [reviewed in ref. 20].

longed febrile seizures [5, 13, 22–24]. However, the occurrence of an initial precipitating injury may not always be followed by seizures, MTS or MTLE. The current thinking considers epileptogenesis as a multifactorial process, resulting from the interplay of a permissive genetic substrate, initial precipitating event(s) occurring at a vulnerable age, and may perhaps be further sustained by the occurrence of seizures. It is thus hypothesized that when the initial precipitating event occurs at a sensitive age in an individual who has a permissive genetic substrate, it will trigger functional or pathological changes, which will facilitate the development of the first provoked or unprovoked seizure. Those patients, who are genetically predisposed towards epilepsy, and who experience the first seizure at a sensitive age, will further develop the characteristic structural (i.e. MTS) and functional changes associated with the specific epileptic syndrome (i.e. MTLE).

When Does MTS Occur?

A volume of studies has dealt with the question of what leads to the development of MTLE [13, 21]. The occurrence of a first seizure, whether brief or prolonged, may or may not be followed by epilepsy. The presence of MTS is not always diagnostic or predictive of the development of MTLE. Most MTLE cases are not linked to a genetic locus. In retrospective studies, 50–90% of MTLE patients have experienced antecedent events, which may represent an ‘initial precipitating injury’, such as perinatal complications, central nervous system infection, ischemia or trauma, neonatal or infantile seizures, including pro-

Retrospective studies in humans support the notion that seizures early in life may produce MTS. Early-life convulsions, usually complex febrile seizures, are very common in MTLE patients [5, 13, 25, 26]. Febrile seizures longer than 90 min have been correlated with the development of MTLE and MTS [27], but such seizures are very uncommon [23]. Prospective studies do not provide any link between one seizure early in life and MTS. In a community-based study of children with epilepsy, febrile seizures were not a risk factor for MTLE [23]. Treatments that prevented recurrence of febrile seizures had no impact upon the development of epilepsy [28]. A clear causal association between febrile seizures and MTLE cannot yet be established. It is possible that the same factors predisposing to MTLE may also lower the threshold to febrile seizures [23]. Although MTS is the most common pathological finding in adult patients with intractable MTLE, it is less frequently encountered in children [reviewed in ref. 6]. The age differences in the prevalence of MTS in MTLE patients may reflect the relative resistance of the immature brain to hippocampal injury or the existence of agespecific syndromic variants of MTLE. The preexisting pathologic abnormalities, which are more common in childhood MTLE, may precipitate the development of epilepsy [6]. The progressive development of MTS after occasional brief unprovoked generalized tonic-clonic seizures, or status epilepticus (SE) has been documented

Age-Specific Consequences of Seizures

Dev Neurosci 2002;24:355–363

Epileptogenesis and Epilepsy

357

with serial MRIs or autopsy reports [27, 29–31]. It is unclear whether this reflects the effects of the precipitating factors, the parallel progression of epilepsy and hippocampal pathology, or a cause-effect relationship.

Studies in Animal Models

To determine possible causalities and seizure-induced age-specific hippocampal changes, animal models are needed. Normal rodents exposed to chemoconvulsants such as kainic acid [32, 33] or pilocarpine [34, 35], or electrical stimulation [36, 37] may develop seizures. These models show age-specific differences in the seizure threshold, and semiological and pathological features. In addition, studies in rodents with genetic or developmental abnormalities have increased our knowledge about the pathogenesis of MTS. Effects of Seizures on the Adult Hippocampus Acute administration of kainic acid or pilocarpine induces SE, defined as recurrent motor seizures for at least 30 consecutive minutes, without interictal return to baseline behavior. A latent period spanning several weeks without overt clinical seizures precedes the development of the chronic epileptic state [38]. The ensuing neuropathological changes resemble human MTS [34, 39]. Neuronal loss at variable degrees occurs as early as 2.5 h following the onset of SE [40]. Both glial cells and neurons undergo phenotypic changes, which may include altered expression and function of neurotransmitter receptors [41, 42]. Reactive gliosis appears a few days following SE and persists or increases over the following weeks [41, 43, 44]. The collateral connections of mossy fibers with the supragranular dentate neurons may form aberrant excitatory networks [45]. Seizures may induce the birth of new neurons [46], which are thought to contribute to the mossy fiber sprouting or the dispersion of the dentate gyrus [46, 47]. The absence of neoneurogenesis though does not prevent mossy fiber sprouting [48]. Extrahippocampal structures are often involved [39, 49]. The repetitive subconvulsive electrical stimulation of specific brain sites, i.e. kindling, will also eventually elicit more severe limbic seizures [50]. Repetitive severe kindled seizures result in CA4 injury, mossy fiber sprouting, and changes in the neurophysiologic properties of hippocampal neurons [50]. Adult rats subjected to kainic acid or lithium-pilocarpine induced SE, or severe kindled seizures showed forebrain hypermetabolism due to a large increase in the gly-

358

Dev Neurosci 2002;24:355–363

colytic rather than the oxidative pathway [50–52]. During the latent phase, the vulnerable forebrain regions involved in the generation of seizures showed an initial decrease followed by a later hypermetabolism. The end of the latent phase is characterized by hypermetabolism of the substantia nigra pars reticulata and superior colliculus, structures involved in seizure control. In the chronic phase, interictal hypometabolism occurs in the areas that sustained seizure-induced injury. Following SE, the rats may exhibit learning, memory, and behavioral abnormalities [53]. Effects of Seizures on Normal Developing Rats Young animals are more prone to seizures, but more resistant to seizure-induced cell loss compared with adults [35, 54, 55]. In the pilocarpine model, the adult pattern of behavioral and electrographic seizures, and MTS pathology was seen in pups older than 3 weeks [35]. Two-weekold pups had either no injury or occasional CA1 neuronal loss [56]. Similar results were obtained from the kainic acid-induced SE or with kindling [54, 57]. Metabolic studies performed in immature rats during kainic acid- or lithium-pilocarpine-induced SE, or severe kindled seizures showed increases in deoxyglucose accumulation [50, 51]. Dube et al. [51] used deoxyglucose studies to determine factors underlying age-dependent consequences of seizures in the pilocarpine model. They concluded that their results suggest that ‘the process of epileptogenesis and its functional consequences differ in PN21 and adult rats’. Another important finding of the study was that only 24% of PN21 rats developed spontaneous seizures. Indeed, spontaneous seizures occur very rarely in rats that experience SE before the 3rd week of life [51, 58]. Age-specific effects have also been reported for hypoxia-induced seizures. PN10 pups exposed to severe hypoxia (ambient oxygen concentration less than 3.5%) develop seizures acutely, have minimal or no pathological changes, and are rendered more susceptible to seizures in adulthood [59, 60]. Owens et al. [61] demonstrated that hypoxia-induced seizures and long-term effects on seizure susceptibility occur when ambient oxygen levels are less than 3.5%. These effects are not seen in any other age [59] or at higher oxygen levels (more than 5%), which do not induce seizures [61, 62]. The superimposition of seizures has recently been shown to exacerbate the hippocampal injury induced by hypoxia-ischemia [63, 64]. Developmental changes in the expression and function of inhibitory or excitatory neurotransmitter systems may also contribute in the age-specific vulnerability to seizureinduced hippocampal injury [50, 65–69]. Brooks-Kayal et

Galanopoulou/Vidaurre/Moshé

al. [70] reported that the subunit composition of GABAA receptors in immature rats is such that it allows for higher sensitivity to zinc and benzodiazepines compared with adults. GABAA receptors undergo a functional switch during development from depolarizing to hyperpolarizing, due to changes in the chloride gradient [71]. According to the ‘GluR2 hypothesis’ [72], insults that decrease GluR2 will cause neurotoxicity because they render ·-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptors more permeable to calcium. In contrast to adults, immature rat hippocampal neurons sustain GluR2 levels following SE and therefore survive [73]. Other factors may however modify the requirement for GluR2 [74]. The lack of mossy fiber sprouting in immature rats has been attributed to an immature pattern of mossy fiber innervation [75] or to low expression of GAP43 protein [76]. However, mossy fiber sprouting has been observed in pups of less than 2 weeks old, following the creation of hippocampal lesions [77–80], suggesting that sprouting can occur at this age and the lack of sprouting after SE or kindling is an age-specific effect. Furthermore, although increased glycolysis is observed in both adult and immature rats during severe seizures [50], the immature rats may be resistant to the detrimental effects of lactate, unlike the situation observed in adults. Indeed, lactate can be an alternative nutrient in the immature brain, especially during hypoglycemia [81].

The Resistance of the Immature Hippocampus Is Not Absolute

In occasional models, such as daily induction of generalized seizures between PN0–PN15 or administration of corticotropin-releasing hormone may cause hippocampal injury and mossy fiber sprouting [82, 83]. Factors such as underlying neurological abnormalities, seizure etiology, genetic predisposition and age may facilitate seizureinduced hippocampal injury. Early-life seizures may also induce age-specific long-term effects. These include functional changes like transcriptional regulation of hyperpolarization-activated cyclic nucleotide-gated channels in the hippocampus [84], altered hippocampal neoneurogenesis [85], and decrease in dendritic spine density [86], which lower the seizure threshold [84]. In addition, longterm developmental, cognitive or behavioral abnormalities may occur in a model- and age-specific pattern [53, 87, 88]. However, the latter changes are less severe than those observed in older animals. Finally, it should be pointed out, that amygdala kindling, once induced in PN15 rats,

Age-Specific Consequences of Seizures

persists through life even though to date no histological hippocampal changes have been documented [57].

Under What Circumstances Can Seizure-Induced Hippocampal Changes Occur?

Prior Neurologic Abnormality Children with a preexisting neurological abnormality and febrile seizures have higher risk for epilepsy [89]. Altered neurogenesis or abnormal hippocampal architecture, whether genetic or acquired, has been proposed as a factor predisposing to early-onset MTLE associated with MTS [90]. Rats with perinatal ischemia or cortical dysplasias induced by fetal exposure to methylaxozymethanol are more prone to seizures [91–93]. Nevertheless, when exposed to seizures during the first 2 weeks, the extent of injury is still much smaller than that observed in adults with or without prior history of brain injury [94]. The Cause of the Seizure May Produce Brain Injury per se Infections may trigger fever and a spectrum of other host responses, via increase in endogenous pyrogens, such as interleukin-1ß (IL-1ß) [95]. Release of Il-1ß has proconvulsant effects in various experimental models [96]. Increased levels of IL-1ß occur in the cerebrospinal fluid of children with encephalitis but not with febrile seizures [97]. However, a high IL-1ß producer allele has been associated with febrile seizures and MTLE [98, 99]. These support the thesis that febrile seizures do not normally lead to the development of MTLE. However, when the insult is severe or under a permissive genetic substrate, increased IL-1ß production may ensue, promoting the development of epilepsy. Genetic Predisposition Several families with autosomal dominant familial temporal lobe epilepsy (FTLE) have been reported [100– 103]. FTLE may be associated with febrile seizures [104– 108] or MTS [108, 109]. FTLE associated with febrile seizures has been linked to sodium channel SCN1A mutations [104]. MTLE with MTS has been associated with certain HLA class II antigens [110] and a high IL-1ß producer allele [99], at least in certain populations [111]. The transgenic mice research has implicated genes like neuropeptide Y [112], cytokines [113], or the neuronal specific potassium chloride cotransporter KCC2 [114] in seizure susceptibility, although it is not clear how these factors contribute to MTS and MTLE.

Dev Neurosci 2002;24:355–363

359

Increasing Age With increasing age, the seizure-induced hippocampal damage is more readily observed. The occurrence of typical MTS-like changes occurs around adolescence [35, 54, 56]. It is tempting to suggest that it is not the initial seizure (occurring in infancy) that induces MTS, but the seizures that occur later on, after the end of the first decade of life. Indeed, there are experimental data [115, 116] suggesting that a second hit (often a seizure) may augment the extent of age-specific injury.

Are All Seizure-Induced Changes Detrimental to the Brain? Another Point of View!

Reports that certain types of seizures ‘protect’ against subsequent convulsant insults are challenging the existing notion that all seizures are detrimental to the brain. Kindling of the dorsal hippocampus may decrease SEinduced injury [117]. A series of maximal electroconvulsive shocks may delay the appearance of limbic seizures [118]. Rats with posttraumatic kindled seizures performed better in the Morris water maze compared with controls [119].

Treatment Strategies in Epilepsy as a Function of Age

In both humans and rats, therefore, there is a spectrum of age-related seizure-induced hippocampal effects. In the younger ages, short- and long-term functional changes (not necessarily accompanied by neuronal loss or gliosis) may occur in response to prolonged seizures, and are not necessarily detrimental. A second hit, however, may pro-

mote epileptogenesis and MTS-like changes, especially in a susceptible host. A better understanding of the epileptic process in humans as a function of age, sex, and etiology is warranted in order to design individualized therapeutic modalities for epilepsy. Our current armamentarium includes drugs, which act upon neurotransmitter receptors or channels to reduce excitation or enhance inhibition. Knowledge of the age-specific function and expression of these receptors and channels will direct us towards an ageappropriate use of these medications or of novel approaches, such as the vaccination against the NR1 subunit of the NMDA receptors [113, 120, 121]. Elucidation of the age-specific role of subcortical structures in seizure control will improve the therapeutic use of deep brain electrical stimulation in epilepsy [122]. Gonadal hormones have recently been shown to reduce seizureinduced hippocampal injury [123, 124], but this effect depends upon the hormonal milieu [125]. More studies are needed to determine whether and in what form hormonotherapy may be beneficial during development. Immunomodulatory agents may be proven beneficial in the selected population of children with prolonged febrile seizures, early-life meningitis, or excessive release of cytokines [113, 120, 121]. Enrichment of the environment may also protect by reducing neuronal apoptosis [126]. Future approaches should aim not only to stop ongoing seizures but also exert neuroprotective effects.

Acknowledgements This study was supported in part by grant NS-20253 from NINDS (S.L.M.), a grant from the Epilepsy Foundation of America (A.S.G.) and the Heffer Family Foundation. S.L.M. is a Martin A. and Emily L. Fisher Fellow in Neurology and Neuroscience.

References 1 Gloor P: The hippocampal system; in Gloor P (ed): The Temporal Lobe and Limbic System. New York, Oxford University Press, 1997, pp 325–590. 2 Lorente de No R: Studies on the structure of the cerebral cortex. 2. Continuation of the study of the ammonic system. J Psychol Neurol 1934;46:113–177. 3 Margerison JH, Corsellis JAN: Epilepsy in the temporal lobes. Brain 1966;89:499–530. 4 Babb TL, Brown WJ: Pathological findings in epilepsy; in Engel JJ (ed): Surgical Treatment of the Epilepsies. New York, Raven, 1987, pp 511–540.

360

5 French JA, Williamson PD, Thadani VM, Darcey TM, Mattson RH, Spencer SS, Spencer DD: Characteristics of medial temporal lobe epilepsy. 1. Results of history and physical examination. Ann Neurol 1993;34:774–780. 6 Bourgeois BF: Temporal lobe epilepsy in infants and children. Brain Dev 1998;20:135– 141. 7 Engel J Jr: Introduction to temporal lobe epilepsy. Epilepsy Res 1996;26:141–150. 8 Mathern GW, Babb TL, Armstrong DL: Hippocampal sclerosis; in Engel JJ, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Philadelphia, Lippincott-Raven, 1997, pp 133–155.

Dev Neurosci 2002;24:355–363

9 Sommer W: Erkrankung des Ammonshorns als aetiologischer Moment der Epilepsie. Arch Psychiatr Nervenkr 1880;10:631–675. 10 Bratz E: Ammonshornbefunde bei Epileptikern. Arch Psychiatr Nervenkr 1899;32:820– 835. 11 Thom M, Sisodiya SM, Beckett A, Martinian L, Lin WR, Harkness W, Mitchell TN, Craig J, Duncan J, Scaravilli F: Cytoarchitectural abnormalities in hippocampal sclerosis. J Neuropathol Exp Neurol 2002;61:510–519.

Galanopoulou/Vidaurre/Moshé

12 Blumcke I, Zuschratter W, Schewe JC, Suter B, Lie AA, Riederer BM, Meyer B, Schramm J, Elger CE, Wiestler OD: Cellular pathology of hilar neurons in Ammon’s horn sclerosis. J Comp Neurol 1999;414:437–453. 13 Blumcke I, Thom M, Wiestler OD: Ammon’s horn sclerosis: A maldevelopmental disorder associated with temporal lobe epilepsy. Brain Pathol 2002;12:199–211. 14 Van Paesschen W, Connelly A, King MD, Jackson GD, Duncan JS: The spectrum of hippocampal sclerosis: A quantitative magnetic resonance imaging study. Ann Neurol 1997;41:41– 51. 15 Grattan-Smith JD, Harvey AS, Desmond PM, Chow CW: Hippocampal sclerosis in children with intractable temporal lobe epilepsy: Detection with MR imaging. AJR Am J Roentgenol 1993;161:1045–1048. 16 Cascino GD, Jack CR Jr, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, Marsh WR, O’Brien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy: Pathological correlations. Ann Neurol 1991; 30:31–36. 17 Spencer S: Selection of candidates for temporal lobe resection; in Wyllie E (ed): The Treatment of Epilepsy: Principles and Practice. Baltimore, Williams & Wilkins, 1996, pp 1036–1049. 18 Arruda F, Cendes F, Andermann F, Dubeau F, Villemure JG, Jones-Gotman M, Poulin N, Arnold DL, Olivier A: Mesial atrophy and outcome after amygdalohippocampectomy or temporal lobe removal. Ann Neurol 1996;40: 446–450. 19 Engel JJ: Surgery for seizures. N Engl J Med 1996;334:647–652. 20 Benbadis SR, Wallace J, Reed Murtagh F: MRI evidence of mesial temporal sclerosis in subjects without seizures. Seizure 2002;11:340– 343. 21 Mathern GW, Babb TL, Leite JP, Pretorius K, Yeoman KM, Kuhlman PA: The pathogenic and progressive features of chronic human hippocampal epilepsy. Epilepsy Res 1996;26:151– 161. 22 Mathern GW, Pretorius JK, Babb TL: Influence of the type of initial precipitating injury and at what age it occurs on course and outcome in patients with temporal lobe seizures. J Neurosurg 1995;82:220–227. 23 Berg AT, Shinnar S, Levy SR, Testa FM: Childhood-onset epilepsy with and without preceding febrile seizures. Neurology 1999;53:1742– 1748. 24 Sztriha L, Gururaj AK, Bener A, Nork M: Temporal lobe epilepsy in children: Etiology in a cohort with new-onset seizures. Epilepsia 2002;43:75–80. 25 Falconer MA: Mesial temporal (Ammon’s horn) sclerosis as a common cause of epilepsy. Aetiology, treatment, and prevention. Lancet 1974;ii:767–770.

Age-Specific Consequences of Seizures

26 Cendes F, Andermann F, Dubeau F, Gloor P, Evans A, Jones-Gotman M, Olivier A, Andermann E, Robitaille Y, Lopes-Cendes I, et al: Early childhood prolonged febrile convulsions, atrophy and sclerosis of mesial structures, and temporal lobe epilepsy: An MRI volumetric study. Neurology 1993;43:1083–1087. 27 VanLandingham KE, Heinz ER, Cavazos JE, Lewis DV: Magnetic resonance imaging evidence of hippocampal injury after prolonged focal febrile convulsions. Ann Neurol 1998;43: 413–426. 28 Rosman NP, Colton T, Labazzo J, Gilbert PL, Gardella NB, Kaye EM, Van Bennekom C, Winter MR: A controlled trial of diazepam administered during febrile illnesses to prevent recurrence of febrile seizures. N Engl J Med 1993;329:79–84. 29 Briellmann RS, Newton MR, Wellard RM, Jackson GD: Hippocampal sclerosis following brief generalized seizures in adulthood. Neurology 2001;57:315–317. 30 Jackson GD, Chambers BR, Berkovic SF: Hippocampal sclerosis: Development in adult life. Dev Neurosci 1999;21:207–214. 31 Nohria V, Lee N, Tien RD, Heinz ER, Smith JS, DeLong GR, Skeen MB, Resnick TJ, Crain B, Lewis DV: Magnetic resonance imaging evidence of hippocampal sclerosis in progression: A case report. Epilepsia 1994;35:1332–1336. 32 Schwarcz R, Zaczek R, Coyle JT: Microinjection of kainic acid into the rat hippocampus. Eur J Pharmacol 1978;50:209–220. 33 Nadler JV, Perry BW, Cotman CW: Intraventricular kainic acid preferentially destroys hippocampal pyramidal cells. Nature 1978;271: 676–677. 34 Turski WA, Czuczwar SJ, Kleinrok Z, Turski L: Cholinomimetics produce seizures and brain damage in rats. Experientia 1983;39: 1408–1411. 35 Cavalheiro EA, Silva DF, Turski WA, Calderazzo-Filho LS, Bortolotto ZA, Turski L: The susceptibility of rats to pilocarpine-induced seizures is age dependent. Brain Res 1987;465: 43–58. 36 Burnham WM: Primary and ‘transfer’ seizure development in the kindled rat. Can J Neurol Sci 1975;2:417–428. 37 Moshe SL, Albala BJ, Ackermann RF, Engel JJ: Increased seizure susceptibility of the immature brain. Brain Res 1983;283:81–85. 38 Cavalheiro EA, Riche DA, Le Gal La Salle G: Long-term effects of intrahippocampal kainic acid injection in rats: A method for inducing spontaneous recurrent seizures. Electroencephalogr Clin Neurophysiol 1982;53:581– 589. 39 Ben-Ari Y, Tremblay E, Ottersen OP, Meldrum BS: The role of epileptic activity in hippocampal and ‘remote’ cerebral lesions induced by kainic acid. Brain Res 1980;191:79– 97. 40 Covolan L, Mello LE: Temporal profile of neuronal injury following pilocarpine or kainic acid-induced status epilepticus. Epilepsy Res 2000;39:133–152.

41 Niquet J, Ben-Ari Y, Represa A: Glial reaction after seizure induced hippocampal lesion: Immunohistochemical characterization of proliferating glial cells. J Neurocytol 1994;23:641– 656. 42 Schwarzer C, Tsunashima K, Wanzenbock C, Fuchs K, Sieghart W, Sperk G: GABAA receptor subunits in the rat hippocampus. 2. Altered distribution in kainic acid-induced temporal lobe epilepsy. Neuroscience 1997;80:1001– 1017. 43 Roch C, Leroy C, Nehlig A, Namer IJ: Magnetic resonance imaging in the study of the lithium-pilocarpine model of temporal lobe epilepsy in adult rats. Epilepsia 2002;43:325–335. 44 Khurgel M, Ivy GO: Astrocytes in kindling: Relevance to epileptogenesis. Epilepsy Res 1996;26:163–175. 45 Tauck DL, Nadler JV: Evidence of functional mossy fiber sprouting in hippocampal formation of kainic acid-treated rats. J Neurosci 1985;5:1016–1022. 46 Parent JM, Yu TW, Leibowitz RT, Geschwind DH, Sloviter RS, Lowenstein DH: Dentate granule cell neurogenesis is increased by seizures and contributes to aberrant network reorganization in the adult rat hippocampus. J Neurosci 1997;17:3727–3738. 47 Scharfman HE, Goodman JH, Sollas AL: Granule-like neurons at the hilar/CA3 border after status epilepticus and their synchrony with area CA3 pyramidal cells: Functional implications of seizure-induced neurogenesis. J Neurosci 2000;20:6144–6158. 48 Parent JM, Tada E, Fike JR, Lowenstein DH: Inhibition of dentate granule cell neurogenesis with brain irradiation does not prevent seizureinduced mossy fiber synaptic reorganization in the rat. J Neurosci 1999;19:4508–4519. 49 Turski WA, Cavalheiro EA, Bortolotto ZA, Mello LM, Schwarz M, Turski L: Seizures produced by pilocarpine in mice: A behavioral, electroencephalographic and morphological analysis. Brain Res 1984;321:237–253. 50 Sperber EF, Stanton PK, Haas K, Ackermann RF, Moshe SL: Developmental differences in the neurobiology of epileptic brain damage. Epilepsy Res Suppl 1992;9:67–80. 51 Dube C, da Silva Fernandes MJ, Nehlig A: Age-dependent consequences of seizures and the development of temporal lobe epilepsy in the rat. Dev Neurosci 2001;23:219–223. 52 Ackermann RF, Moshe SL, Albala BJ: Restriction of enhanced [2–14C]deoxyglucose utilization to rhinencephalic structures in immature amygdala-kindled rats. Exp Neurol 1989;104: 73–81. 53 Stafstrom CE: Assessing the behavioral and cognitive effects of seizures on the developing brain. Prog Brain Res 2002;135:377–390. 54 Albala BJ, Moshe SL, Okada R: Kainic-acidinduced seizures: A developmental study. Brain Res 1984;315:139–148. 55 Nitecka L, Tremblay E, Charton G, Bouillot JP, Berger ML, Ben-Ari Y: Maturation of kainic acid seizure-brain damage syndrome in the rat. 2. Histopathological sequelae. Neuroscience 1984;13:1073–1094.

Dev Neurosci 2002;24:355–363

361

56 Sankar R, Shin DH, Liu H, Mazarati A, Pereira de Vasconcelos A, Wasterlain CG: Patterns of status epilepticus-induced neuronal injury during development and long-term consequences. J Neurosci 1998;18:8382–8393. 57 Haas KZ, Sperber EF, Opanashuk LA, Stanton PK, Moshe SL: Resistance of immature hippocampus to morphologic and physiologic alterations following status epilepticus or kindling. Hippocampus 2001;11:615–625. 58 Sankar R, Shin D, Mazarati AM, Liu H, Katsumori H, Lezama R, Wasterlain CG: Epileptogenesis after status epilepticus reflects age- and model-dependent plasticity. Ann Neurol 2000; 48:580–589. 59 Jensen FE, Holmes GL, Lombroso CT, Blume HK, Firkusny IR: Age-dependent changes in long-term seizure susceptibility and behavior after hypoxia in rats. Epilepsia 1992;33:971– 980. 60 Chiba S: Long-term effect of postnatal hypoxia on the seizure susceptibility in rats. Life Sci 1985;37:1597–1604. 61 Owens J Jr, Robbins CA, Wenzel HJ, Schwartzkroin PA: Acute and chronic effects of hypoxia on the developing hippocampus. Ann Neurol 1997;41:187–199. 62 Moshe SL, Albala BJ: Perinatal hypoxia and subsequent development of seizures. Physiol Behav 1985;35:819–823. 63 Wirrell EC, Armstrong EA, Osman LD, Yager JY: Prolonged seizures exacerbate perinatal hypoxic-ischemic brain damage. Pediatr Res 2001;50:445–454. 64 Dzhala V, Ben-Ari Y, Khazipov R: Seizures accelerate anoxia-induced neuronal death in the neonatal rat hippocampus. Ann Neurol 2000;48:632–640. 65 Brooks-Kayal AR, Pritchett DB: Developmental changes in human Á-aminobutyric acidA receptor subunit composition. Ann Neurol 1993;34:687–693. 66 Raol YH, Lynch DR, Brooks-Kayal AR: Role of excitatory amino acids in developmental epilepsies. Ment Retard Dev Disabil Res Rev 2001;7:254–260. 67 Marks JD, Friedman JE, Haddad GG: Vulnerability of CA1 neurons to glutamate is developmentally regulated. Brain Res Dev Brain Res 1996;97:194–206. 68 Miller LP, Johnson AE, Gelhard RE, Insel TR: The ontogeny of excitatory amino acid receptors in the rat forebrain. 2. Kainic acid receptors. Neuroscience 1990;35:45–51. 69 Tremblay E, Roisin MP, Represa A, CharriautMarlangue C, Ben-Ari Y: Transient increased density of NMDA binding sites in the developing rat hippocampus. Brain Res 1988;461:393– 396. 70 Brooks-Kayal AR, Shumate MD, Jin H, Rikhter TY, Kelly ME, Coulter DA: Á-Aminobutyric acidA receptor subunit expression predicts functional changes in hippocampal dentate granule cells during postnatal development. J Neurochem 2001;77:1266–1278. 71 Ben-Ari Y: Excitatory actions of GABA during development: The nature of the nurture. Nat Rev Neurosci 2002;3:728–739.

362

72 Bennett MV, Pellegrini-Giampietro DE, Gorter JA, Aronica E, Connor JA, Zukin RS: The GluR2 hypothesis: Ca2+-permeable AMPA receptors in delayed neurodegeneration. Cold Spring Harb Symp Quant Biol 1996;61:373– 384. 73 Friedman LK, Sperber EF, Moshe SL, Bennett MV, Zukin RS: Developmental regulation of glutamate and GABAA receptor gene expression in rat hippocampus following kainateinduced status epilepticus. Dev Neurosci 1997; 19:529–542. 74 Sanchez RM, Koh S, Rio C, Wang C, Lamperti ED, Sharma D, Corfas G, Jensen FE: Decreased glutamate receptor 2 expression and enhanced epileptogenesis in immature rat hippocampus after perinatal hypoxia-induced seizures. J Neurosci 2001;21:8154–8163. 75 Ribak CE, Navetta MS: An immature mossy fiber innervation of hilar neurons may explain their resistance to kainate-induced cell death in 15-day-old rats. Brain Res Dev Brain Res 1994; 79:47–62. 76 Holmes GL, McCabe B: Brain development and generation of brain pathologies. Int Rev Neurobiol 2001;45:17–41. 77 Zimmer J: Changes in the Timm sulfide silver staining pattern of the rat hippocampus and fascia dentata following early postnatal deafferentation. Brain Res 1973;64:313–326. 78 Zimmer J: Long-term synaptic reorganization in rat fascia dentata deafferented at adolescent and adult stages: Observations with the Timm method. Brain Res 1974;76:336–342. 79 Laurberg S, Zimmer J: Lesion-induced sprouting of hippocampal mossy fiber collaterals to the fascia dentata in developing and adult rats. J Comp Neurol 1981;200:433–459. 80 Laurberg S, Zimmer J: Lesion-induced rerouting of hippocompal mossy fibers in developing but not in adult rats. J Comp Neurol 1980;190: 627–650. 81 Vannucci RC, Vannucci SJ: Glucose metabolism in the developing brain. Semin Perinatol 2000;24:107–115. 82 Holmes GL, Sarkisian M, Ben-Ari Y, Chevassus-Au-Louis N: Mossy fiber sprouting after recurrent seizures during early development in rats. J Comp Neurol 1999;404:537–553. 83 Baram TZ, Ribak CE: Peptide-induced infant status epilepticus causes neuronal death and synaptic reorganization. Neuroreport 1995;6: 277–280. 84 Brewster A, Bender RA, Chen Y, Dube C, Eghbal-Ahmadi M, Baram TZ: Developmental febrile seizures modulate hippocampal gene expression of hyperpolarization-activated channels in an isoform- and cell-specific manner. J Neurosci 2002;22:4591–4599. 85 McCabe BK, Silveira DC, Cilio MR, Cha BH, Liu X, Sogawa Y, Holmes GL: Reduced neurogenesis after neonatal seizures. J Neurosci 2001;21:2094–2103. 86 Swann JW, Smith KL, Lee CL: Neuronal activity and the establishment of normal and epileptic circuits during brain development. Int Rev Neurobiol 2001;45:89–118.

Dev Neurosci 2002;24:355–363

87 Lee CL, Hannay J, Hrachovy R, Rashid S, Antalffy B, Swann JW: Spatial learning deficits without hippocampal neuronal loss in a model of early-onset epilepsy. Neuroscience 2001;107:71–84. 88 Lynch M, Sayin U, Bownds J, Janumpalli S, Sutula T: Long-term consequences of early postnatal seizures on hippocampal learning and plasticity. Eur J Neurosci 2000;12:2252– 2264. 89 Shinnar S, Berg AT, Moshe SL, Petix M, Maytal J, Kang H, Goldensohn ES, Hauser WA: Risk of seizure recurrence following a first unprovoked seizure in childhood: A prospective study. Pediatrics 1990;85:1076– 1085. 90 Sarkisian MR, Holmes GL, Carmant L, Liu Z, Yang Y, Stafstrom CE: Effects of hyperthermia and continuous hippocampal stimulation on the immature and adult brain. Brain Dev 1999;21:318–325. 91 Olson JE, Horne DS, Holtzman D, Miller M: Hyperthermia-induced seizures in rat pups with preexisting ischemic brain injury. Epilepsia 1985;26:360–364. 92 Germano IM, Zhang YF, Sperber EF, Moshe SL: Neuronal migration disorders increase susceptibility to hyperthermia-induced seizures in developing rats. Epilepsia 1996;37: 902–910. 93 Baraban SC, Schwartzkroin PA: Flurothyl seizure susceptibility in rats following prenatal methylazoxymethanol treatment. Epilepsy Res 1996;23:189–194. 94 Germano IM, Sperber EF, Ahuja S, Moshe SL: Evidence of enhanced kindling and hippocampal neuronal injury in immature rats with neuronal migration disorders. Epilepsia 1998;39:1253–1260. 95 Dinarello CA, Cannon JG, Wolff SM: New concepts on the pathogenesis of fever. Rev Infect Dis 1988;10:168–189. 96 Vezzani A, Conti M, De Luigi A, Ravizza T, Moneta D, Marchesi F, De Simoni MG: Interleukin-1ß immunoreactivity and microglia are enhanced in the rat hippocampus by focal kainate application: Functional evidence for enhancement of electrographic seizures. J Neurosci 1999;19:5054–5065. 97 Lahat E, Livne M, Barr J, Katz Y: Interleukin-1ß levels in serum and cerebrospinal fluid of children with febrile seizures. Pediatr Neurol 1997;17:34–36. 98 Virta M, Hurme M, Helminen M: Increased frequency of interleukin-1ß (-511) allele 2 in febrile seizures. Pediatr Neurol 2002;26:192– 195. 99 Kanemoto K, Kawasaki J, Miyamoto T, Obayashi H, Nishimura M: Interleukin (IL)1ß, IL-1·, and IL-1 receptor antagonist gene polymorphisms in patients with temporal lobe epilepsy. Ann Neurol 2000;47:571– 574. 100 Berkovic SF, McIntosh A, Howell RA, Mitchell A, Sheffield LJ, Hopper JL: Familial temporal lobe epilepsy: A common disorder identified in twins. Ann Neurol 1996;40:227– 235.

Galanopoulou/Vidaurre/Moshé

101 Cendes F, Lopes-Cendes I, Andermann E, Andermann F: Familial temporal lobe epilepsy: A clinically heterogeneous syndrome. Neurology 1998;50:554–557. 102 Gambardella A, Messina D, Le Piane E, Oliveri RL, Annesi G, Zappia M, Andermann E, Quattrone A, Aguglia U: Familial temporal lobe epilepsy autosomal dominant inheritance in a large pedigree from southern Italy. Epilepsy Res 2000;38:127–132. 103 Ikeda A, Kunieda T, Miyamoto S, Fukuyama H, Shibasaki H: Autosomal dominant temporal lobe epilepsy in a Japanese family. J Neurol Sci 2000;176:162–165. 104 Abou-Khalil B, Ge Q, Desai R, Ryther R, Bazyk A, Bailey R, Haines JL, Sutcliffe JS, George AL Jr: Partial and generalized epilepsy with febrile seizures plus and a novel SCN1A mutation. Neurology 2001;57:2265– 2272. 105 Baulac S, Picard F, Herman A, Feingold J, Genin E, Hirsch E, Prud’homme JF, Baulac M, Brice A, LeGuern E: Evidence for digenic inheritance in a family with both febrile convulsions and temporal lobe epilepsy implicating chromosomes 18qter and 1q25–q31. Ann Neurol 2001;49:786–792. 106 Depondt C, Van Paesschen W, Matthijs G, Legius E, Martens K, Demaerel P, Wilms G: Familial temporal lobe epilepsy with febrile seizures. Neurology 2002;58:1429–1433. 107 Ward N, Evanson J,Cockerell OC: Idiopathic familial temporal lobe epilepsy with febrile convulsions. Seizure 2002;11:16–19. 108 Briellmann RS, Torn-Broers Y, Jackson GD, Berkovic SF: Seizures in family members of patients with hippocampal sclerosis. Neurology 2001;57:1800–1804. 109 Kobayashi E, Lopes-Cendes I, Guerreiro CA, Sousa SC, Guerreiro MM, Cendes F: Seizure outcome and hippocampal atrophy in familial mesial temporal lobe epilepsy. Neurology 2001;56:166–172.

Age-Specific Consequences of Seizures

110 Ozkara C, Altintas A, Yilmaz E, Eskazan E, Erkol G, Ozyurt E, Erdogan E, Kuday C: An association between mesial temporal lobe epilepsy with hippocampal sclerosis and human leukocyte antigens. Epilepsia 2002;43:236– 239. 111 Buono RJ, Ferraro TN, O’Connor MJ, Sperling MR, Ryan SG, Scattergood T, Mulholland N, Gilmore J, Lohoff FW, Berrettini WH: Lack of association between an interleukin 1ß(IL-1ß) gene variation and refractory temporal lobe epilepsy. Epilepsia 2001;42: 782–784. 112 Vezzani A, Michalkiewicz M, Michalkiewicz T, Moneta D, Ravizza T, Richichi C, Aliprandi M, Mule F, Pirona L, Gobbi M, Schwarzer C, Sperk G: Seizure susceptibility and epileptogenesis are decreased in transgenic rats overexpressing neuropeptide Y. Neuroscience 2002;110:237–243. 113 Vezzani A, Moneta D, Conti M, Richichi C, Ravizza T, De Luigi A, De Simoni MG, Sperk G, Andell-Jonsson S, Lundkvist J, Iverfeldt K, Bartfai T: Powerful anticonvulsant action of IL-1 receptor antagonist on intracerebral injection and astrocytic overexpression in mice. Proc Natl Acad Sci USA 2000;97: 11534–11539. 114 Woo NS, Lu J, England R, McClellan R, Dufour S, Mount DB, Deutch AY, Lovinger DM, Delpire E: Hyperexcitability and epilepsy associated with disruption of the mouse neuronal-specific K-Cl cotransporter gene. Hippocampus 2002;12:258–268. 115 Koh S, Storey TW, Santos TC, Mian AY, Cole AJ: Early-life seizures in rats increase susceptibility to seizure-induced brain injury in adulthood. Neurology 1999;53:915–921. 116 Liu Z, Mikati M, Holmes GL: Mesial temporal sclerosis: Pathogenesis and significance. Pediatr Neurol 1995;12:5–16. 117 Kelly ME, McIntyre DC: Hippocampal kindling protects several structures from the neuronal damage resulting from kainic acidinduced status epilepticus. Brain Res 1994; 634:245–256.

118 Andre V, Ferrandon A, Marescaux C, Nehlig A: Electroshocks delay seizures and subsequent epileptogenesis but do not prevent neuronal damage in the lithium-pilocarpine model of epilepsy. Epilepsy Res 2000;42:7–22. 119 Hamm RJ, Pike BR, Temple MD, O’Dell DM, Lyeth BG: The effect of postinjury kindled seizures on cognitive performance of traumatically brain-injured rats. Exp Neurol 1995;136:143–148. 120 During MJ, Symes CW, Lawlor PA, Lin J, Dunning J, Fitzsimons HL, Poulsen D, Leone P, Xu R, Dicker BL, Lipski J, Young D: An oral vaccine against NMDAR1 with efficacy in experimental stroke and epilepsy. Science 2000;287:1453–1460. 121 Etzioni A, Jaffe M, Pollack S, Zelnik N, Benderly A, Tal Y: High dose intravenous Á-globulin in intractable epilepsy of childhood. Eur J Pediatr 1991;150:681–683. 122 Velisek L,Veliskova J, Moshe SL: Electrical stimulation of substantia nigra pars reticulata is anticonvulsant in adult and young male rats. Exp Neurol 2002;173:145–152. 123 Veliskova J, Velisek L, Galanopoulou AS, Sperber EF: Neuroprotective effects of estrogens on hippocampal cells in adult female rats after status epilepticus. Epilepsia 2000;41 (suppl 6):S30–S35. 124 Reibel S, Andre V, Chassagnon S, Andre G, Marescaux C, Nehlig A, Depaulis A: Neuroprotective effects of chronic estradiol benzoate treatment on hippocampal cell loss induced by status epilepticus in the female rat. Neurosci Lett 2000;281:79–82. 125 Galanopoulou AS, Veliskova J, Moshe SL: Gender-specific effects of estradiol administration on hippocampal injury following lithium-pilocarpine-induced status epilepticus. Epilepsia 2000;41:46. 126 Young D, Lawlor PA, Leone P, Dragunow M, During MJ: Environmental enrichment inhibits spontaneous apoptosis, prevents seizures and is neuroprotective. Nat Med 1999; 5:448–453.

Dev Neurosci 2002;24:355–363

363

Perspectives Dev Neurosci 2002;24:364–366 DOI: 10.1159/000069040

Animal Models of Developmental Brain Injury: Relevance to Human Disease A Summary of the Panel Discussion from the Third Hershey Conference on Developmental Cerebral Blood Flow and Metabolism

Henrik Hagberg (Göteborg) Rebecca Ichord (Philadelphia, Pa.) Charles Palmer (Hershey, Pa.) Jerome Y. Yager (Saskatoon) Susan J. Vannucci (New York, N.Y.)

A major issue that arose from the Second Hershey Conference in June 2000, which has been echoed at several meetings since, relates to specific characteristics of the various animal models currently employed in the study of injury to the immature brain, their relation to each other, and their relevance to the human infant. Aspects of this ongoing debate were addressed throughout the conference in a variety of presentations. Thus, the final session was designed to serve as an open forum to discuss, summarize and synthesize the range of information. The format of the panel discussion was based on 5 general questions, each of which had previously been posed to the panel members as well as the participants. Panel members were requested to formulate their responses to these questions and to present 1–2 slides in not more than 5 min. Meeting participants who wished to make a comment were asked to do so with 1 slide in less than 5 min. The questions covered 5 general topics: (1) Cerebral maturation. What is the developmental stage of the species at birth, i.e. how mature is the brain at birth and how does this correlate with a term human infant? With a premature infant? What is the rate of

Susan Vannucci was the moderator and Henrik Hagberg, Rebecca Ichord, Charles Palmer and Jerome Yager were the discussants.

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0364$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

development/cerebral maturation, i.e. which ages correspond most closely to the neonate/infant/child/adolescent? (2) Cerebral blood flow. What are the species-specific characteristics of cerebral blood flow that would impact on an experimental ischemia model? (3) Cerebral metabolism. What are the important cerebral metabolic considerations at different stages of development, i.e. prevailing substrate availability/utilization; the level of oxidative metabolism/energy utilization, and neurotransmitter status? (4) White matter injury. What is the timing and extent of myelination? What are the pros/cons for using any given species in a study of white matter injury? (5) Different paradigms. Compare/contrast different experimental models of injury, i.e. permanent vs. transient, focal vs. global ischemia; trauma, epilepsy, hemorrhage, and hypoglycemia. Finally, is it really important that any given model truly mimics a human disease or is it more important to define general mechanisms of injury and neuroprotection? Many current models of developmental brain injury, and especially hypoxic-ischemic encephalopathy, utilize rats during the early postnatal period, from postnatal day 7 (P7) to P14, as being roughly comparable with the human infant at term. Henrik Hagberg of the University

of Göteborg, Sweden [1], was the first to address the topic of cerebral maturation and specifically the validity of this assumption by presenting a summary slide comparing various aspects of cerebral maturation in the rat at birth, 7 days of postnatal age, 21 days and adult with the human being at 20 weeks of gestation, birth, 1 year of age, and adult. His general comment was that it is very difficult to compare the maturational age of the CNS between different species, and especially given the very different anatomy of the rat brain, which has a lissencephalic smooth surface, compared with the human brain, which is generally gyrencephalic. Rather, a consideration of the timing of different indices of development, such as neuronal proliferation, glial proliferation and growth, myelination, synaptogenesis, and neurotransmitter receptor activities, would affect the impact of specific insults at different ages. Thus, a comparison of results obtained at different ages could then generate a developmental spectrum of specific vulnerability and outcome. Rebecca Ichord of the University of Pennsylvania presented the classic slide from Dobbing and Sands [2] as a graphic depiction of periods of cerebral development as a function of gestation and postnatal age across a wide variety of species. Of these species, she pointed out that the piglet, another commonly used animal species for developmental studies, is very similar to the human infant at birth relative to its degree of brain maturation. However, each animal species has something to offer as an experimental model, based not solely on its cerebral maturation, but also on its size and the ability to perform several physiological manipulations. Investigators need to appreciate the limitations, as well as the strengths, of any species. When the discussion was opened to the audience, Brian Richardson of the University of Western Ontario, Canada, presented a curve describing perinatal brain development in terms of a velocity growth curve to make the point that brain development can be characterized in several different contexts, depending on the questions being asked. Thus, brain development can be characterized in terms of neurophysiological and electrophysiological development, or in terms of metabolic development. Such functional assessments have anatomic associations as well. Robert C. Vannucci of the Hershey Medical Center/Pennsylvania State University College of Medicine discussed the neuroanatomic rationale of choosing the P7 rat as the immature animal for the development of the model of unilateral cerebral hypoxia-ischemia. In collaboration with the British neuropathologist, James Brierly, they noted that at P7, the germinal matrix was small and cortical layering was complete in all 6 layers, making the brain of the P7 rat pup

most similar to the 34- to 35-week gestation human infant, with the P10 rat approximating the human infant at term, as discussed in the original description of this model [3]. Donna Ferriero of the University of California San Francisco, Calif., USA, made the very relevant point about the comparison of immature rats and mice. In extending the studies previously done in the immature rat model of unilateral cerebral hypoxia-ischemia to the mouse to utilize the number of transgenic lines, the tendency has been to equate the postnatal development of these two rodent species, which may not be valid. Clearly, a careful description of the mouse cerebral development is essential to meaningful interpretation of the genetically altered strains. Although rodents, and especially mice, offer tremendous opportunities to address questions of gene and proteins involved in developmental brain injury, larger animal species such as pigs and sheep offer distinct advantages, especially in terms of physiological monitoring. R.C. Vannucci discussed some of the important physiological studies that have been performed in the fetal sheep because they are easily monitored. The piglet, another model offering comprehensive physiological monitoring, is quite advanced at birth, perhaps even more developed than the human infant; for example, newborn piglets can walk, vocalize, and function as members of a social group. Dean Kurth of the University of Pennsylvania indicated that the newborn piglet has higher rates of cerebral blood flow (CBF) and metabolism than the term human infant. Although histologically the newborn piglet brain closely resembles that of the term human infant, physiologically it is more mature. Furthermore, the maturation and rate of growth are more rapid, in that a 2-week-old piglet is comparable with a young child; beyond that, they quickly reach maturity and are able to reproduce. Marvin Cornblath raised a very important issue regarding whole body metabolism, which again differs from the human situation. Newborn piglets have only 1% body fat and are essentially incapable of mobilizing body fat, but this changes dramatically with postnatal age. A 2- to 4-day-old piglet has approximately 4% body fat, and demonstrates rapid changes in rates of glucose metabolism. These unique characteristics of whole body substrate utilization and metabolism must be accounted for in the design of metabolic studies, such as the effects of hypoglycemia. The expansive discussion of the first topic of speciesspecific cerebral maturation thus included aspects of the ensuing questions, and occupied the majority of the time allotted on the final day of the conference. However, in the remaining time, several specific comments were made

Animal Models of Developmental Brain Injury

Dev Neurosci 2002;24:364–366

365

in direct response to questions 2–5. On the issue of CBF, it is clear that the relationship of the flow-metabolism couple is present in all species, at all ages. However, rates of normal cerebral metabolism differ markedly across species and it is therefore invalid to measure CBF in absolute terms and compare those numbers across species. The valid measurement is the rate of CBF in the immature animal relative to the adult of that species. Richard Traystman of the Johns Hopkins School of Medicine clarified an additional problem arising from the exclusive use of Laser Doppler techniques to study CBF in that all measurements determined this way are only relative to baseline, and do not provide a valid comparison among animals. Stephen Back of the University of Oregon Health Sciences Center presented data on a new methodology to determine CBF using fluorescent microspheres, specifically designed to measure the heterogeneity of blood flow, especially in very low flow areas such as cerebral white matter. Continued sophistication of MRI techniques to measure CBF will also help in addressing heterogeneity of flow and changes with both normal and pathologic development. Although species-specific differences in white matter were briefly addressed during the initial discussion, the panel returned to this topic and to the actual relevance of models of white matter injury to mimic human injury. There was guarded agreement that pure white matter injury, i.e. cystic lesions only in white matter, may not be very relevant to the human pathology and current efforts towards developing such a model in animals may not be valuable. Jerry Yager of the University of Saskatchewan gave his opinion from the perspective of the pediatric neurologist, i.e. what is seen clinically is the entire spectrum of damage from global cell death to specific white

366

Dev Neurosci 2002;24:364–366

matter injury. His suggestion was that perhaps a more fruitful approach to our experimental designs would be to be less restrictive in our animal models with respect to the endpoints we are seeking and actually accept, and learn from, the spectrum of damage seen in our animal models as well. In summary, there was good agreement that no single model can ever replicate the human condition. However, that does not preclude valuable information and insights being derived from animal experimentation. In attempting to address the prevalent phenotype we see following brain injury in infants and children, it becomes increasingly important to match imaging and functional outcome studies, with more static assessments of mechanisms of cell death. Neonatal brain injury is a diffuse process with a range of developmental outcomes. Understanding the characteristics, strengths and limitations of each of the relevant animal models will enable future studies to address mechanistic issues in smaller species, i.e. rodents, and apply the information obtained to larger animals. From these studies on animal models, researchers hope to better understand, and hopefully treat the human condition arising from developmental brain injury.

References

1 Hagberg H, Bona E, et al: Hypoxia-ischemia model in the 7-day-old rat: Possibilities and shortcomings. Acta Pediatr Suppl 1997;422: 85–88. 2 Dobbing J, Sands J: Comparative aspects of the brain growth spurt. Early Hum Dev 1979;3: 79–83. 3 Rice JE, Vannucci RC, Bierley JB: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9: 131–141.

Hagberg/Ichord/Palmer/Yager/Vannucci

Original Paper Dev Neurosci 2002;24:367–381 DOI: 10.1159/000069049

Received: August 26, 2002 Accepted: October 24, 2002

Prolonged Neonatal Seizures Exacerbate Hypoxic-Ischemic Brain Damage: Correlation with Cerebral Energy Metabolism and Excitatory Amino Acid Release Jerome Y. Yager a Edward A. Armstrong a Hero Miyashita a Elaine C. Wirrell b Departments of Pediatrics, a University of Saskatchewan, Saskatoon and b University of Calgary, Calgary, Canada

Key Words Neonatal seizures W Hypoxic-ischemic brain damage W Cerebral energy metabolism W Excitatory amino acid release

Abstract Background: Perinatal hypoxia-ischemia (HI) is the most common precipitant of seizures in the first 24–48 h of a newborn’s life. In a previous study, our laboratory developed a model of prolonged, continuous electrographic seizures in 10-day-old rat pups using kainic acid (KA) as a proconvulsant. Groups of animals included those receiving only KA, or HI for 15 or 30 min, followed by KA infusion. Our results showed that prolonged electrographic seizures following 30 min of HI resulted in a marked exacerbation of brain damage. We have undertaken studies to determine alterations in hippocampal highenergy phosphate reserves and the extracellular release of hippocampal amino acids in an attempt to ascertain the underlying mechanisms responsible for the damage promoted by the combination of HI and KA seizures. Methods: All studies were performed on 10-day-old rats. Five groups were identified: (1) group I – KA alone, (2) group II – 15 min of HI plus KA, (3) group III – 15 min of HI alone, (4) group IV – 30 min of HI plus KA, and

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0367$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

(5) group VI – 30 min of HI alone. HI was induced by right common carotid artery ligation and exposure to 8% oxygen/balance nitrogen. Glycolytic intermediates and highenergy phosphates were measured. Prior to treatment, at the end of HI (both 15 and 30 min), prior to KA injection, and at 1 (onset of seizures), 3, 5 (end of seizures), 7, 24 and 48 h, blood samples were taken for glucose, lactate and ß-hydroxybutyrate. At the same time points, animals were sacrificed by decapitation and brains were rapidly frozen for subsequent dissection of the hippocampus and measurement of glucose, lactate, ß-hydroxybutyrate, adenosine triphosphate (ATP) and phosphocreatine (PCr). In separate groups of rats as defined above, microdialysis probes (CMA) were stereotactically implanted into the CA2–3 region of the ipsilateral hippocampus for measurement of extracellular amino acid release. Dialysate was collected prior to any treatment, at the end of HI (15 and 30 min), prior to KA injection, and at 1 (onset of seizures), 3, 5 (end of seizures), 7 and 9 h. Determination of glutamate, serine, glutamine, glycine, taurine, alanine, and GABA was accomplished using high-performance liquid chromatography with EC detection. Results: Blood and hippocampal glucose concentrations in all groups receiving KA were significantly lower than control during seizures (p ! 0.05). ß-Hydroxybutyrate values displayed the inverse, in that values were sig-

Jerome Y. Yager, MD Room 3717, Royal University Hospital 103 Hospital Drive Saskatoon, Saskatchewan S7N 0W8 (Canada) Tel. +1 306 966 815, Fax +1 306 975 3767, E-Mail [email protected]

nificantly higher (p ! 0.01) in all KA groups compared with pretreatment controls during seizure activity. Values returned to control by 2 h following the cessation of seizures. Lactate concentrations in brain and blood mimicked those of ß-hydroxybutyrate. ATP values declined to 0.36 mmol/l in both the 15 and 30 min hypoxia groups compared with 1.85 mmol/l for controls (p ! 0.01). During seizures, ATP and PCr values declined significantly below their homologous controls. Following seizures, ATP values only for those animals receiving KA plus HI for 30 min remained below their homologous controls for at least 24 h. Determination of amino acid release revealed elevations of glutamate, glycine, taurine, alanine and GABA above pretreatment control during HI, with a return to normal prior to KA injections. During seizures and for the 4 h of recovery monitored, only glutamate in the combined HI and KA group rose significantly above both the 15 min of HI plus KA and the KA alone group (p ! 0.05). Conclusion: Under circumstances in which there is a protracted depletion of high-energy phosphate reserves, as occurs with a combination of HI- and KAinduced seizures, excess amounts of glutamate become toxic to the brain. The latter may account for the exacerbation of damage to the newborn hippocampus, and serve as a target for future therapeutic intervention. Copyright © 2002 S. Karger AG, Basel

Introduction

Seizures remain the most common neurologic symptom in the newborn, often heralding significant underlying neuropathology [Sheth et al., 1999; Volpe, 2002]. The most common cause of seizures, particularly in the term infant, is a hypoxic-ischemic insult, accounting for 640% of all cases. When combined with other forms of perinatal cerebral ischemic events such as bland and hemorrhagic infarction, cerebrovascular disease accounts for 50–60% of all seizures in the term and preterm infant [Holmes, 1997; Mizrahi and Kellaway, 1998; Sheth et al., 1999]. Because of the difficulty in recognizing and diagnosing seizures in this age group [Mizrahi and Kellaway, 1987], their exact incidence is uncertain. The National Collaborative Perinatal Project [Holden et al., 1982], a large population-based study, estimated the incidence to be 0.5% among all newborns. In premature and term infants admitted to the neonatal intensive care unit, Sheth et al. [1999] found the incidence to rise dramatically to 12– 14%.

368

Dev Neurosci 2002;24:367–381

Though neonates have the greatest predisposition to seizures compared with any other age group [Wasterlain and Shirasaka, 1994; Volpe, 2002; Moshe, 1993], controversy continues as to whether seizures per se cause cell death in the immature brain [Camfield, 1997; Wasterlain, 1997; Wasterlain and Shirasaka, 1994; Holmes and BenAri, 2001]. To a large extent, this may be due to the variability in animal models and species utilized in studies of neonatal seizures, and in particular the age at which research groups have defined the term immature or newborn [Holmes, 1997; Wasterlain and Shirasaka, 1994; Albala et al., 1984; Hirsch et al., 1992; Nitecka et al., 1984]. While damage has been induced in the 10-day-old rabbit [Franck and Schwartzkroin, 1984], the latter is a precocial animal whose development is advanced to that of the human [Dobbing and Sands, 1979]. In the rat, a nonprecocial animal, 7–10 days of age has been likened to the term human [Hagberg et al., 1997; Dobbing and Sands, 1979]. In this regard, seizures modeled on otherwise healthy brains have not been shown to cause neuronal death in rats !15 days of age [Wasterlain and Shirasaka, 1994; Holmes, 1997; Wasterlain, 1997; Albala et al., 1984], though clearly alterations in synaptogenesis, mossy fiber sprouting and epileptogenesis do occur [Holmes and Ben-Ari, 2001; Holmes et al., 1998; Stafrom et al., 1992]. Recently, our laboratory has established a model which more closely resembles the clinical situation in which neonatal seizures occur following a hypoxic-ischemic insult in the term infant [Wirrell et al., 2001]. Ten-day-old rats (equivalent to a 36- to 40-week term infant) underwent unilateral common carotid artery ligation and exposure to 8% oxygen for 30 min [Rice et al., 1981; Yager et al., 1993, 1996; Wirrell et al., 2001], producing a minimal lesion with cell death occurring within the hippocampus and neocortex. Prolonged and continuous seizures, induced by the systemic injection of kainic acid (KA), superimposed on this experimental paradigm dramatically increased neuronal necrosis in a topographically specific manner within the hippocampus and particularly in subfield CA3. To our knowledge, this was the first report of seizures in the immature newborn brain exacerbating hypoxic-ischemic damage. In the present study, we determined the effect of prolonged neonatal seizures following a hypoxic-ischemic injury on cerebral energy reserves and glycolytic intermediates. We then correlated these alterations with the release of excitatory amino acids, in an effort to determine the underlying mechanisms whereby seizures exacerbate hypoxic-ischemic brain damage, yet do not cause damage to the otherwise ‘healthy’ neonatal brain.

Yager/Armstrong/Miyashita/Wirrell

Methods Female Wistar rats (Charles River, Montreal, Canada) were bred in our laboratory. The rat pups were reared with their dams in a temperature-controlled environment at 21 B 0.5 ° C, with a relative humidity of 30%, and a 12-hour on/off lighting schedule. The date of experimentation was determined to be postnatal day 10 (P10; date of birth: day 1). This study received ethical approval from the Animal Care Committee at the University of Saskatchewan, Canada. Surgical Methods Cerebral hypoxia-ischemia (HI) was induced in 10-day-old rat pups using the modified Levine preparation described by Rice et al. [1981]. Rat pups were lightly anesthetized with halothane (4% induction; 1% maintenance), and placed in the supine position. The right common carotid artery was separated from contiguous structures and permanently ligated through a midline neck incision measuring no more than 1 cm. The incision was then sutured, and a PE10 flexible tubing was inserted subcutaneously between the scapula of each rat pup, and held in place with acrocyanate adhesive. Those rats which required cerebral microdialysis for the measurement of extracellular fluid amino acid analysis were subsequently transferred to a neonatal stereotactic unit (Kopf), still under light anesthesia. Following the production of a 1-cm midline-sagittal incision of the scalp, a burr hole was made with a 25-gauge needle using bregma as a reference (AP –2.9 mm, lateral –2.8 mm). Using these coordinates, a CMA/7 (CMA microdialysis AB, Stockholm, Sweden) guide and probe was lowered to a depth of 3.3 mm (from the calvarium) into the CA3 region of the right hippocampus (ipsilateral to the common carotid artery ligation). The guide was then fixed in place using a Fuji II light-cured glass ionomer (GC Corporation, Tokyo, Japan). The rat pup was then gently swaddled, and placed in an incubator thermocontrolled to nesting temperature (34 B 0.2 ° C). The entire surgical procedure lasted no longer than 20 min, after which the animals were allowed to recover. Induction of HI Each animal was allowed to recover for 2 h, after which HI was induced by placing individual animals in 500-ml glass jars through which a gas mixture of 8% oxygen/balance nitrogen was delivered via inlet and outlet portals. The pups were exposed to hypoxia for either 15 or 30 min. They were allowed to recover for an additional 30 min prior to the induction of seizures. Thermoregulation was maintained by placing each animal within the glass jars, in a single-walled neonatal incubator (Air Shields), thermocontrolled to 34 B 0.2 ° C (nesting temperature). Induction of Seizures Seizures were induced by providing a single subcutaneous injection of 3 mg/kg KA (Ocean Produce International) followed by a continuous subcutaneous infusion of 2 mg/kg/h for 3 h. Previous studies [Wirrell et al., 2001] have shown that the induction of seizures in this fashion consistently produces prolonged electrographic and clinical seizures lasting a mean duration of 4 h and 42 min. Seizure onset in all animals occurred at approximately 30 min after injection. Quantitation of Excitatory Amino Acids Modified Ringer’s solution (NaCl 145 mM, KCl 2.7 mM, CaCl2 1.2 mM, MgCl2 1.2 mM, pH 7.4) was continuously pumped through the microdialysis probe at a rate of 1.0 Ìl/min. Samples of the per-

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

fusate were collected on ice for 20 min/sample at the following intervals: (1) immediately before hypoxia (pretreatment control), (2) at the end of hypoxia, (3) prior to KA injection, (4) at 1, 3, 5 h (terminus of seizures), and at 7, and 9 h (2 and 4 h of recovery) following KA injection. After the last time point, the animals were sacrificed to ensure appropriate probe placement, and the microdialysis probe was placed in a standard solution of amino acids. All amino acid values were corrected for percent probe recovery. All amino acid samples were immediately transferred to –70 ° C for storage prior to analysis. Amino acids were measured by highperformance liquid chromatography by the method of Donazanti and Yamamoto [1998], as previously described [Owen et al., 1997; Shuaib et al., 1994; Ravindran et al., 1994; Kanthan et al., 1995; Kanthan et al., 1996]. Precolumn (Novapak C18; 3.9 ! 150 mm, Waters, Toronto, Canada) derivation of amino acids was performed with o-phthaldialdehyde/2-mercaptoethanol prior to electrochemical detection with the 715 Wisp sample processor (Waters). A fully automated system for the derivation procedure was used to reduce the inherent instability of amino acids in the o-phthaldialdehyde mixture and the variability of their half-lives. The Ultra Wisp dispenses the agent into the microdialysis perfusate and mixes it thoroughly in a 200-Ìl injection loop. The injection is made precisely 2.0 min after the reaction time. The data analysis and calculations of amino acid concentrations were made using the baseline 810 chromatography workstation data-processing system (Waters). The mobile phase comprised of 0.10 M disodium hydrogen orthophosphate, 0.14 M ethylene diaminetetra-acetic acid disodium salt and 28% methanol. External samples of 0.625, 1.25, 2.5, 25, 100, and 300 pmol/10 Ìl were used for analyzing all samples. Quantitation of Glycolytic Intermediates and Cerebral High-Energy Phosphates Glycolytic intermediates (glucose, lactate, ß-hydroxybutyrate) and cerebral high-energy phosphates [adenosine triphosphate (ATP), phosphocreatine (PCr)] were measured in hippocampal brain samples extracted from the right cerebral hemisphere ipsilateral to the common carotid artery, according to methods previously described [Yager and Asselin, 1996; Yager et al., 1992b]. Briefly, at time points which coincided with (1) immediately prior to HI (pretreatment controls), (2) at the terminus of HI, (3) following 30 min of recovery – immediately prior to KA injection, (4) at 1, 3, and 5 h (terminus of seizures), and at (5) 7, 24 and 48 h following KA injection (recovery), rat pups were rapidly decapitated and their brains immediately frozen in liquid nitrogen [Lowry et al., 1964]. Blood samples for glucose, lactate, and ß-hydroxybutyrate were collected from the severed neck vessels at the same time points in a glass capillary tube. Glucose and lactate were immediately analyzed using glucose and lactate analyzers (YSI). Twenty microliters of blood were diluted 1/10 in 0.5% perchloric acid and frozen at –70 ° C for later determination of ß-hydroxybutyrate. All brains were maintained at –70 ° C until tissue extraction. Portions of the ipsilateral hippocampus were dissected at –20 ° C and powdered under liquid nitrogen. 10–20 mg of tissue were then extracted with perchloric acid and neutralized with bicarbonate. Concentrations of glucose, lactate, ß-hydroxybutyrate, ATP and PCr were determined fluorometrically, using the appropriate enzymes, as previously described [Yager et al., 1992b, 1996; Lowry et al., 1964]. Study Groups Five groups of animals were assessed: group I consisted of animals receiving KA alone without hypoxia and served as the experi-

Dev Neurosci 2002;24:367–381

369

Fig. 1. Depiction of experimental paradigm. Groups are outlined in the text.

mental controls. Group II consisted of rat pups that received 15 min of HI plus KA injection. Group III were those animals that received only 15 min of HI without KA injection. Group IV rat pups received 30 min of HI and KA injection, and group V received only 30 min of HI. Those pups that did not receive KA injections were infused with normal saline in the same fashion, and of equal volume to those that received subcutaneous injections of KA (150 Ìl followed by an infusion of 100 Ìl/h for 3 h). Groups of rat pups were analyzed prior to any treatment and acted as controls. Experimental Paradigm Figure 1 displays the organizational timetable for the experimental paradigms in either of the experiments. Pretreatment control values were obtained for both of the experiments after a 2-hour recovery period following surgery. In those groups receiving HI, recovery occurred for 1/2 h prior to the injection of KA. In all groups receiving KA, seizures began both clinically and electrographically [Wirrell et al., 2001] approximately 30 min following the administration of KA. Statistical Analysis Comparisons of glycolytic intermediates and high-energy phosphate reserves were done using one-way ANOVA with Dunnett’s correction. Between-group comparisons were done using two-tailed Student t tests. Excitatory amino acid measurements within groups were done using repeated-measures ANOVA with Dunnett’s correction. For between-group comparisons, we utilized one-way ANOVA with Tukey’s multiple comparison post test. Significance was determined to be p ! 0.05.

370

Dev Neurosci 2002;24:367–381

Results

There was no mortality in groups III and V, which received HI alone. In those groups that received KAinduced seizures, mortality ranged from 14 to 25%, though no significant differences existed between the groups receiving KA. Glycolytic Intermediates Blood and brain concentrations of glycolytic intermediates are shown in tables 1 and 2, respectively. Blood glucose concentrations in all groups remained normal to the point of seizure onset, even during the brief periods of HI in groups II–V. In groups III and V, which only received 15 and 30 min of HI without the induction of seizures, glucose values dipped slightly below control values at the 5-hour mark of the experimental paradigm, likely as a result of hypoglycemic fasting for that length of time. In groups I, II, and IV, all of which experienced prolonged seizures, glucose concentrations fell significantly below pretreatment control values to the terminus of the clinical seizures, after which values rose to control. In group IV, glucose concentrations remained significantly below control values for at least 2 h of recovery, returning to normal by 24 h after seizure onset. Brain concentrations of glucose followed a similar pattern to that seen in blood. Comparison of the blood/brain concentrations between groups II and IV during seizures shows that the ratios are lower in the latter group, and that this is due to higher concentra-

Yager/Armstrong/Miyashita/Wirrell

Table 1. Blood concentrations of glycolytic intermediates

KA

15 min HI + KA 15 min HI

30 min HI + KA 30 min HI

5.95B0.26

5.90B0.18 5.79B0.12 6.03B0.15

5.90B0.18 5.79B0.12 6.03B0.15

5.90B0.18 5.39B0.30 6.31B0.11

5.90B0.18 5.39B0.30 6.31B0.11

Seizures 1h 3h 5h

5.64B0.40 3.92B0.40a 4.15B0.38a

3.59B0.62a, b, c 4.02B0.51c 4.73B0.11a

5.70B0.12 5.44B0.27 4.85B0.33a

6.61B0.66d 4.12B0.31a 4.78B0.27a

5.64B0.22 5.17B0.19 4.64B0.26a

Recovery 7h 24 h 48 h

6.28B0.21 5.65B0.19 5.57B0.14

6.08B0.31 5.79B0.22 5.71B0.15

6.26B0.22 6.04B0.11 5.87B0.25

3.13B0.17a, b, c, d 5.69B0.40 6.15B0.19 5.90B0.09 5.55B0.23 5.98B0.28

0.76B0.03

0.68B0.06 11.27B0.84a 1.30B0.37

0.68B0.06 11.27B0.84a 1.30B0.37

Glucose Pretreatment control HI Pre-KA injection

Lactate Pretreatment control HI Pre-KA injection

0.68B0.06 13.19B0.74a 2.30B0.42a

0.68B0.06 13.19B0.74a 2.30B0.42a

Seizures 1h 3h 5h

5.06B0.52a 1.69B0.34a 1.50B0.08

4.35B0.33a, c 2.55B0.13a, b, c 1.99B0.14a, b, c

0.73B0.05 0.76B0.03 8.80B0.08

3.83B0.51a, c 1.64B0.18c, d 1.48B0.17b, c, d

0.74B0.06 0.86B0.07 0.96B0.11

Recovery 7h 24 h 48 h

1.04B0.09 0.98B0.07 0.82B0.04

1.30B0.07b, c 0.88B0.05 0.79B0.05

0.81B0.05 0.87B0.07 0.78B0.05

1.02B0.06d 1.04B0.07 0.88B0.06

0.91B0.09 0.99B0.05 0.83B0.05

0.39B0.03

0.52B0.10 0.58B0.03 0.28B0.04

0.52B0.10 0.58B0.03 0.28B0.04

0.52B0.10 0.84B0.07d 0.46B0.19

0.52B0.10 0.84B0.07 0.46B0.19

Seizures 1h 3h 5h

0.32B0.02 2.17B0.22a 1.92B0.15a

0.51B0.05 2.06B0.23a, c 1.60B0.21a, c

0.59B0.12 0.48B0.06 0.54B0.04

0.44B0.03 1.97B0.32a, c 1.36B0.19a, c

0.57B0.05 0.61B0.07 0.61B0.08

Recovery 7h 24 h 48 h

1.11B0.12a 0.34B0.03 0.32B0.03

1.54B0.27a, c 0.43B0.04 0.36B0.02

0.43B0.08 0.43B0.04 0.36B0.02

0.73B0.13d 0.25B0.04d 0.30B0.05

0.51B0.09 0.42B0.06 0.32B0.03

ß-Hydroxybutyrate Pretreatment control HI Pre-KA injection

Values are depicted as the mean of 6–8 animals per time B SEM in millimoles per liter. a p ! 0.05 compared with homologous pretreatment control; b p ! 0.05 compared with animals receiving KA alone (group I); c p ! 0.05 compared with homologous HI group, and d p ! 0.05 compared with 15 min HI plus KA (group II).

tions of glucose in brain. This finding persists through recovery. Lactate concentrations in blood and brain rose dramatically during HI in groups II–V. Interestingly, whereas concentrations were higher in those animals that received

30 min of HI compared with 15 min, during the seizures, blood lactate values in group II (15 min of HI and KA) remained above pretreatment control values, and significantly above values in group IV (30 min of HI and KA). Lactate in the KA alone group saw a modest, though sig-

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

Dev Neurosci 2002;24:367–381

371

Table 2. Brain (hippocampal) concentrations of glycolytic intermediates

KA

15 min HI + KA 15 min HI

30 min HI + KA 30 min HI

2.01B0.13

1.90B0.15 0.45B0.10a 1.52B0.27

1.90B0.15 0.45B0.10a 1.52B0.27

1.90B0.15 0.50B0.05a 1.40B0.17

1.90B0.15 0.50B0.05a 1.40B0.17

Seizures 1h 3h 5h

1.92B0.17 0.84B0.13a 1.29B0.17a

1.10B0.18a, b 0.86B0.17a, c 1.52B0.11

1.76B0.08 1.71B0.21 1.66B0.13

2.35B0.19 1.03B0.10a, c 1.72B0.04

1.86B0.13 1.85B0.14 1.98B0.26

Recovery 7h 24 h 48 h

1.42B0.06a 1.59B0.10 1.59B0.09

1.74B0.08 2.08B0.09 1.65B0.17

1.89B0.08 1.88B0.11 1.91B0.17

1.89B0.28 2.04B0.09 2.12B0.16

1.95B0.12 1.97B0.14 2.27B0.10

1.66B0.06

1.77B0.16 10.48B0.38a 1.30B0.17

1.77B0.16 10.48B0.38a 1.30B0.17

1.77B0.16 11.46B0.23a 3.91B0.58a

1.77B0.16 11.46B0.23a 3.91B0.58a

Glucose Pretreatment control HI Pre-KA injection

Lactate Pretreatment control HI Pre-KA injection Seizures 1h 3h 5h

4.10B0.27a 2.58B0.13a 2.13B0.11

4.01B0.34a, c 3.16B0.27a, c 3.03B0.26a, b, c

1.32B0.11 1.54B0.12 1.30B0.13

3.70B0.40a, c 3.38B0.27a, c 2.81B0.30c

2.04B0.43 1.98B0.28 1.63B0.13

Recovery 7h 24 h 48 h

1.22B0.11 1.13B0.06 1.60B0.21

1.21B0.04 1.12B0.08 1.18B0.06

1.10B0.09 1.32B0.17 1.21B0.10

1.68B0.10b, c 1.54B0.16 1.33B0.11

1.11B0.09 1.73B0.20 1.61B0.16

0.09B0.02

0.15B0.03 0.24B0.02 0.19B0.02

0.15B0.03 0.24B0.02 0.19B0.02

0.15B0.03 0.58B0.04a 0.12B0.01

0.15B0.03 0.58B0.04a 0.12B0.01

Seizures 1h 3h 5h

0.13B0.01a 0.72B0.10a 0.60B0.04a

0.14B0.03 0.68B0.11a, c 0.54B0.10a, c

0.15B0.02 0.17B0.04 0.17B0.02

0.17B0.03 0.67B0.07a, c 0.60B0.14a, c

0.12B0.02 0.15B0.02 0.16B0.02

Recovery 7h 24 h 48 h

0.14B0.01 0.09B0.02 0.08B0.01

0.26B0.05 0.12B0.02 0.11B0.02

0.15B0.02 0.14B0.03 0.12B0.03

0.21B0.03 0.16B0.02 0.12B0.02

0.14B0.02 0.15B0.03 0.12B0.02

ß-Hydroxybutyrate Pretreatment control HI Pre-KA injection

Values are depicted as the mean of 6–8 animals per time B SEM millimoles per liter. a p ! 0.05 compared with homologous pretreatment control; b p ! 0.05 compared with animals receiving KA alone (group I); c p ! 0.05 compared with homologous HI group, and d p ! 0.05 compared with 15 min HI plus KA (group II).

nificant, increase during seizures, followed by a slow return to normal values by 24 h of recovery. Once again, brain concentrations of lactate followed those seen in blood. Comparisons of blood/brain lactate concentrations indicate again that those pups in group IV (30 min of HI-

372

Dev Neurosci 2002;24:367–381

and KA-induced seizures) have lower ratios during seizures in comparison with group II and I, largely on the basis of relatively higher brain concentrations of the substrate compared with the latter 2 groups.

Yager/Armstrong/Miyashita/Wirrell

Blood ß-hydroxybutyrate concentrations remained unchanged from normal in those pups receiving only HI. Brain concentrations of ß-hydroxybutyrate rose significantly above control during HI in group V, and rapidly returned to normal. All groups experiencing KA-induced seizures witnessed a significant increase in both blood and brain ß-hydroxybutyrate concentrations during the event, after which they returned to levels not significantly different from their pretreatment control values. In group II animals (15 min of HI- and KA-induced seizures), the rise in ß-hydroxybutyrate mimicked exactly that seen in group I. In group IV (30 min of HI- and KA-induced seizures), ß-hydroxybutyrate increased during seizures but to a lesser extent than group II, and during recovery had levels significantly below values seen in group II, for up to 24 h of recovery. Brain concentrations of ß-hydroxybutyrate followed closely those changes that occurred in blood during active seizures. During recovery, no significant differences between groups existed. Blood/brain concentration comparisons again showed a much lower ratio in group IV compared with the others; however, this time, it appears to be due to a relatively lower value for blood concentration, rather than a relatively higher value for brain. Graphs depicting the relationship between blood concentrations of glucose, lactate, and ß-hydroxybutyrate are shown in figure 2. As indicated, a close, compensatory relationship is seen between the 3 intermediates, each of which is known to substantially serve as substrates to the neonatal brain. In both groups I and II (KA alone and 15 min of HI plus KA seizures, respectively), neither of which experienced any brain damage, as glucose values decreased, both lactate and ß-hydroxybutyrate increased in an inverse relationship, and only returned to normal as concentrations of glucose rose following the episode of seizures. In group IV (30 min of HI- plus KA-induced seizures), blood glucose concentrations remained below normal concentrations, even during the first 2 h of recovery from seizures, and when both lactate and ß-hydroxybutyrate were also declining. Indeed, blood concentrations in group IV animals were significantly below those seen in group II for up 2 h of recovery for lactate and ß-hydroxybutyrate, and up to 24 h for ß-hydroxybutyrate (table 1). Cerebral Energy Reserves Table 3 contains the measurements for cerebral highenergy phosphate reserves (ATP and PCr) extracted from the hippocampus. Concentrations of PCr fell to between 15–25% of control values during HI. Recovery prior to the onset of seizures was universal. In all groups experiencing seizures, PCr again fell to levels significantly

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

below control for the first 2.5–3.0 h of seizure activity, but recovered to control values before the terminus of seizure activity and continued at control values for 48 h of recovery. Values of ATP likewise were significantly depleted during HI. As anticipated, complete recovery occurred by 1 h following HI in groups III and V which did not experience seizures. Interestingly, in group I, in which only KA seizures were induced, there was no perturbation of ATP throughout the seizure episode. In group II (15 min of HI and KA seizures), ATP did not recover completely prior to seizure onset, but did gradually replenish to normal values by 3 h, half-way into the seizure episode. Group IV (30 min of HI and KA seizures) displayed a prolonged reduction in ATP concentrations, remaining significantly below control values following HI and not recovering until the terminus of seizure activity. In this group, ATP concentrations remained below normal, and significantly below those concentrations measured in all other groups for beyond the first 2 h of recovery from seizure activity. Alterations in Excitatory Amino Acid Release Table 4 contains the concentrations of amino acids measured at specific time points during the experimental paradigm. Serine, glutamine, and alanine showed very little and inconsistent alterations during either seizures, HI, or the combination of the two. Glycine levels increased significantly above pretreatment control values during the first 3 h of seizures, returning to normal by the end of the seizure event. GABA levels increased in the extracellular fluid of the hippocampus, during the entirety of the seizures, but were significantly higher than controls in an inconsistent fashion. Both glycine and GABA increased significantly during 30 min of HI, returning rapidly to control values prior to the onset of seizures. Glutamate displayed a rapid rise above normal values during 30 min of HI, but did not do so after only 15 min. During seizures, glutamate values increased significantly above control values in all 3 groups experiencing seizures (groups I, II, and IV). This occurred in a stepwise fashion (fig. 4) such that the increase was greatest in those animals receiving 30 min of HI- and KA-induced seizures (group IV), and was significantly higher than in either groups II or I (p ! 0.01). This enhanced release of glutamate was seen throughout the episode of seizures, after which they declined in those groups which did not show any damage due to seizure activity. In group IV, concentrations of glutamate continued to rise throughout the seizures, and during the 4 h of recovery that were monitored during the experiment.

Dev Neurosci 2002;24:367–381

373

Fig. 2. Flow line graphs indicating alter-

ations in glycolytic intermediates during the course of the experimental paradigm in those animals having had KA-induced seizures. Rat pups having undergone HI alone are not shown.

374

Dev Neurosci 2002;24:367–381

Yager/Armstrong/Miyashita/Wirrell

Fig. 3. Brain (hippocampal) ATP concentra-

tions during the course of HI and seizures. Bars represent mean values for 6–8 animals B SEM. a p ! 0.05 compared with pretreatment control; bp ! 0.05 compared with animals receiving KA alone (group I), and c p ! 0.05 compared with 15 min of HI plus KA (group II).

Table 3. Brain (hippocampal) concentrations of high-energy phosphates

KA

15 min HI + KA 15 min HI

30 min HI + KA 30 min HI

2.90B0.13

2.81B0.17 0.37B0.04a 2.45B0.28

2.81B0.17 0.37B0.04a 2.45B0.28

2.81B0.17 0.60B0.19a 2.61B0.28

2.81B0.17 0.60B0.19a 2.61B0.28

Seizures 1h 3h 5h

2.12B0.08a 2.36B0.15a 2.66B0.15

1.89B0.15a, c 1.95B0.21a, c 2.79B0.21

2.82B0.10a 2.79B0.10 3.15B0.14

1.96B0.11a, c 2.10B0.23a, c 2.58B0.17

2.72B0.21b 2.84B0.18 2.94B0.16

Recovery 7h 24 h 48 h

2.59B0.08 2.49B0.12 2.65B0.17

2.91B0.18 2.88B0.22 2.50B0.19

3.23B0.15 3.18B0.26 3.25B0.12

2.81B0.17 3.04B0.19 2.98B0.24

2.93B0.17 3.07B0.24 2.52B0.31

1.84B0.15

1.85B0.11 0.36B0.05a 1.34B0.18a, b

1.85B0.11 0.36B0.05a 1.34B0.18a, b

1.85B0.11 0.36B0.12a 1.30B0.18a, b

1.85B0.11 0.36B0.12a 1.30B0.18a, b

Seizures 1h 3h 5h

1.37B0.15 1.55B0.09 2.12B0.11

1.33B0.06a, c 1.66B0.09 2.11B0.12

2.04B0.09 1.95B0.14 1.97B0.09

1.33B0.06a, c 1.36B0.10a, c 1.63B0.06b, c, d

2.13B0.06 2.08B0.05 1.92B0.10

Recovery 7h 24 h 48 h

2.33B0.11 2.21B0.08 2.30B0.15

2.27B0.12 2.10B0.12 2.05B0.07

2.19B0.16 2.11B0.12 2.12B0.08

1.68B0.04b, c, d 1.85B0.14b 1.98B0.14

2.07B0.14 1.88B0.11 1.95B0.15

PCr Pretreatment control HI Pre-KA injection

ATP Pretreatment control HI Pre-KA injection

Values are depicted as the mean of 6–8 animals per time B SEM millimoles per liter. a p ! 0.05 compared with homologous pretreatment control; b p ! 0.05 compard with animals receiving KA alone (group I); c p ! 0.05 compared with homologous HI group, and d p ! 0.05 compared with 15 min HI plus KA (group II).

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

Dev Neurosci 2002;24:367–381

375

Table 4. Excitatory amino acid release during experimental paradigm Glutamate

Serine

Glutamine

Glycine

Taurine

Alanine

GABA

1.53B0.09

33.17B3.12

278.81B22.38

7.51B0.79

10.83B0.80

47.08B7.20

0.41B0.07

3.38B0.66a 4.99B0.38a 4.05B0.25a

41.79B6.48 36.99B3.86 33.48B4.94

303.31B37.00 309.09B27.79 274.85B22.12

11.28B1.00 17.40B1.28a 16.54B1.-80a

23.77B3.08a 25.11B2.55a 21.43B1.41a

55.77B10.73 57.27B6.81 56.72B6.92

0.76B0.04 0.96B0.12a 1.17B0.20a

3.86B0.18a 3.01B0.37a

32.49B4.36 28.18B2.90

236.78B22.97 264.94B30.90

15.57B1.56a 9.08B1.23

17.46B0.77 12.59B1.14

53.99B6.77 43.91B7.50

0.68B0.07 0.62B0.12

1.73B0.16 2.51B0.34 2.57B0.37

34.05B3.23 35.78B3.61 33.92B4.34

244.45B17.97 229.79B21.33 225.59B28.19

6.99B1.07 11.03B1.84 7.60B0.87

8.45B0.90 16.10B3.24a 9.49B1.45

44.89B4.60 62.65B13.17 56.24B6.69

0.39B0.08 0.81B0.16 0.47B0.13

Seizures 1h 3h 5h

5.05B1.24a, d 4.84B0.91a 4.20B0.79a

35.61B4.28 37.62B6.68 35.10B6.64

235.96B35.56 236.94B31.47 226.80B42.95

16.81B2.88a, c 15.82B2.84a 14.06B2.04

16.93B2.16a 24 49B2.15a 24.70B2.73a

63.66B8.00 52.15B6.94 80.34B12.43a

0.87B0.13 1.28B0.25a, c 0.84B0.23

Recovery 7h 9h

3.55B0.48a 3.72B1.12

34.03B6.26 27.82B4.49

223.31B34.62 199.89B25.97

14.83B4.34 11.63B1 44

21.00B199a 18.05B2.04a

68.40B12.23a 60.24B8.92

1.08B0.24 0.33B0.08

1.57B0.19 2.25B0.24 2.11B0.17

31.57B2.28 46.25B4.46 32.38B4.74

228.27B13.56 262.95B30.67 244.61B34.35

6.25B0.86 11.40B2.51 5.72B0.55

9.13B1.07 14.44B2.15a 9.26B1.46

47.70B6.66 71.13B10.35a 62.63B11.06

0.32B0.07 0.83B0.48 0.41B0.07

Seizures 1h 3h 5h

1.55B0.22 1.74B0.23b 2.31B0.39

36.08B3.53 34.81B3.00 33.71B5.03

252.60B15.60 261.95B24.24 236.94B23.10

6.52B0.97 10.05B2.90 8.49B1.66b

10.47B1.56 11.28B1.32 10.80B1.35

53.62B5.98 55.94B6.66 60.77B11.44

0.34B0.07 0.47B0.07 0.30B0.06

Recovery 7h 9h

1.72B0.56 1.99B0.36

29.94B4.08 30.87B2.44

242.57B51.02 239.33B38.37

10.42B2.42 11.03B2.35

11.83B1.47 11.26B1.19

65.92B9.08 63.61B9.35

0.40B0.08 0.27B0.04

1.72B0.07 8.68B1.35a, d 2.24B0.23

41.87B4.38 52.09B6.36 39.37B6.12

241.69B12.39 241.51B13.67 190.14B15.96

7.23B0.78 18.75B3.17a 8.50B1.03

10.84B1.48 71.94B12.15a, d 44.40B7.67a, c, d

41.71B2.82 74.40B11.23a, c 60.72B10.75a

0.38B0.06 2.45B0.47a, d 0.84B0.21

Group I Pretreatment control Seizures 1h 3h 5h Recovery 7h 9h Group II Pretreatment control HI Pre-KA injection

Group III Pretreatment control HI Pre-KA injection

Group IV Pretreatment control HI Pre-KA injection Seizures 1h 3h 5h

51.59B5.82 5.79B0.63c 9.72B0.90a, c, d 49.00B5.42 10.82B2.13a, b, c, d 50.59B5.33

273.89B36.48 290.84B18.71 321.73B33.36a

19.13B2.57a, c 14.83B1.99a 14.08B1.79

39.27B7.66a, b, c, d 64.77B8.37 42.88B6.99a, b, c, d 45.11B5.56 37.45B5.90a, b, c 62.45B7.96

1.44B0.46a, d 1.28B0.15d 0.98B0.12d

Recovery 7h 9h

13.21B3.15a, b, c, d 48.20B6.35 11.52B2.81a, b, c, d 41.38B2.80

353.44B29.03a 340.84B11.92

17.87B3.35a 14.93B3.18

29.31B5.55b, c 23.53B4.83b, c

57.77B6.11 48.69B5.79

0.78B0.27 0.93B0.38

Group V Pretreatment control HI Pre-KA injection

1.54B0.16 6.59B0.74a 1.65B0.26

40.36B6.66 45.73B4.88 46.60B4.92

258.58B27.33 270.95B31.36 228.20B19.00

9.07B1.01 17.82B2.33a 12.58B1.95

11.53B1.65 50.45B7.31a 27.37B3.70a

40.88B6.21 58.48B8.22a 59.00B13.17

0.39B0.08 1.35B0.41a 0.46B0.25

Seizures 1h 3h 5h

1.55B0.24 1.55B0.24b 1.91B0.32

44.50B5.43 43.19B8.27 35.14B3.79

269.30B23.10 297.46B33.09 272.83B21.06

10.82B0.67 10.12B0.68 11.60B0.75

16.79B3.69 15.16B2.88 16.13B3.84

44.77B8.96 39.27B7.11 34.44B7.22

0.35B0.05 0.43B0.10 0.33B0.08

Recovery 7h 9h

2.91B0.63 2.95B0.54

41.11B6.74 35.47B5.86

328.47B36.49 308.54B15.71

12.54B0.85 11.37B1.14

15.34B4.25 13.23B2.68

27.62B2.66 23.84B3.83

0.44B0.12 0.34B0.08

Extracellular fluid concentrations as determined by in vivo microdialysis. Values are depicted as the mean of 6–8 animals per time B SEM in millimoles per liter. a p ! 0.05 compared with homologous pretreatment control; b p ! 0.05 compared with animals receiving KA alone (group I); c p ! 0.05 compared with homologous HI group, and d p ! 0.05 compared with 15 min HI plus KA (group II).

376

Dev Neurosci 2002;24:367–381

Yager/Armstrong/Miyashita/Wirrell

Fig. 4. Line graph depicting representative alterations in extracellular concentrations of amino acids (glutamate) during the course of HI and seizures in groups I, II, and IV. Animals receiving HI alone are not included. Lines represent means of 6–8 animals B SEM. a p ! 0.05 compared with pretreatment control; bp ! 0.05 compared with animals receiving KA alone (group I), and c p ! 0.05 compared with 15 min of HI plus KA (group II).

Taurine concentrations likewise increased above control values during HI, and again during seizure activity. The highest concentrations of extracellular taurine were reached during HI in group IV. In contrast to the alterations seen with glutamate, taurine levels in this group remained stable during the seizure activity, and then slowly declined, remaining, however, above values seen in group II.

Discussion

The present findings contribute importantly to our understanding of the damaging effects neonatal seizures have on the compromised hypoxic-ischemic brain. The data confirm previous work by others that prolonged seizures in an otherwise healthy newborn brain do not compromise high-energy phosphate reserves [Young et al., 1984, 1985, 1991]. They further suggest that for seizures to cause cell death, a reduction of high-energy phosphate reserves, significant enough in depth and duration to begin cellular disruption is required. In this regard, our previous study [Wirrell et al., 2001] indicated that pups exposed to either KA alone or in combination with 15 min of HI experienced no cell death. Animals exposed to 30 min of HI displayed cell death in specific regions of the cortex and hippocampus. When combined with sei-

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

zures, the energy depletion seen in this group (table 3, fig. 3) lingers throughout the ictus, and remains below levels seen in the brains of recovering nondamaged animals. In association with these perturbations in high-energy phosphate reserves, elevations of taurine and glutamate reach levels 2- and 3-fold those seen in either of the other groups experiencing KA-induced seizures. The combination of metabolic alterations that occurs with hypoxicischemic injury followed by prolonged seizures results in the enhanced damage displayed in the hippocampus. In the immature brain, normally resistant to the cell death following KA-induced seizures [Carmant et al., 1995; Wasterlain and Duffy, 1976; Wirrell et al., 2001; Fernandes et al., 1999], a prolonged depletion of high-energy phosphates is required as a prerequisite to the release of high enough concentrations of glutamate during seizures to enhance neuronal damage. The depletion of high-energy phosphate reserves as a prerequisite for cell death is consistent with investigations which have found the immature brain to be resistant to damage during prolonged seizures alone. In this regard, Young et al. [1984, 1991, 1995] found no reduction of ATP, even with status epilepticus induced in neonatal dogs for up to 3 h. Since ATP is the critical regulator of cell function and is responsible for cell membrane integrity through the active transport of Na+, K+, and Ca2+ ions, it has been suggested by many investigators that it is the

Dev Neurosci 2002;24:367–381

377

depletion of high-energy phosphates that must occur as a necessary prerequisite to those mechanisms underlying cellular dysfunction and death [Yager et al., 1994; Lipton and Whittingham, 1982; Novelli et al., 1988]. Indeed, Novelli et al. [1988] showed that despite similar concentrations of glutamate in cerebellar cultures of 8-day-old rats, it only became neurotoxic when energy depletion occurred as well. Hence, whereas prolonged seizures alone do not deplete the energy reserves of the immature brain, HI certainly does [Yager et al., 1991; Palmer at al., 1990] and likely acts to not only trigger, but enable further metabolic disruption and cell death via the excitatory amino acid pathway. The most striking finding associated with the current study is that the preceding hypoxic-ischemic stress of 30 min resulted in a marked enhancement of glutamate accumulation during the subsequent KA-induced seizures. An elevation above control of extracellular concentrations of glutamate also occurred in animals that were exposed to seizures alone or in combination with a shorter ‘subthreshold’ duration of HI. In these latter groups, however, glutamate increased only 2- to 3-fold above control, whereas animals displaying enhanced damage as a result of the seizures showed increases that were 7-fold greater than normal. Cataltepe et al. [1996] measured CSF levels of glutamate and GABA during HI and following seizures in 7-day-old rats. In previous work done by this group [Cataltepe et al., 1995] in neonatal rats, seizures did not exacerbate hypoxic-ischemic brain damage. In keeping with these latter findings, concentrations of amino acids in the group experiencing HI and seizures did not increase above those experiencing seizures alone. Young et al. [1992] utilized in vivo microdialysis to measure the release of several excitatory and inhibitory amino acids in the brains of juvenile rabbits that had been exposed to brief hypoxia followed by 45 min of bicuculline-induced seizures. As in our study, glutamate rose substantially, as did GABA and taurine during the seizures that followed a preceding hypoxic stress. In this group, values of glutamate only rose approximately 150% above control. Whether these latter animals experienced any brain damage is not clear. That the excitatory amino acid, glutamate, plays a major role in seizure-induced brain damage is certainly well recognized. The immature brain, however, appears to be more resistant to the toxic effects of glutamate than the mature brain [Bickler et al., 1993; Liu et al., 1996]. Several possible explanations for the resistance to damage in immature brains have been proposed. Among these is the finding that the immature brain has low concentra-

378

Dev Neurosci 2002;24:367–381

tions of glutamate and glutamate-related enzymes [Bickler et al., 1993]. In addition, the postsynaptic action of glutamate differs in that there is less Ca2+ entry through NMDA receptors and greater Ca2+-buffering capabilities [Carmant et al., 1995; Marks et al., 1996]. Finally, the immature brain is less metabolically active than its more mature counterpart, particularly during prolonged seizures [Fernandes et al., 1999]. Several questions are raised by the current data. What mechanisms underly the more prolonged decrease in hippocampal ATP concentrations in rat pups where hypoxicischemic brain damage is exacerbated by seizures? Several studies have now shown that, despite the lower metabolic rate of the newborn [Vannucci and Duffy, 1976], during seizures, a hypermetabolic state arises in specific regions of the brain. Fernandes et al. [1999] compared regional cerebral metabolic rate for glucose (rCMRglu) in rats aged 10 and 21 days, and adults during pilocarpineinduced seizures. Marked regional hypermetabolism occurred in all age groups, but only the adult and 21-day-old rats displayed brain damage. The P10 rats increased their rCMRglu by up to 700% in the hippocampi, without any alterations in blood or brain glucose concentrations, suggesting that in this group, metabolic demands were met by adequate substrate supply. In neonatal dogs rendered hypoglycemic, in vivo 31P NMR studies were performed during prolonged seizures. In these studies, Young et al. [1987] found that brain ATP concentrations were minimally (not significantly) compromised compared with control, despite significant systemic hypoglycemia. Blood concentrations of lactate were elevated in seizure-induced groups, whereas only the hypoglycemic group showed a significant elevation of ß-hydroxybutyrate. The authors concluded that under circumstances of substrate depletion and high metabolic demands, alternate substrates such as lactate or fatty acids may be used to preserve highenergy phosphate production. In the current study, lactate rose dramatically during HI, whereas glucose and ß-hydroxybutyrate remained within normal limits. During seizures, blood glucose fell significantly below normal in all groups experiencing seizures, and remained significantly below normal for at least the first 4 h of recovery in group IV, which experienced the seizure-exacerbated brain damage. Evaluation of the blood/brain ratios of the glycolytic intermediates (not shown) indicates that for this latter group, ratios were lower for all three intermediates (glucose, lactate, and ßhydroxybutyrate) than for either of the other two seizure groups. Lower ratios can either be due to lower blood levels or higher brain concentrations between the 3 groups.

Yager/Armstrong/Miyashita/Wirrell

In our studies, brain glucose concentrations remained relatively high in group IV compared with group II, during the 4 h of recovery, suggesting an inability to metabolize this substrate, despite a presumed hypermetabolic recovering state. Similarly, brain lactate concentrations showed a similar pattern, either because of glucose shunting towards lactate, or an inability of the substrate to enter aerobic metabolism. Both suggest a relative deficiency of the brain to utilize these substrates in the efficient recovery of ATP. The blood/brain concentrations for ß-hydroxybutyrate, however, suggest that the ratio for this substrate is lower in group IV animals because of relatively lower values in blood compared with brain. The latter suggest an increased utilization of this substrate through aerobic metabolism, in an attempt to restore ATP. These findings are consistent with our previous work which indicated that the brain of the neonate is particularly sensitive to substrate deficiency, and that under the stress of enhanced metabolic demands may be compromised. Hence, in previous studies, we showed that under circumstances of HI, the mitochondria of the newborn rat brain becomes paradoxically oxidized, suggesting more a limitation of substrate than of oxygen supply to the brain [Yager et al., 1991, 1996]. That providing alternate or enhanced substrates can protect the newborn brain has also been shown by several studies displaying a protective effect of glucose supplementation [Vannucci et al., 1992], fasting ketonemia [Yager et al., 1992c], and fatty acid supplementation [Yager and Thornhill, 1997]. More recently, Ying et al. [2002] have shown the importance of substrate supply in the prevention of cell death in an in vitro model of neurons and astrocytes. These authors investigated the role of poly (ADP-ribose) polymerase-1 (PARP1) in astrocyte-neuron cocultures. PARP1 is a DNA repair enzyme that, when extensively activated, as might occur during HI, may actually contribute to cell death by impairing glycolysis and causing further energy depletion. By administering supplemental amounts of the mitochondrial substrates ·-ketoglutarate or pyruvate, and bypassing glycolysis, cell death was decreased by 70%. It seems likely that during the initial phase of HI, in our model of neonatal seizures, PARP1 becomes activated and with the onset of a seizure-induced hypermetabolic state impedes the ability of cells to reestablish their energy requirements, causing further cell death [Harkness, 1997]. Other possible explanations for the enhanced damage seen in this model relates to the role of HI in altering glutamate metabolism. As shown in this study, glutamate concentration as measured in the extracellular fluid surrounding CA3 is profoundly increased during the seizures

and early recovery, in those animals known to experience enhanced damage. This can only be due to either an increase in the efflux of excitatory amino acid, or a reduction in uptake. Given that glutamate is normally taken up by astroglia and transformed to glutamine by an energydependant mechanism, one would have anticipated an increase in measured glutamine as well. As there was none, it is justifiable to assume that the increased glutamate levels were due to an impairment of glutamate reuptake. In this regard, Martin et al. [1997] showed that following HI, glutamate transporter 1 and excitatory amino acid carrier are diminished by 15 and 55%, respectively. Tanaka et al. [1997] likewise showed enhanced damage following seizures in mice genetically deficient of glutamate transporter 1. Considerable evidence suggests that Ca2+ permeable AMPA receptors are critical mediators of delayed selective neuronal death associated with sustained seizures [Friedman et al., 1994, 1997; Friedman and Veliskova, 1998]. Furthermore, Oguro et al. [1999] showed that a reduction of GluR2 mRNA, and enhancement of AMPA receptor mediated Ca2+ influx, caused by a brief sublethal ischemic insult resulted in delayed neuronal death of specific neuronal populations within the hippocampus. Sanchez et al. [2001] also found that hypoxic-induced seizures in P10 rats caused a reduction in GluR2 protein at 48 and 96 h following the insult. In summary, the findings of our study reaffirm the resistant nature of the immature healthy neuron to seizure-induced damage. This intrinsic protection appears to be in the first instance due to the ability of the newborn brain to maintain adequate energy reserves during seizures such that the increased release of glutamate remains as a neurotransmitter and not a neurotoxin. When the brain is compromised, and in particular when high-energy phosphate compounds are diminished, the further hypermetabolic stress of seizures reduces the ability of the brain to resynthesize ATP such that recovery can occur. The latter may be due to a relative lack of substrate supply or an inability of the cell to properly utilize substrates. Once in this state, glutamate accumulates in the extracellular space, and in the presence of diminished ATP becomes neurotoxic, leading to an enhancement of cell damage resulting from seizures. The findings have clear correlates to neonatal seizures in the human. Though studies are few, Younkin et al. [1986] reported on 8 infants with seizures in whom 31P NMR studies were done. Of these, 5 had perinatal asphyxia as a cause of the neonatal seizures. When NMR studies were done during the seizures in 4 of the infants, high-energy metabolites were consistently compromised in each of those who had previously experi-

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

Dev Neurosci 2002;24:367–381

379

enced HI, but not in the infant who had seizures for other reasons. Similarly, Miller et al. [2002], using proton spectroscopy in human newborns experiencing seizures as a result of asphyxia, showed alterations in cerebral metabolism during seizures, independent of the previous asphyxial insult. The authors suggested that the data supported the contributing role of seizures to the damage seen in newborns with perinatal asphyxia. Our findings not only support an aggressive approach to the treatment of neonatal seizures which compromise perinatal HI, but suggest alternate approaches to therapeutic intervention and neuronal rescue.

Acknowledgements The authors gratefully acknowledge the support for this research by grants from the Toronto Hospital for Sick Children Foundation, and the Health Services Utilization and Research Commission of Saskatchewan. Dr. J.Y. Yager is the recipient of the Henry J. M. Barnett Scholarship from the Heart and Stroke Foundation of Canada.

References Albala BJM, Moshe SL, Okada R (1984): Kainicacid-induced seizures: A development study. Brain Res 13:139–148. Bickler PE, Gallego SM, Hansen BM (1993): Developmental changes in intracellular calcium regulation in rat cerebral cortex during hypoxia. J Cereb Blood Flow Metab 13:811–819. Camfield PR (1997): Recurrent seizures in the developing brain are not harmful. Epilepsia 38: 735–737. Carmant L, Liu Z, Werner SJ, Mikati MA, Holmes GL (1995): Effect of kainic acid-induced status epilepticus on inositiol-trisphosphate and seizure-induced brain damage in mature and immature animals. Brain Res Dev Brain Res 89: 67–72. Cataltepe O, Towfighi J, Vannucci RC (1996): Cerebrospinal fluid concentrations of glutamate and GABA during perinatal cerebral hypoxiaischemia and seizures. Brain Res 709:326– 330. Cataltepe O, Vannucci RC, Hietjan DF, Towfighi J (1995): Effect of status epilepticus on hypoxicischemic brain damage in the immature rat. Pediatr Res 38:251–257. Dobbing J, Sands J (1979): Comparative aspects of the brain growth spurt. Early Hum Dev 3:79– 83. Donazanti BA, Yamamoto BK (1998): An improved rapid HPLC-EC method for the isocratic separation of amino acid neurotransmitters from brain tissue and microdialysis perfusate. Life Sci 43:913–922. Fernandes MJS, Dube C, Boyet S, Marescaux C, Nehlig A (1999): Correlation between hypermetabolism and neuronal damage during status epilepticus induced by lithium and pilocarpine in immature and adult rats. J Cereb Blood Metab 19:195–209. Franck JE, Schwartzkroin PA (1984): Immature rabbit hippocampus is damaged by systemic but not intraventricular kainic acid. Brain Res 13:219–227.

380

Friedman LK, Pellegrini-Giampietro DE, Sperber EF, Bennett MVL, Moshe SL, Zukin RS (1994): Kainate-induced status epilepticus alters glutamate and GABA-A receptor gene expression. J Neurosci 14:2697–2707. Friedman LK, Sperber EF, Bennett MVL, Moshe SL, Zukin RS (1997): Developmental regulation of glutamate and GABAA receptor gene expression in rat hippocampus following kainate-induced status epilepticus. Dev Neurosci 19:529–542. Friedman LK, Veliskova J (1998): GluR2 hippocampal knockdown reveals developmental regulation of epileptogenicity and neurodegeneration. Brain Res Mol Brain Res 61:224–231. Hagberg H, Bona E, Gilland E, Puka-Sundvall M (1997): Hypoxia-ischaemia model in the 7-dayold rat: Possibilities and shortcomings. Acta Paediatr Suppl 422:85–88. Harkness RA (1997): Is post-hypoxic-ischemic cell damage associated with excessive ATP consumption rather than a failure of ATP production? Acta Paediatr 86:1–5. Hirsch E, Baram TZ, Snead OC (1992): Ontogenic study of lithium-pilocarpine-induced status epilepticus in rats. Brain Res 583:120–126. Holden KR, Mellits ED, Freeman JM (1982): Neonatal seizures. 1. Correlation of prenatal and perinatal events with outcomes. Pediatrics 70: 165–176. Holmes GL (1997): Epilepsy in the developing brain: Lessons for the laboratory and clinic. Epilepsia 38:12–30. Holmes GL, Ben-Ari Y (2001): The neurobiology and consequences of epilepsy in the developing brain. Pediatr Res 49:320–325. Holmes GL, Gairsa J-L, Chevassus-Au-Louis N, Ben-Ari Y (1998): Consequence of neonatal seizures in the rat: Morphological and behavioral effects. Ann Neurol 44:845–857. Kanthan R, Shuaib A, Griebel R, El-Alazounni H, Miyashita H, Kalra J (1996): Evaluation of monoaminergic neurotransmitters in acute focal ischemic human brain model by intracerebral in vivo microdialysis. Neurochem Res 21: 563–566.

Dev Neurosci 2002;24:367–381

Kanthan R, Shuaib A, Griebel R, Miyashita H (1995): Intracerebral human microdialysis. In vivo study of an acute focal ischemic model of the human brain. Stroke 26:870–873. Lipton P, Whittingham TS (1982): Reduced ATP concentration as a basis for synaptic transmission failure during hypoxia in the in vitro guinea-pig hippocampus. J Physiol 325:51–65. Liu Z, Stafstrom CE, Sarkisian M, Tandon P, Yang Y, Hori A, Holmes GL (1996): Age-dependent effects of glutamate toxicity in the hippocampus. Brain Res Dev Brain Res 97:178–184. Lowry OH, Passonneau JV, Hasselberger FX, Schulz DW (1964): Effect of ischemia on known substrates and cofactors of the glycolytic pathway in brain. J Biol Chem 239:18–30. Marks JD, Friedman JE, Haddad GG (1996): Vulnerability of CA1 neurons to glutamate is developmentally regulated. Brain Res Dev Brain Res 97:194–206. Martin LJ, Brambrink AM, Lehmann C, PorteraCailliau C, Koehler R, Rothstein J, Traystman RJ (1997): Hypoxia-ischemia causes abnormalities in glutamate transporters and death of astroglia and neurons in newborn striatum. Ann Neurol 42:335–348. Miller SP, Weiss BS, Barnwell A, Ferriero DM, Lalta-Hajnal B, Ferrer-Rogers A, Newton N, Partridge JC, Glidden DV, Vigneron DB, Barkovich AJ (2002): Seizure-associated brain injury in term newborns with perinatal asphyxia. Neurology 58:542–548. Mizrahi EM, Kellaway P (1987): Characterization and classification of neonatal seizures. Neurology 37:1837–1844. Mizrahi EM, Kellaway P (1998): Diagnosis and Management of Neonatal Seizures. New York, Lippincott-Raven, pp 47–60. Moshe SL (1993): Seizures in the developing brain. Neurology 43(suppl 5):S3–S7. Nitecka L, Tremblay E, Charton G, Bouillot JP, Berger ML, Ben-Ari Y (1984): Maturation of kainic acid seizure-brain damage syndrome in the rat. 2. Histopathological sequelae. Neuroscience 13:1073–1094.

Yager/Armstrong/Miyashita/Wirrell

Novelli A, Reilly JA, Lysko PG, Henneberry RC (1988): Glutamate becomes neurotoxic via the N-methyl-D-aspartate receptor when intracellular energy levels are reduced. Brain Res 451: 205–212. Oguro K, Oguro N, Kojima T, Grooms SY, Calderone A, Zheng X, Bennett MVL, Zukin RS (1999): Knockdown of AMPA receptor GluR2 expression causes delayed neurodegeneration and increases damage by sublethal ischemia in hippocampal CA1 and CA3 neurons. J Neurosci 19:9218–9227. Owen JA, Ijaz S, Miyashita H, Wishart T, Howlett W, Shuaib A. (1997): Zonisamide as a neuroprotective agent in an adult gerbil model of global forebrain ischemia: A histological, in vivo microdialysis and behavioral study. Brain Res 770:155–122. Palmer C, Brucklacher RM, Christensen MA, Vannucci RC (1990): Carbohydrate and energy metabolism during the evolution of hypoxic-ischemic brain damage in the immature rat. J Cereb Blood Flow Metab 10:227–235. Ravindran J, Shuaib A, Ijaz S, Galazka P, Ishaqzay R, Miyashita H (1994): High extracellular GABA levels in the hippocampus as a mechanism of neuronal protection in cerebral ischemia in adrenalectomized gerbils. Neurosci Lett 176:209–211. Rice JE, Vannucci RC, Brierly JB (1981): The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 9:131– 141. Sanchez RM, Koh S, Rio C, Wang C, Lamperti ED, Sharma D, Corfas G, Jensen FE (2001): Decreased glutamate receptor 2 expression and enhanced epileptogenesis in immature rat hippocampus after perinatal hypoxia-induced seizures. J Neurosci 21:8154–8163. Sheth RD, Hobbs GR, Mullett M (1999): Neonatal seizures: Incidence, onset, and etiology by gestational age. J Perinatol 19:40–43. Shuaib A, Ijaz S, Hemmings S, Galazka P, Ishaqzay R, Miyashita H (1994): Decreased glutamate release during hypothyroidism may contribute to protection in cerebral ischemia. Exp Neurol 128:260–265. Stafstrom CE, Thompson JL, Holmes GL (1992): Kainic acid seizures in the developing brain: Status epilepticus and spontaneous recurrent seizures. Brain Res Dev Brain Res 65:227– 236.

Neonatal Seizures and Hypoxic-Ischemic Brain Damage

Tanaka K, Watase K, Manabe T, Yamada K, Watanabe M, Takahashi K, Iwama H, Nishikawa T, Ichihara N, Kikuchi T, Okuyama S, Kawashima N, Hori S, Takimoto M, Wada K (1997): Epilepsy and exacerbation of brain injury in mice lacking the glutamate transporter GLT-1. Science 276:1699–1702. Van Gelder NM (1992): Pathologies of the CNS and associated taurine changes; in Lombardini JB (ed): Taurine. New York, Plenum Press, pp 253–261. Vannucci RC, Duffy TE (1976): Cerebral oxidative and energy metabolism of fetal and neonatal rats during anoxia and recovery. Am J Physiol 230:1269–1275. Vannucci RC, Mujsce DJ (1992): Effect of glucose on perinatal hypoxic-ischemic brain damage. Biol Neonate 62:215–222. Volpe JJ (2002): Neurology of the Newborn, ed 4. Philadelphia, WB Saunders, pp 178–214. Wasterlain CG (1997): Recurrent seizures in the developing brain are harmful. Epilepsia 38: 728–734. Wasterlain CG, Duffy TE (1976): Status epilepticus in immature rats. Arch Neurol 33:821– 827. Wasterlain CG, Shirasaka Y (1994): Seizures, brain damage and brain development. Brain Dev 16: 279–295. Wirrell EC, Armstrong EA, Osman LD, Yager JY (2001): Prolonged seizures exacerbate perinatal hypoxic-ischemic brain damage. Pediatr Res 50:445–454. Yager JY, Asselin J (1996): The effect of mild hypothermia on cerebral energy metabolism during the evolution of hypoxic-ischemic brain damage in the immature rat. Stroke 27:919–926. Yager JY, Brucklacher RM, Mujsce DJ, Vannucci RC (1992a): Cerebral oxidative metabolism during hypothermia and circulatory arrest in newborn dogs. Pediatr Res 32:547–553. Yager JY, Brucklacher RM, Vannucci RC (1991): Cerebral oxidative metabolism and redox state during hypoxia-ischemia and early recovery in immature rats. Am J Physiol 261:H1102– H1108. Yager JY, Brucklacher RM, Vannucci RC (1992b): Cerebral energy metabolism during hypoxiaischemia and early recovery in immature rats. Am J Physiol 262:H672–H677. Yager JY, Brucklacher RM, Vannucci RC (1996): Paradoxical mitochondrial oxidation in perinatal hypoxic-ischemic brain damage. Brain Res 712:230–238.

Yager JY, Heitjan DF, Towfighi J, Vannucci RC (1992c): Effect of insulin-induced and fasting hypoglycemia on perinatal hypoxic-ischemic brain damage. Pediatr Res 31:138–142. Yager JY, Kala G, Hertz L, Juurlink BHJ (1994): Correlation between high-energy phosphates and hypoxic-ischemic damage in immature and mature astrocytes. Brain Res Dev Brain Res 82:62–68. Yager JY, Thornhill JA (1997): The effect of age on susceptibility to hypoxic-ischemic brain damage. Neurosci Biobehav Rev 21:167–174. Yager JY, Towfighi J, Vannucci RC (1993): Influence of mild hypothermia on hypoxic-ischemic brain damage in the immature rat. Pediatr Res 34:525–529. Ying W, Chen Y, Alano CC, Swanson RA (2002): Tricarboxylic acid cycle substrates prevent PARP-mediated death of neurons and astrocytes. J Cereb Blood Flow Metab 22:774–779. Young RSK, Cowan BE, Petroff OAC, Novotny E, Dunham SL, Briggs RW (1987): In vivo 31P and in vitro 1H nuclear magnetic resonance study of hypoglycemia during neonatal seizure. Ann Neurol 22:622–628. Young RSK, During MJ, Aquila WJ, Tendler D, Ley E (1992): Hypoxia increases extracellular concentrations of excitatory and inhibitory neurotransmitters in subsequently induced seizures: In vivo microdialysis study in the rabbit. Exp Neurol 117:204–209. Young RSK, Osbakken MD, Briggs RW, Yagel S K, Rice DW, Goldberg S (1985): 31P NMR study of cerebral metabolism during prolonged seizures in the neonatal dog. Ann Neurol 18: 14–20. Young RSK, Petroff OAC, Chen B, Gore JC, Aquila WJ (1991): Brain energy state and lactate metabolism during status epilepticus in the neonatal dog. In vivo 31P and 1H nuclear magnetic resonance study. Pediatr Res 29:191– 195. Young RSK, Yagel SK, Gonya-Magee T (1984): Physiologic and metabolic alterations associated with seizures in normoxic and asphyxiated neonatal dogs. Epilepsia 25:69–573. Younkin DP, Delivoria-Papadopoulos M, Maris J, Donlon E, Clancy R, Chance B (1986): Cerebral metabolic effects of neonatal seizures measured with in vivo 31P NMR spectroscopy. Ann Neurol 20:513–519.

Dev Neurosci 2002;24:367–381

381

Original Paper Received: June 8, 2002 Accepted: August 19, 2002

Dev Neurosci 2002;24:382–388 DOI: 10.1159/000069043

Neuroprotection of Creatine Supplementation in Neonatal Rats with Transient Cerebral Hypoxia-Ischemia Kathryn H. Adcock a Johann Nedelcu a, † Thomas Loenneker a Ernst Martin a Theo Wallimann b Bendicht P. Wagner c a Neuroradiology

and Magnetic Resonance Research, University Children’s Hospital, b Department of Cell Biology, Swiss Federal Institute of Technology (ETH), Zurich, and c Pediatric Intensive Care, University Children’s Hospital, Berne, Switzerland

Key Words Hypoxia-ischemia W Neonatal rat W Neuroprotection W Creatine W Supplementation W Brain W Rat

Abstract We hypothesized that creatine (Cr) supplementation would preserve energy metabolism and thus ameliorate the energy failure and the extent of brain edema seen after severe but transient cerebral hypoxia-ischemia (HI) in the neonatal rat model. Six-day-old (P6) rats received subcutaneous Cr monohydrate injections for 3 consecutive days (3 g/kg body weight/day), followed by 31P-magnetic resonance spectroscopy (MRS) at P9. In a second group, P4 rats received the same Cr dose as above for 3 days prior to unilateral common carotid artery ligation followed 1 h later by 100 min of hypoxia (8% O2) at P7. Rats were maintained at 37 ° C rectal temperature until magnetic resonance imaging was performed 24 h after HI. Cr supplementation for 3 days significantly increased the energy potential, i.e. the ratio of phosphocreatine to ß-nucleotide triphosphate (PCr/ßNTP) and PCr/inorganic phosphate (PCr/Pi) as measured by 31P-MRS. Rats with hemispheric cerebral hypoxic-ischemic insult that had

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0382$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

received Cr showed a significant reduction (25%) of the volume of edemic brain tissue compared with controls as calculated from diffusion-weighted images (DWI). Thus, prophylactic Cr supplementation demonstrated a significant neuroprotective effect 24 h after transient cerebral HI. We hypothesize that neuroprotection is probably due to the availability of a larger metabolic substrate pool leading to a reduction of the secondary energy failure because DWI has been reported to correlate with the PCr/Pi ratio in the acute phase of injury. Additional protection by Cr may be related to prevention of calcium overload, prevention of mitochondrial permeability transition pore opening and direct antioxidant effects. Copyright © 2002 S. Karger AG, Basel

Introduction

Perinatal cerebral hypoxia-ischemia (HI) leads to acute energy failure and cell death, resulting in chronic neurological symptoms. HI occurs in 2–4/1,000 full-term births and it is estimated that 10% of these infants will develop neurological disability such as cerebral palsy [1–3].

Bendicht Wagner Pediatric Intensive Care, Inselspital CH–3010 Berne (Switzerland) Tel. +41 31 632 9310, Fax +41 31 632 9748 E-Mail [email protected]

Studies of animal models of HI have shed some light onto the functional and temporal aspects of the pathogenesis of HI. A widely used model of neonatal HI is an adaptation of that developed by Levine [4], whereby unilateral common carotid artery ligation in the 7-day-old (P7) rat is followed by exposure to hypoxic conditions after a short recovery [5]. The model has a number of advantages: it represents a transient hemispheric cerebral insult reproducibly inducing a small core region within a large penumbra and is followed by the so-called secondary energy failure (SEF) [6, 7]. Histological and noninvasive magnetic resonance spectroscopy (MRS) and imaging (MRI) using this rat model have shown that there is, in fact, a multiphase pattern of the brain pathology [6–11]. (1) During HI, aerobic respiration fails; the end products of anaerobic respiration [creatine (Cr), inorganic phosphate (Pi) and lactate] exceed adenosine triphosphate (ATP) and phosphocreatine (PCr) and the pH falls [8, 11]. Hemispheric cytotoxic edema is visible on diffusionweighted images (DWI) and apparent diffusion coefficient (ADC) maps, at a time when depletion of cellular energy stores causes neuronal swelling [6, 10]. (2) For a few hours after HI, energy metabolism recovers and cytotoxic edema vanishes except for a small core lesion in the parietal cortex [6, 11]. (3) However, a delayed and secondary rise in the levels of lactate, Cr and Pi and a fall in ATP and PCr begins within hours of the insult while brain pH and cerebral blood flow remain constant [8–11]. This phenomenon is called the SEF. It is associated with glial activation and a reappearance of edema, the progression of which can be traced with MRI [6]. Astrocytic swelling induces further cytotoxic edema, which is visible with DWI and ADC, whereas necrotic and apoptotic neurons enhance vasogenic edema, as is depicted by T2-weighted imaging (T2WI) [6, 7, 10]. (4) Abnormal histological findings, representing ongoing neurodegenerative cell death in the adult rat, continue for much longer than was previously thought [12, 13]. Eventually, the tissue is resorbed and any hope of recovery is lost. In clinical settings, this multiphase pattern of brain pathology is also observed. After a transient cerebral HI insult, the fetus or newly born infant quickly recovers to a normal cardiovascular state. However, MRI and MRS examinations of the brain correspond to the SEF, and suggest, retrospectively, that perinatal cerebral HI had occurred [14–17]. In addition, the MRS observation of cerebral lactic alkalosis persisting months after HI may be indicative of an ongoing postasphyxial process, paralleling the slow progressing cell death in experimental models. Finally, the ratios and concentrations of several

Neuroprotection by Creatine in Hypoxia-Ischemia

metabolites derived from spectroscopic investigations of asphyxiated infants within the first week of life have been found to correlate with their long-term outcome, such as the PCr/Pi ratio [14, 15, 18, 19], and the ATP/total exchangeable Pi and lactate/N-acetylaspartate ratios [16]. Therapeutic interventions aimed at lessening SEF may, therefore, reduce the long-term neuropsychological handicap caused by perinatal cerebral HI and may not share the fate of other neuroprotective agents, which were discontinued due to a lack of long-term improvements [3]. For example, postischemic hypothermia ameliorates the size of infarct and the drop in PCr/Pi in the neonatal rat and furthermore provides long-term improvement of the behavioral outcome [7, 20]. Other, more direct approaches to prevent energy metabolism failure may also prove beneficial after HI. Cr is a simple guanidine compound, either synthesized endogenously from arginine, glycine and S-adenosylmethionine or ingested with fish and meat and is found throughout the body, including the brain [for a review, see ref. 21]. Its phosphorylated form, PCr, is the source of high-energy phosphate in the conversion of ADP to ATP by the enzyme Cr kinase at times of high cellular energy requirement [for a review, see ref. 22]. Cr supplementation has been shown to increase both Cr and PCr levels in the brain [23–26] and protect neurons [27], and provide functional benefits in several studies [28–32]. In light of these findings, we hypothesized that Cr supplementation would reduce the effect of SEF and the extent of brain edema observed after severe, transient cerebral HI in the neonatal rat model.

Methods The animal studies reported here have been approved by the Animal Care and Experimentation Committees of the Cantons of Zurich and Berne, Switzerland. Guidelines for animal care, research and ethics published by the Swiss National Academy of Medical Sciences were strictly adhered to. Cr Supplementation and 31P-MRS P6 Sprague Dawley rats of 16.5 B 1.2 g (mean B SD; n = 16) were injected subcutaneously with Cr monohydrate (3 g/kg body weight/ day; 28 mg/ml in 0.9% sodium chloride) for 3 days. At P9, noninvasive 31P-MRS under halothane anesthesia (4% induction, 1.5% maintenance in 70% N2O, 30% O2) was carried out with a purposebuilt surface coil made to fit one hemisphere in a specially designed water-warmed rat holder. The 31P spectra were recorded in pulseacquire mode using an adiabatic 90 ° pulse, a bandwidth of 5 kHz, 512 signal averages and repetition time of 8 s. The spectra were fitted fully automatically in the time domain with prior knowledge using a combination of Lorentzian and Gaussian line shapes and 10-Hz line

Dev Neurosci 2002;24:382–388

383

broadening [6, 7]. Classical and clinically relevant indices of energy within the brain were calculated from the ratios of PCr to Pi or ßnucleotide triphosphate (ßNTP) from the spectra [9, 14]. Saturation effects due to the different T1 relaxation times of PCr (F4 s) or ßNTP and Pi (F2 s) as reported elsewhere [33] do slightly underestimate the metabolite ratios, but do not affect the qualitative comparison of the ratios between control and Cr-supplemented animals. Animal Model of HI We used the modified Levine model of transient hemispheric cerebral HI [4, 5], which has been described in detail before [6, 10]. Briefly, P7 (n = 16) rats of 11.1 B 0.7 g (mean B SD), which had previously received Cr monohydrate supplementation as above for 3 days, were subjected to a right common carotid artery ligation under general anesthesia (halothane 4% induction, 1.5% maintenance in 70% N2O, 30% O2) and returned to their dams in a temperatureregulated cage for 1 h. The rats were then placed into a hypoxic chamber (8% O2, 0.5% halothane) for 100 min during which time breathing was monitored. The chamber was also heated to ensure a constant body temperature of 37 ° C for the duration of the hypoxia. The rats were subsequently returned to their dams and maintained at 37 ° C in a specially designed cage for 24 h after the insult, at which time MRI was carried out. It is our previous experience that P7 rats died 5 days after HI due to poor suckling [7, 20]. Although it was not our aim to maintain the rats for a long time, we routinely gavage fed HI rats every 6 h after HI in order to maintain their health for the duration of the experiment in accordance with the principle of permissive under-feeding in acute illness [34]. The feed consisted of 0.5 ml unsweetened condensed milk. Magnetic Resonance Imaging MRI was carried out in accordance with our previous studies [6, 20, 35]. Briefly, T2WI was acquired with a multislice RARE technique [36] with TR = 4 s, RARE factor = 16 and an interecho interval of 22 ms, resulting in an effective echo time (TE eff) of 252 ms. The matrix size was 256 ! 128 (pixel dimensions 156 ! 312 Ìm) and the field of view was 4 cm2. A stimulated echo sequence was used to acquire DWI: TE = 18 ms, TR = 2 s, diffusion weighting factor b = 1,290 s/mm2, matrix size = 128 ! 64, field of view = 4 cm2. The slice packages for both T2WI and DWI comprised 8 1.5-mm-thick slices in the axial plane (coronal with respect to the rat) interleaved by an 0.3-mm gap, which encompassed the whole brain. All measurements were carried out on a 2-tesla (T) whole body MR system (Bruker Medical, Faellanden, Switzerland) equipped with an actively shielded gradient insert with a 33-cm bore, maximum gradient strength of 30 mT/m and 150 Ìs rise time. Fractional volume of tissue with edema was calculated from 8 DWI slices per animal as described previously [6, 7, 20]. Edema was defined as a hyperintense signal within the HI-affected hemisphere compared with the unaffected contralateral hemisphere. The area of edema from each slice was measured by automated software (Paravision, Bruker Medical) and summed to yield the entire edemic tissue volume, which was expressed as a fraction of whole brain volume. Statistical Analysis Comparisons between control and Cr-treated groups were carried out by two-sided exact tests (Mann-Whitney). All values are mean B SEM.

384

Dev Neurosci 2002;24:382–388

Results

The rats tolerated subcutaneous injections well. There were no fatalities during the 3 days of injections, or during HI. However, there was 1 Cr-treated pup fatality due to aspiration after gavage feeding. Cr-supplemented rats did not differ from controls with respect to weight gain or rectal temperature (data not shown). High-Energy Phosphorous-Containing Metabolite Ratios after Cr Supplementation The levels of phosphorous-containing metabolites were measured in the hemisphere ipsilateral to the right common carotid artery ligation after 3 days of subcutaneous Cr supplementation in P9 rats. Figure 1 shows typical spectra for control and Cr-supplemented animals. The ratio of PCr/Pi was found to be 1.50 B 0.12 (mean B SEM) in control rats, but was significantly higher (2.25 B 0.26) in Cr-supplemented animals (p = 0.028). The ratio of PCr/ßNTP in control rats was 1.09 B 0.14 (n = 8), which rose significantly to 1.52 B 0.10 (n = 8) upon Cr supplementation (p = 0.021; fig. 2). Brain Edema after Cr Supplementation Figure 3 shows typical T2WI and DWI obtained from P8 control or Cr-supplemented rats 24 h after HI. DWI and T2WI hyperintensity represent cytotoxic and vasogenic edema, respectively. The MR images were hyperintense in the ipsilateral cortex, the hippocampus and basal ganglia in nonsupplemented animals, indicating widespread combined cytotoxic and vasogenic edema. Cytotoxic injury size was calculated to represent 42.5 B 2.4% (n = 8) of total brain volume in control animals from DWI. In contrast, Cr-supplemented rats showed a marked reduction in the volume of tissue with edema, 32.3 B 2.8% brain volume (n = 7) that was significantly different to controls (p = 0.0401; fig. 4). The secondary edema was mainly limited to the ipsilateral cortex in Cr-supplemented animals.

Discussion

We hypothesized that the subcutaneous Cr supplementation of neonatal rats would be neuroprotective after transient hemispheric HI. Indeed, it could be shown here that injection of Cr monohydrate for 3 days increased the PCr/ßNTP and PCr/Pi ratios in the brain, by nearly 40 and 50%, respectively, in P9 animals compared with controls. Since the signal to noise ratio for ßNTP and Pi

Adcock/Nedelcu/Loenneker/Martin/ Wallimann/Wagner

T2WI

DWI

Control

Cr

Fig. 1. Representative in vivo 31P-MRS spectra of P9 rat brains from control (top) or Cr-supplemented (bottom) pups. PME = Phosphomonoester; PDE = phosphodiester. The smoothed lines in the figure represent spectra fitted fully automatically in the time domain with prior knowledge using a combination of Lorentzian and Gaussian line shapes and 10-Hz line broadening [6, 7].

Fig. 3. T2WIs (left-hand column) and DWIs (right-hand column) of 2 representative P8 rats 24 h after HI. The control animal (top) displayed marked DWI and T2WI hyperintensity in the ipsilateral cortex, hippocampus and basal ganglia. In contrast, the Cr-supplemented animal (bottom) showed a marked reduction in the extent of hyperintensity in MR images; while the ipsilateral cortex was affected by HI, the hippocampus and basal ganglia were spared from edema.

Fig. 2. Changes in PCr/ßNTP and PCr/Pi after 3 days of Cr supplementation in P9 rats. Open bars = Control; filled bars = Cr-supplemented rats. Cr significantly increased both PCr/ßNTP and PCr/Pi compared with control animals. (* p ! 0.05).

Fig. 4. Changes in volume of tissue with edema in P8 rats 24 h after

Neuroprotection by Creatine in Hypoxia-Ischemia

Dev Neurosci 2002;24:382–388

HI subsequent to 3 days Cr supplementation. Open bar = Control + HI rats; filled bar = Cr-supplemented + HI rats. Cr significantly decreased the size of cytotoxic edema compared with control rats (* p ! 0.05).

385

remained almost constant between supplemented and nonsupplemented animals, the increase in the PCr/ßNTP and PCr/Pi ratios appeared to be due to an increase in PCr within the brain. Furthermore, it is shown here that Cr-supplemented rats had 25% smaller HI lesions 24 h after HI compared with nonsupplemented litter mates. These data suggest that after just 3 days of Cr supplementation, sufficient Cr was able to cross the blood brain barrier to ameliorate the severe damage inflicted by HI. Effects of Cr Supplementation on Brain Metabolism 1H-MRS reveals that 4 weeks of oral supplementation increases total Cr in the adult human brain [37]. Cr has also been used successfully in the treatment of infants with Cr deficiency; after oral supplementation, not only was there a significant increase in cerebral Cr concentration, but also improvements of neurological symptoms [38]. To date, there is no literature on the effects of exogenous Cr on newborn infants. We have shown here that PCr/Pi and PCr/ßNTP increase upon Cr administration from P6 to P8. In addition, we have found increased PCr/Pi and PCr/ßNTP after Cr injections in a small group of younger rats (P4– P6, data not shown). These data are in agreement with the observation that Cr supplementation increases PCr/ ßNTP in P10 rats [25]. It is interesting to note that augmentation of the energetic state by Cr is lost by P20 [25], which may be related to the fact that the Cr transporter is significantly downregulated during brain differentiation [39]. Other studies of Cr supplementation for more extended periods of time have shown increases, albeit small ones, with variability in different brain regions, of both total Cr and PCr in the brain of adult rabbits, rats, mice and guinea pigs [26, 30, 40]. This suggests that younger, immature rats are more responsive to Cr supplementation, perhaps due to a higher expression of the Cr transporter [39]. Since P7 rats are, in many ways, equivalent to near-term humans with respect to metabolism and maturation of the brain [41], we speculate that Cr supplementation may be very effective in increasing cerebral PCr in human newborns as well. The rate of reaction of Cr kinase, i.e. the interconversion of PCr and ATP in the brain is not significantly altered by Cr injections in the immature rat and rabbit, suggesting that excess Cr is sufficient to exert the functional consequences of Cr supplementation [25, 26].

386

Dev Neurosci 2002;24:382–388

Effects of Prophylactic Cr Supplementation on Cerebral HI DWI hyperintensity returns 2 h after HI in the neonatal rat and reaches a maximum between 20 and 30 h [6, 35]. The secondary edema profile shown here at 24 h after HI correlated well with previous studies [6, 7, 20, 35]. Therefore, 24 h after the HI insult, we would expect to see the effect of Cr supplementation on cytotoxic edema. Indeed, Cr significantly reduced the extent of hyperintensity seen in DWI at this time. We have previously shown that DWI hyperintensity is proportional to the extent of the reduction in PCr/Pi, i.e. the SEF [6]. Given the correlation between DWI and PCr/Pi during SEF, we suspect that Cr acts by lessening the SEF due to the increased preischemic PCr/Pi observed in age-matched rats. Thus, reduction of SEF is likely to be due to the availability of a large substrate pool as the neuroprotective source. In addition, there is a linear relationship between the cerebral ADC of water and PCr/Pi and hence SEF [42]. DWI is, therefore, an appropriate tool for measuring SEF after HI; moreover, DWI has a shorter sample time, a higher spatial resolution and shows fewer movement artifacts than 31P-MRS. In concert with increased PCr/ßNTP and PCr/Pi, Cr administration may affect energy metabolism by enhancing mitochondria function via Cr-stimulated respiration [43]. Cr administration also increases the ADP sensitivity of mitochondrial respiration [44]; ADP clearly increases as ATP is hydrolyzed in times of energy crises, e.g. during SEF. The mdx mouse has impaired energy metabolism in muscle cells and is routinely used as a model of Duchenne muscular dystrophy. If these mice are fed a Cr-enriched diet, their mitochondrial function is rescued and elevated to wild-type levels [45]. In addition, the energetics of calcium handling in mdx myotubes was improved [46] and calcium-induced muscle necrosis significantly reduced [45]. Upon energy failure, cells usually display increasing difficulties with calcium handling, which may eventually lead to chronic calcium overload, a condition shown to result in further deterioration of cellular energy status, in calcium uptake by mitochondria and opening of the mitochondrial permeability transition pore [47]. It has been shown, with transgenic mice expressing mitochondrial Cr kinase in their liver mitochondria, that Cr protected these mitochondria from undergoing calcium-induced permeability transition pore opening [48], an early event in apoptosis. Therefore, it has been proposed that Cr could protect mitochondria from failure by stabilizing the multienzyme complex of mitochondrial Cr kinase, porin and ATP/ADP translocase [49], thus providing continued

Adcock/Nedelcu/Loenneker/Martin/ Wallimann/Wagner

transport of high-energy phosphate into the cytoplasm. It is possible that in our model of SEF, Cr could be acting in such a manner. Finally, recently discovered properties of Cr are that it can directly act as an antioxidant against aqueous reactive oxygen species, e.g. peroxynitrite in vitro [50] or as a protective agent against excitotoxic damage of neurons [27, 51]. Intracellular calcium overload, free radicals and excitatory amino acids have been found to play an important role in brain injury following HI. It is possible that Cr supplementation would have longlasting effects. There is a strong correlation between SEF and the pathological outcome; the minimum value of PCr/Pi is proportional to the severity of the infarct [52]. Despite there being no long-term animal investigations into the relationship between SEF and the neurological outcome, reports from the clinic indicate that the severity of SEF (i.e. PCr/Pi) in the first weeks after birth correlate with the neurological outcome [14–16, 53, 54]. Therefore, we predict that by reducing the extent of SEF, one would see an improvement of neurological symptoms both in the short and long term. Holtzman et al. [25] have shown that subcutaneous injection of Cr into the 10-day-old rat, a time when energy metabolism is immature, increases PCr/ßNTP, survival and prevents hypoxia-induced seizures. Feeding Cr-enriched chow to pregnant mice increased ATP concentrations in brain stem slices of the neonates, which was sufficient to attenuate both firing duration and strength of hypoglossal neurons during and after anoxia in vitro [55,

56]. Adult rats, despite having a slight increase in overall Cr after supplementation, do not show a significant improvement of ADC during transient global hypoxia [24, 57], suggesting again that the metabolic maturity of the animal is critical to its response to Cr and HI. In concert with our findings of neuroprotection, Cr supplementation has shown potential benefits in animal models of brain pathologies other than HI, such as amyotrophic lateral sclerosis [28], Parkinson’s disease [29], Huntington’s disease [30, 31, 51] and traumatic brain injury [32]. We have shown that metabolically immature rats benefit from Cr supplementation by exhibiting an increased PCr/ßNTP ratio, indicating that an elevation of highenergy phosphates in the brain seems to be able to protect the brain after HI in the newborn rat. The long-term benefits of Cr supplementation are currently under investigation prior to evaluation of its effect in the clinical setting. It is our belief that supplementation with exogenous Cr during and after HI, by preserving brain energy levels, with its positive consequences on calcium homeostasis and mitochondrial function, could limit, if not prevent, damage caused by HI by ameliorating SEF.

Acknowledgements This work was supported by the Swiss National Science Foundation (project No. 32-52647-97 and 32-64110-00).

References 1 Patel J, Edwards AD: Prediction of outcome after perinatal asphyxia. Curr Opin Pediatr 1997;9:128–132. 2 Thornberg E, Thiringer K, Odeback A, Milsom I: Birth asphyxia: Incidence, clinical course and outcome in a Swedish population. Acta Paediatr 1995;84:927–932. 3 Vannucci RC, Perlman JM: Interventions for perinatal hypoxic-ischemic encephalopathy. Pediatrics 1997;100:1004–1014. 4 Levine S: Anoxic-Ischemic encephalopathy in rats. Am J Pathol 1960;36:1–17. 5 Rice JE 3rd, Vannucci RC, Brierley JB: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9: 131–141. 6 Nedelcu J, Klein MA, Aguzzi A, et al: Biphasic edema after hypoxic-ischemic brain injury in neonatal rats reflects early neuronal and late glial damage. Pediatr Res 1999;46:297–304.

Neuroprotection by Creatine in Hypoxia-Ischemia

7 Nedelcu J, Klein MA, Aguzzi A, Martin E: Resuscitative hypothermia protects the neonatal rat brain from hypoxic-ischemic injury. Brain Pathol 2000;10:61–71. 8 Palmer C, Brucklacher RM, Christensen MA, Vannucci RC: Carbohydrate and energy metabolism during the evolution of hypoxic-ischemic brain damage in the immature rat. J Cereb Blood Flow Metab 1990;10:227–235. 9 Lorek A, Takei Y, Cady EB, et al: Delayed (‘secondary’) cerebral energy failure after acute hypoxia-ischemia in the newborn piglet: Continuous 48-hour studies by phosphorus magnetic resonance spectroscopy. Pediatr Res 1994;36:699–706. 10 Rumpel H, Buchli R, Gehrmann J, et al: Magnetic resonance imaging of brain edema in the neonatal rat: A comparison of short- and longterm hypoxia-ischemia. Pediatr Res 1995;38: 113–118.

11 Penrice J, Lorek A, Cady EB, et al: Proton magnetic resonance spectroscopy of the brain during acute hypoxia-ischemia and delayed cerebral energy failure in the newborn piglet. Pediatr Res 1997;41:795–802. 12 Coimbra C, Drake M, Boris-Moller F, Wieloch T: Long-lasting neuroprotective effect of postischemic hypothermia and treatment with an anti-inflammatory/antipyretic drug. Evidence for chronic encephalopathic processes following ischemia. Stroke 1996;27:1578–1585. 13 Corbett D, Hamilton M, Colbourne F: Persistent neuroprotection with prolonged postischemic hypothermia in adult rats subjected to transient middle cerebral artery occlusion. Exp Neurol 2000;163:200–206. 14 Cady EB: Magnetic resonance spectroscopy in neonatal hypoxic-ischaemic insults. Childs Nerv Syst 2001;17:145–149. 15 Martin E, Buchli R, Ritter S, et al: Diagnostic and prognostic value of cerebral 31P magnetic resonance spectroscopy in neonates with perinatal asphyxia. Pediatr Res 1996;40:749–758.

Dev Neurosci 2002;24:382–388

387

16 Penrice J, Cady EB, Lorek A, et al: Proton magnetic resonance spectroscopy of the brain in normal preterm and term infants, and early changes after perinatal hypoxia-ischemia. Pediatr Res 1996;40:6–14. 17 Robertson RL, Maier SE, Robson CD, et al: MR line scan diffusion imaging of the brain in children. AJNR Am J Neuroradiol 1999;20: 419–425. 18 Azzopardi D, Wyatt JS, Hamilton PA, et al: Phosphorus metabolites and intracellular pH in the brains of normal and small for gestational age infants investigated by magnetic resonance spectroscopy. Pediatr Res 1989;25:440– 444. 19 Hope PL, Costello AM, Cady EB, et al: Cerebral energy metabolism studied with phosphorus NMR spectroscopy in normal and birthasphyxiated infants. Lancet 1984;ii:366–370. 20 Wagner BP, Nedelcu J, Martin E: Delayed postischemic hypothermia improves long-term behavioral outcome after cerebral hypoxiaischemia in neonatal rats. Pediatr Res 2002;51: 354–360. 21 Persky AM, Brazeau GA: Clinical pharmacology of the dietary supplement creatine monohydrate. Pharmacol Rev 2001;53:161–176. 22 Wyss M, Kaddurah-Daouk R: Creatine and creatine metabolism. Physiol Rev 2000;80: 1107–1213. 23 Balestrino M, Rebaudo R, Lunardi G: Exogenous creatine delays anoxic depolarization and protects from hypoxic damage: Dose-effect relationship. Brain Res 1999;816:124–130. 24 Michaelis T, Wick M, Fujimori H, et al: Proton MRS of oral creatine supplementation in rats. Cerebral metabolite concentrations and ischemic challenge. NMR Biomed 1999;12:309– 314. 25 Holtzman D, Togliatti A, Khait I, Jensen F: Creatine increases survival and suppresses seizures in the hypoxic immature rat. Pediatr Res 1998;44:410–414. 26 Holtzman D, Khait I, Mulkern R, et al: In vivo development of brain phosphocreatine in normal and creatine-treated rabbit pups. J Neurochem 1999;73:2477–2484. 27 Brewer GJ, Wallimann TW: Protective effect of the energy precursor creatine against toxicity of glutamate and beta-amyloid in rat hippocampal neurons. J Neurochem 2000;74:1968– 1978. 28 Klivenyi P, Ferrante RJ, Matthews RT, et al: Neuroprotective effects of creatine in a transgenic animal model of amyotrophic lateral sclerosis. Nat Med 1999;5:347–350. 29 Matthews RT, Ferrante RJ, Klivenyi P, et al: Creatine and cyclocreatine attenuate MPTP neurotoxicity. Exp Neurol 1999;157:142–149.

388

30 Matthews RT, Yang L, Jenkins BG, et al: Neuroprotective effects of creatine and cyclocreatine in animal models of Huntington’s disease. J Neurosci 1998;18:156–163. 31 Shear DA, Haik KL, Dunbar GL: Creatine reduces 3-nitropropionic-acid-induced cognitive and motor abnormalities in rats. Neuroreport 2000;11:1833–1837. 32 Sullivan PG, Geiger JD, Mattson MP, Scheff SW: Dietary supplement creatine protects against traumatic brain injury. Ann Neurol 2000;48:723–729. 33 Gruetter R, Fusch C, Martin E, Boesch C: Determination of saturation factors in 31P NMR spectra of the developing human brain. Magn Reson Med 1993;29:7–11. 34 Zaloga GP, Roberts P: Permissive underfeeding. New Horiz 1994;2:257–263. 35 Rumpel H, Nedelcu J, Aguzzi A, Martin E: Late glial swelling after acute cerebral hypoxiaischemia in the neonatal rat: A combined magnetic resonance and histochemical study. Pediatr Res 1997;42:54–59. 36 Hennig J, Nauerth A, Friedburg H: RARE imaging: A fast imaging method for clinical MR. Magn Reson Med 1986;3:823–833. 37 Dechent P, Pouwels PJ, Wilken B, et al: Increase of total creatine in human brain after oral supplementation of creatine-monohydrate. Am J Physiol 1999;277:R698–R704. 38 Stockler S, Holzbach U, Hanefeld F, et al: Creatine deficiency in the brain: A new, treatable inborn error of metabolism. Pediatr Res 1994;36:409–413. 39 Kekelidze T, Khait I, Togliatti A, Holtzman D: Brain creatine kinase and creatine transporter proteins in normal and creatine-treated rabbit pups. Dev Neurosci 2000;22:437–443. 40 Ipsiroglu OS, Stromberger C, Ilas J, et al: Changes of tissue creatine concentrations upon oral supplementation of creatine-monohydrate in various animal species. Life Sci 2001;69: 1805–1815. 41 Hagberg H, Bona E, Gilland E, Puka-Sundvall M: Hypoxia-ischaemia model in the 7-day-old rat: Possibilities and shortcomings. Acta Paediatr Suppl 1997;422:85–88. 42 Thornton JS, Ordidge RJ, Penrice J, et al: Temporal and anatomical variations of brain water apparent diffusion coefficient in perinatal cerebral hypoxic-ischemic injury: Relationships to cerebral energy metabolism. Magn Reson Med 1998;39:920–927. 43 Kay L, Nicolay K, Wieringa B, et al: Direct evidence for the control of mitochondrial respiration by mitochondrial creatine kinase in oxidative muscle cells in situ. J Biol Chem 2000;275: 6937–6944. 44 Walsh B, Tonkonogi M, Soderlund K, et al: The role of phosphorylcreatine and creatine in the regulation of mitochondrial respiration in human skeletal muscle. J Physiol 2001;537: 971–978.

Dev Neurosci 2002;24:382–388

45 Passaquin AC, Renard M, Kay L, et al: Creatine supplementation reduces skeletal muscle degeneration and enhances mitochondrial function in mdx mice. Neuromuscul Disord 2002;12:174–182. 46 Pulido SM, Passaquin AC, Leijendekker WJ, et al: Creatine supplementation improves intracellular Ca2+ handling and survival in mdx skeletal muscle cells. FEBS Lett 1998;439:357– 362. 47 Kruman II, Mattson MP: Pivotal role of mitochondrial calcium uptake in neural cell apoptosis and necrosis. J Neurochem 1999;72:529– 540. 48 O’Gorman E, Beutner G, Dolder M, et al: The role of creatine kinase in inhibition of mitochondrial permeability transition. FEBS Lett 1997;414:253–257. 49 Schlattner U, Forstner M, Eder M, et al: Functional aspects of the X-ray structure of mitochondrial creatine kinase: A molecular physiology approach. Mol Cell Biochem 1998;184: 125–140. 50 Lawler JM, Barnes WS, Wu G, et al: Direct antioxidant properties of creatine. Biochem Biophys Res Commun 2002;290:47–52. 51 Malcon C, Kaddurah-Daouk R, Beal MF: Neuroprotective effects of creatine administration against NMDA and malonate toxicity. Brain Res 2000;860:195–198. 52 Blumberg RM, Cady EB, Wigglesworth JS, et al: Relation between delayed impairment of cerebral energy metabolism and infarction following transient focal hypoxia-ischaemia in the developing brain. Exp Brain Res 1997;113: 130–137. 53 Azzopardi D, Wyatt JS, Cady EB, et al: Prognosis of newborn infants with hypoxic-ischemic brain injury assessed by phosphorus magnetic resonance spectroscopy. Pediatr Res 1989;25: 445–451. 54 Hanrahan JD, Cox IJ, Azzopardi D, et al: Relation between proton magnetic resonance spectroscopy within 18 hours of birth asphyxia and neurodevelopment at 1 year of age. Dev Med Child Neurol 1999;41:76–82. 55 Wilken B, Ramirez JM, Probst I, et al: Creatine protects the central respiratory network of mammals under anoxic conditions. Pediatr Res 1998;43:8–14. 56 Wilken B, Ramirez JM, Probst I, et al: Anoxic ATP depletion in neonatal mice brainstem is prevented by creatine supplementation. Arch Dis Child Fetal Neonatal Ed 2000;82:F224– F227. 57 Wick M, Fujimori H, Michaelis T, Frahm J: Brain water diffusion in normal and creatinesupplemented rats during transient global ischemia. Magn Reson Med 1999;42:798–802.

Adcock/Nedelcu/Loenneker/Martin/ Wallimann/Wagner

Original Paper Dev Neurosci 2002;24:389–395 DOI: 10.1159/000069044

Received: June 8, 2002 Accepted: September 24, 2002

Inhibition of nNOS and iNOS following Hypoxia-Ischaemia Improves Long-Term Outcome but Does Not Influence the Inflammatory Response in the Neonatal Rat Brain Evelyn R.W. van den Tweel a, b Cacha M.P.C.D. Peeters-Scholte a Frank van Bel a Cobi J. Heijnen b Floris Groenendaal a a Department

of Neonatology, b Laboratory for Psycho-Neuro-Immunology, Wilhelmina Children’s Hospital, University Medical Center, Utrecht, The Netherlands

Key Words Hypoxia-ischaemia W iNOS W nNOS W Inhibition W Cytokine W Heat shock protein 70 W Neonatal rat brain W Rat

Abstract In this study, we tested the hypothesis that combined inhibition of nNOS and iNOS will reduce neuronal damage and the inflammatory response induced by perinatal hypoxia-ischaemia (HI). In 12-day-old rats, HI was induced by right carotid artery occlusion followed by 90 min of 8% O2. Immediately upon reoxygenation, the rats were treated with NOS inhibitors (n = 24) or placebo (n = 24). Neuropathology was scored at 6 weeks after HI on a 4-point scale (n = 12 per group). The expression of heat shock protein 70 (HSP70) and mRNA expression for cytokines were measured 12 h after HI (n = 12 per group). Histopathological analysis showed that the ipsilateral hemisphere in the NOS inhibition group was less damaged than in the placebo group (p ! 0.05). HI induced a significant increase in HSP70 levels (p ! 0.05) in the ipsilateral hemispheres, which tended to be lower in the NOS inhibition group (p = 0.07). HI induced an increase in

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0389$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

mRNA expression for IL-1ß, TNF-· and TNF-ß, but there was no difference between the ipsi- and contralateral hemispheres. Combined inhibition of nNOS and iNOS did not induce any change in cytokine expression. We conclude that the long-term neuroprotective effects of combined nNOS and iNOS inhibition were not achieved by an altered cytokine response. Copyright © 2002 S. Karger AG, Basel

Introduction

Many interconnected pathways are involved in the pathogenesis of perinatal hypoxia-ischaemia (HI) brain injury, including excessive Ca2+ influx, overproduction of excitatory amino acids and NO release, formation of reactive oxygen species, lipid peroxidation and induction of cytokine production [1, 2]. There is increasing evidence that NO is one of the most important initiators of the neuronal damage following HI [3, 4]. The production of NO during HI and upon reoxygenation is followed by the production of peroxynitrite and induction of membrane damage due to lipid peroxidation [3, 5]. In addition, NO

Floris Groenendaal, MD, PhD Wilhelmina Children’s Hospital, University Medical Center Utrecht Department of Neonatology, Room KE.04.123.1 PO Box 85090, NL–3508 AB Utrecht (The Netherlands) Tel. +31 30 2504545, Fax +31 30 2505320, E-Mail [email protected]

may stimulate the release of cytochrome C from mitochondria leading to apoptosis [6]. In previous studies, it has been shown that nNOS or iNOS knockout mice are less susceptible to HI-induced neuronal injury [7, 8]. Moreover, in neonatal HI animal models, it has been shown that inhibition of nNOS in a pretreatment protocol can have neuroprotective effects [9–11]. Others have demonstrated that inhibition of iNOS both before HI and after HI also resulted in neuroprotection [12]. However, for clinical purpose, treatment following HI is most relevant. Hypoxic-ischaemic stress induces the release of heat shock protein 70 (HSP70), a chaperone protein assumed to protect cells by binding to denaturated proteins and prevent further denaturation or degradation [13]. HI is associated with an early induction of HSP70 in several regions of the brain and increased HSP70 expression is considered a marker of cell damage during the first days after HI [14–16]. In the hours following HI, cytokines are produced that are also thought to contribute significantly to additional neuronal damage [17, 18]. NO can be the cause of this increased cytokine production, since it has been described that NO can induce production of cytokines as TNF-·, IL-1, IL-8 and TGF-ß1 [19–21]. The induction of cytokine production can be caused by the NO-induced NF-ÎB activation following HI [22]. In 7-day-old rats, an increase in the cytokines IL-1ß, IL-6 and TNF-· and chemokines MIP-1· and MIP-2 have been demonstrated following HI [23–25]. Treatment with IL-1 receptor antagonist reduced brain damage in neonatal rats after right carotid artery occlusion with subsequent exposure to 8% O2 and in adult rats upon stroke [24, 26]. The influence of the inflammatory response induced by HI on the neuronal damage is not well established and better understanding of the role of cytokines in the neonatal brain following HI can provide further insight necessary for the development of novel intervention strategies. In the present study, we tested the hypothesis that combined inhibition of nNOS and iNOS in 12-day-old rats following HI will result in: (1) long-term neuroprotection, (2) reduction of HSP70 production as an early marker of cell damage, and (3) reduction of cytokine expression.

Material and Methods Animal Protocol The animal ethical committee of the UMC-Utrecht approved all animal experiments. A total of 52 Wistar rat pups of both sexes were used (Wistar, Central Laboratory Animal Institute, Utrecht Univer-

390

Dev Neurosci 2002;24:389–395

sity, Utrecht, The Netherlands). At postnatal day 12, pups were exposed to a hypoxic-ischaemic insult according to an adaptation of the model of Rice et al. [27]. The pups were anaesthetised with halothane (5.0% induction, 2.0% maintenance) in a mixture of N2O and O2 (1:1). The right common carotid artery was exposed and ligated followed by closure of the wound. The duration of the surgery was 5–7 min. Thereafter, the pups were allowed to recover for at least 1 h. Subsequently, the pups were placed for 90 min in an incubator with a constant temperature (37 ° C) and humidity which was perfused with a gas mixture containing 8% O2 in N2. After the hypoxic-ischaemic insult, the pups were returned to their dams. Four rats underwent anaesthesia and neck incision without carotid artery occlusion; these animals served as sham-operated controls. The animals were kept at room temperature with a light:dark cycle of 12:12 h and food and water ad libitum. Drug Treatment The 48 animals that were subjected to HI were randomly assigned to two groups (n = 24 for each group) and were treated directly after HI, i.e. within the first 2 min of reoxygenation. Group 1: 7-nitroindazole (7NI) + aminoguanidine hemisulphate (AG); 7NI (SigmaAldrich, Steinheim, Germany) at a dose of 50 mg/kg intraperitoneally in peanut oil once, plus AG (Sigma-Aldrich) at a dose of 100 mg/kg intraperitoneally in saline once every 12 h for 48 h. Group 2: placebo, equal volumes and gifts of peanut oil and saline. Animals from both groups were subsequently randomly selected for evaluation of neuroprotection at 6 weeks after HI (n = 12 per treatment group) or cytokine and HSP70 levels at 12 h after HI (n = 12 per treatment group). Histology Histological studies were performed at 6 weeks after HI in 24 rats (n = 12 from each treatment group). Following sedation with pentobarbital, rats were intracardially perfused with 4% paraformaldehyde and brains were removed, postfixed and embedded in paraffin according to standard histological procedures. Sections (8 Ìm) were cut, mounted on coated slides and stained with haematoxylin and eosin (Klinipath, Duiven, The Netherlands). The histopathology was scored, by two investigators blinded to the treatment, on a 4-point scale as previously described [28] (1 = normal; 2 = few neurons damaged, mainly in hippocampus; 3 = moderate number of neurons damaged, mainly in hippocampus and cortex, and 4 = cystic infarction). Examples are presented in figure 1. RNA and Protein Extraction Pups were sacrificed by decapitation 12 h following HI (n = 12 per group). The brain was rapidly removed, divided into two hemispheres, snap-frozen in liquid nitrogen and stored at –80 ° C. Tissue samples were homogenised in Trizol reagent (Life Technologies, Grand Island, N.Y., USA) using 1 ml of Trizol per 100 mg of tissue. RNA and protein were extracted from the different hemispheres according to manufacturer’s instructions. RNA was dissolved in ddH2O, concentration was calculated by A260 measurement and quality was assessed by electrophoresis on a 1% agarose gel. Protein was dissolved in 1% SDS and concentrations determined using BioRad protein assay reagent with BSA as standard. Western Blotting Proteins from the two hemispheres from both groups were separated by electrophoresis through a 10% SDS-PAGE gel. Proteins

van den Tweel/Peeters-Scholte/van Bel/ Heijnen/Groenendaal

Fig. 1. Representative HE-stained sections of the histological damage in the hippocampus. A Score 1, normal. B Score 2, few neurons damaged. C Score 3, moderate number of neurons damaged. D Score 4, cystic infarction.

were transferred to nitrocellulose membranes (Hybond-C, Amersham, UK) by electroblotting. Membranes were stained with specific anti-HSP70 mouse monoclonal antibody (Stressgen Biotech, Victoria, Canada) at a final dilution of 1:1,000 in TBST and developed with a horseradish peroxidase-conjugated goat antimouse secondary antibody (A3682, Sigma-Aldrich). Specific bands were visualised by using enhanced chemiluminescence detection system (ECL, Amersham, UK) and exposure to X-ray film. Autoradiographs were scanned using a GS-700 Imaging Densitometer (Bio-Rad, Hercules, Calif., USA). RNase Protection Assay The rat cytokine multiprobe template set rCK-1 (Riboquant, Pharmingen, San Diego, Calif., USA) containing probes for the cytokines IL-1·, IL-1ß, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, TNF-·, TNF-ß, and IFN-Á and housekeeping genes L32 and GAPDH was used to perform a ribonuclease protection assay according to manufacturer’s instructions. Autoradiographs were scanned using a GS-700 Imaging Densitometer (Bio-Rad,) and values for the cytokines were normalised on the basis of L32 intensities in the sample.

NOS Inhibition, Cytokines, and Neuroprotection after Hypoxia-Ischaemia

Statistics Wilcoxon signed rank test was used for comparisons of contralateral vs. ipsilateral hemisphere and the Mann-Whitney U tests for comparisons between the placebo and the 7NI + AG group, using SPSS software version 10. With expected group differences of 15%, power analysis revealed that groups of 10 animals each would be sufficient to demonstrate these differences with a ß value of 0.10 and an · value of 0.05

Results

Histology Scoring of the histopathology at 6 weeks after HI showed that there was significantly more neurodegeneration in the ipsilateral hemisphere than in the contralateral hemisphere (p ! 0.05; fig. 2). Secondly, the degree of brain damage in the ipsilateral hemisphere in the 7NI + AG group was significantly reduced compared with the ipsi-

Dev Neurosci 2002;24:389–395

391

Fig. 2. Histological evaluation of the neurological damage 6 weeks after HI in 12-day-old rats. Evaluated are ipsi- and contralateral hemispheres of placebo-treated rats (n = 12) and rats treated with 7NI + AG (n = 12). Statistical analysis showed p ! 0.05 ipsilateral vs. contralateral and p ! 0.05 ipsilateral 7NI + AG vs. ipsilateral placebo.

lateral hemisphere in the placebo group (p ! 0.05; fig. 2). In all cases with histological damage, the hippocampus appeared to be the area most affected. HSP70 HSP70 was significantly increased in the ipsilateral hemisphere compared with the contralateral hemisphere in the placebo group at 12 h following HI (p ! 0.05; fig. 3). In the 7NI + AG group, a significant increase was also observed in the ipsilateral hemisphere compared with the contralateral hemisphere (p ! 0.05). However, the levels of HSP70 in the ipsilateral hemisphere of the 7NI + AG group tended to be lower than those in the ipsilateral hemisphere of the placebo group (p = 0.07; fig. 4).

Fig. 3. Expression of HSP70 in ipsi- and contralateral hemispheres of placebo-treated rats (n = 12) and rats treated with 7NI + AG (n = 12) at 12 h after HI as detected with Western blotting. In arbitrary density units, * p ! 0.05 ipsilateral vs. contralateral.

Discussion

Cytokines In sham-operated animals, no mRNA for cytokines could be detected in the RNase protection assay. Twelve hours after HI, mRNA for the cytokines IL-1ß, TNF-· and TNF-ß were detected in both the ipsi- and contralateral hemispheres (fig. 4). Using this assay, mRNA for the cytokines IL-1·, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10 and IFN-Á could not be detected in the brain tissue at 12 h following HI. The observed amounts of mRNA for the cytokines IL-1ß, TNF-· and TNF-ß were not significantly different between both hemispheres at 12 h following HI. Moreover, there was no difference observed between the 7NI + AG and the placebo group.

392

Dev Neurosci 2002;24:389–395

In this study, we demonstrate that combined inhibition of nNOS and iNOS following HI in P12 rats resulted in long-term neuroprotection observed at 6 weeks after HI. Previous studies demonstrated neuroprotective effects of either nNOS or iNOS inhibition after HI and reoxygenation [9–12]. A recent study from our institute has combined nNOS and iNOS inhibition after HI and has shown neuroprotection at 24 h after HI in a piglet model [3]. Others demonstrated that NO is produced in the brain by activated nNOS in the early phase following HI and by activated iNOS from about 12 h of reperfusion [4, 29]. Based on these publications, it was decided to inhibit both nNOS and iNOS in the present study. Although nNOS and iNOS activities were not measured,

van den Tweel/Peeters-Scholte/van Bel/ Heijnen/Groenendaal

Fig. 4. Expression of cytokines as measured with RNase protection assays in ipsi- and contralateral hemispheres of

placebo-treated rats (n = 12) and rats treated with 7NI + AG (n = 12) at 12 h following HI. mRNA for the cytokines TNF-· (A), TNF-ß (B) and IL-1ß (C) are corrected for L32 and expressed in arbitrary units (a.u.).

others showed that the doses of 7NI and AG are sufficient to inhibit nNOS and iNOS, respectively [9, 30]. This study is the first to demonstrate long-term neuroprotection with combined inhibition of both nNOS and iNOS for 48 h after HI. Inhibition of NO may have exerted its

neuroprotective effects through different mechanisms. First, reduced NO formation, and thereby reduced peroxynitrite production may have decreased lipid peroxidation of cell membranes [5]. Secondly, NO can stimulate activation of p38-MAP kinase, cytochrome C release

NOS Inhibition, Cytokines, and Neuroprotection after Hypoxia-Ischaemia

Dev Neurosci 2002;24:389–395

393

from mitochondria, and apoptosis [6] or NO-mediated stimulation of calpains can result in cell death [31]. In the present study, the exact mechanisms cannot be ascertained. We also showed that HI resulted in increased production of HSP70 in the brain and that this up-regulation of HSP70 tended to be reduced by inhibition of nNOS and iNOS. Others have also observed an up-regulation of HSP70 in neonatal rat brain following HI. The level of HSP70 correlated to the extent of brain damage [14–16]. It has been suggested that HSP70 has anti-apoptotic effects since it can inhibit caspase-3, Apaf-1, NF-ÎB, iNOS and cytokines [13, 32]. In contrast with this finding, HSP70 over-expressing transgenic mice were not protected against HI-induced neuronal cell death [33]. Based on these findings and our present observations, we consider HSP70 production an early marker of neuronal injury. Finally, we detected an increased expression of mRNA for the cytokines TNF-·, TNF-ß and IL-1ß at 12 h after HI. In a pilot study, we observed that cytokine mRNA expression started to increase 6 h after HI, and peaked at 12 h after HI. The cytokine mRNA expression was almost abolished 48 h after HI. Previous studies in P7 rats have shown an up-regulation of the cytokines IL-1ß, IL-6 and TNF-· with RT-PCR and increased bioactivity of IL-1 and IL-6 with a maximum at 4–12 h after the insult [23– 25]. It is possible that we did not detect mRNA for IL-6 in this study because RT-PCR is more sensitive than the RNase protection assay. In the present study, the increase in cytokine mRNA expression was observed in both the ipsi- and contralateral hemispheres. This is in line with observations by others [23–25] and may suggest that hypoxia alone is already sufficient to induce cytokine mRNA expression in the brain. Despite marked differences in neuronal damage between the 7NI + AG and placebo group, no differences in cytokine mRNA expression between the groups were found. Taken together, these data suggest that the observed expression levels of the cytokines TNF-·, TNF-ß and IL-1ß as such play a minor role in the induction of the neuronal damage. In line with this observation, there is evidence that IL-1ß has neurotoxic effects in an injured brain, whereas this cytokine has no effects when administered to healthy brain tissue [34–36]. However, it has also been shown in previous studies that inhibition of TNF-· or administration of IL-1 receptor antagonist can result in reduced neuronal damage in brain tissue subjected to HI [24, 26]. Therefore, it may be possible that in the present study, the cytokines add to the neuronal damage in the ipsilateral hemisphere, but cause no harm to the contralateral hemisphere.

394

Dev Neurosci 2002;24:389–395

In contrast with studies demonstrating that NO induces production of cytokines [19–21], in our study, inhibition of nNOS and iNOS did not influence cytokine expression at 12 h following HI, indicating that NO production is not directly linked to brain cytokine production 12 h after HI in neonatal rats. In line with this, it is described that in an astrocyte culture inhibition of NOS following hypoxia/reoxygenation did not reduce the IL-6 production [37]. The long-term neuroprotection obtained by combined nNOS and iNOS inhibition appears to be achieved by other mechanisms than a reduced cytokine expression. In the present study, 12-day-old rats were used, since the developing cerebral cortex of the rat at that stage is comparable with that of the full-term human neonate [38]. In conclusion, inhibition of both nNOS and iNOS with 7NI + AG following HI results in long-term neuroprotection. At 12 h after HI, HSP70 is induced in the ipsilateral hemisphere indicating neuronal damage, which tends to be lower in the NOS inhibition-treated group. The expression of the cytokines TNF-·, TNF-ß and IL-1ß is up-regulated in both the ipsi- and contralateral hemispheres following HI and is not inhibited by combined administration of nNOS and iNOS inhibitors. The relation between NO production and cytokine expression following HI needs to be further examined.

Acknowledgement This study has been funded by the University Medical Center Utrecht.

van den Tweel/Peeters-Scholte/van Bel/ Heijnen/Groenendaal

References 1 Peeters C, van Bel F: Pharmacotherapeutical reduction of post-hypoxic-ischemic brain injury in the newborn. Biol Neonate 2001;79:274– 280. 2 Volpe JJ: Perinatal brain injury: From pathogenesis to neuroprotection. Ment Retard Dev Disabil Res Rev 2001;7:56–64. 3 Peeters-Scholte C, Koster J, Veldhuis W, Van Den TE, Zhu C, Kops N, Blomgren K, Bar D, Buul-Offers S, Hagberg H, Nicolay K, van Bel F, Groenendaal F: Neuroprotection by selective nitric oxide synthase inhibition at 24 hours after perinatal hypoxia-ischemia. Stroke 2002;33:2304–2310. 4 Samdani AF, Dawson TM, Dawson VL: Nitric oxide synthase in models of focal ischemia. Stroke 1997;28:1283–1288. 5 Numagami Y, Zubrow AB, Mishra OP, Delivoria-Papadopoulos M: Lipid free radical generation and brain cell membrane alteration following nitric oxide synthase inhibition during cerebral hypoxia in the newborn piglet. J Neurochem 1997;69:1542–1547. 6 Cheng A, Chan SL, Milhavet O, Wang S, Mattson MP: p38 MAP kinase mediates nitric oxide-induced apoptosis of neural progenitor cells. J Biol Chem 2001;276:43320–43327. 7 Ferriero DM, Holtzman DM, Black SM, Sheldon RA: Neonatal mice lacking neuronal nitric oxide synthase are less vulnerable to hypoxicischemic injury. Neurobiol Dis 1996;3:64–71. 8 Iadecola C, Zhang F, Casey R, Nagayama M, Ross ME: Delayed reduction of ischemic brain injury and neurological deficits in mice lacking the inducible nitric oxide synthase gene. J Neurosci 1997;17:9157–9164. 9 Higuchi Y, Hattori H, Kume T, Tsuji M, Akaike A, Furusho K: Increase in nitric oxide in the hypoxic-ischemic neonatal rat brain and suppression by 7-nitroindazole and aminoguanidine. Eur J Pharmacol 1998;19;342:47– 49. 10 Ishida A, Trescher WH, Lange MS, Johnston MV: Prolonged suppression of brain nitric oxide synthase activity by 7-nitroindazole protects against cerebral hypoxic-ischemic injury in neonatal rat. Brain Dev 2001;23:349–354. 11 Muramatsu K, Sheldon RA, Black SM, Tauber M, Ferriero DM: Nitric oxide synthase activity and inhibition after neonatal hypoxia ischemia in the mouse brain. Brain Res Dev Brain Res 2000;123:119–127. 12 Tsuji M, Higuchi Y, Shiraishi K, Kume T, Akaike A, Hattori H: Protective effect of aminoguanidine on hypoxic-ischemic brain damage and temporal profile of brain nitric oxide in neonatal rat. Pediatr Res 2000;47:79–83. 13 Garrido C, Gurbuxani S, Ravagnan L, Kroemer G: Heat shock proteins: Endogenous modulators of apoptotic cell death. Biochem Biophys Res Commun 2001;286:433–442.

NOS Inhibition, Cytokines, and Neuroprotection after Hypoxia-Ischaemia

14 Blumenfeld KS, Welsh FA, Harris VA, Pesenson MA: Regional expression of c-fos and heat shock protein-70 mRNA following hypoxiaischemia in immature rat brain. J Cereb Blood Flow Metab 1992;12:987–995. 15 Munell F, Burke RE, Bandele A, Gubits RM: Localization of c-fos, c-jun, and HSP70 mRNA expression in brain after neonatal hypoxiaischemia. Brain Res Dev Brain Res 1994;77: 111–121. 16 Gilby KL, Armstrong JN, Currie RW, Robertson HA: The effects of hypoxia-ischemia on expression of c-Fos, c-Jun and HSP70 in the young rat hippocampus. Brain Res Mol Brain Res 1997;48:87–96. 17 Oygur N, Sonmez O, Saka O, Yegin O: Predictive value of plasma and cerebrospinal fluid tumour necrosis factor-alpha and interleukin1beta concentrations on outcome of full-term infants with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal Ed 1998; 79:F190–F193. 18 Savman K, Blennow M, Gustafson K, Tarkowski E, Hagberg H: Cytokine response in cerebrospinal fluid after birth asphyxia. Pediatr Res 1998;43:746–751. 19 Villarete LH Remick DG: Nitric oxide regulation of IL-8 expression in human endothelial cells. Biochem Biophys Res Commun 1995; 211:671–676. 20 Vodovotz Y, Chesler L, Chong H, Kim SJ, Simpson JT, DeGraff W, Cox GW, Roberts AB, Wink DA, Barcellos-Hoff MH: Regulation of transforming growth factor beta1 by nitric oxide. Cancer Res 1999;59:2142–2149. 21 Marcinkiewicz J, Grabowska A, Chain B: Nitric oxide up-regulates the release of inflammatory mediators by mouse macrophages. Eur J Immunol 1995;25:947–951. 22 O’Neill LA Kaltschmidt C: NF-kappa B: A crucial transcription factor for glial and neuronal cell function. Trends Neurosci 1997;20:252– 258. 23 Bona E, Andersson AL, Blomgren K, Gilland E, Puka-Sundvall M, Gustafson K, Hagberg H: Chemokine and inflammatory cell response to hypoxia-ischemia in immature rats. Pediatr Res 1999;45:500–509. 24 Hagberg H, Gilland E, Bona E, Hanson LA, Hahin-Zoric M, Blennow M, Holst M, McRae A, Soder O: Enhanced expression of interleukin (IL)-1 and IL-6 messenger RNA and bioactive protein after hypoxia-ischemia in neonatal rats. Pediatr Res 1996;40:603–609. 25 Szaflarski J, Burtrum D, Silverstein FS: Cerebral hypoxia-ischemia stimulates cytokine gene expression in perinatal rats. Stroke 1995;26: 1093–1100. 26 Stroemer RP Rothwell NJ: Cortical protection by localized striatal injection of IL-1ra following cerebral ischemia in the rat. J Cereb Blood Flow Metab 1997;17:597–604.

27 Rice JE III, Vannucci RC, Brierley JB: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9: 131–141. 28 Geddes R, Vannucci RC, Vannucci SJ: Delayed cerebral atrophy following moderate hypoxia-ischemia in the immature rat. Dev Neurosci 2001;23:180–185. 29 Iadecola C, Zhang F, Xu S, Casey R, Ross ME: Inducible nitric oxide synthase gene expression in brain following cerebral ischemia. J Cereb Blood Flow Metab 1995;15:378–384. 30 Nagayama M, Zhang F, Iadecola C: Delayed treatment with aminoguanidine decreases focal cerebral ischemic damage and enhances neurologic recovery in rats. J Cereb Blood Flow Metab 1998;18:1107–1113. 31 Volbracht C, Fava E, Leist M, Nicotera P: Calpain inhibitors prevent nitric oxide-triggered excitotoxic apoptosis. Neuroreport 2001;12: 3645–3648. 32 Feinstein DL, Galea E, Aquino DA, Li GC, Xu H, Reis DJ: Heat shock protein 70 suppresses astroglial-inducible nitric-oxide synthase expression by decreasing NFkappaB activation. J Biol Chem 1996;271:17724–17732. 33 Lee JE, Yenari MA, Sun GH, Xu L, Emond MR, Cheng D, Steinberg GK, Giffard RG: Differential neuroprotection from human heat shock protein 70 overexpression in in vitro and in vivo models of ischemia and ischemia-like conditions. Exp Neurol 2001;170:129–139. 34 Stroemer RP Rothwell NJ: Exacerbation of ischemic brain damage by localized striatal injection of interleukin-1beta in the rat. J Cereb Blood Flow Metab 1998;18:833–839. 35 Pringle AK, Niyadurupola N, Johns P, Anthony DC, Iannotti F: Interleukin-1beta exacerbates hypoxia-induced neuronal damage, but attenuates toxicity produced by simulated ischaemia and excitotoxicity in rat organotypic hippocampal slice cultures. Neurosci Lett 2001;305:29–32. 36 Loddick SA Rothwell NJ: Neuroprotective effects of human recombinant interleukin-1 receptor antagonist in focal cerebral ischaemia in the rat. J Cereb Blood Flow Metab 1996;16: 932–940. 37 Maeda Y, Matsumoto M, Hori O, Kuwabara K, Ogawa S, Yan SD, Ohtsuki T, Kinoshita T, Kamada T, Stern DM: Hypoxia/reoxygenation-mediated induction of astrocyte interleukin 6: A paracrine mechanism potentially enhancing neuron survival. J Exp Med 1994;180: 2297–2308. 38 Romijn HJ, Hofman MA, Gramsbergen A: At what age is the developing cerebral cortex of the rat comparable to that of the full-term newborn human baby? Early Hum Dev 1991;26: 61–67.

Dev Neurosci 2002;24:389–395

395

Original Paper Received: June 8, 2002 Accepted: September 24, 2002

Dev Neurosci 2002;24:396–404 DOI: 10.1159/000069045

Effects of Selective Nitric Oxide Synthase Inhibition on IGF-1, Caspases and Cytokines in a Newborn Piglet Model of Perinatal Hypoxia-Ischaemia Cacha Peeters-Scholte a Johanna Koster a Evelyn van den Tweel a Klas Blomgren d Nicole Hamers c Changlian Zhu d Sylvia van Buul-Offers c Henrik Hagberg d Frank van Bel a Cobi Heijnen b Floris Groenendaal a Departments of a Neonatology, b Psycho-Neuro-Immunology and c Paediatric Endocrinology, Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht, The Netherlands; d Perinatal Center, Institute of Physiology and Pharmacology, Göteborg University, Göteborg, Sweden

Key Words Perinatal hypoxia-ischaemia W Reperfusion injury W Caspases W IGF-1 W In situ nick end labelling W Cytokines W Pig

Abstract Selective inhibition of neuronal and inducible nitric oxide synthase (NOS) with 2-iminobiotin previously showed a reduction in brain cell injury. In the present study, we investigated the effects of 2-iminobiotin treatment on insulin-like growth factor-1 (IGF-1) expression, caspase activity and cytokine expression in a newborn piglet model of perinatal hypoxia-ischaemia. Newborn piglets were subjected to 1 h of hypoxia-ischaemia and were treated intravenously with vehicle or 2-iminobiotin. Vehicle-treated piglets showed reduced IGF-1 mRNA expression and increased caspase-3 activity and DNA fragmentation. 2-Iminobiotin treatment, administered immediately upon reperfusion, prevented these observations. No differences in caspase-8 and -9 activity and cytokine [interleukin (IL)-1·/ß, IL-6, tumour necrosis factor

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0396$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

(TNF)-·, transforming growth factor (TGF)-ß] mRNA expression were demonstrated between vehicle- and 2-iminobiotin-treated piglets at 24 h following hypoxia-ischaemia. IGF-1 mRNA correlated inversely with caspase-3 and transferase-mediated dUTP-biotin in situ nick end labelling score in the cortex, but positively with caspase8. Cytokine mRNA did not correlate with IGF-1 mRNA, caspase-3 activity or DNA fragmentation. The present results indicate that the previously demonstrated neuroprotective effect of 2-iminobiotin treatment after perinatal hypoxia-ischaemia coincided with a preservation of the endogenous IGF-1 production and reduced caspase3 activity, but not with a significant decrease in cytokine production. Copyright © 2002 S. Karger AG, Basel

Introduction

Selective inhibition of neuronal and inducible nitric oxide synthase (NOS) with 2-iminobiotin has previously been shown to reduce neuronal cell damage and the

Floris Groenendaal, MD, PhD Wilhelmina Children’s Hospital/University Medical Center Utrecht Department of Neonatology; Room KE.04.123.1 PO Box 85090, NL–3508 AB Utrecht (The Netherlands) Tel. +31 30 2504545, Fax +31 30 2505320, E-Mail [email protected]

amount of transferase-mediated dUTP-biotin in situ nick end labelling (TUNEL)-positive neuronal cells following perinatal hypoxia-ischaemia and reperfusion in piglets [1]. Reduced amounts of oxidative stress, as demonstrated by a reduction in the nitration of tyrosine, provided additional evidence for the proposed mechanism of action of 2-iminobiotin via peroxynitrite inhibition. However, reduction of growth factors and expression of caspase-induced programmed cell death, possibly activated via the cytokine pathway, might also be important pathways in the determination of cell death following hypoxia-ischaemia. Insulin-like growth factor-1 (IGF-1) is an important anabolic pleiotropic factor, essential for the postnatal brain development. In a neonatal rat model of hypoxiaischaemia, it was shown that neuronal IGF-1 and IGFbinding protein (IGFBP)-5 mRNA decreased immediately after onset of hypoxia-ischaemia [2]. The observed decrease was concurrent with an increase in the number of apoptotic cells and contributed to the increased vulnerability of neuronal cells following hypoxia-ischaemia. In the same model of slightly older rats, it was shown that endogenous IGF-1 mRNA was up-regulated in the brain of vehicle-treated rats by 3 days after transient hypoxiaischaemia compared with control rats [3]. Furthermore, exogenous IGF-1, injected in the cerebral spinal fluid of adult rats following a similar hypoxic-ischaemic insult, reduced the amount of neuronal loss [4]. It was demonstrated that the IGFBPs are involved in altering the bioavailability and effect of IGF-1 and/or IGF-2 in the late phase of neuronal recovery and repair, suggesting a modulating role of IGFBPs in hypoxic-ischaemic brain injury [5]. According to these data, preservation of endogenous IGF-1 levels might be of benefit in the treatment of perinatal hypoxia-ischaemia. Caspases, which are cystein proteases, play a crucial role in the development of programmed cell death. Activation of caspase-3 appears to be a key event in the initiation of this process [6]. Programmed cell death can be initiated by at least three major pathways. The intrinsic caspase pathway is activated when cytochrome C leaks from the mitochondrion into the cytosol and complexes with APAF-1. The complex can activate caspase-9, which on its turn activates caspase-3. In the extrinsic caspase pathway, caspase-8 is activated via the Fas or tumour necrosis factor (TNF)-· membrane receptors and activates on its turn caspase-3. The third pathway through which programmed cell death can occur is caspase independent and is mediated through apoptosis-inducing factor [7].

Selective NOS Inhibition, IGF-1, Caspases, and Cytokines after Perinatal Asphyxia

Until now, no studies have been available that correlate the effect of IGF-1 mRNA expression with caspase activation in a neonatal animal model of hypoxia-ischaemia. It has previously been demonstrated that chemokines and pro-inflammatory cytokines, predominantly TNF-· and interleukin (IL)-1ß, play a role in the inflammatory response in the immature brain and might contribute to the progression of perinatal brain injury for instance through the above-mentioned caspase activation [8, 9]. Since IGF-1 and pro-inflammatory cytokines are expressed in close proximity of each other after hypoxicischaemic events in vivo, a relationship was suggested [10, 11]. In the present study, we tested the hypothesis that 2iminobiotin, a selective NOS inhibitor, could prevent the posthypoxic-ischaemic decrease in IGF-1 mRNA expression and the increase in caspase activity and cytokine expression. Furthermore, we aimed at getting more insight in the relationship between the neuroprotective effects of 2-iminobiotin and IGF-1 mRNA expression, caspase activation, DNA fragmentation and cytokine expression.

Materials and Methods Experimental Design Twenty-three 1- to 3-day-old Dutch store piglets were anaesthetised by inhalation of a N2O/O2/isoflurane mixture (75/21/4%) and were subsequently intubated and ventilated. Venous lines were inserted for glucose administration (5%; 5 ml/kg/h) and drug infusion. An arterial line was inserted in the right femoral artery for continuous registration of blood pressure and blood gas analysis. After regaining haemodynamic stability, the piglets were subjected to 1 h of hypoxia-ischaemia by occlusion of both carotid arteries with hypoxia, controlled by frequent 31P-MRS measurements as previously reported [12]. Briefly, 31P-MR spectra were made continuously during hypoxia-ischaemia until severe energy failure was evident, defined as a reduction in the phosphocreatine/inorganic phosphate (PCr/Pi) ratio to at least 30% of baseline values for minimally 45 min [13]. Peak amplitudes of PCr and Pi were determined with timedomain-fitting procedures using prior knowledge (Magnetic Resonance User Interface 97.2, Barcelona, Spain [14]). Metabolite ratios of PCr/Pi were calculated. Rectal temperature was maintained between 38.5–39.5 ° C during the surgical procedure and the hypoxicischaemic insult with a heating lamp or water blanket. Directly upon reperfusion and reoxygenation, 12 piglets were vehicle-treated (normal saline) and 11 piglets received 2-iminobiotin treatment (SigmaAldrich, St. Louis, Mo., USA; 0.2 mg/kg i.v. immediately upon reperfusion and a repeated dose every 4 h). Piglets received buprenorphine (0.03 mg/kg bid) for analgesia, starting from 3 h after hypoxia-ischaemia. Five additional piglets served as sham-operated controls.

Dev Neurosci 2002;24:396–404

397

Caspase Activity At 24 h after hypoxia-ischaemia, piglets were killed with an overdose of pentobarbital and the brain was rapidly perfused with normal saline to remove the excess of blood. Right hemispheres were dissected into cortex, hippocampus and striatum and were snap frozen in liquid nitrogen and stored at –80 ° C until further analysis of caspase activity and cytokine expression. Left hemispheres were used for determination of cell death and IGF-1 mRNA expression, as described later on. Caspase-3, -8 and -9 activities were measured in the cortex and striatum as substrate-specific cleavage (Ac-DEVDAMC/Ac-IETD-AMC/Ac-LEHD-AMC) and were expressed as picomole AMC released per milligram protein per minute according to previous reports [15]. The cortex consists of both parietal and temporal cortex and the value displayed indicates the mean caspase activity of these brain parts. Cytokine Expression RNA was isolated from brain cells in the cortex and hippocampus of right hemispheres using Tripure (Roche, Palo Alto, Calif., USA) according to the manufacturer’s instructions. For determination of the cytokine mRNA expression in the cortex and hippocampus, RNase protection assays (Pharmingen, San Diego, Calif., USA) were performed according to the manufacturer’s instructions using a custom-made multiprobe template especially designed for porcine RNA. The template consisted of probes for IL-1·, IL-1ß, IL-6, TNF-·, TGF-ß2, RANTES, L32 and GAPDH. Autoradiographs were scanned using a GS-700 Imaging Densitometer (Bio-Rad, Hercules, Calif., USA) and intensities of cytokines were normalised on the basis of L32 expression and were expressed in arbitrary units. Furthermore, the intensities of L32 and GAPDH, the housekeeping genes, were measured in all groups. IGF-1 in situ Hybridisation and DNA Fragmentation Left-brain halves from all piglets were taken out after intracardial perfusion with 4% phosphate-buffered formaldehyde and processed for light microscopy. The hemisphere was cut into 3 parts using a leaden mall and was processed for histology and embedded in paraffin. Standardised coronal sections (8 Ìm) through the striatum, cortex and hippocampus were cut [1] and mounted onto 2% aminopropyl-tri-ethoxy-silane-coated glass slides. DNA fragmentation was assessed with terminal deoxynucleotidyl TUNEL using an in situ detection kit (ApopTag peroxidase kit; Intergen, Purchase, N.Y., USA). In closely adjacent slides at maximally 10 Ìm distance, nonradioactive in situ hybridisation for IGF-1 mRNA was performed on 5 Ìm paraffin sections as described previously [16]. The degree of DNA fragmentation in the cortex, striatum and hippocampus and IGF-1 mRNA expression in the cortex and hippocampus was scored on a 4-point scale. Score 1 indicated that ! 25% of the neurons were stained; score 2: 25–50%; score 3: 50–75%, and score 4 indicated that 1 75% of the neurons were stained. When no viable neurons were seen anymore, a score of 0 was assigned. The value of the hippocampus represents the mean score in cornu ammonis 1–4 and the dentate gyrus. Statistical Analysis Data are presented as mean B SEM. Sham-operated piglets were used to describe the normal situation, but were not used for statistical analysis. Vehicle- and 2-iminobiotin-treated groups were analysed with Mann-Whitney U tests for continuous data and cross-tabulation (Gamma test) for ordinal data. Assessment of correlation was per-

398

Dev Neurosci 2002;24:396–404

formed for the vehicle- and 2-iminobiotin-treated piglets using the Spearman correlation test for ordinal data. A p value ! 0.05 was determined statistically significant.

Results

IGF-1 in situ Hybridisation In the sham-operated piglets, IGF-1 mRNA was clearly demonstrated using in situ hybridisation in the Purkinje cells of the cerebellum, in layer III and V of the parietal cortex, in the pyramidal layer of the hippocampus and dentate gyrus, and in the piriform cortex (fig. 1A–D). IGF-1 mRNA was significantly preserved in the cortex and hippocampus of 2-iminobiotin-treated piglets compared with the reduction observed in the vehicle-treated piglets at 24 h after hypoxia-ischaemia (p ! 0.05; table 1). For representative examples of IGF-1 mRNA expression in the piriform cortex and hippocampus in vehicle- and 2-iminobiotin-treated piglets, see figure 1E–H. Caspase Activity Caspase-3 activity was very low in sham-operated piglets. It was significantly increased in the cortex and striatum of vehicle-treated piglets, but 2-iminobiotin-treated piglets prevented this increase following hypoxia-ischaemia (p ! 0.05; table 1). Caspase-8 and -9 were both present in the cortex and striatum of sham-operated animals. Hypoxia-ischaemia did not affect the levels of caspase-8 and -9 activity in vehicle- and 2-iminobiotintreated piglets in the investigated brain regions (table 1). DNA Fragmentation The degree of DNA fragmentation was low in the cortex and hippocampus of sham-operated piglets, but was increased in the cortex and hippocampus of vehicletreated piglets (table 1). 2-Iminobiotin treatment following hypoxia-ischaemia prevented the increase in DNA fragmentation in the cortex and hippocampus (p ! 0.05), but not in the striatum (p = 0.079), as compared with the vehicle-treated piglets. Cytokine Expression In the present study, no increase in IL-1·, IL-1ß mRNA and RANTES were found in any group at 24 h following hypoxia-ischaemia using RNase protection assays (data not shown). IL-6, TNF-· and TGF-ß mRNA tended to be preserved in the cortex and hippocampus of 2-iminobiotin-treated piglets compared with vehicletreated piglets at 24 h after hypoxia-ischaemia (fig. 2A, B),

Peeters-Scholte et al.

Fig. 1. Normal distribution of IGF-1 mRNA

expression, as measured with in situ hybridisation, in sham-operated newborn piglets in Purkinje cells of the cerebellum (A), layer III and V of the parietal cortex (B), pyramid layer of the hippocampus (C) and in the piriform cortex (D). Representative example of decreased IGF-1 mRNA expression in the hippocampus (E) and the piriform cortex (F) of vehicle-treated piglets and preserved IGF-1 mRNA expression in the hippocampus (G) and the piriform cortex (H) of 2-iminobiotin-treated piglets at 24 h following hypoxia-ischaemia.

Selective NOS Inhibition, IGF-1, Caspases, and Cytokines after Perinatal Asphyxia

Dev Neurosci 2002;24:396–404

399

Table 1. IGF-1 mRNA, caspase

activity and DNA fragmentation at 24 h after hypoxia-ischaemia

IGF-1 mRNA Caspase-3 activity Caspase-8 activity Caspase-9 activity TUNEL

cortex hippocampus cortex striatum cortex striatum cortex striatum cortex striatum hippocampus

Vehicle

2-Iminobiotin

Sham

1.0B0.4 1.0B0.4 5.0B1.4 5.3B1.7 3.4B0.7 3.5B0.6 0.9B0.1 0.8B0.1 2.9B0.4 2.6B0.4 2.8B0.5

2.9B0.6* 2.9B0.6* 0.3B0.1* 1.6B0.8* 4.4B0.7 4.6B0.5 0.9B0.1 0.9B0.1 1.9B0.4* 2.0B0.6 1.3B0.3*

3.6B0.3 3.4B0.2 0.2B0.0 0.3B0.0 4.5B0.8 3.9B1.0 0.8B0.2 0.7B0.2 1.6B0.2 2.2B0.5 1.0B0.0

IGF-1 mRNA and TUNEL are scored on a 4-point scale. Caspase activity is expressed in picomole AMC released per milligram protein per min. Although IGF-1 and TUNEL data are ordinal data, mean B SEM is reported for reasons of clarity. * p ! 0.05 versus vehicletreated piglets.

There is a significant, inverse correlation between IGF-1 mRNA and caspase-3 activity (r = 0.71; p ! 0.01) and TUNEL score (r = 0.81; p ! 0.005). A positive correlation was present between IGF-1 mRNA and caspase-8 activity (r = 0.78; p ! 0.005). No correlation with caspase-9 activity was present. As expected, caspase-3 activity correlated positively with the TUNEL score (r = 0.69; p ! 0.005) and caspase-8 activity correlated inversely with the TUNEL score (r = 0.68; p ! 0.005), whereas caspase-9 activity did not correlate with the TUNEL score. Cytokine mRNA, as assessed with RNase protection assays, did not correlate with IGF-1 mRNA, caspase activity or DNA fragmentation at 24 h after hypoxia-ischaemia.

Fig. 2. IL-6, TNF-· and TGF-ß mRNA, as measured with RNase

protection assays (normalised to percentage of L32 and expressed in arbitrary units) in the cortex (A) and hippocampus (B) of vehicletreated piglets, 2-iminobiotin-treated piglets and sham-operated piglets at 24 h following hypoxia-ischaemia. No significant differences were detected between vehicle- and 2-iminobiotin-treated groups.

but no significant differences could be demonstrated. The intensities of the housekeeping genes, L32 and GAPDH, were equally high in vehicle- and 2-iminobiotin-treated piglets and sham-operated piglets (data not shown).

400

Dev Neurosci 2002;24:396–404

Discussion

The present study showed that neuronal IGF-1 mRNA expression was severely decreased in both cortex and hippocampus of vehicle-treated piglets at 24 h after hypoxiaischaemia. However, in piglets treated with 2-iminobiotin upon reperfusion, the endogenous IGF-1 mRNA production was preserved in the cortex and hippocampus. Since the mRNA expression of the housekeeping genes L32 and GAPDH was equally high among treatment groups, the preservation in IGF-1 mRNA in 2-iminobiotin-treated piglets is not likely to be caused by an altered, overall mRNA expression following hypoxia-ischaemia.

Peeters-Scholte et al.

IL-1 IL-1R

Nucleus

MEK1

DNA fragmentation AIF

TNF-a

FAS-L

ICAD

CAD

IGF-1R

PI3K

IGF-1

Akt

Caspase-3

TNFR

FAS

Transcription factors (NF-κB)

Caspase-8

Caspase-9

Procaspase -8

Procaspase -9

cyt C APAF-1 bid

bcl-2

Mitochon drion

bax bad

Fig. 3. Schematic presentation between relation of IGF-1, caspases and cytokines following hypoxia-ischaemia. AIF =

Apoptosis-inducing factor; Akt = protein kinase B; APAF-1 = mammalian homologue of ced-4; bid, bad, bax = proapoptotic proteins; bcl-2 = anti-apoptotic proteins; cyt C = cytochrome C; FAS-L = FAS ligand; (I)CAD = (inhibitor of) caspase-activated DNAse; IGF-1R = insulin-like growth factor-1 receptor; IL-1R = interleukin-1 receptor; MEK1 = mitogen-activated protein kinase/extracellular signal-related kinase kinase 1; NF-ÎB = nuclear factor ÎB; PI3K = phosphatidylinositol-3 kinase; TNFR = tumour necrosis factor-· receptor.

Furthermore, it was shown that the IGF-1 mRNA expression in this newborn piglet model at 24 h following hypoxia-ischaemia correlated with the activity of caspase-3 and with the TUNEL score. This observation is in line with an earlier study in 7day-old rats, reporting that IGF-1 mRNA expression was reduced directly after cerebral ischaemia and was concurrent with an increase in the number of apoptotic neurons [2]. Although the presence of TUNEL-positive cells is not merely a characteristic of cells subjected to programmed cell death and can also be present during necrotic or oxidative stress-induced cell death [6], the presence of an increased activity of caspase-3 together with the presence of TUNEL-positive cells will enhance the chance on programmed cell death instead of necrotic cell death. Indeed, it has previously been suggested that the IGF1-mediated neuroprotection is mediated through an antiapoptotic effect [17]. More recently, it has been shown in cell-signalling studies following ischaemia that the antiapoptotic effects of IGF-1 might be regulated via mitogen-activated protein kinase and phosphatidyl-3-inositol

(PI3k)/Akt pathways [18], after which bcl-2 expression, an anti-apoptotic protein, is promoted and caspase-3 activation is inhibited [19]. For a schematic summary of the pathways involved, see figure 3. The inverse correlation between IGF-1 expression and caspase-3 activity observed in this study supports the hypothesis that IGF-1 can inhibit caspase-3 activation. An additional role of IGF-1 has been postulated in providing trophic support for neurons after hypoxia-ischaemia and during development. Embryonic cortical neurons showed a marked increase in survival time in vitro, when co-cultured with (precursor or differentiated) oligodendrocytes, which are able to produce IGF-1 [20]. In experimental studies in 21-day-old rats exposed to hypoxia-ischaemia, it was confirmed that IGF-1 mRNA was expressed by reactive microglia in regions of delayed neuronal death from 3 days after hypoxia-ischaemia onwards. So, it might be concluded that IGF-1 mRNA can be produced by glial cells after the hypoxic-ischaemic insult in order to provide neurotrophic support to neurons. In the present study, we studied the piglets for only 24 h after

Selective NOS Inhibition, IGF-1, Caspases, and Cytokines after Perinatal Asphyxia

Dev Neurosci 2002;24:396–404

401

hypoxia-ischaemia. Therefore, we were not able to look at the secondary up-regulation of IGF-1 mRNA in the vehicle-treated piglets. In contrast to IGF-1, caspase-3 activity was significantly increased in the brains of the vehicle-treated piglets compared with the 2-iminobiotin-treated piglets at 24 h upon reperfusion. This observation is in line with the findings in a 7-day-old rat model of perinatal hypoxiaischaemia, where caspase-3 activity and the caspase-3 protein expression was maximally increased at 24 h following hypoxia-ischaemia [15]. Following focal cerebral ischaemia in adult rats, caspase-3 mRNA, as assessed with RT-PCR, was also maximally up-regulated at 24 h following the insult. In a permanent focal ischaemia model in 8-week-old mice, however, caspase-3 protein showed a biphasic expression with a maximal increase at 1 and 12 h following middle cerebral artery occlusion (MCAO) [21]. Although we did not perform a time profile of caspase-3 activation to investigate the time point of maximal activation, we did see a significant elevation of caspase-3 activity at 24 h following hypoxia-ischaemia in the vehicle-treated piglets. Caspase-8 and -9 were not elevated after perinatal hypoxia-ischaemia in vehicle- and 2-iminobiotin-treated piglets and were as high as in sham-operated piglets at 24 h following hypoxia-ischaemia. Recently, it has been shown that caspase-9 gene expression and protease activity are developmentally up-regulated in newborn rats and guinea pigs [22, 23]. It was suggested that caspase-9 plays an essential role in neuronal apoptosis during the perinatal brain development and might explain why detectable levels of caspase-9 were found in sham-operated animals in our study. It is more difficult to understand the mechanisms through which caspase-8 activity was as high in sham-operated piglets as in the vehicle-treated piglets, while caspase-3 was virtually absent and DNA fragmentation was minimal in the cortex and hippocampus of shamoperated piglets. A plausible explanation might be that caspase-8 was activated in the sham-operated piglets for instance through surgery with the use of general anaesthesia with small amounts of oxidative stress [24], but that it is a reversible process that does not have to lead to caspase-3 activation or programmed cell death. Both caspase-8 as well as caspase-9 have been shown to be up-regulated in experimental animal models following cerebral ischaemia. Caspase-8 expression was increased at 24 h following ischaemia in adult rats and mice [25, 26] and reduced in transgenic mice, overexpressing the cytosolic copper/zinc superoxide dismutase, suggesting a positive relation with the amount of reactive oxygen species

402

Dev Neurosci 2002;24:396–404

production [26]. This cannot explain the observation that caspase-8 was equally high in vehicle- and 2-iminobiotintreated piglets in the present study, whereas the latter showed reduced amounts of oxidative stress, as measured via the amount of nitration of tyrosine [1]. In a rat focal ischaemia model, it was shown that caspase-9 mRNA was decreased between 3–12 h following MCAO, but recovered to normal values by 24 h following MCAO [25]. This latter study might explain why no differences in caspase-9 were observed between vehicle- and 2-iminobiotintreated piglets at 24 h after hypoxia-ischaemia. In the present study, no significant differences were demonstrated between IL-6, TNF-· and TGF-ß mRNA in the vehicle- and 2-iminobiotin-treated piglets, although these cytokines tended to be increased in vehicle-treated piglets at 24 h upon reperfusion. Using RNase protection assays, IL-1·, IL-1ß, TNF-ß and RANTES were not detected at 24 h following hypoxia-ischaemia. Although little information is present on time course changes of cytokine formation in the newborn piglet model of hypoxia-ischaemia, some data are available from the perinatal rat model. IL-1ß and IL-6 mRNA attained maximal expression at 3 h after hypoxia-ischaemia and IL-1 and IL-6 bioactive protein peaked at 6 h [27]. In the same model, Szaflarski et al. [28] showed that IL-1ß and TNF-· mRNA was maximally up-regulated at 4 h following ischaemia, and returned to normal values by 24 h. Although one has to be cautious in comparing rodent data with piglet data, these studies might explain that no IL-1 mRNA was detected at 24 h in our model. The detection of IL-6 and TNF-· mRNA with the less sensitive RNA protection assays at 24 h following hypoxia-ischaemia might be explained by the use of a different species in the present study. Although we only investigated one single time point, i.e. 24 h after hypoxia-ischaemia, the strength of the present study is that a piglet model was applied, which has many similarities to a situation of human hypoxia-ischaemia. The maturation and the grey/white matter ratio of the neonatal gyrencephalic brain of piglets correspond well to that of the human brain at term. Furthermore, the model allows evaluation of brain lesions with MRI, which is used clinically, and the distribution of brain injury bears great resemblance to that seen after birth asphyxia in humans. We believe that it is crucial that some mechanistic information is also retrieved in such large animal models in spite of their limitations, and further dissection of mechanisms will depend on rodent models and in vitro studies.

Peeters-Scholte et al.

IL-1 and TNF-· are commonly regarded as neurotoxic, since IL-1 receptor antagonists reduce hypoxic-ischaemic brain injury [27] and neonatal mice deficient in IL-1ß converting enzyme (caspase-1) are more resistant to hypoxic-ischaemic insults [29]. There are also some studies suggesting that IL-1ß might be crucial to the repair of the central nervous system, presumably through the induction of astrocyte-, microglia- or macrophage-derived IGF-1 [30]. TNF-· can also be modestly neuroprotective for granular neurons when phosphatidylinositol-3-kinase is up-regulated [10]. Increasing concentrations of TNF-·, however, can be neurotoxic by inhibiting the ability of IGF-1 to promote neuronal survival [31, 10]. Therefore, it was concluded that cytokines are not neurotoxic per se, but can be deleterious in the injured brain by inhibiting the protective effects of endogenous growth factors in the brain [11]. In our study, IGF-1 correlated with caspase-3 activity and TUNEL, but not with cytokine expression at 24 h following hypoxia-ischaemia. This is not contradictory to the hypothesis that cytokines may only be toxic in the absence of IGF-1. TGF-ß, another member of the cytokine family, is present in cortical astrocytes and neurons and plays a role in the process of angiogenesis, tissue inflammation, and fibrosis [32]. Three isoforms of TGF-ß are present in the central nervous system: whereas astrocytes in culture express all three isoforms, neurons in culture express only TGF-ß2 [33]. It was reported that TGF-ß1 was up-regulated at 24 h following hypoxia-ischaemia for about 3 days in a mouse model of permanent MCAO [33]. Surprising-

ly, in a rat model of transient global ischaemia, bioactive TGF-ß1 was up-regulated at 3 and 6 h following ischaemia, mainly in the pyramidal neurons of the hippocampus, and increased further in the days following ischaemia [32]. It was concluded that the endogenous TGF-ß1 expressed in neurons might play a role in the pathological process of DNA degradation and delayed neuronal death after ischaemia. For the TGF-ß2 component, as investigated in our study, the effects on neuronal death are less clear. So, we can conclude that IGF-1 mRNA was significantly preserved in piglets after 2-iminobiotin treatment. IGF-1 production forms an endogenous protection against perinatal hypoxia-ischaemia by preventing programmed cell death, as assessed with caspase-3 activation and TUNEL staining. Further studies are warranted to explain the results of caspase-8 and -9 activation. Cytokine mRNA did not correlate with the degree of brain injury in newborn piglets at 24 h following perinatal hypoxia-ischaemia. This suggests a minor role for cytokines in the determination of cell death following perinatal hypoxia-ischaemia.

Acknowledgements We thank Ilka Post, Tomoaki Ioroi and the biotechnicians of the Central Laboratory Animal Institute of the Utrecht University for their enthusiastic help during the experiments. Finally, we greatly appreciated the help of Nicole Kops for carrying out the histological procedures.

References 1 Peeters-Scholte C, Koster JG, Veldhuis W, van den Tweel E, Zhu C, Blomgren K, Bar PR, Van Buul-Offers SC, Hagberg H, Nicolay K, Van Bel F, Groenendaal F: Neuroprotection by selective nitric oxide synthase inhibition after perinatal hypoxia-ischemia. Stroke 2002;33: 2304–2310. 2 Clawson TF, Vannucci SJ, Wang GM, Seaman LB, Yang XL, Lee WH: Hypoxia-ischemiainduced apoptotic cell death correlates with IGF-1 mRNA decrease in neonatal rat brain. Biol Signals Recept 1999;8:281–293. 3 Beilharz EJ, Russo VC, Butler G, Baker NL, Connor B, Sirimanne ES, Dragunow M, Werther GA, Gluckman PD, Williams CE, Scheepens A: Co-ordinated and cellular specific induction of the components of the IGF/ IGFBP axis in the rat brain following hypoxicischemic injury. Brain Res Mol Brain Res 1998;59:119–134.

Selective NOS Inhibition, IGF-1, Caspases, and Cytokines after Perinatal Asphyxia

4 Gluckman P, Klempt N, Guan J, Mallard C, Sirimanne E, Dragunow M, Klempt M, Singh K, Williams C, Nikolics K: A role for IGF-1 in the rescue of CNS neurons following hypoxicischemic injury. Biochem Biophys Res Commun 1992;182:593–599. 5 Klempt ND, Klempt M, Gunn AJ, Singh K, Gluckman PD: Expression of insulin-like growth factor-binding protein 2 (IGF-BP 2) following transient hypoxia-ischemia in the infant rat brain. Brain Res Mol Brain Res 1992;15: 55–61. 6 Graham SH, Chen J: Programmed cell death in cerebral ischemia. J Cereb Blood Flow Metab 2001;21:99–109. 7 Susin SA, Daugas E, Ravagnan L, Samejima K, Zamzami N, Loeffler M, Costantini P, Ferri KF, Irinopoulou T, Prevost MC, Brothers G, Mak TW, Penninger J, Earnshaw WC, Kroemer G: Two distinct pathways leading to nuclear apoptosis. J Exp Med 2000;192:571–580.

8 Bona E, Andersson AL, Blomgren K, Gilland E, Puka-Sundvall M, Gustafson K, Hagberg H: Chemokine and inflammatory cell response to hypoxia-ischemia in immature rats. Pediatr Res 1999;45:500–509. 9 Silverstein FS, Barks JD, Hagan P, Liu XH, Ivacko J, Szaflarski J: Cytokines and perinatal brain injury. Neurochem Int 1997;30:375– 383. 10 Venters HD, Broussard SR, Zhou JH, Bluthe RM, Freund GG, Johnson RW, Dantzer R, Kelley KW: Tumor necrosis factor-· and insulin-like growth factor-1 in the brain: Is the whole greater than the sum of its parts? J Neuroimmunol 2001;119:151–165. 11 Loddick SA, Rothwell NJ: Mechanisms of tumor necrosis factor alpha action on neurodegeneration: Interaction with insulin-like growth factor-1. Proc Natl Acad Sci USA 1999;96: 9449–9451.

Dev Neurosci 2002;24:396–404

403

12 Groenendaal F, de Graaf RA, van Vliet G, Nicolay K: Effects of hypoxia-ischemia and inhibition of nitric oxide synthase on cerebral energy metabolism in newborn piglets. Pediatr Res 1999;45:827–833. 13 Lorek A, Takei Y, Cady EB, Wyatt JS, Penrice J, Edwards AD, Peebles D, Wylezinska M, Owen-Reece H, Kirkbride V: Delayed (‘secondary’) cerebral energy failure after acute hypoxia-ischemia in the newborn piglet: Continuous 48-hour studies by phosphorus magnetic resonance spectroscopy. Pediatr Res 1994;36: 699–706. 14 Naressi A, Couturier C, Devos JM, Janssen M, Mangeat C, de Beer R, Graveron-Demilly D: Java-based graphical user interface for the MRUI quantitation package. MAGMA 2001; 12:141–152. 15 Wang X, Karlsson JO, Zhu C, Bahr BA, Hagberg H, Blomgren K: Caspase-3 activation after neonatal rat cerebral hypoxia-ischemia. Biol Neonate 2001;79:172–179. 16 Smink JJ, Koster JG, Hendriks-Stegeman BI, Buul-Offers SC: Insulin-like growth factor (IGF) II induced changes in expression of IGF binding proteins in lymphoid tissues of hIGFII transgenic mice. Endocrinology 1999;140: 5876–5882. 17 Baker NL, Carlo R, V, Bernard O, D’Ercole AJ, Werther GA: Interactions between bcl-2 and the IGF system control apoptosis in the developing mouse brain. Brain Res Dev Brain Res 1999;118:109–118. 18 Gallaher BW, Hille R, Raile K, Kiess W: Apoptosis: Live or die – hard work either way! Horm Metab Res 2001;33:511–519.

404

19 Tamatani M, Ogawa S, Tohyama M: Roles of Bcl-2 and caspases in hypoxia-induced neuronal cell death: A possible neuroprotective mechanism of peptide growth factors. Brain Res Mol Brain Res 1998;58:27–39. 20 Wilkins A, Chandran S, Compston A: A role for oligodendrocyte-derived IGF-1 in trophic support of cortical neurons. Glia 2001;36:48– 57. 21 Benchoua A, Guegan C, Couriaud C, Hosseini H, Sampaio N, Morin D, Onteniente B: Specific caspase pathways are activated in the two stages of cerebral infarction. J Neurosci 2001; 21:7127–7134. 22 Cao G, Luo Y, Nagayama T, Pei W, Stetler RA, Graham SH, Chen J: Cloning and characterization of rat caspase-9: Implications for a role in mediating caspase-3 activation and hippocampal cell death after transient cerebral ischemia. J Cereb Blood Flow Metab 2002;22:534–546. 23 Ticinelli T, Chawla K, Fung C, Mishra OP, Delivoria-Papadopoulos M: Cytosolic caspase9 activity in the developing fetal guinea pig brain following hypoxia. Pediatr Res 2002;51: 2592A. 24 Delogu G, Moretti S, Famularo G, Marcellini S, Santini G, Antonucci A, Marandola M, Signore L: Mitochondrial perturbations and oxidant stress in lymphocytes from patients undergoing surgery and general anesthesia. Arch Surg 2001;136:1190–1196. 25 Harrison DC, Davis RP, Bond BC, Campbell CA, James MF, Parsons AA, Philpott KL: Caspase mRNA expression in a rat model of focal cerebral ischemia. Brain Res Mol Brain Res 2001;89:133–146. 26 Morita-Fujimura Y, Fujimura M, Yoshimoto T, Chan PH: Superoxide during reperfusion contributes to caspase-8 expression and apoptosis after transient focal stroke. Stroke 2001; 32:2356–2361.

Dev Neurosci 2002;24:396–404

27 Hagberg H, Gilland E, Bona E, Hanson LA, Hahin Zoric M, Blennow M, Holst M, McRae A, Soder O: Enhanced expression of interleukin (IL)-1 and IL-6 messenger RNA and bioactive protein after hypoxia-ischemia in neonatal rats. Pediatr Res 1996;40:603–609. 28 Szaflarski J, Burtrum D, Silverstein FS: Cerebral hypoxia-ischemia stimulates cytokine gene expression in perinatal rats. Stroke 1995;26: 1093–1100. 29 Liu XH, Kwon D, Schielke GP, Yang GY, Silverstein FS, Barks JD: Mice deficient in interleukin-1 converting enzyme are resistant to neonatal hypoxic-ischemic brain damage. J Cereb Blood Flow Metab 1999;19:1099–1108. 30 Mason JL, Suzuki K, Chaplin DD, Matsushima GK: Interleukin-1beta promotes repair of the CNS. J Neurosci 2001;21:7046–7052. 31 Wu Y, Tewari M, Cui S, Rubin R: Activation of the insulin-like growth factor-1 receptor inhibits tumor necrosis factor-induced cell death. J Cell Physiol 1996;168:499–509. 32 Zhu Y, Roth-Eichhorn S, Braun N, Culmsee C, Rami A, Krieglstein J: The expression of transforming growth factor-ß1 (TGF-ß1) in hippocampal neurons: A temporary upregulated protein level after transient forebrain ischemia in the rat. Brain Res 2000;866:286–298. 33 Vivien D, Bernaudin M, Buisson A, Divoux D, MacKenzie ET, Nouvelot A: Evidence of type I and type II transforming growth factor-ß receptors in central nervous tissues: Changes induced by focal cerebral ischemia. J Neurochem 1998;70:2296–2304.

Peeters-Scholte et al.

Original Paper Dev Neurosci 2002;24:405–410 DOI: 10.1159/000069046

Received: August 1, 2002 Accepted: September 29, 2002

Evidence that p38 Mitogen-Activated Protein Kinase Contributes to Neonatal Hypoxic-Ischemic Brain Injury Byung Hee Han a, b Junjeong Choi b David M. Holtzman b, c, d a Department

of Pharmacy, Ewha Women’s University College of Pharmacy, Seoul, Korea; of Neurology, c Center for the Study of Nervous System Injury, and d Molecular Biology and Pharmacology, Washington University, St. Louis, Mo., USA b Department

Key Words p38 W Mitogen-activated protein kinase W Hypoxia-ischemia W Neurodegeneration W Apoptosis W Caspase-3 W Neuroprotection

Abstract We tested the response of stress-activated mitogen-activated protein kinases (MAPKs) – p38 MAPK and c-JUN NH2-terminal kinase (JNK) – following hypoxia-ischemia (H-I) induced by unilateral carotid artery ligation and hypoxia (8% O2 and 92% N2) for 2.5 h in postnatal-day-7 rats. Phosphorylation of p38 MAPK increased in the hippocampus and cortex immediately following H-I and returned to a basal level 6 h later. In contrast to p38 MAPK, phosphorylation of JNK decreased in the hippocampus and cortex immediately following H-I. Intracerebroventricular administration of two different p38 MAPK inhibitors prior to H-I significantly protected the neonatal brain from H-I injury. Interestingly, p38 MAPK inhibitors did not attenuate caspase-3 activation 24 h after H-I. Thus, these data suggest that p38 MAPKs contribute to the rapid, early component of brain injury following neonatal H-I. Copyright © 2002 S. Karger AG, Basel

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2002 S. Karger AG, Basel

Accessible online at: www.karger.com/dne

Introduction

Hypoxic-ischemic (H-I) encephalopathy in survivors of perinatal asphyxia is a major contributor to morbidity and mortality [Vannucci, 1990; Volpe, 1995]. Experimental models of H-I in neonatal rodents have been shown to mimic many aspects of the pathological and cognitive abnormalities seen in children who have suffered from an H-I insult [Almli et al., 2000]. The well-characterized Levine procedure includes unilateral carotid ligation in postnatal-day-7 (P7) rats followed by exposure to hypoxia [Levine, 1960]. This H-I insult results in reproducible brain injury ipsilateral to carotid ligation [Rice et al., 1981; Johnston, 1983; Ferriero et al., 1988] with either no or minimal damage to the contralateral hemisphere depending on the length of exposure to hypoxia [Towfighi et al., 1995]. We and other groups have demonstrated that there are prominent features of apoptotic as well as necrotic neuronal cell death following H-I [Vannucci, 1990; Han and Holtzman, 2000; Yamashima, 2000; Blomgren et al., 2001; Han et al., 2001; Northington et al., 2001]. While there is delayed caspase-3 activation in the lesioned hemisphere, calpain activation, seemingly marking excitotoxic/necrotic cell death, occurs immediately after H-I injury [Han et al., 2002]. In a recent study, we have demonstrated that the protein clusterin contributes to caspase-3-independent brain injury and clusterin-deficient mice have 50% less brain injury as compared with

David M. Holtzman, MD Washington University School of Medicine Department of Neurology, 660 S. Euclid Avenue, Box 8111 St. Louis, MO 63110 (USA) Tel. +1 314 362 9872, Fax +1 314 362 2826, E-Mail [email protected]

wild-type mice; yet there was no attenuation of caspase-3 activation [Han et al., 2001]. Mitogen-activated protein kinases (MAPKs) are serine/threonine kinases that control cellular signal transduction from the cell surface to the nucleus. Three main MAPKs include the extracellular signal-regulated kinases (ERKs), the c-JUN NH2-terminal kinases (JNKs), and the p38 MAPKs. ERKs (ERK1/2) are mitogenic and activated by mitogens and growth factors in most mammalian cells [Robinson and Cobb, 1997; Kyriakis and Avruch, 2001]. In addition, growth factor-mediated ERK1/2 activation mediates neuronal survival effects in many settings of cell death including in PC12 cells [Xia et al., 1995], cerebellar granule neurons [Bonni et al., 1999], and following neonatal H-I in vivo [Han and Holtzman, 2000]. In contrast to ERKs, the JNK and p38 MAPKs are activated by a variety of cellular stresses including UV light, hyperosmolarity, heat shock, and proinflammatory cytokines [Martin-Blanco, 2000; Nebreda and Porras, 2000]. p38 MAPKs have been implicated in neuronal cell death [Mielke and Herdegen, 2000; Harper and LoGrasso, 2001; Irving and Bamford, 2002]. For example, activation of p38 MAP kinase is associated with NGF-deprived cell death in PC12 cells [Kummer et al., 1997], as well as with excitotoxicity in cerebellar granule neurons [Kawasaki et al., 1997]. More recent reports have demonstrated that p38 MAPK is activated following focal cerebral ischemia in adult rats. Furthermore, systemic administration of p38 MAPK inhibitors significantly attenuated infarct size in an adult stroke model [Barone et al., 2001; Legos et al., 2001]. Whether activation of p38 MAPKs contributes to neonatal brain injury following H-I has not been clarified. In this study, we found that p38 MAPK phosphorylation, indicating activation, was strongly increased in the hippocampus and cortex immediately following H-I and returned to a basal level 6 h later. Pharmacological inhibition of p38 MAPK resulted in neuroprotection against H-I-induced brain injury in neonatal rats. Materials and Methods Animals and the Surgical Procedure Newborn Sprague-Dawley rats (dam plus 10 pups per litter) were obtained from Sasco Breeders when the pups were 3–4 days of age. The pups were housed with their dam in the home cage under a 12:12-hour light:dark cycle, with food and water freely available throughout the study. The neonatal H-I brain injury model was performed based on the Levine procedure [Levine, 1960; Rice et al., 1981; Vannucci, 1990] as described previously [Cheng et al., 1998; Han et al., 2000; Han and Holtzman, 2000]. In all animals that

406

Dev Neurosci 2002;24:405–410

underwent unilateral carotid ligation, the ligation was performed on the left carotid artery. p38 MAPK inhibitors, SB203580 and SB220025 (Calbiochem) were dissolved in phosphate-buffered saline containing 20% dimethylsulfoxide. For in vivo studies, either vehicle or compound (10 nmol in 5 Ìl vehicle) was intracerebroventricularly injected into the left hemisphere of P7 rats as described previously [Cheng et al., 1998; Han and Holtzman, 2000]. DEVD-AMC Cleavage Assay Following H-I in P7 rats, tissues from the hippocampus and cortex, both ipsi- and contralateral to the ligation, were dissected and frozen in dry ice. Tissue samples were homogenized in lysis buffer [10 mM Hepes, pH 7.4, 5 mM MgCl2, 1 mM DTT, 1% Triton X-100, 2 mM EGTA, 2 mM EDTA, 1 mM PMSF, and protease inhibitor cocktail (Boehringer Mannheim, Germany, catalog No. 1697498)] and centrifuged at 12,000 g for 10 min at 4 ° C. Caspase-3-like activity was determined by DEVD-AMC cleavage assay as described previously [Han et al., 2000; Han and Holtzman, 2000]. Western Blotting Brain tissue lysates prepared as described above for the caspase assay were subjected to Western blotting. Proteins (40 Ìg per lane) were separated by SDS-PAGE and transferred to nitrocellulose membranes (Bio-Rad) as described previously [Han et al., 2000; Han and Holtzman, 2000]. Blots were blocked with 3% dried milk in TBS containing 0.05% Tween 20 overnight. Blots were then incubated for 2–3 h with primary antibody, followed by incubation with antirabbit or antimouse HRP-conjugated IgG, and visualized with enhanced chemiluminescence (Amersham). Primary antibodies used were as follows: rabbit antiprocaspase-3 at 1:1,000 (Santa Cruz), rabbit antiactive caspase-3 at 1:1,000 (Cell Signaling Technology), rabbit antiphospho-p38 MAPK at 1:2,000 (Cell Signaling Technology), and rabbit antiphospho-JNK at 1:2,000 (Cell Signaling Technology). Assessment of Brain Damage due to H-I To assess regional area loss, 1 week after H-I, brain sections were prepared as described above and damage due to H-I was determined by calculating the amount of surviving tissue in coronal sections as described previously [Cheng et al., 1998; Han and Holtzman, 2000]. Briefly, coronal sections from the genus of the corpus callosum to the end of the dorsal hippocampus were stained with cresyl violet. The cross-sectional areas of the striatum, cortex, and hippocampus in each of eight equally spaced reference planes were photo-scanned and the area of each brain region was calculated using SigmaScan Pro (Jandel Scientific Software). The sections utilized for quantification corresponded approximately to plates 12, 15, 17, 20, 23, 28, 31, and 34 in the rat brain atlas [Paxinos and Watson, 1986]. Data are presented as the mean B SEM and were analyzed by t test.

Results

p38 MAPK Is Transiently Activated following H-I We and other groups have found that there are prominent biochemical and morphological features of apoptotic as well as excitotoxic/necrotic neuronal cell death following H-I in P7 rats [Vannucci, 1990; Han and Holtzman,

Han/Choi/Holtzman

2000; Yamashima, 2000; Northington et al., 2001]. While calpain activation is associated with excitotoxic/necrotic neuronal injury and occurs immediately after H-I, caspase-3 activation contributes to apoptotic-like brain injury in a delayed fashion [Cheng et al., 1998; Han et al., 2002]. To determine whether different MAPK signaling pathways contribute to H-I-induced brain injury, brain tissues from the hippocampus and cortex prepared at different time points following H-I were subjected to biochemical assay as well as Western blot. Consistent with previous results [Han et al., 2000; Han et al., 2002], caspase-3-like activity as assessed by DEVD-AMC cleavage activity was notable at 6 h after H-I in the hippocampus and cortex ipsilateral to the carotid ligation (fig. 1). This increase in caspase-3-like activity was paralleled by the appearance of the 18-kD fragment of active caspase-3 when analyzed by Western blot (fig. 2). p38 MAPK phosphorylation was strongly and transiently increased in the hippocampus and cortex 30 min to 1 h following H-I, which was more prominent in the hemisphere ipsilateral to the carotid ligation (fig. 2). In contrast to p38 MAPK, JNK phosphorylation was decreased in the hemisphere ipsilateral to the ligation immediately following H-I (data now shown). p38 MAPK Inhibitors SB203580 and SB220025 Are Neuroprotective To determine whether p38 MAPK activation contributes to brain injury, we utilized the p38 MAPK inhibitors SB220025 and SB203580. These 2 inhibitors have distinct chemical structures. P7 rats received an intracerebroventricular injection of either vehicle or p38 MAPK inhibitor (10 nmol/pup) just prior to hypoxia exposure, and brain injury was assessed 1 week after the insult (fig. 3). We found that a single administration of both SB220025 (fig. 3a) and SB203580 (fig. 3b) strongly decreased brain tissue loss in all brain regions examined. p38 MAPK inhibitors decreased tissue loss in the striatum, hippocampus, and cortex by approximately 50–70% compared with the vehicle-treated groups. p38 MAPK Inhibition Does Not Influence Caspase-3 Activation following H-I Since p38 MAPK inhibition is neuroprotective in this model, we next tested whether caspase-3 activation, an apoptotic marker, was influenced by p38 MAPK inhibition. Surprisingly, despite the neuroprotective action of this compound, an intracerebroventricular injection of SB220025 (10 nmol/pup) prior to H-I insult did not influence DEVD-AMC cleavage activity 24 h following H-I as compared with the vehicle-treated group (fig. 4).

Neuroprotective Effects of p38 MAPK Inhibitors

Fig. 1. Time course of caspase-3 activation following H-I injury. P7

rats received unilateral carotid ligation followed by exposure to 8% O2 for 2.5 h. At different time points following H-I, animals were sacrificed and the brain tissues from the hippocampus (a) and cortex (b) ipsilateral to the ligation were subjected to DEVD-AMC cleavage assay. Data are presented as mean B SEM (n = 8/time point). * p ! 0.05 compared with sham-operated group when analyzed by one-way ANOVA followed by Dunnett’s post hoc test.

Discussion

The present study demonstrated that there was an increase in phosphorylation of p38 which remained at an elevated level up to 1 h and then returned to basal levels by 6 h after neonatal H-I. Pharmacological inhibition of p38 MAPKs resulted in a significant reduction in brain

Dev Neurosci 2002;24:405–410

407

Sham Active caspase-3

R

L

0.5 R

6 L

R

12 L

R

18 h L

R

L 17 kD

p-p38 MAPK

Fig. 2. Western blot analysis of stress-kinase activity following H-I injury. P7 rats received unilateral (left) carotid ligation followed by exposure to 8% O2 for 2.5 h. At different time points following H-I, animals were sacrificed and the brain tissues from the right hippocampus (R) contralateral and left hippocampus (L) ipsilateral to

carotid ligation were dissected. Tissues were lysed and soluble proteins (40 Ìg/lane) were separated by SDS-PAGE followed by immunoblotting with the following antibodies: anticaspase-3 antibody and antiphospho-p38 MAPK (p-p38 MAPK) antibody. Data are representative of three similar experiments.

injury without affecting caspase-3 activation following H-I compared with the vehicle-treated group. These data suggest that p38 MAPKs contribute to neonatal brain injury via a caspase-3-independent pathway in this model. Accumulating evidence has demonstrated that there are prominent biochemical and morphological features of apoptosis as well as necrosis following neonatal H-I brain injury [Vannucci, 1990; Han and Holtzman, 2000; Yamashima, 2000; Blomgren et al., 2001; Han et al., 2001; Northington et al., 2001]. Following neonatal H-I, apoptosis as assessed by caspase-3 activation is detectable in the lesioned hemisphere after 6 h with a peak occurring 12– 24 h after H-I. In contrast, necrosis as assessed by morphological criteria and calpain activation occurs immediately after H-I [Olney et al., 1989; Northington et al., 2001; Han et al., 2002]. In this study, we sought to determine whether stress-activated kinases were involved in neonatal H-I brain injury. Western blot analysis revealed that phosphorylation of p38 MAPK was increased by several fold in the brain by 1 h after H-I, a time point when excitotoxic/necrotic cell death, but not apoptotic-like death, has been initiated. Recent reports have demonstrated that in adult rats, activated p38 MAPK is rapidly elevated within 1 h and returned to a baseline value by 4 h following focal ischemic brain injury [Barone et al., 2001; Legos et al., 2001]. The fact that necrosis predominantly occurs following focal ischemic brain injury and that inhibition of p38 MAPK activity is neuroprotective suggests a direct contribution of p38 MAPKs to necrotic brain injury after ischemic stroke. Our data support the idea that inhibition of p38 MAPK reduced the brain tissue loss induced by neonatal H-I via a caspase-3-independent

manner. Further studies are necessary to determine how activation of p38 MAPKs contributes to excitotoxic/ necrotic mechanisms (e.g. calpain activation) following H-I in neonatal rats. We have recently demonstrated that the selective, reversible caspase-3 inhibitor M826 blocks caspase-3 activation and its substrate cleavage, resulting in significant neuroprotection against neonatal H-I brain injury. However, inhibition of caspase-3 activity did not affect excitotoxic/necrotic cell death associated with calpain activation occurring immediately following neonatal H-I [Han et al., 2002]. It will be interesting to determine whether blockade of both caspase-3 and p38 MAPK activities may be additively neuroprotective against neonatal H-I brain injury. The exact mechanism by which activation of p38 MAPK contributes to neonatal H-I brain injury is not known. Evidence that p38 MAPK directly contributes to neuronal cell death comes from findings that p38 MAPK activation occurs following NGF deprivation of PC12 cells [Kummer et al., 1997] and following excitotoxicity in cerebellar granule neurons [Kawasaki et al., 1997]. In addition to possible direct effects of p38 in neurons, activation of the p38 MAPK pathway plays an essential role in the production of proinflammatory cytokines such as IL-1ß and TNF-· [Perregaux et al., 1995; Irving and Bamford, 2002]. It has been demonstrated that inflammation responses including inflammatory cytokine synthesis and release, and glial cell activation contribute to ischemiainduced neuronal cell death [Liu et al., 1993, 1994; Barone et al., 2001; Hedtjarn et al., 2002]. Recent studies with p38 MAPK inhibitors suggest that inhibition of p38 MAPK protects neurons from ischemic injury via a direct mechanism as well as via inhibition of pro-inflammatory

408

Dev Neurosci 2002;24:405–410

Han/Choi/Holtzman

Fig. 4. p38 MAPK inhibition does not influence caspase-3 activation

following H-I. P7 rats received unilateral carotid ligation followed by exposure to 8% O2 for 2.5 h. Just prior to exposure to hypoxia, animals received intracerebroventricular injection of either vehicle or 10 nmol of SB220025. Twenty-four hours later, animals were sacrificed, and brain tissues (hippocampus and cortex) from the hemisphere contra- and ipsilateral to carotid ligation were dissected. Tissues were lysed and DEVD-AMC cleavage activity was assessed. Data are presented as mean B SEM. There was no significant difference between the two groups when analyzed by t test.

Fig. 3. p38 MAPK inhibitors attenuate H-I-induced brain injury. P7

rats received unilateral (left) carotid ligation followed by exposure to 8% O2 for 2.5 h. Just prior to exposure to hypoxia, animals received an intracerebroventricular injection of either vehicle or 10 nmoles of the p38 MAPK inhibitor, SB220025 (a) or SB203580 (b). One week later at P14, animals were sacrificed and brain sections were stained with cresyl violet. Regional area loss from the striatum, hippocampus, and cortex was assessed. Data are presented as mean B SEM. * p ! 0.01 compared with vehicle treated group analyzed by t test.

cytokine release [Barone et al., 2001; Legos et al., 2001]. Although our findings that rapid phosphorylation of p38 MAPK occurs following H-I suggest the direct contribution of p38 MAPK to neuronal injury, further studies are required to determine the cell type in which p38 is activated as well as how rapid activation of p38 MAPK

Neuroprotective Effects of p38 MAPK Inhibitors

affects delayed inflammatory processes following neonatal H-I. In contrast to p38 MAPKs, phosphorylation of JNKs was markedly reduced in the hemisphere ipsilateral to the ligation immediately after neonatal H-I. It is interesting to note that activation of JNKs has been shown to be increased after 90 min of transient focal cerebral ischemia [Hayashi et al., 2000] and global ischemia in the adult rat [Hu et al., 2000]. Immunohistochemical studies have demonstrated that increased phospho-JNK immunoreactivity is present within the ischemic region, suggesting that JNK activation may contribute to ischemia-induced cell death. The present data, however, suggest that unlike brain injury in adult ischemic stroke models, activation of JNKs appear less likely to contribute to H-I brain injury in the developing rat brain. We have recently found that phosphorylation of ELK-1 and ATF-2, which are phosphorylated by both p38 and JNKs [Mielke and Herdegen, 2000; Irving and Bamford, 2002], are markedly reduced in the lesioned hemisphere immediately after and up to

Dev Neurosci 2002;24:405–410

409

18 h following neonatal H-I [Han and Holtzman, unpubl. data]. A search for downstream signaling molecules specific to p38 MAPKs may provide new clues to delineate the exact role of the p38 MAPK and other downstream pathways which contribute to neonatal H-I brain injury.

Acknowledgement This work was supported by NIH grant NS35902 (DMH).

References Almli CR, Levy TJ, Han BH, Shah AR, Gidday JM, Holtzman DM (2000): BDNF protects. Against spatial memory deficits following neonatal hypoxia-ischemia. Exp Neurol 166:99– 114. Barone FC, Irving EA, Ray AM, Lee JC, Kassis S, Kumar S, Badger AM, White RF, McVey MJ, Legos JJ, Erhardt JA, Nelson AH, Ohlstein EH, Hunter AJ, Ward K, Smith BR, Adams JL, Parsons AA (2001): SB 239063, a second-generation p38 mitogen-activated protein kinase inhibitor, reduces brain injury and neurological deficits in cerebral focal ischemia. J Pharmacol Exp Ther 296:312–321. Blomgren K, Zhu C, Wang X, Karlsson JO, Leverin AL, Bahr BA, Mallard C, Hagberg H (2001): Synergistic activation of caspase-3 by m-calpain after neonatal hypoxia-ischemia: A mechanism of ‘pathological apoptosis’? J Biol Chem 276:10191–10198. Bonni A, Brunet A, West AE, Datta SR, Takasu MA, Greenberg ME (1999): Cell survival promoted by the Ras-MAPK signaling pathway by transcription-dependent and -independent mechanisms. Science 286:1358–1362. Cheng Y, Deshmukh M, D’Costa A, Demaro JA, Gidday JM, Shah A, Sun Y, Jacquin MF, Johnson EM, Holtzman DM (1998): Caspase inhibitor affords neuroprotection with delayed administration in a rat model of neonatal hypoxic-ischemic brain injury. J Clin Invest 101: 1992–1999. Ferriero DM, Arcavi LJ, Sagar SM, McIntosh TK, Simon RP (1988): Selective sparing of NADPH-diaphorase neurons in neonatal hypoxia-ischemia. Ann Neurol 24:670–676. Han BH, D’Costa A, Back SA, Parsadanian M, Patel S, Shah AR, Gidday JM, Srinivasan A, Deshmukh M, Holtzman DM (2000): BDNF blocks caspase-3 activation in neonatal hypoxia-ischemia. Neurobiol Dis 7:38–53. Han BH, DeMattos RB, Dugan LL, Kim-Han JS, Brendza RP, Fryer JD, Kierson M, Cirrito J, Quick K, Harmony JA, Aronow BJ, Holtzman DM (2001): Clusterin contributes to caspase-3independent brain injury following neonatal hypoxia-ischemia. Nat Med 7:338–343. Han BH, Holtzman DM (2000): BDNF protects the neonatal brain from hypoxic-ischemic injury in vivo via the ERK pathway. J Neurosci 20: 5775–5781.

410

Han BH, Xu D, Choi J, Han Y, Xanthoudakis S, Roy S, Tam J, Vaillancourt J, Colucci J, Siman R, Giroux A, Robertson GS, Zamboni R, Nicholson DW, Holtzman DM (2002): Selective, reversible caspase-3 inhibitor is neuroprotective and reveals distinct pathways of cell death after neonatal hypoxic-ischemic brain injury. J Biol Chem 277:30128–30136. Harper SJ, LoGrasso P (2001): Signalling for survival and death in neurones: The role of stressactivated kinases, JNK and p38. Cell Signal 13: 299–310. Hayashi T, Sakai K, Sasaki C, Zhang WR, Warita H, Abe K (2000): c-Jun N-terminal kinase (JNK) and JNK interacting protein response in rat brain after transient middle cerebral artery occlusion. Neurosci Lett 284:195–199. Hedtjarn M, Leverin AL, Eriksson K, Blomgren K, Mallard C, Hagberg H (2002): Interleukin-18 involvement in hypoxic-ischemic brain injury. J Neurosci 22:5910–5919. Hu BR, Liu CL, Park DJ (2000): Alteration of MAP kinase pathways after transient forebrain ischemia. J Cereb Blood Flow Metab 20:1089– 1095. Irving EA, Bamford M (2002): Role of mitogenand stress-activated kinases in ischemic injury. J Cereb Blood Flow Metab 22:631–647. Johnston MV (1983): Neurotransmitter alterations in a model of perinatal hypoxic-ischemic brain injury. Ann Neurol 13:511–518. Kawasaki H, Morooka T, Shimohama S, Kimura J, Hirano T, Gotoh Y, Nishida E (1997): Activation and involvement of p38 mitogen-activated protein kinase in glutamate-induced apoptosis in rat cerebellar granule cells. J Biol Chem 272: 18518–18521. Kummer JL, Rao PK, Heidenreich KA (1997): Apoptosis induced by withdrawal of trophic factors is mediated by p38 mitogen-activated protein kinase. J Biol Chem 272:20490– 20494. Kyriakis JM, Avruch J (2001): Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation. Physiol Rev 81:807–869. Legos JJ, Erhardt JA, White RF, Lenhard SC, Chandra S, Parsons AA, Tuma RF, Barone FC (2001): SB 239063, a novel p38 inhibitor, attenuates early neuronal injury following ischemia. Brain Res 892:70–77. Levine S (1960): Anoxic-ischemic encephalopathy in rats. Am J Pathol 36:1–17. Liu T, Clark RK, McDonnell PC, Young PR, White RF, Barone FC, Feuerstein GZ (1995): Tumor necrosis factor-alpha expression in ischemic neurons. Stroke 25:1481–1488.

Dev Neurosci 2002;24:405–410

Liu T, McDonnell PC, Young PR, White RF, Siren AL, Hallenbeck JM, Barone FC, Feurestein GZ (1993): Interleukin-1 beta mRNA expression in ischemic rat cortex. Stroke 24:1746–1751. Martin-Blanco E (2000): p38 MAPK signalling cascades: Ancient roles and new functions. Bioessays 22:637–645. Mielke K, Herdegen T (2000): JNK and p38 stresskinases – degenerative effectors of signal- transduction-cascades in the nervous system. Prog Neurobiol 61:45–60. Nebreda AR, Porras A (2000): p38 MAP kinases: Beyond the stress response. Trends Biochem Sci 25:257–260. Northington FJ, Ferriero DM, Graham EM, Traystman RJ, Martin LJ (2001): Early neurodegeneration after hypoxia-ischemia in neonatal rat is necrosis while delayed neuronal death is apoptosis. Neurobiol Dis 8:207–219. Olney JW, Ikonomidou C, Mosinger JL, Frierdich G (1989): MK-801 prevents hypobaric-ischemic neuronal degeneration in infant rat brain. J Neurosci 9:1701–1704. Paxinos G, Watson C (1986): The Rat Brain in Stereotaxic Coordinates, ed 2. New York, Academic Press. Perregaux DG, Dean D, Cronan M, Connelly P, Gabel CA (1995): Inhibition of interleukin-1 beta production by SKF86002: Evidence of two sites of in vitro activity and of a time and system dependence. Mol Pharmacol 48:433– 442. Rice JE, Vannucci RC, Brierley JB (1981): The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 9:131– 141. Robinson MJ, Cobb MH (1997): Mitogen-activated protein kinase pathways. Curr Opin Cell Biol 9:180–186. Towfighi J, Zec N, Yager J, Housman C, Vannucci RC (1995): Temporal evolution of neuropathologic changes in an immature rat model of cerebral hypoxia: A light microscopic study. Acta Neuropathol 90:375–386. Vannucci RC (1990): Experimental biology of cerebral hypoxia-ischemia: Relation to perinatal brain damage. Pediatr Res 27:317–326. Volpe JJ (1995): Neurology of the Newborn, ed 3. Philadelphia, Saunders. Xia X, Dickens M, Raingeaud J, Davis R, Greenberg M (1995): Opposing effects of ERK and JNK-p38 MAP kinases on apoptosis. Science 270:1326–1331. Yamashima T (2000): Implication of cysteine proteases calpain, cathepsin and caspase in ischemic neuronal death of primates. Prog Neurobiol 62:273–295.

Han/Choi/Holtzman

Original Paper Dev Neurosci 2002;24:411–417 DOI: 10.1159/000069051

Received: September 19, 2002 Accepted: October 7, 2002

Hypoxic Preconditioning Increases Brain Glycogen and Delays Energy Depletion from Hypoxia-Ischemia in the Immature Rat Robert M. Brucklacher a Robert C. Vannucci a Susan J. Vannucci b a Department of Pediatrics (Pediatric Neurology), Pennsylvania State University College of Medicine, Hershey, Pa., and b Department of Pediatric Critical Care Medicine, Columbia University, New York, N.Y., USA

Key Words Brain W Hypoxia-ischemia W Neuroprotection W Immaturity W Glycogen W Energy reserves

Abstract Recent studies have shown a protection from cerebral hypoxic-ischemic (HI) brain damage in the immature rat following a prior systemic hypoxic exposure when compared with those not exposed previously. To investigate the mechanism(s) of hypoxic preconditioning, brain glycogen and high-energy phosphate reserves were measured in naı¨ve and preconditioned rat pups subjected to HI. Groups in this study included untouched (naı¨ve) controls, preconditioned controls (i.e., hypoxia only), preconditioned with HI insult, and naı¨ve pups with HI insult. Hypoxic preconditioning was achieved in postnatal-day6 rats subjected to 8% systemic hypoxia for 2.5 h at 37 ° C. Twenty-four hours later, they were subjected to unilateral common carotid artery ligation and systemic hypoxia with 8% oxygen at 37 ° C for 90 min. Animals were allowed to recover from HI for up to 24 h. At specific intervals, animals in each group were frozen in liquid nitrogen for determination of cerebral metabolites. Preconditioned animals showed a significant increase in

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0411$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

brain glycogen 24 h following the initial hypoxic exposure, corresponding to the beginning of the HI insult. Measurement at the end of 90 min of HI showed a depletion of high-energy phosphates, ATP and phosphocreatine, in all animals although ATP remained significantly higher in the preconditioned animals. Thus, the energy from increased glycogen following preconditioning slowed high-energy phosphate depletion during HI, thereby allowing for long-term protection. Copyright © 2002 S. Karger AG, Basel

Introduction

Preconditioning involves treatment with a sublethal stress to provide protection from a subsequent brain injury. Several studies have documented that a period of cerebral hypoxia or ischemia insufficient to produce brain damage provides tolerance to injury in adult animal models of experimental ischemia [Kato et al., 1992; Kirino, 1982; Liu et al., 1992; Chen and Simon, 1997]. We, and others, have demonstrated that hypoxia provides a preconditioning tolerance to hypoxic-ischemic (HI) brain damage in the immature rat [Gidday et al., 1994; Vannucci et al., 1998; Bergeron et al., 2000; Jones and Bergeron,

Susan J. Vannucci, PhD Research Director, Pediatric Critical Care Medicine Morgan Stanley Children’s Hospital of New York/Columbia University 3959 Broadway, BHN 10–24, New York, NY 10032 (USA) Tel. +1 212 342 0275, Fax +1 212 342 2293, E-Mail [email protected]

2001]. Immature rats, when subjected to a systemic hypoxic event, are protected from a subsequent HI insult, with the peak protective influence occurring at 24 h. The mechanism(s) underlying hypoxic preconditioning in the immature brain are still unknown. We previously demonstrated that levels of high-energy phosphate compounds following 2.5 h of HI were similarly depleted in preconditioned and naı¨ve animals. Therefore, preconditioning did not appear to improve the metabolic state during HI, although the preconditioned rat pups did not exhibit the secondary depletion of ATP and phosphocreatine (PFCr) observed in naı¨ve animals during the early recovery phase [Vannucci et al., 1998]. Several recent studies have suggested that astrocytic glycogen provides an important energy source to the brain during periods of ischemia and glucose deprivation [Swanson and Choi, 1993; Wender et al., 2000]. Hypoxic preconditioning in the immature brain is associated with increased levels of hypoxia-inducible factor, Hif-1, with the concomitant increase in several enzymes involved in glycolysis and energy metabolism [Bergeron et al., 2000; Jones and Bergeron, 2001], which could indirectly stimulate glycogen metabolism. We decided to test the hypothesis that hypoxic preconditioning might extend the HI interval necessary for high-energy phosphate compound depletion, with brain glycogen providing the necessary substrate for energy production. In the present experiments, we chose a 90-min insult, as high-energy phosphate compounds are close to depletion at this interval, which would allow for the observation of any energy preservation in the preconditioned animals [Palmer et al., 1990; Yager et al., 1992].

Materials and Methods The experiments described here were reviewed by the Animal Care and Use Committee of the Pennsylvania State University College of Medicine and approved on May 28, 1998. Dated, pregnant Wistar rats were purchased from a commercial breeder (Charles River Laboratories, Wilmington, Mass., USA) and housed in individual cages. Offspring were delivered vaginally, and litters were reduced to 10 and kept with their dams until the time of the experiment at postnatal days P6 and P7. Induction of Cerebral Hypoxia and HI At P6, half of the rat pups randomized among 11 litters (n = 60/ 109) were placed in 500-ml air-tight jars partially submerged in a 37 ° C water bath to maintain a constant thermal environment at 35 ° C ambient temperature. A gas mixture of 8% oxygen-92% nitrogen was delivered through the jars via inlet and outlet portals at a flow rate of 100 ml/min. The pups were exposed to this hypoxic condition for 2.5 h and then allowed to recover in the jars open to room

412

Dev Neurosci 2002;24:411–417

air for 15 min, with the exception of 5 pups decapitated directly into liquid nitrogen for metabolite determination. The remaining pups (n = 59/109) were placed in open-air jars in the water bath for 2.75 h. Five pups from this group were also decapitated into liquid nitrogen as controls. All remaining pups were returned to the dams for 24 h. It has previously been shown that there is no brain damage resulting from hypoxia alone, or ligation alone, in the immature rat [Rice et al., 1981]. For the induction of unilateral cerebral HI, P7 pups from both the preconditioned and naı¨ve groups were subjected to permanent right common carotid artery ligation, followed 2–3 h later by systemic hypoxia with 8% O2 as previously described [Rice et al., 1981; Vannucci et al., 1995]. A control (C) group underwent neither preconditioning nor HI (n = 33). A preconditioned only (PC) group consisted of pups which underwent 2.5 h of systemic hypoxia on P6 but no HI on P7 (n = 36). A naı¨ve (HI) group underwent HI on P7 (n = 20). The final (PC-HI) group consisted of preconditioned animals, which also underwent HI on P7 (n = 20). Following 90 min of 8% O2-92% N2, a minimum of 6 pups in each group were immediately decapitated into liquid nitrogen. The remaining animals were allowed to recover with their dams; pups from each group were sacrificed at 4 or 24 h and heads immediately frozen as described. Preparation of Samples Prior to processing, all specimens were maintained at –80 ° C. The brain of each rat pup was removed from its skull in a cold box set at –20 ° C, and a portion of the right and left cerebral hemisphere in the distribution of the middle cerebral artery was dissected, powdered under liquid nitrogen, separated and weighed on a microanalytical balance into 2 separate tubes (approximately 50 mg each). One tube of each sample was extracted with perchloric acid and neutralized, as previously described [Vannucci et al., 1974]. Concentrations of ATP and P F Cr were measured fluorometrically by specific enzymatic techniques [Lowry et al., 1972; Vannucci et al., 1974; Hernandez et al., 1980]. The second tube was used for the measurement of glycogen. The frozen brain tissue was suspended in 50 volumes of 0.03 N HCl at 0 ° C and heated for 45 min at 100 ° C in sealed tubes [Passonneau et al., 1967; Vannucci et al., 1974]. Glycogen was determined as glucose, following the preincubation of a brain tissue extract with amylo-·-1,4-·-1,6-glucosidase, followed by reaction with hexokinase. Glucose was determined in this assay as well, with the hexokinase reaction and no preincubation. Enzymes and Reagents Enzymes were purchased from Roche Applied Science (Indianapolis, Ind., USA). Substrates used in the assays were purchased from Sigma Corporation (St. Louis. Mo., USA). Statistical Analyses Statistical analyses of the data included unpaired t tests for comparisons between preconditioned and naı¨ve groups; ANOVA with Tukey for multiple comparisons.

Brucklacher/Vannucci/Vannucci

Fig. 1. Effect of hypoxia on glycogen concen-

trations in immature rat brain. Rats were subjected to 2.5 h of hypoxia on P6 and brains were frozen at 4 and 24 h, and 1 week of recovery (rec.) for subsequent determination of glycogen. The nonhypoxic control group consisted of 12 age-matched littermates frozen at each interval. Each hypoxic recovery group had 6–8 animals. Values are mean B SD. * p ! 0.05, ** p ! 0.001, using AVOVA as compared with the naı¨ve control.

Results

2.5 h of hypoxia resulted in a significant reduction, but not complete depletion, of brain glycogen in the P6 rat (fig. 1, p ! 0.001). Glycogen resynthesis during recovery was rapid and continual, such that it was significantly elevated at 4 h with a further increase at 24 h; the 28% increase is significant in relation to the control value at 24 h (p ! 0.05). By 1 week, the values are not different from age-matched controls (data not shown). Figure 2 shows the concentrations of glycogen and glucose in the ipsilateral hemisphere at 0, 4 and 24 h of recovery from HI in preconditioned and naı¨ve pups. As in figure 1, preconditioning alone resulted in significant increases in brain glycogen and glucose at 24 h after hypoxia (p ! 0.05, both glycogen and glucose), such that the preconditioned animals had higher levels of brain glycogen than their naı¨ve littermates at the onset of HI. Glycogen was partially depleted in both groups by the end of HI (20% of the C group, p ! 0.05 and 28% of the PC control group, p ! 0.001, respectively). Despite the trend towards better preservation of brain glycogen at the end of HI in the PC group, the values were not significantly different during any of the recovery intervals when compared between groups, although there were significant increases as compared with controls (p ! 0.001). Glucose followed similar trends in both preconditioned and naı¨ve groups, with no statistical differences between groups, although significant differences from control groups were evident. Following glycogen breakdown during HI, reoxygenation resulted in significant increases in brain glycogen in ipsilateral hemispheres of both groups at 24 h of recovery

Effects of Hypoxic Preconditioning in the Immature Rat

(+80%). The contralateral hemisphere is primarily hypoxic, and the changes in brain glycogen mirror what was seen during hypoxic preconditioning, i.e. a reduction in glycogen at the end of the hypoxic interval (0.95 B 0.29, HI animals vs. 1.37 B 0.38, PC-HI animals), and a resynthesis to levels exceeding baseline by 24 h of recovery (3.24 B 0.28 HI group vs. 2.98 B 0.15, PC-HI group), or approximately a 35% increase in both groups. The effect of preconditioning and HI on ipsilateral cerebral hemispheric ATP and PFCr at 0, 4 and 24 h of recovery from HI in preconditioned and naı¨ve immature rats is shown in figure 3. The depletion of the high-energy reserves, ATP and PFCr, did not occur as rapidly in the preconditioned animals; at the end of the 90-min HI, ATP values were 69% of preconditioned control compared with the naı¨ve group (25% of control). All were significantly different from the respective control (p ! 0.001). The difference between the two groups at 90 min of HI is significant for ATP (p ! 0.05), and is apparent in both ipsilateral (shown) and contralateral (not shown) hemispheres. The seemingly better preservation of PFCr in the ipsilateral hemisphere of the PC pups did not reach significance (p = 0.12), although the improved energy state in the contralateral hemisphere was significant (1.57 B 0.17, HI animals vs. 2.06 B 0.15, PC-HI animals; p ! 0.05). Both groups recovered by 4 h, but at 24 h, a secondary depletion occurred only in the naı¨ve animals, both in ATP (73% of control, p ! 0.01) and PFCr (56% of control, p ! 0.01). In the absence of sufficient oxygen and substrate delivery, the energy available to maintain cerebral metabolism is due to endogenous stores of ATP and PFCr [energy

Dev Neurosci 2002;24:411–417

413

Fig. 2. Effect of preconditioning and HI on brain glucose and glycogen. Brain glycogen and glucose were measured in ipsilateral hemispheres of brains of preconditioned and naı¨ve immature rats prior to, and following HI. Control animals were randomly selected at each time interval (n = 25); PC animals represent 24 and 29.5 h of recovery (rec.) from hypoxia, corresponding to pre-HI and 4 h of recovery (n = 14). Five to seven animals were used for each recovery interval. Values are mean B SD. a p ! 0.05 (unpaired t test between baseline preconditioned versus naı¨ve control); b p ! 0.05; c p ! 0.01, and d p ! 0.001, using AVOVA with Tukey multiple comparison analysis as compared with the naı¨ve control.

equivalents (EE)], as well as glucose and glycogen [glucose equivalents (GE)]. In complete ischemia (i.e. no oxygen), glucose and glycogen are degraded to lactate. The energy utilization can be estimated in terms of high-energy phosphate, or FP, by the changes in these four compounds, as described by Lowry et al. [1964]. Thus, for each mole of lactate formed from free glucose, 1 FP is generated (2/ glucose), and for each mole of lactate generated from glycogen, there are 1.45 FP, which is based on the assumption of 10% branching. To estimate the amount of energy available from GE, calculated as [(glucose ! 2) + (glyco-

414

Dev Neurosci 2002;24:411–417

Fig. 3. Effect of preconditioning and HI on brain ATP and PFCr.

ATP and PFCr were analyzed in the ipsilateral hemispheres of animals described in figure 2. Values are mean B SD for numbers of animals described in figure 2. * p ! 0.05 [unpaired t test between 0 h recovery (rec.) of the HI animals vs. 0 h recovery of the PC-HI animals]; ** p ! 0.01, and *** p ! 0.001, using AVOVA with Tukey multiple comparison analysis as compared with the naı¨ve control.

gen ! 2.9)], and EE, calculated as [ATP ! 2 + PFCr], in the preconditioned and naı¨ve brains, GE and EE were calculated, and the values are presented in figure 4. Hypoxic preconditioning significantly increased the potential energy available from glucose and glycogen (GE) relative to naı¨ve controls at P7 (p ! 0.05), and 90 min of HI substantially depleted these reserves, with a trend towards better preservation in the PC pups, although this did not reach significance. Reoxygenation resulted in a complete recovery and increase in GE in both groups of rats, such that values at 4 and 24 h were significantly elevated over agematched controls with no specific effect from preconditioning. Because ATP and PFCr were slightly reduced in the PC group, the available energy (EE) was slightly, but

Brucklacher/Vannucci/Vannucci

Discussion

Fig. 4. Effect of preconditioning and HI on GE and EE. GE were calculated from the data of figures 2 and 3 as 2[Glu] +2.9[Gly], and EE as 2[ATP] + [PFCr] as originally described by Lowry et al. [1964]. * p ! 0.05 [unpaired t test between baseline preconditioned control versus naı¨ve control in both GE and EE and also 0 h recovery (rec.) of HI animals versus 0 h recovery of PC-HI animals in EE]; ** p ! 0.01, and *** p ! 0.001, using AVOVA with Tukey multiple comparison analysis as compared with the naı¨ve control.

significantly, reduced in these animals at the outset of HI. Despite this, total energy was better preserved such that at the end of 90 min of HI, there was significantly greater energy reserve in the preconditioned brains than in the naı¨ve animals, which had utilized 76% of their EE, relative to 40% used in the PC group (p ! 0.05). The secondary energy depletion was evident at 24 h of recovery in the naı¨ve HI animals (p ! 0.01 compared with the C animals), whereas hypoxic preconditioning completely prevented a secondary energy failure.

Effects of Hypoxic Preconditioning in the Immature Rat

Several studies have reported that hypoxic preconditioning provides tolerance to cerebral HI in the immature rat, although the mechanisms are still unclear. A previous study from our laboratory found no significant differences in levels of glycolytic and tricarboxylic intermediates or high-energy reserves at the end of 2.5 h of HI between naı¨ve and hypoxia-preconditioned immature rats. Accordingly, we reported that protection was not the result of differences in anaerobic glycolysis, tissue acidosis or depletion in high-energy reserves [Vannucci et al., 1998]. The results of the present study support a modification of our original conclusions, to indicate that hypoxic preconditioning does contribute to the preservation of high-energy reserves during HI in the immature rat brain by providing additional metabolic fuel as cerebral glycogen. Glycogen is the major energy reserve in the brain and accounts for approximately 65% of ATP that is generated under ischemic conditions [Lowry et al., 1964]. Studies have shown that the glycogen accumulation in the brain which follows trauma or a metabolic insult, while previously considered a sign of tissue damage, may represent a metabolic perturbation triggering enhanced glycogen synthesis and/or reduced degradation which might be protective [Folbergrova´ et al., 1996]. The majority of glycogen is localized to astrocytes [Koizumi, 1974; Phelps, 1975; Cataldo and Broadwell, 1986], and astrocytes have been shown to protect cultured neurons from anoxia or hypoglycemia-induced degeneration [Vibulsreth et al., 1987; Swanson and Choi, 1993]. Experimentally induced increases in astrocytic glycogen are protective in in vitro models of ischemia and glucose deprivation [Swanson and Choi, 1993; Wender et al., 2000]. We have previously observed increased glycogen content in the contralateral, hypoxic, hemisphere of the P7 rat brain at 24 h of recovery from HI [unpubl. obs.], suggesting that cerebral hypoxia induces a resynthesis of brain glycogen to levels above baseline. The results of figure 1 confirm this observation in the P6 rat brain as well, with the reduction of brain glycogen by 2.5 h of hypoxia followed by a rapid and robust resynthesis, representing a 28% increase over control at 24 h of recovery. As this is the interval between the preconditioning hypoxic stimulus and the subsequent HI, we propose that this newly synthesized glycogen represents an additional energy reserve available to the immature brain during HI. This is reflected in a significant preservation of high-energy reserves ATP and, to a lesser degree, PFCr in the preconditioned pups at 90 min of HI (fig. 3). Our previous metabolic study involved a severe

Dev Neurosci 2002;24:411–417

415

HI insult of 2.5 h duration which showed a similar energy depletion in both preconditioned and nonpreconditioned animals, although shorter periods of HI were not evaluated [Vannucci et al., 1998]. Thus, it is likely that preconditioning results in a delay in the depletion of the high-energy reserves during HI and constitutes a significant protective effect such that the secondary energy failure is prevented. The effectiveness of hypoxic preconditioning in preventing the secondary energy failure following HI in the immature rat brain is demonstrated in figures 3 and 4, and confirms our original report [Vannucci et al., 1998]. That this is due, at least in part, to a hypoxia-induced increase in brain glycogen prior to HI is a new observation, unique to this study. The questions that now arise relate to the way in which an increase in glycogen, presumably in astrocytes, protects the neurons from HI, and the mechanism by which hypoxia induces enhanced glycogen resynthesis. Two potential mechanisms by which glycogen in astrocytes could protect neurons from conditions of glucose deprivation, such as HI and hypoglycemia, were proposed by Swanson and Choi [1993]: (1) astrocytes utilize the energy derived from glycogen breakdown themselves to continue to fuel the energy-requiring process of glutamate uptake, thus protecting the neurons from glutamate excitotoxicity, or (2) astrocytes degrade the glycogen to lactate which they provide to the neurons to fuel their energy demands, especially during initial re-oxygenation. The first mechanism could clearly be operative during both HI and hypoglycemia, whereas the second mechanism could really only be useful during hypoglycemia since brain lactate is substantially elevated during HI, with no apparent need for additional lactate from astrocytes. A third possibility, however, would be degradation of astrocytic glycogen to glucose and release of glucose to be available to the neuron. Of greater relevance to issues of neuroprotection is the mechanism by which hypoxia actually induces enhanced glycogen resynthesis to achieve levels that are significantly increased relative to control. Hypoxia initiates a wide range of adaptive responses, largely through stabilization and activation of the transcription factor Hif-1 and the subsequent activation of several target genes involved in erythropoiesis, angiogenesis, vasodilation, glucose transport, and stimulation of anaerobic glycolysis [Semenza, 2000]. Hypoxic preconditioning increases Hif-1 expression in the immature rat brain, and is associated with increased gene and protein expression of the GLUT1 glucose transporter and the enzymes aldolase, phosphofructokinase, and lactate dehydrogenase [Bergeron et al., 2000; Jones and Bergeron, 2001]. An intracellular signal-

416

Dev Neurosci 2002;24:411–417

ing pathway upstream of these events involves hypoxiamediated activation of phosphatidylinositol 3-kinase (PI 3-K)/Akt which promotes stabilization of Hif-1; Akt also phosphorylates and inactivates the enzyme glycogen synthase kinase-3 (GSK-3) [reviewed in Cohen and Frame, 2001]. Although GSK-3 is now widely studied due to its ability to regulate a wide range of cellular functions involving gene expression, cytoskeletal integrity, as well as metabolism, it was first described for its ability to inhibit glycogen synthesis via phosphorylation of glycogen synthase [Cohen and Frame, 2001]. PI 3-K/Akt inactivation of GSK-3 is integral to insulin-mediated stimulation of glycogen synthesis in insulin-sensitive cells in the periphery; the possibility that this occurs in the brain in response to hypoxia provides a plausible mechanism of hypoxia-induced glycogen resynthesis reported here. In conclusion, the results reported here indicate that brain glycogen contributes to energy preservation during periods of cerebral hypoxia and is resynthesized in excess of control levels during the 24 h following the hypoxia. We propose that this involves activation of PI 3-K/Akt and inactivation of GSK-3, primarily in glycogen-containing astrocytes. This additional pool of brain glycogen is then available to provide sufficient metabolic fuel to delay the depletion of high-energy reserves during a subsequent period of HI and prevent the secondary energy failure that is consistent with tissue damage. Whether this is due to the maintenance of astrocytic energy balance and continued glutamate uptake, or delivery of metabolic energy to neurons in the form of glucose, or both was not determined in this study.

Brucklacher/Vannucci/Vannucci

References Bergeron M, Gidday JM, Yu AY, Semenza GL, Ferriero DM, Sharp FR (2000): Role of hypoxia-inducible factor-1 in hypoxia-induced ischemic tolerance in neonatal rat brain. Ann Neurol 48:285–296. Cataldo AM, Broadwell RD: Cytochemical identification of cerebral glycogen and glucose-6phosphatase activity under normal and experimental conditions. Neurons and glia. J Electron Microsc Tech 3:413–437. Chen J, Simon R (1997): Ischemic tolerance in the brain. Neurology 48:306–311. Cohen P, Frame S: The renaissance of GSK3. Nat Rev Mol Cell Biol 2001;2:769–776. Folbergrova´ J, Katsura, K, Siesjö BK (1996): Glycogen accumulated in the brain following insults is not degraded during a subsequent period of ischemia. J Neurol Sci 137:7–13. Gidday JM, Fitzgibbons JC, Shah AR, Park TS (1994): Neuroprotection from ischemic brain injury by hypoxic preconditioning in the neonatal rat. Neurosci Lett 168:221–224. Hernandez MJ, Vannucci RC, Salcedo A, Brennan RW (1980): Cerebral blood flow and metabolism during hypoglycemia in newborn dogs. J Neurochem 354:622–628. Jones NM, Bergeron M (2001): Hypoxic preconditioning induces changes in HIF-1 target genes in neonatal rat brain. J Cereb Blood Flow Metab 21:1105–1114. Kato H, Araki T, Kogure K (1992): Repeated focal cerebral ischemia in gerbils is associated with development of infarction. Brain Res 596:315– 319.

Effects of Hypoxic Preconditioning in the Immature Rat

Kirino T (1982): Delayed neuronal death in the gerbil hippocampus following ischemia. Brain Res 239:57–69. Koizumi J (1974): Glycogen in the central nervous system. Prog Histochem Cytochem 6:1–35. Liu Y, Kato H, Nakata N, Kogure K (1992): Protection of rat hippocampus against ischemic neuronal damage by pretreatment with sublethal ischemia. Brain Res 586:121–124. Lowry OH, Passonneau JY, Hasselberger FH, Schulz DW (1964): Effect of ischemia on known substrates and cofactors of the glycolytic pathway in brain. J Biol Chem 239:18–30. Palmer C, Brucklacher RM, Christensen MA, and Vannucci RC (1990): Carbohydrate and energy metabolism during the evolution of hypoxicischemic brain damage in the immature rat. J Cereb Blood Flow Metab 10:227–235. Phelps CH (1975): An ultrastructural study of methionine sulphoximine-induced glycogen accumulation in astrocytes of the mouse cerebral cortex. J Neurocytol 4:479–490. Rice JE, Vannucci RC, Bierly JB (1981): The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 9:131– 141. Semenza GL (2000): HIF-1:Mediator of physiological and pathophysiological responses to hypoxia. J Appl Physiol 88:1474–1480. Swanson RA, Choi DW (1993): Glial glycogen stores affect neuronal survival during glucose deprivation in vitro. J Cereb Blood Flow Metab 13:162–169.

Vannucci RC, Duffy TE (1974): The influence of birth on carbohydrate and energy metabolism in rat brain. Am J Physiol 226:933–940. Vannucci SJ, Seaman LB, Vannucci RC (1996): Effects of hypoxia-ischemia on GLUT1 and GLUT3 glucose transporters in immature rat brain. J Cereb Blood Flow Metab 16:77–81. Vannucci RC, Towfighi J, Vannucci SJ (1998): Hypoxic preconditioning and hypoxic-ischemic brain damage in the immature rat: Pathologic and metabolic correlates. J Neurochem 71:1215–1220. Vibulsreth S, Hefti F, Ginsberg MD, Dietrich WD, Busto R (1987): Astrocytes protect cultured neurons from degeneration induced by anoxia. Brain Res 422:303–311. Wender R, Brown AM, Fern R, Swanson RA, Farrell K, Ransom BR (2000): Astrocytic glycogen influences axon function and survival during glucose deprivation in cerebral white matter. J Neurosci 20:6804–6810. Williams GD, Pamer C, Roberts RC, Heitjan DF, Smith MB (1992): 31P NMR spectroscopy of perinatal hypoxic-ischemic brain damage: A model to evaluate neuroprotective drugs in immature rats. NMR Biomed 5:145–153. Yager JY, Brucklacher RM, Vannucci RC (1992): Cerebral energy metabolism during hypoxiaischemia and early recovery in the immature rat. Am J Physiol 262:H672–H677.

Dev Neurosci 2002;24:411–417

417

Original Paper Received: October 28, 2002 Accepted: October 31, 2002

Dev Neurosci 2002;24:418–425 DOI: 10.1159/000069053

Early Appearance of Functional Deficits after Neonatal Excitotoxic and Hypoxic-Ischemic Injury: Fragile Recovery after Development and Role of the NMDA Receptor Barbara T. Felt a, b Timothy Schallert b–d Jie Shao b Yiqing Liu a Xiaoling Li b John D.E. Barks a, b a Department

of Pediatrics, b Center for Human Growth and Development, c Department of Neurosurgery, University of Michigan, Ann Arbor, Mich.; d Department of Psychology and Institute for Neuroscience, University of Texas at Austin, Austin, Tex., USA

Key Words Cerebral ischemia W Stroke W Cerebral hypoxia-ischemia W Neonatal brain injury W Sensorimotor deficit W Plasticity W Excitotoxicity W N-methyl-D-aspartate W MK-801

Abstract We sought to determine whether neonatal rats that sustain unilateral cerebral hypoxic-ischemic or excitotoxic insults (1) manifest contralateral sensorimotor deficits during development or in adulthood and (2) recover from those deficits. Seven-day-old (P7) rats received a right intrastriatal injection of the glutamate analog N-methyl-D-aspartate (NMDA). Unilateral hypoxia-ischemia (HI) was induced by right carotid ligation followed by 1.5 h in 8% O2. Both procedures produce neuronal loss in the striatum and sensorimotor cortex. Nonlesioned controls were included. We scored percent forepaw placement on the edge of a horizontal surface, with lateral vibrissa stimulation, from P9 to P19, and at P33 and P50. Then, on P50, rats were treated with the NMDA antagonist MK-801 to determine whether deficits could be reinstated. NMDA- and HI-lesioned rats exhibited a

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0418$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

deficit in contralateral vibrissa-stimulated forepaw placing that emerged during the second week of life. Yet, by P33 and P50, the lesioned groups and controls were indistinguishable. MK-801 injection on P50 resulted in transient reinstatement of the placing deficit. After unilateral neonatal excitotoxic or HI brain injury, contralateral sensorimotor deficits are detected, but in many animals, these deficits have resolved by adulthood. Thus, timing of sensorimotor tests may influence their sensitivity for detection of focal neuropathology originating in the neonatal period. Copyright © 2002 S. Karger AG, Basel

Introduction

Human trials of promising neuroprotective therapies (e.g. hypothermia) utilize neurologic function as their primary outcome measure. In rodent trials of potential neuroprotective therapies, there is increasing use of functional outcomes to complement neuropathologic outcomes. Questions that still need to be answered regarding such testing in developing rodents include: ‘what are the appro-

Timothy Schallert, PhD Department of Psychology, Seay Building Campus Mail Code: A8000, University of Texas at Austin Austin, TX 78712 (USA) Tel. +1 512 471 6141, E-Mail [email protected] or [email protected]

priate tests?’ and, ‘when is the optimal time to test?’ An additional question, closely related to the latter question, is ‘do deficits change over time?’ In adult rodents, there is evidence, in several injury models, that sensorimotor deficits become less pronounced over time [Schallert et al., 2000]. Such deficits can, however, be reinstated pharmacologically using central nervous system (CNS) depressant drugs [Barth et al., 1990; Schallert et al., 1986]. A variety of cognitive and sensorimotor deficits have been reported in immature rats that undergo hypoxic-ischemic lesions [Almli et al., 2000; Altemus and Almli, 1997; Balduini et al., 2000; Bona et al., 1998; Bona et al., 1997; Chou et al., 2001; Ford et al., 1989; Ikeda et al., 2001; Jansen and Low, 1996a; Jansen and Low, 1996b; Liu et al., 2001; Tomimatsu et al., 2002; Wagner et al., 2002; Young et al., 1986]. Many reports of functional deficits after neonatal hypoxia-ischemia (HI) involve only a single time point; few incorporate serial testing. Rats with moderately severe hypoxic-ischemic lesions on day 7 of life (P7) had a deficit in Rota-Rod sensorimotor performance at 3 weeks that persisted in weekly testing until 9 weeks of life [Jansen and Low, 1996a]. The same group of rats also manifested asymmetry in apomorphine-stimulated rotation from 3 to 15 weeks of age [Jansen and Low, 1996a], consistent with severe striatal damage [Schallert and Whishaw, 1985]. In an immature (P17) rat model of traumatic brain injury, sensorimotor deficits resolved over time [Adelson et al., 1997]. Kolb [1999] has shown that after neonatal cortical aspiration lesions, the extent and time course of functional recovery parallels changes in dendritic morphology in the remaining intact cortex. In a preliminary experiment to characterize the evolution of sensorimotor deficits after an episode of neonatal cerebral HI, we found evidence of recovery or compensation from a unilateral deficit. Specifically, in a test of paw placement while exploring a Plexiglas cylinder, rats that underwent unilateral (right) HI on P7 exhibited a rightsided paw weight-bearing preference, peaking on P35. Yet these same rats no longer had this ipsilateral preference as adults [Barks et al., 2002]. In the experiment reported here, we injected the glutamate analog N-methyl-D-aspartate (NMDA) into the right striatum. Our goal was to create lesions that were less heterogeneous than HI lesions and that were sufficiently localized to produce predominantly motor deficits. These animals were serially tested over the next 2 months to evaluate reflex ontogeny, forelimb placing and strength, and sensorimotor integration.

Functional Deficits in Neonatal Brain Injury

Methods NMDA Lesioning Isoflurane-anesthetized P7 Sprague-Dawley rats from two litters (n = 10) underwent a midline scalp incision followed by stereotaxic right intrastriatal injections of NMDA (Sigma) 10 nmol in 0.5 Ìl (pH 7.4), using coordinates (relative to bregma) AP 0, ML 2.0, V 3.8 mm [Hagan et al., 1996]. Pups recovered in a 37 ° C incubator and were returned to the dam when active. To begin to evaluate whether recovery of function was possible after other neonatal brain injuries, unilateral cerebral HI was elicited in a small number of littermates in one experiment. In 3 rats, double ligation and division of the right carotid artery was followed 1 h later by 1.5 h in 8% O2 (balance nitrogen). Littermate controls (n = 12) did not undergo surgery. Both the NMDA lesioning and HI lesioning protocols were designed to produce lesions of moderate severity. NMDA Antagonist Injection On P50, after sensorimotor testing, all NMDA-injected and HI rats received an intraperitoneal injection of the noncompetitive NMDA receptor antagonist MK-801 (dizocilpine maleate, SigmaRBI, 1 mg/kg in PBS). The purpose of these injections was to determine whether, as in adult-lesioned rats, a sensorimotor deficit could be reinstated after behavioral recovery had occurred [Barth et al., 1990]. Developmental Assessment Measures of development that were assessed at P7 (pre-NMDA lesion), and P9, 11, 13, and 15, using a 0- to 5-point Likert scale, included: growth (fur, ear, eye opening); general development (auditory startle, bar hold, surface righting, negative geotaxis), and the sensorimotor skill of forelimb placement [Altman and Bayer, 1997; Lapointe and Nosal, 1979]. Forelimb placement was assessed at the above ages and at P17 and P19, by the number placed of 10 trials/ limb in response to ipsilateral vibrissa or head-on (chin) stimulus on a horizontal surface. Postdevelopmental Sensorimotor Assessment At P33, P50 (pre-MK-801) and P51–P53 (post-MK-801), forelimb placement in response to ipsilateral vibrissae and head-on stimuli were assessed as above for both litters [Hua et al., 2002; Schallert et al., 2000; Schallert et al., 2002b; Wu et al., 2002]. Other sensorimotor assessments included: cylinder test [number of placements of each forelimb on rearing in a clear Plexiglas cylinder (R – L)/total]; sticker test (number of and latency to contact and removal of an adhesive dot placed on each upper extremity); cross-vibrissa placing (number placed of 10 trials/limb in response to contralateral vibrissa stimulus), and parachute test (posture in response to tail suspension; number of left-turn postures per 10 trials) (fig. 1) [Whishaw et al., 1981]. The cylinder and sticker tests were completed at P33, P50 and P53 in the first litter. The cross-vibrissa and parachute tests were completed at P50 and P51–P53 for the second litter. Brain Pathology After the completion of all testing, rats were weighed and euthanized. Brains from the first litter were weighed and then frozen under powdered dry ice, for coronal sectioning and cresyl violet staining. Brains from the second litter were fixed in situ with PBS followed by 4% paraformaldehyde under deep chloral hydrate anesthesia; in this litter, 1 NMDA-injected and 1 control rat were sacrificed on P21 to

Dev Neurosci 2002;24:418–425

419

severity, the percent atrophy (i.e. percent damage) was calculated from the bilateral mean regional cross-sectional areas for each animal, using the formula % damage = 100 ! (L – R)/L. The relationship between the ultimate striatal and cortical percent damage and percent success in contralateral forelimb placing was evaluated by linear regression at P11, P13, P15, P17, P19, P33 and P50–P53.

Results

Growth and Physical Development There were no differences in day of eye opening or ear opening, nor in fur development between the NMDAinjected, HI and control pups. There was no difference in mean body weight between groups, at the end of testing (P54). Reflex Development There were no differences among groups in auditory startle, surface righting, negative geotaxis or bar hold during development.

Fig. 1. Tail-suspension response in normal and NMDA-lesioned

rats. At P50–P53, rats were suspended by the tail above a horizontal surface and lowered toward the surface. In the normal response, a nonlesioned animal reaches toward the surface below with both forelimbs extended and abducted; the hindlimbs are also extended and abducted (a). b Abnormal response frequently seen in the neonatally NMDA-lesioned rats. The contralateral forelimb (in this case, the left) is flexed, adducted and internally rotated, and both hindlimbs are adducted and internally rotated, with the hindpaws clasping each other.

assess the extent of injury and microgyria at this age. Cortical and striatal cross-sectional areas in 20-Ìm coronal sections (5/brain through striatum and dorsal hippocampus) were measured by microcomputer-based image analysis (NIH image). As an index of atrophy, percent damage of each ipsilateral structure, compared with the nonlesioned side, was calculated using bilateral regional areas, by the formula % damage = 100 ! (L – R)/L. Sections from NMDA-lesioned rats in the second litter, at P54, were also stained by the von Kossa method with a neutral red counterstain to demonstrate calcium deposition [Pathology, 1992]. Data Analysis Measures of development and sensorimotor skill emergence, and percent forelimb placement, were compared among NMDA-injected and control rats, and HI littermates, by ANOVA or t test for continuous and ¯2 for categorical variables. Repeated-measures ANOVA was also applied as appropriate. Significance was set at p ! 0.05. Body and brain weights were compared between groups by t test or ANOVA. As an index of ipsilateral striatal and cortical damage

420

Dev Neurosci 2002;24:418–425

Forelimb Placing In nonlesioned controls, vibrissa-stimulated lateral forepaw placement was not detected until P11, and was fully established (10/10 trials in all subjects) by P19 (fig. 2). In the NMDA-lesioned rats, the ontogeny of ipsilateral (right-sided) forepaw placement was indistinguishable from nonlesioned controls. In the NMDA-lesioned rats, there was no delay in onset of contralateral (leftsided) vibrissa-stimulated forepaw placement, but there was an overall performance deficit compared with controls during the preweaning period (fig. 2; p ! 0.05, repeated-measures ANOVA). When evaluating results on individual testing days, between-group differences in contralateral placing were not detected on P11 or P13, but they were detected on P15, P17 and P19 (fig. 2; p ! 0.001, Student t test). On P33 and P50, contralateral performance in the NMDA-lesioned rats was no longer different from controls (fig. 2). The 3 HI rats exhibited a transient contralateral deficit in forepaw placing, on P15 and P17, which had resolved by P19 (fig. 2; p ! 0.02, Student t test). Twenty-four hours after MK-801 injection on P50, control rats exhibited a mild bilaterally symmetric reduction in forepaw placing that resolved at 72 h after MK-801. Among the NMDA-injected rats, treatment with MK-801 on P50 resulted in a total (0/10 trials) deficit in contralateral vibrissa-stimulated placing 24 h later in all animals that had resolved by 72 h after MK-801 injection (fig. 2). Among the HI comparison group, 2/3 exhibited marked deterioration (0–1/10 trials) in contralateral placing 24 h

Felt/Schallert/Shao/Liu/Li/Barks

Fig. 2. Neonatal NMDA and HI lesioning induce deficits in forepaw placing ontogeny. a, b The mean B SE of vibrissa-stimulated lateral forelimb percent placement (of 10 trials/side), from P9 to P33 for both the contralateral (left, a) and ipsilateral (right, b) sides, in rats that received either intrastriatal NMDA injections (10 nmol, n = 10) on P7 (NMDA), noninjected littermate controls (n = 12), and a comparison group of 3 littermates that underwent unilateral right cerebral HI (HI; see Methods). As the lateral vibrissa placing response developed in controls, the NMDA-injected animals exhibited a contralateral placing deficit (p ! 0.01, repeated measures ANOVA). When group data from individual days were compared, the NMDA group had a lateral placing deficit on P15, P17 and P19 (** p ! 0.001, t test), and the HI group had a placing deficit on P15 and P17 (* p ! 0.02, t test). c, d By P33 and P50, there was no significant difference

in vibrissa-stimulated contralateral forelimb placing among groups (mean B SE). No deficit in vibrissa-stimulated placing was detected ipsilateral to the NMDA or HI lesions. After vibrissa-stimulated forepaw placing testing on P50, all rats received an injection of the NMDA antagonist and CNS depressant MK-801 (1 mg/kg, i.p.). Twenty-four hours later, controls had a modest bilateral deterioration in placing, which had resolved by 72 h. In contrast, the entire NMDA-lesioned group was unable to place contralaterally (c), 24 h after MK-801 (* p ! 0.02, ANOVA with Fisher PLSD, NMDA vs. control). The ipsilateral placing performance of NMDA-lesioned rats 24 h after MK-801 was similar to controls (d). The MK-801-induced forepaw-placing deficit in the NMDA-lesioned animals had resolved by 72 h after MK-801 treatment. Similar trends were detected in the HI group (not shown).

after MK-801 injection. No lateralizing deficit in head-on (chin-stimulated) forepaw placing was detected either during development, prior to or following MK-801 treatment. To attempt to determine whether placing deficits were attributable to sensory vs. motor dysfunction, we tested cross-midline vibrissa-stimulated placing in animals of the second litter. On P50, 1 each of the NMDA and HI animals had a residual contralateral vibrissa-stimulated placing deficit (2 left forepaw placings/10 trials after left vibrissa stimulation). When these animals were tested by crossed vibrissa placing, i.e. the left vibrissae were stimu-

lated by the edge of the bench, with the left forelimb restrained, the right forepaw was successfully placed in 10/10 trials. The latter finding implies that left vibrissa sensation was intact, and that the lesion-induced contralateral (left) forelimb dysfunction was primarily motoric. The sensitivity of lateral vibrissa-stimulated forelimb placing to both striatal and cortical lesion severity during development was evaluated by linear regression, which included all subjects, both lesioned and controls. There was a linear relationship between striatal percent atrophy and contralateral percent placement on P15, P17 and P19 (R2 = 0.426, 0.254 and 0.179, respectively, p ! 0.05), but

Functional Deficits in Neonatal Brain Injury

Dev Neurosci 2002;24:418–425

421

not at earlier ages nor at P33. There was a linear relationship between cortical percent atrophy and contralateral percent placement only on P15 (R2 = 0.2790, p ! 0.05). On P50, prior to MK-801 treatment, and again on P53, 72 h after MK-801 treatment, there was a linear relationship between both striatal and cortical percent atrophy and percent contralateral placement (R2 = 0.32 and 0.493, respectively, on P50, and 0.194 and 0.36, respectively, on P53, p ! 0.05). Twenty-four hours after MK-801 treatment, when contralateral placing was at its nadir in lesioned animals, there was no relationship between damage severity and percent placement. Tail Suspension In all (4/4) controls in the second litter, the normal symmetric hindlimb response (fig. 1a) was elicited on P35. In all (3/3) NMDA-injected rats, and in 2/3 HI rats in the second litter, hindlimb adduction and clasping was elicited. The number of abnormal forelimb and hindlimb responses in the NMDA-lesioned rats was greater than in the controls (median, control vs. NMDA: forelimb 2.5 vs. 8, hindlimb 0 vs. 8, p ! 0.05, Mann-Whitney). Other Assessments There was no significant difference between the experimental groups for latency to touch or remove the stickers at P33, before MK-801 (P50) or at 72 h after MK-801 (P53). Similarly, there was no significant group difference for upper extremity use in the cylinder task at P33, P50 and P51.

Fig. 3. Histopathology in NMDA-injected rats. Coronal cresyl violet-stained sections at the level of the rostral striatum (left column) and anterior commisure (right column) are presented for each of 6 rats from a single experiment that received right intrastriatal injections of NMDA (10 nmol) 7 week earlier. Note ipsilateral striatal atrophy in all cases, and dilatation of the ipsilateral ventricle (black asterisk) in 5/6 cases (a–e). Abnormal cortical morphology, including microgyria (black arrows) and abnormal layering (white asterisk) is evident in 4/6 (a–d). C = Cortex; Str = striatum. Scale bar = 1 mm.

422

Dev Neurosci 2002;24:418–425

Neuropathology In the first litter, the brains of the NMDA-injected rats were 7% smaller than those of controls (mean B SD: NMDA-injected 1.64 B 0.08 g; control 1.77 B 0.06 g, p ! 0.01). Based on our previous experience evaluating NMDA-induced neuropathology 5 days after P7 lesioning, we expected to find ipsilateral striatal atrophy and thinning of the overlying neocortex, as a result of excitotoxic neuronal loss in the striatum and cortex, with ex vacuo dilatation of the right lateral cerebral ventricle [Hagan et al., 1996]. These expectations were confirmed on examination at 47 days after intrastriatal NMDA lesioning; there was a considerable degree of heterogeneity in damage severity (fig. 3). Two additional findings were noteworthy. First, areas of microgyria dorsolateral to the lesioned striatum, primarily seen rostral to the anterior genu of the corpus callosum, were noted in 8/10 NMDAinjected rats. Calcium deposition was assessed in the second litter at P54 and noted in the ipsilateral hemisphere

Felt/Schallert/Shao/Liu/Li/Barks

in the NMDA-injected rats. Foci of ipsilateral calcification included the dorsolateral striatum and deep layers of the overlying residual cortex (fig. 4). Of the 3 HI animals, all followed to P54, 2 had moderate ipsilateral striatal and cortical infarction with foci of microgyria and 1 had complete ipsilateral hemisphere infarction.

Discussion

Based on the location of the excitotoxic lesions in these experiments, with damage to the striatum and overlying cortex, a sensorimotor deficit, such as a contralateral deficit in vibrissa-stimulated lateral forepaw placing is not unexpected. Analogous functional deficits have been reported in adult rats with similarly located lesions [Schallert et al., 2000]. Similar deficits were detected in the small number of HI animals. Crossed vibrissa-stimulation testing suggested that these performance deficits reflected more motor than sensory dysfunction [Schallert et al., 2002a]. The deficits in lateral vibrissa-stimulated placement were dependent on lesion severity both during development and in adulthood. The functional deficits that we detected in the preweaning period were transient, consistent with previous reports of plasticity in the primate developing nervous system after injury [Goldman and Galkin, 1978]. The deficit in contralateral vibrissastimulated lateral forelimb placing demonstrated substantial recovery during the preweaning period, i.e. within 2 weeks of lesioning. Recovery from sensorimotor and cognitive deficits has been reported after traumatic brain injury in immature rats; the time course of functional recovery was 3–10 days, depending on lesion severity [Adelson et al., 1997]. Recovery of function has also been described after ischemic and ablation cortical injury in adult rats, but in mature animals, reports suggest less recovery of function when the lesion includes deeper structures [Schallert et al., 2000]. The placing deficit was not detected immediately after lesioning. The most likely explanation for the latter finding is that a deficit could not be detected until the age at which rats normally develop a reliable placing response during the second week of life. That is, the animals developed into the impairment [Goldman, 1978]. In clinical practice, similar situations arise; a hemiparesis may not become evident until the age at which an infant would be expected to have voluntary control of reaching and grasping. Our inability to detect a deficit in head-on forepaw placement indicates that the motor program for contralateral forelimb placing per se is not impaired but rather the

Functional Deficits in Neonatal Brain Injury

Fig. 4. Calcium deposition in NMDA-lesioned brain. This representative section from a P54 rat that received intrastriatal injections of NMDA (10 nmol) on P7 was stained by the von Kossa method (see Methods) to detect deposition of calcium in brain tissue. Note the black staining (silver grains), indicative of calcium deposition, in the lesioned dorsolateral striatum (white asterisk) and in the deep layer of the residual, damaged overlying cortex (arrowheads). There is no calcium deposition in the remaining upper cortical layers of a lesioninduced microgyrus (arrow). Scale bar = 1 mm.

capacity for the vibrissae on the contralateral side to initiate placing in that forelimb. The head-on procedure stimulates both ipsilateral and contralateral vibrissae. To the best of our knowledge, only when there is severe damage to striatal or nigrostriatal neurons is there a lasting deficit in the capacity of the nonimpaired vibrissae, when stimulated, to initiate placing in the impaired forelimb (as though sensory input to the intact hemisphere can activate motor programs in the injured hemisphere). Therefore, it is not unexpected that the head-on procedure would not detect impairment in the neonatal HI and NMDA-injected animals at time points when sensory input into the injured hemisphere is unable to consistently activate motor programs in that same injured hemisphere. It is also unlikely that there is sensory impairment in the contralateral vibrissae, based on our crossed stimulation studies. In every unilateral adult injury model examined, including severe striatal and cortical ischemic or neurotoxic injuries from which forelimb placing does not recover, we have found in cross-midline placing tests that the vibrissae on the contralateral side of the body can readily activate placing of the ipsilateral forelimb [Schallert et al., 2002a; Schallert et al., 2002b].

Dev Neurosci 2002;24:418–425

423

Microgyria has been described previously after focal cortical freeze or excitotoxic lesions on the first or second day of life in rats or mice [Humphreys et al., 1991; Marret et al., 1995; Redecker et al., 1998]. In our previous experience with striatal NMDA lesioning, we did not note microgyria, but did not evaluate neuropathology later than P12 [Hagan et al., 1996]. In younger animals, acquired microgyria takes between 5 and 10 days to develop after freeze-induced cortical necrosis [Rosen et al., 1992]. In our previous studies with striatal NMDA injections, we did not evaluate for calcium deposition at P12. Yet, the calcium deposits at P54 in NMDA-lesioned animals are not an unexpected finding. Calcium deposition is frequently seen in human infants after neonatal-perinatal hypoxic-ischemic brain injury [Ansari et al., 1990], and has been described in neonatal and adult cortical aspiration lesions [Braun et al., 1972; Whishaw et al., 1981] and in excitotoxic lesions of adult rats [Mahy et al., 1999]. Our findings suggest that in developing rodents, depending on the lesion and the test used, the time window for detection of functional deficits and for detection of behavioral evidence of plasticity may be quite narrow. Thus, the persistence of Rota-Rod sensorimotor performance deficits from 2–8 weeks after neonatal HI lesioning [Jansen and Low, 1996a] could represent a missed opportunity to detect earlier evidence of recovery rather than absence of plasticity. The phenomenon of recovery from functional deficits might also underlie some reported difficulties in detecting functional evidence of neuroprotection using a battery of sensorimotor tests 5 weeks after treatment of neonatal rats with post-HI hypothermia [Bona et al., 1998]. Our findings suggest that the time course of functional deficits in immature rats should be carefully characterized to facilitate optimal choice of time for testing. The forelimb placing deficit in the neonatally NMDAinjected rats was transiently reinstated by treatment with the noncompetitive NMDA receptor antagonist MK-801.

Reinstatement of injury-induced functional deficits in adult rats after MK-801 or ethanol is associated with alteration of postinjury dendritic plasticity [Kozlowski et al., 1997; Kozlowski et al., 1994]. Our findings indicate that behavioral plasticity in response to neonatal brain injury is fragile, and NMDA receptor dependent. During normal CNS development, excitatory amino acid (EAA) receptors play a key role in use-dependent plasticity, and chronic treatment with EAA receptor antagonists can disrupt developmental plasticity [Gu et al., 1989; Ramoa et al., 2001; Schlaggar and O’Leary, 1993]. Thus, we speculate that EAA receptor-dependent neuroplasticity may also play a role in the recovery of function after neonatal brain injury. Regardless of the mechanism underlying the reinstatement of the functional deficit by MK-801, from a practical viewpoint, the injection of MK-801 in adult rats might be a useful adjunct to increase the sensitivity of testing for sensorimotor deficits related to neonatal lesioning. In conclusion, we have demonstrated sensorimotor deficits and patterns of recovery using a battery of tests that assess distinct functional capabilities after neonatal brain injury. In our own experience and in published reports, adult animals with similar severity of damage would not be expected to demonstrate the degree of functional sparing or recovery seen in our developing animals. If we had only examined these animals behaviorally early after injury, we would have failed to appreciate the amount of functional recovery that could occur. Using multiple behavioral tests at multiple time points was important in detecting the degree of injury as well as the extent of recovery/compensation. Knowledge about the onset of a development-sensitive behavioral deficit and its temporal pattern of recovery, as well as of residual deficits and spared functions or compensations, would be useful for targeting studies of time-dependent molecular, cellular and structural events that might be involved in the mechanisms underlying recovery from early brain injury.

References Adelson PD, Dixon CE, Robichaud P, Kochanek PM (1997): Motor and cognitive functional deficits following diffuse traumatic brain injury in the immature rat. J Neurotrauma 14:99– 108. Almli CR, Levy TJ, Han BH, Shah AR, Gidday JM, Holtzman DM (2000): BDNF protects against spatial memory deficits following neonatal hypoxia-ischemia. Exp Neurol 166:99– 114.

424

Altemus KL, Almli CR (1997): Neonatal hippocampal damage in rats: Long-term spatial memory deficits and associations with magnitude of hippocampal damage. Hippocampus 7: 403–415. Altman J, Bayer SA (1997): Epilogue: Behavior consequences of experimental interference with cerebellar development; in Altman J, Bayer SA: Development of the Cerebellar System: In Relation to Its Evolution, Structure

Dev Neurosci 2002;24:418–425

and Function. Boca Raton, CRC Press, pp 726–751. Ansari MQ, Chincanchan CA, Armstrong DL (1990): Brain calcification in hypoxic-ischemic lesions: An autopsy review. Pediatr Neurol 6: 94–101. Balduini W, De Angelis V, Mazzoni E, Cimino M (2000): Long-lasting behavioral alterations following a hypoxic/ischemic brain injury in neonatal rats. Brain Res 859:318–325.

Felt/Schallert/Shao/Liu/Li/Barks

Barks JD, Li X, Pecina S, Sloan KE, Liu Y, Schallert T, Felt BT (2002): Delayed appearance of functional deficits after neonatal hypoxic-ischemic and excitotoxic brain injury, and recovery during late development. Pediatr Res 51: 456A. Barth TM, Grant ML, Schallert T (1990): Effects of MK-801 on recovery from sensorimotor cortex lesions. Stroke 21(11 suppl):153–157. Bona E, Hagberg H, Loberg EM, Bagenholm R, Thoresen M (1998): Protective effects of moderate hypothermia after neonatal hypoxia-ischemia: Short- and long-term outcome. Pediatr Res 43:738–745. Bona E, Johansson BB, Hagberg H (1997): Sensorimotor function and neuropathology five to six weeks after hypoxia-ischemia in seven-day-old rats. Pediatr Res 42:678–683. Braun JJ, Slick TB, Lorden JF (1972): Involvement of gustatory neocortex in the learning of taste aversions. Physiol Behav 9:637–641. Chou I-C, Trakht T, Signori C, Smith J, Felt BT, Vazquez DM, Barks JDE (2001): Behavioral/ environmental intervention improves learning after cerebral hypoxia-ischemia in rats. Stroke 32:2192–2197. Ford LM, Sanberg PR, Norman AB, Fogelson MH (1989): MK-801 prevents hippocampal neurodegeneration in neonatal hypoxic-ischemic rats. Arch Neurol 46:1090–1096. Goldman PS (1978): Neuronal plasticity in primate telencephalon: Anomalous projections induced by prenatal removal of frontal cortex. Science 202:768–770. Goldman PS, Galkin TW (1978): Prenatal removal of frontal association cortex in the fetal rhesus monkey: Anatomical and functional consequences in postnatal life. Brain Res 152:451– 485. Gu QA, Bear MF, Singer W (1989): Blockade of NMDA-receptors prevents ocularity changes in kitten visual cortex after reversed monocular deprivation. Brain Res Dev Brain Res 47:281– 288. Hagan P, Barks JDE, Yabut M, Davidson BL, Roessler B, Silverstein FS (1996): Adenovirusmediated over-expression of interleukin-1 receptor antagonist reduces susceptibility to excitotoxic injury in perinatal rats. Neuroscience 75:1033–1045. Hua Y, Schallert T, Keep RF, Wu J, Hoff JT, Xi G (2002): Behavioral tests after intracerebral hemorrhage in the rat. Stroke 33:2478–2484. Humphreys P, Rosen GD, Press DM, Sherman GF, Galaburda AM (1991): Freezing lesions of the developing rat brain: A model for cerebrocortical microgyria. J Neuropathol Exp Neurol 50:145–160.

Functional Deficits in Neonatal Brain Injury

Ikeda T, Mishima K, Yoshikawa T, Iwasaki K, Fujiwara M, Xia YX, Ikenoue T (2001): Selective and long-term learning impairment following neonatal hypoxic-ischemic brain insult in rats. Behav Brain Res 118:17–25. Jansen EM, Low WC (1996a): Long-term effects of neonatal ischemic-hypoxic brain injury on sensorimotor and locomotor tasks in rats. Behav Brain Res 78:189–194. Jansen EM, Low WC (1996b): Quantitative analysis of contralateral hemisphere hypertrophy and sensorimotor performance in adult rats following unilateral neonatal ischemic-hypoxic brain injury. Brain Res 708:93–99. Kolb B (1999): Synaptic plasticity and the organization of behaviour after early and late brain injury. Can J Exp Psychol 53:62–76. Kozlowski DA, Hilliard S, Schallert T (1997): Ethanol consumption following recovery from unilateral damage to the forelimb area of the sensorimotor cortex: Reinstatement of deficits and prevention of dendritic pruning. Brain Res 763:159–166. Kozlowski DA, Jones TA, Schallert T (1994): Pruning of dendrites and restoration of function after brain damage: Role of the NMDA receptor. Restor Neurol Neurosci 7:119–126. Lapointe G, Nosal G (1979): A rat model of neurobehavioral development. Experientia 35:205– 207. Liu XH, Eun BL, Barks JDE (2001): Platelet-activating factor antagonist BN 50730 attenuates hypoxic-ischemic brain injury in neonatal rats. Pediatr Res 49:804–811. Mahy N, Prats A, Riveros A, Andres N, Bernal F (1999): Basal ganglia calcification induced by excitotoxicity: An experimental model characterised by electron microscopy and X-ray microanalysis. Acta Neuropathol (Berl) 98:217– 225. Marret S, Mukendi R, Gadisseux J-F, Gressens P, Evrard P (1995): Effect of ibotenate on brain development: An excitotoxic mouse model of microgyria and posthypoxic-like lesions. J Neuropathol Exp Neurol 54:358–370. Pathology AFIO (1992): Laboratory Methods in Histotechnology. Washington, American Registry of Pathology. Ramoa AS, Mower AF, Liao D, Jafri SI (2001): Suppression of cortical NMDA receptor function prevents development of orientation selectivity in the primary visual cortex. J Neurosci 21:4299–4309. Redecker C, Hagemann G, Witte OW, Marret S, Evrard P, Gressens P (1998): Long-term evolution of excitotoxic cortical dysgenesis induced in the developing rat brain. Brain Res Dev Brain Res 109:109–113.

Rosen GD, Press DM, Sherman GF, Galaburda AM (1992): The development of induced cerebrocortical microgyria in the rat. J Neuropathol Exp Neurol 51:601–611. Schallert T, Fleming SM, Leasure JL, Tillerson JL, Bland ST (2000): CNS plasticity and assessment of forelimb sensorimotor outcome in unilateral rat models of stroke, cortical ablation, parkinsonism and spinal cord injury. Neuropharmacology 39:777–787. Schallert T, Fleming SM, Woodlee ST (2002a): Should the injured and intact hemispheres be treated differently during the early phases of physical restorative therapy in experimental stroke or parkinsonism? Clin Phys Med Rehab, in press. Schallert T, Hernandez TD, Barth TM (1986): Recovery of function after brain damage: Severe and chronic disruption by diazepam. Brain Res 379:104–111. Schallert T, Whishaw IQ (1985): Neonatal hemidecortication and bilateral cutaneous stimulation in rats. Dev Psychobiol 18:501–514. Schallert T, Woodlee ST, Fleming SM (2002b): Pre-clinical behavioral analysis of drug treatments for stroke; in Krieglstein J, Klumpp S (eds): Pharmacology of Cerebral Ischemia. Stuttgart, Medpharm Scientific Publishers, in press. Schlaggar BL, O’Leary DD (1993): Patterning of the barrel field in somatosensory cortex with implications for the specification of neocortical areas. Perspect Dev Neurobiol 1:81–91. Tomimatsu T, Fukuda H, Endoh M, Mu J, Watanabe N, Kohzuki M, Fujii E, Kanzaki T, Oshima K, Doi K, Kubo T, Murata Y (2002): Effects of neonatal hypoxic-ischemic brain injury on skilled motor tasks and brainstem function in adult rats. Brain Res 926:108–117. Wagner BP, Nedelcu J, Martin E (2002): Delayed postischemic hypothermia improves long-term behavioral outcome after cerebral hypoxiaischemia in neonatal rats. Pediatr Res 51:354– 360. Whishaw IQ, Schallert T, Kolb B (1981): An analysis of feeding and sensorimotor abilities of rats after decortication. J Comp Physiol Psychol 95: 85–103. Wu J, Hua Y, Keep R, Schallert T, Hoff J, Xi G (2002): Oxidative brain injury from extravasated erythrocytes after intracerebral hemorrhage. Brain Res 953:45. Young RS, Kolonich J, Woods CL, Yagel SK (1986): Behavioral performance of rats following neonatal hypoxia-ischemia. Stroke 17: 1313–1316.

Dev Neurosci 2002;24:418–425

425

Original Paper Dev Neurosci 2002;24:426–436 DOI: 10.1159/000069052

Damage to the Choroid Plexus, Ependyma and Subependyma as a Consequence of Perinatal Hypoxia/Ischemia Raymond P. Rothstein Steven W. Levison Department of Neuroscience and Anatomy, Pennsylvania State University, College of Medicine, Hershey, Pa., USA

Key Words Necrosis W Apoptosis W Cell death W Cerebral palsy W Stroke W Subventricular zone

Abstract Cerebral hypoxia/ischemia (H/I) of the premature infant is a major cause of cerebral palsy and mental retardation. An important determinant of the ultimate outcome from this insult is the extent to which the stem cells and progenitors in the brain are affected. Irreversible injury to these cells will impair normal development of the infant’s brain and, hence, its function. In the present study, we examine early intervals after H/I to identify which cells in the periventricular region are most vulnerable. At 0 h of recovery from a perinatal H/I insult, the choroid plexus shows extensive necrotic damage. The adjacent ependymal and subependymal cells are also affected. Swelling of the ependymal and medial subependymal cells is observed; however, these cells rarely sustain permanent damage. By contrast, cells in the most lateral aspect of the subventricular zone (SVZ) show more delayed, but extensive apoptotic and hybrid cell deaths. Interestingly, activated macrophages/microglia are observed adjacent to the swollen ependymal cells as well as within the affected subependyma. We conclude that the choroid plexus is an especially vulnera-

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2002 S. Karger AG, Basel

Accessible online at: www.karger.com/dne

ble structure in the immature brain, whereas the ependymal and adjacent subependymal cells are relatively resistant to damage. As the medial aspect of the SVZ contains neural stem cells, we predict that neural stem cells will be especially resistant to perinatal H/I brain damage. Copyright © 2002 S. Karger AG, Basel

Obstetrical and neonatal intensive care have improved dramatically over the past decades; however, the incidence of premature births has not changed. Each year, approximately 60,000 infants are born prematurely weighing less than 1,500 g. Premature babies have a higher than normal risk of cerebral H/I or ‘birth asphyxia’ and this birth complication is a major contributor to perinatal mortality and morbidity. Of the survivors, up to 50% exhibit permanent neuropsychological handicaps in the form of mental retardation, cerebral palsy, learning disability or epilepsy [1]. Despite major advances in our understanding of the development of the central nervous system (CNS) and perinatal diseases, data are still incomplete regarding the mechanisms leading to permanent brain injury. A prominent structure in the immature human brain is the germinal matrix – which harbors neural stem cells and a variety of progenitors including multipotential, bipoten-

Steven W. Levison, PhD Department of Neuroscience and Anatomy, H109 Penn State College of Medicine, PO Box 850 Hershey, PA 17033 (USA) Tel. +1 717 531 8650, Fax +1 717 531 0714, E-Mail [email protected]

tial and unipotential cells. In rodents, this region is known as the subventricular zone (SVZ). We previously reported that a moderate perinatal H/I insult depletes the SVZ and adjacent periventricular white matter of oligodendrocyte progenitors, and that by 4 h of recovery, there is a 20% reduction in the cell population within the SVZ [2, 3]. In the present study, we looked at earlier intervals after a moderate H/I insult to determine the time course and location of cellular degeneration within the periventricular region with the goal of identifying which cells were most vulnerable. Here we report that the choroid plexus is especially vulnerable to an H/I insult, whereas the ependymal and immediately subjacent subependymal cells are relatively resistant to damage. By contrast, cells in the most lateral aspect of the SVZ are extremely vulnerable. As that part of the SVZ that is immediately subjacent to the ependymal layer is enriched in neural stem cells and the lateral aspect is enriched in progenitors, these observations predict that neural stem cells are resistant to perinatal H/I brain damage, whereas progenitors are more vulnerable. Thus, these data provide hope that therapies can be designed to amplify the neural stem cell population to effect repair of the damaged infant brain. Materials and Methods Reagents Unless specified, standard laboratory reagents were obtained from VWR Scientific (Bridgeport, N.J., USA). Glutaraldehyde, lead citrate, sodium cacodylate and osmium tetroxide, uranyl acetate and embed 812 epoxy resin were purchased from Electron Microscopy Sciences (Fort Washington, Pa., USA). Perinatal H/I Timed pregnant Wistar rats were purchased from Charles River Laboratories (Charles River, Wilmington, Del., USA), housed in individual cages and fed high-fat lab chow. Offspring were delivered vaginally, and the litter size was adjusted to 10 pups per litter on the day of delivery. Three different groups of animals were used: controls, hypoxia controls, and experimental H/I animals. Cerebral H/I was produced in 7-day-old rats (day of birth being P1) by a permanent unilateral common carotid ligation followed by hypoxia [4]. Briefly, pups were lightly anesthetized with halothane (4% induction, 1.5% maintenance). Once fully anesthetized, a midline neck incision was made and the right common carotid artery (CCA) was identified. The CCA was separated from the vagus nerve and then ligated using 3-0 silk. Animals were returned to the dam for 2 h. Before exposure to hypoxia (8% O2/92% N2), the pups were prewarmed in jars for 20 min in a 37 ° C waterbath. The pups were then exposed to 1.5 h of hypoxia. After hypoxia, the pups were returned to their dam for recovery periods of 0, 2, and 4 h, at which time they were anesthetized and sacrificed by intracardiac perfusion. Controls were warmed for 1.5 h at 37 ° C with the jar lids off for 1.5 h.

Acute Damage to Periventricular Structures

Histology and Lectin Histochemistry Animals were perfusion fixed with 4% paraformaldehyde in 0.1 M phosphate buffer, and brains were removed and immersion fixed in the same fixative overnight at 4 ° C. The brains were then coronally blocked and prepared for paraffin embedding. Six-micrometer-thick coronal sections were cut at the level of the anterior commissure, –0.3 to –0.6 bregma, and stained with Harris’ hematoxylin and eosin. Other sections were stained with Giffonia Simplicifolia IB4 lectin (GSA-IB4; Sigma) as follows. Sections were baked for 30 min at 50 ° C to melt the paraffin, then deparaffinized and rehydrated through xylenes and ethanols to Tris-buffered saline (TBS). Sections were digested for 12 min with proteinase K (5 Ìg/ml; Roche, Indianapolis, Ind., USA) in 10 mM Tris (pH 8.0) at 37 ° C, then rinsed in TBS. Sections were incubated in 1% hydrogen peroxide in methanol for 10 min to inactivate endogenous peroxidases, rinsed in TBS and then incubated in a buffer containing 0.1% Triton-X-100, 10% normal goat serum and 10% BSA in 0.5 M Tris for 45 min at room temperature. Sections were then incubated in biotinylated GSA-IB4 lectin at 10 Ìg/ml (diluted in TBS w/1 mM each of Mg2+, Mn2+ and Ca2+) overnight at 4 ° C. After extensive rinsing in TBS, the sections were incubated in streptavidin-horseradish peroxidase diluted 1/500 in 10% BSA for 1 h at 37 ° C. The sections were rinsed extensively and then the color reaction was developed using the Nova Red Substrate (Vector Laboratories, Burlingame, Calif., USA). Sections were rinsed in water, counterstained in hematoxylin for 3–5 min, dehydrated through ethanols to xylenes and then coverslipped. Electron Microscopy Brain ultrastructure was evaluated 0, 2 and 4 h of recovery from perinatal H/I. Animals were anesthetized and perfused with 30 ml of 2% glutaraldehyde in 0.1 M cacodylate buffer with 3% sucrose and 0.02% CaCl2. The brains were then removed and placed in the same fixative for 12 h at 4 ° C. Coronal blocks were cut and the brains were split into the two hemispheres. The periventricular region of each hemisphere was isolated using a dissecting microscope. Samples were postfixed for 2 h in 1% osmium tetroxide in 0.1 M cacodylate buffer, washed in buffer, dehydrated, and embedded with embed 812 epoxy resin. Semithin sections (1 Ìm) were stained with methylene blue/ azure II and examined with a light microscope. Ultrathin sections, 60–90 nm were collected on copper grids and stained with 2% aqueous uranyl acetate for 30 min, followed by 5 min of lead citrate. Sections were examined and photographed using a Phillips TEM 400 electron microscope.

Results

Ninety minutes of H/I damaged many regions of the hemisphere ipsilateral to the common carotid ligation. In particular, the choroid plexus, SVZ, and striatum at the level of the anterior commissure sustained dramatic changes at early recovery time points (fig. 1). Other areas, such as the cortex and white matter also showed considerable changes, but they will not be discussed in this article as they have been studied in depth previously. Surprisingly, the choroid plexus was the structure that was most susceptible to damage in the hemisphere ipsilat-

Dev Neurosci 2002;24:426–436

427

Fig. 1. In this drawing of the left hemisphere of the neonate rat brain, the corpus callosum (WM) appears superior to the SVZ (gray stippled area) which is subjacent to the ependymal layer (Ep) that lines the lateral ventricle and which contains one of the choroid plexi (CP). The SVZ can be subdivided into three distinct regions: the medial region is a very dense cellular region that extends approximately 80 Ìm from the lateral ventricle; the mediolateral region is less cellular and extends from approximately 80 to 200 Ìm from the lateral ventricle, and the lateral tail is another densely cellular region that extends from 200 Ìm to the most lateral tip of the SVZ.

eral to the ligation of the regions we analyzed. The choroid plexus is comprised of lining cells, stromal cells and endothelial cells. In the normal choroid plexus, the lining cells have pale round nuclei surrounded by lightly stained cytoplasm having abundant mitochondria as well as other organelles. Located on the luminal surface of the lining cells are irregular surface projections that resemble microvilli (fig. 2A, B). Beneath the ependyma and the basal lamina, there is a highly vascularized connective tissue that is composed of a loose network of collagen fibers, fibroblasts, and blood vessels. The lining cells of the choroid plexus had already sustained significant damage in animals sacrificed immediately following hypoxia, i.e. the zero-hour time point (fig. 2C). Their nuclei were smaller and their chromatin was dispersed to the margins of the nuclear envelope. Additionally, their organelles were swollen and the microvillus border on their luminal surface was lost. The lining cells of the choroid plexus continued to degenerate over the following 2 h during recovery from H/I (fig. 2D). Some of the lining cells lost the integrity of their plasma membranes and there were clear areas where the tight junctions that normally connected these

428

Dev Neurosci 2002;24:426–436

cells were lost. By 4 h of recovery, many lining cells had begun to disintegrate (fig. 2E, F). Large gaps were seen in their plasma membranes, and their cytoskeleton and cellular organelles were expelled into the adjacent extracellular space. Tight junctions were no longer present. In addition to changes in the lining cells of the choroid plexus, many fibroblasts within the vascularized connective core showed necrotic changes. By contrast, the vascular endothelial cells within the choroid plexus remained intact through all recovery time points, and hemorrhagic infarcts were never observed. The majority of the brains had no damage on the contralateral hemisphere, but some damage to the choroid plexus on the contralateral side was observed in animals that had sustained extensive ipsilateral damage. In a previous study, we reported that a moderate perinatal H/I insult in Wistar rats depleted the SVZ and adjacent periventricular white matter of oligodendrocyte progenitors, and that by 4 h of recovery, already 20% of the total cells within the SVZ had been deleted [2]. To establish whether the SVZ was affected earlier, and if so, which regions were most vulnerable, we studied 3 different regions of the SVZ at 0, 2, and 4 h of recovery. For this analysis, the SVZ was subdivided into three compartments: a medial region, a mediolateral region and a lateral tail (fig. 1). These regions were defined as follows: the medial region is the highly dense region of SVZ cells that extends approximately 80 Ìm from the lateral ventricle; the mediolateral region extends from approximately 80 to 200 Ìm from the lateral ventricle, and the lateral tail extends from 200 Ìm to the most lateral tip of the SVZ. At 2 h of recovery, the ependymal cells as well as the medial SVZ cells that are adjacent to the ependyma appeared edematous (fig. 3B). The nuclei of these cells had enFig. 2. The choroid plexus is damaged during a moderate hypoxicischemic insult. A Semithin (1 Ìm) section of the choroid plexus

stained with toluidine blue/azure II from the contralateral hemisphere of a 7-day-old rat pup. B Thin section (120 nm) from the contralateral section stained with lead citrate and uranyl acetate. The ependyma (Ep) in these two images has lightly stained, round nuclei, abundant mitochondria, and microvilli (mv) on the luminal surfaces. C At 0 h of recovery from H/I, the lining cells (arrow) show early signs of necrosis, with marginalized chromatin, swollen organelles, and loss of microvilli. D At 2 h of recovery, the nuclei of the lining cells (arrow) exhibit more marked marginalization of their chromatin, vacuolization and loss of microvilli. E A semithin section at 4 h of recovery demonstrates a loss of cellular integrity of the lining cells (arrowhead). F Disintegration of lining cells is seen more convincingly in this thin section. No signs of blood vessel endothelial cell (BV) damage was observed at any recovery time point. Scale bar A, C, D, E = 12 Ìm; B, F = 6 Ìm.

Rothstein/Levison

Acute Damage to Periventricular Structures

Dev Neurosci 2002;24:426–436

429

Fig. 3. Cells within the SVZ are affected differently by H/I, consistent with this region being comprised of a heterogeneous cell population. Semithin sections from the contralateral or ipsilateral SVZs were stained with toluidine blue/azure II. A The medial SVZ on the contralateral hemisphere contains a heterogeneous cell population.

430

Dev Neurosci 2002;24:426–436

B A representative section from the lateral tail of the SVZ on the contralateral hemisphere. C The medial region of the SVZ from the

ipsilateral hemisphere at 2 h of recovery. Many cells near the lateral ventricle (top right) contain swollen nuclei. D The lateral tail from the ipsilateral hemisphere at 2 h of recovery. Many cells within the

Rothstein/Levison

Fig. 4. Electron micrographs of swollen cells within the medial SVZ at 4 h of recovery. Cells within the wedges of swollen cells within the medial SVZ when viewed under the electron microscope have watery cytoplasm compared with their counterparts in the undamaged SVZ; however, their mitochondria and endoplasmic reticulum do not appear to be swollen. Arrowheads in A and B depict the sites shown at higher magnification with the insets. C depicts cells from the contralateral hemisphere, where there was no evidence of damage at this time point. Scale bars represent 3 Ìm, 0.75 Ìm (inset) (A); 1.5 Ìm, 0.75 Ìm (inset) (B).

larged, and the chromatin had become more evenly distributed, when compared with the ependymal cells of the contralateral hemisphere and to ependymal cells in shamoperated controls. Within 4 h of recovery, groups of these swollen cells were apparent such that they appeared as columns or wedge-shaped aggregations perpendicular to

lateral tail display the features of necrotic cells (arrowheads). E The medial region of the ipsilateral SVZ at 4 h of recovery. Cellular swelling continues with the formation of lightly stained wedges of cells perpendicular to the lateral ventricle. F The lateral tail from the ipsilateral hemisphere at 4 h of recovery. Cells with hybrid (arrows) and more classic apoptotic features (arrowheads) are observed. Scale bar = 15 Ìm.

the lateral ventricle. Surrounding these wedges of large light cells were areas that appeared relatively unaffected. The unaffected regions contained an assortment of different cells, including many small, dark cells (fig. 3E). Although these cells appeared swollen, very little death was noticed in the region, suggesting that the cells in the medial aspect of the SVZ are relatively resistant to damage. When examined by electron microscopy, the majority of the swollen cells in the medial aspect of the SVZ had normal organelles within their processes (fig. 4A) and their cell bodies (fig. 4B). However, occasionally, an apoptotic or necrotic cell was observed within the medial aspect of the SVZ. Such cells were especially evident in those brains that were severely damaged. In addition, in

Acute Damage to Periventricular Structures

Dev Neurosci 2002;24:426–436

431

432

Dev Neurosci 2002;24:426–436

Rothstein/Levison

those severely damaged brains, necrotic ependymal cells were observed, but these cells were a rare occurrence (fig. 5E). Finally, there was no evidence that the increased volume of these cells was a consequence of hypertrophy, rather than swelling. The complement of intracellular organelles, such as mitochondria and rough endoplasmic reticulum, was normal. Whereas the medial aspect of the SVZ appeared relatively resistant, the lateral region of the SVZ had a large number of cell deaths. At 0 h of recovery, few damaged cells were observed. However, by 2 h of recovery, necrotic cells could be found (fig. 3D). Swollen cells were also observed in this region at both 2 and 4 h of recovery (fig. 3F). At 4 h of recovery, apoptotic cells (arrowheads) and cells exhibiting features of both necrotic and apoptotic cell deaths were observed (arrows). Thus, by 4 h of recovery, the lateral tail, which is normally very cellular, appeared depleted of cells. At the ependymal wall of the lateral ventricle, many macrophages were observed inside the lateral ventricle ipsilateral to the ligation. Figure 5A depicts 2 macrophages (arrowheads) and one proliferating macrophage at the luminal surface of the lateral ventricle ependyma. Their identity was confirmed using GSA IB4 lectin histochemistry (fig. 5C, D), as well at the electron microscopic level (fig. 5B). Few activated microglia were present immediately adjacent to the ventricle within the SVZ of the contralateral hemisphere, whereas activated microglia/ macrophages were abundant within the SVZ of the damaged hemisphere. Occasionally, phagocytic cells could be found interspersed amongst the ependyma (fig. 5F). The identity of these cells remains unclear. These cells might

be parenchymal microglia, they might be invading macrophages, or they might be phagocytic ependymal cells. The cell depicted in figure 5F has different features from those on either side of it, suggesting that it is a phagocyte that entered the brain parenchyma from the ventricle. One reason for surmising that this cell is not an ependymal cell engulfing a dying neighbor is that it does not have any junctions with neighboring ependymal cells. While the lateral tail of the SVZ is vulnerable to H/I, SVZ cells are overall less vulnerable than neurons. Within the adjacent basal ganglia, 50–75% of the striatal neurons showed degenerative changes after 90 min of H/I. Figure 6A shows a semithin section of normal neurons (arrowheads) from the hemisphere contralateral to ligation. Figure 6B depicts neurons from the ipsilateral hemisphere of that brain. As can be seen, many of the neurons in the ipsilateral striatum show evidence of necrosis (arrowheads). Thus, relative to striatal neurons, SVZ cells, and even those in the most vulnerable lateral tail region are more resistant to damage.

Discussion

phages (arrowhead) adhered to the luminal surface of the ependyma of the right lateral ventricle (rLV). B Electron micrograph of a phagocytic macrophage at the ependymal luminal surface. The primary lysosomes within this macrophage contain the debris of necrotic cells. C, D GSA-IB4 histochemistry for macrophages and microglia (brown reaction product). C SVZ of the contralateral hemisphere, where few microglia are observed next to the lateral ventricle (the stained cells are blood vessels, which also stain using GSA IB4). D SVZ of the damaged hemisphere at 4 h of recovery. Activated macrophages can be seen adjacent to the ependymal cells and activated microglia are seen within the SVZ. E While rare, necrotic ependymal cells were observed. F At 4 h of recovery, a macrophage appears to have migrated through the ependyma of the lateral ventricle where it has phagocytosed a necrotic ependymal cell. Scale bar: 40 Ìm (A), 5 Ìm (B), 10 Ìm (C, D), 6 Ìm (E), and 3 Ìm (F).

We previously reported that a moderate perinatal H/I insult depleted the SVZ and adjacent periventricular white matter of oligodendrocyte progenitors, and that by 4 h of recovery, there is already a 20% reduction in the cell population of the SVZ [2, 3]. In the present study, we examined earlier intervals after H/I to determine the time course and location of those cells that were degenerating within the SVZ as well as within adjacent periventricular structures. The goal of these studies was to identify which cells within these structures were most vulnerable to a 90 min H/I insult. At 0 h of recovery, the lining cells of the choroid plexus demonstrated extensive pathology that an electron microscopic analysis showed was necrotic. More subtle changes were seen in the adjacent ependymal and subependymal cells. Ependymal and medial subependymal cell swelling was observed, with little cell death. By contrast, extensive apoptotic and hybrid cell deaths were observed in the most lateral aspect of the SVZ. Interestingly, these changes were not apparent at 0 h. Another important observation was the appearance of activated macrophages/microglia adjacent to the swollen ependymal cells as well as within the affected subependyma. Altogether, these results show that the lining cells of the choroid plexus are especially vulnerable to H/I, whereas the ependymal and adjacent medial subependymal cells are relatively resistant to damage. Between these two

Acute Damage to Periventricular Structures

Dev Neurosci 2002;24:426–436

Fig. 5. Activated macrophages are present along the ependyma and

within the subependymal zone where swollen cells are apparent. A Semithin section at 4 h of recovery which depicts activated macro-

433

Fig. 6. Cell death in the striatum. Semithin sections were stained with toluidine blue/azure II. A Neurons (arrowheads) within the striatum in the contralateral hemisphere have euchromatic nuclei. B Damaged striatal neurons

(arrowheads) in the ipsilateral hemisphere have condensed, marginalized chromatin and swollen cytoplasm, consistent with a necrotic death process. Scale bar = 15 Ìm.

extremes, cells in the lateral tail of the SVZ are more vulnerable than more medial SVZ cells, yet still not as vulnerable as cells in the choroid plexus or as vulnerable as neurons in the adjacent striatum. Vulnerability of the Choroid Plexus The vulnerability of the choroid plexus to ischemia is not a new observation. Cell death in the choroid plexus was observed at 6 h following global forebrain ischemia in gerbils [5]. These authors described the cells in the choroid plexus as necrotic; however, more recent studies have concluded that the cells in the choroid plexus undergo apoptotic death following ischemia in the adult. Using TUNEL to assess cell death after transient forebrain ischemia, two groups have reported that TUNEL+ cells are observed within the choroid plexus within 6 h of recovery [6, 7]. Our data demonstrate that the newborn choroid plexus is affected earlier and perhaps qualitatively differently than the adult choroid plexus. By electron microscopy, we find that at 0 h of recovery, the lining cells of the choroid plexus show signs of classic necrosis with no apoptotic features. Thus, these data suggest that the ependymal cells lining the choroid plexus of the immature animal are different from those of the adult. Alternatively, it is possible that the adult choroid plexus cells also die by

434

Dev Neurosci 2002;24:426–436

necrosis and that these necrotic cells labeled using the TUNEL method as a consequence of nonspecific DNA degradation rather than internucleosomal cleavage. Here we have methodically characterized the demise of the immature choroid plexus; however, a preliminary characterization of the changes to the choroid was reported by Towfighi et al. [8], where they indicated that the choroid plexus showed signs of deterioration at 24 h of recovery from H/I in a 7-day-old rat. The damage to the choroid is likely a consequence of reduced blood flow through the anterior choroidal artery as a consequence of decreased blood flow from the ligation of the CCA. Significance of Death of the Choroid Plexus Lining Cells The demise of the choroid plexus ependymal cells may directly be responsible for some of the histopathological changes we have observed in periventricular structures following recovery from H/I. Since the junctions between the lining cells of the choroid are the physical site of the blood-cerebrospinal fluid (CSF) barrier, the demise would result in an opening of this barrier and hence plasma proteins, cytokines, excitatory amino acids, calcium and other substances would be elevated in or gain access to the CSF [9–11]. Indeed it has been postulated that the dam-

Rothstein/Levison

age resulting from insulin-induced hypoglycemia (which shows a somewhat similar pattern of damage to that seen here) is a consequence of a neurotoxic substance, possibly aspartate, which is borne by the CSF to the brain parenchyma [12, 13]. As the ependymal cells lining the ventricles are permeable to relatively large proteins, such as horseradish peroxidase, substances that accumulate in the CSF could affect the adjacent subependymal cells [14]. Alternatively, since the choroid plexus is a source of over 28 secreted polypeptides that include growth and trophic factors, the dysfunction of the choroid following H/I would result in a deficiency of these factors, which could adversely affect the adjacent ependymal and subependymal cells [15]. In particular, the choroid produces large quantities of IGF-II which has been shown to enhance SVZ cell proliferation [16–18]. Activated Microglia Likely Exacerbate Damage to the Ependymal and Subependymal Cells It was not surprising to find increased numbers of phagocytes within the damaged SVZ. Ivacko et al. [19] reported that microglia were activated within 10 min of recovery from perinatal H/I and they also observed an increase in the number of ED-1+ microglia in periventricular regions. However, it was surprising to see activated phagocytes within the ventricles adjacent to the swollen ependymal cells. This observation suggests that the microglia/macrophages might be responsible for some of the changes seen within the SVZ. Microglia and macrophages are activated by insults to the CNS where they coordinate tissue remodeling by ingesting dying cells and releasing growth factors that stimulate cell proliferation. Additionally, they recruit phagocytic cells to the site of tissue injury through the release of cytokines, chemokines and other inflammatory mediators. Therefore, it is plausible that their activity will exacerbate the extent of damage following perinatal H/I. Macrophages and activated microglia release a diverse set of cytokines that include IL-1, IL-6, IL-8, IL-12, and TNF-·. These cells also release a variety of other molecules including toxic oxygen radicals, peroxides, nitric oxide and lipid-derived mediators of inflammation, such as prostaglandins, leukotrienes, and platelet-activating factor. In infected tissues, these mediators will kill infectious organisms and infected cells, increase local blood flow and stimulate the production of molecules to amplify the mounting cellular response to damage. However, many of these same responses are triggered after H/I where their net activity likely increases infarct size. Recently, neuroprotection has been achieved after global

Acute Damage to Periventricular Structures

and focal ischemia by administering minocycline [20, 21]. Additional studies suggest that minocycline affords neuroprotection by inhibiting the proliferation and activation of microglia [22]. The hypothesis that activated microglia modify the responses of SVZ cells to H/I is worthy of further investigation. Neural Stem Cells within the SVZ May Be Resistant to Damage Recent studies have suggested that neural stem cells reside in a niche subjacent to the ependymal cells in the most medial aspect of the SVZ and that they may extend a single process that interdigitates between two ependymal cells to contact the cerebral spinal fluid. We noted that cells situated in the most medial region of the SVZ became swollen during recovery from perinatal H/I, but that few of those cells died. These observations support the prediction that neural stem cells are resistant to brain damaging stimuli. To test that hypothesis, we have recently stained the SVZ for both active caspase-3 and nestin. Confirming the results reported here, we could not find any cells that double-labeled, suggesting that, indeed, neural stem cells are resistant to damage [Romanko and Levison, unpubl. results]. Furthermore, studies in our lab have shown that neural stem cells in vitro are resistant to proapoptotic stimuli especially when compared with the vulnerability of late oligodendrocyte progenitors [Brazel and Levison, unpubl. results]. A corollary to the prediction that neural stem cells are resistant to brain damaging stimuli is that progenitors in the SVZ are more sensitive. We find active caspase-3 positive cells in the more lateral aspects of the SVZ [Romanko and Levison, unpubl. results]. Supporting these data, time lapse video studies have shown that cells within the medial aspect of the perinatal SVZ tend to migrate towards the more lateral aspects [23], suggesting that more lateral SVZ cells are more mature. At this developmental age, the majority of those progenitors would be bipotential glial progenitors as well as early oligodendrocyte progenitors. That progenitors are vulnerable to H/I is consistent with our previous studies and with the literature showing the sensitivity of oligodendrocyte progenitors to perinatal H/I, oxidative stress and glutamate [2, 3, 24–27]. Additional studies are clearly warranted to test the predictions made from these observations.

Acknowledgement This work was supported by MH 59950 and HD 30705.

Dev Neurosci 2002;24:426–436

435

References 1 Volpe JJ: Neurology of the Newborn, ed 3. Philadelphia, WB Saunders, 1995. 2 Levison SW, Rothstein RP, Romanko MJ, Snyder MJ, Meyers RL, Vannucci SJ: Hypoxic/ ischemia depletes the perinatal subventricular zone of oligodendrocyte progenitors and neural stem cells. Dev Neurosci 2001;23:234–247. 3 Ness JK, Romanko MJ, Rothstein RP, Wood TL, Levison SW: Perinatal hypoxia-ischemia induces apoptotic and excitotoxic death of periventricular white matter oligodendrocyte progenitors. Dev Neurosci 2001;23:203–208. 4 Rice JE, Vannucci RC, Brierley JB: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurology 1981; 9:131–141. 5 Pulsinelli WA, Brierley JB, Plum F: Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann Neurol 1982;11: 491–498. 6 Gillardon F, Lenz C, Kuschinsky W, Zimmermann M: Evidence for apoptotic cell death in the choroid plexus following focal cerebral ischemia. Neurosci Lett 1996;207:113–116. 7 Ferrand-Drake M: Cell death in the choroid plexus following transient forebrain global ischemia in the rat. Microsc Res Tech 2001;52: 130–136. 8 Towfighi J, Zec N, Yager J, Housman C, Vannucci RC: Temporal evolution of neuropathologic changes in an immature rat model of cerebral hypoxia: A light microscopic study. Acta Neuropathol (Berl) 1995;90:375–386. 9 Nagahiro S, Goto S, Korematsu K, Sumi M, Takahashi M, Ushio Y: Disruption of the blood-cerebrospinal fluid barrier by transient cerebral ischemia. Brain Res 1994;633:305– 311.

436

10 Garabedian BV, Lemaigre-Dubreuil Y, Mariani J: Central origin of IL-1ß produced during peripheral inflammation: Role of meninges. Brain Res Mol Brain Res 2000;75:259–263. 11 Ikeda J, Mies G, Nowak TS Jr, Joo F, Klatzo I: Evidence for increased calcium influx across the choroid plexus following brief ischemia of gerbil brain. Neurosci Lett 1992;142:257–259. 12 Auer RN, Wieloch T, Olsson Y, Siesjo BK: The distribution of hypoglycemic brain damage. Acta Neuropathol (Berl) 1984;64:177–191. 13 Auer RN, Kalimo H, Olsson Y, Siesjo BK: The temporal evolution of hypoglycemic brain damage. 2. Light- and electron-microscopic findings in the hippocampal gyrus and subiculum of the rat. Acta Neuropathol (Berl) 1985; 67:25–36. 14 Brightman MW, Reese TS: Junctions between intimately apposed cell membranes in the vertebrate brain. J Cell Biol 1969;40:648–677. 15 Chodobski A, Szmydynger-Chodobska J: Choroid plexus: Target for polypeptides and site of their synthesis. Microsc Res Tech 2001;52:65– 82. 16 Bartlett WP, Li X, Williams M: Expression of IGF-1 mRNA in the murine subventricular zone during postnatal development. Mol Brain Res 1992;12:285–291. 17 Arsenijevic Y, Weiss S, Schneider B, Aebischer P: Insulin-like growth factor-I is necessary for neural stem cell proliferation and demonstrates distinct actions of epidermal growth factor and fibroblast growth factor-2. J Neurosci 2001;21: 194–202. 18 Arsenijevic Y, Weiss S: Insulin-like growth factor-I is a differentiation factor for postmitotic CNS stem cell-derived neuronal precursors: Distinct actions from those of brain-derived neurotrophic factor. J Neurosci 1998;18:2118– 2128. 19 Ivacko JA, Sun R, Silverstein FS: Hypoxicischemic brain injury induces an acute microglial reaction in perinatal rats. Pediatr Res 1996;39:39–47.

Dev Neurosci 2002;24:426–436

20 Yrjanheikki J, Keinanen R, Pellikka M, Hokfelt T, Koistinaho J: Tetracyclines inhibit microglial activation and are neuroprotective in global brain ischemia. Proc Natl Acad Sci USA 1998;95:15769–15774. 21 Yrjanheikki J, Tikka T, Keinanen R, Goldsteins G, Chan PH, Koistinaho J: A tetracycline derivative, minocycline, reduces inflammation and protects against focal cerebral ischemia with a wide therapeutic window. Proc Natl Acad Sci USA 1999;96:13496–13500. 22 Tikka T, Fiebich BL, Goldsteins G, Keinanen R, Koistinaho J: Minocycline, a tetracycline derivative, is neuroprotective against excitotoxicity by inhibiting activation and proliferation of microglia. J Neurosci 2001;21:2580– 2588. 23 Kakita A, Goldman JE: Patterns and dynamics of SVZ cell migration in the postnatal forebrain: monitoring living progenitors in slice preparations. Neuron 1999;23:461–472. 24 Oka A, Belliveau MJ, Rosenberg PA, Volpe JJ: Vulnerability of oligodendroglia to glutamate: Pharmacology, mechanisms, and prevention. J Neurosci 1993;13:1441–1453. 25 Back SA, Gan X, Li Y, Rosenberg PA, Volpe JJ: Maturation-dependent vulnerability of oligodendrocytes to oxidative stress-induced death caused by glutathione depletion. J Neurosci 1998;18:6241–6253. 26 Volpe J: Brain injury in the premature infant: Neuropathology, clinical aspects, and pathogenesis. Ment Retard Dev Disabil Res Rev 1997;3:3–12. 27 Ness JK, Mitchell NE, Wood TL: IGF-1 and NT-3 signaling pathways in developing oligodendrocytes: Differential regulation and activation of receptors and the downstram effector Akt. Dev Neurosci, in press.

Rothstein/Levison

Original Paper Dev Neurosci 2002;24:437–445 DOI: 10.1159/000069050

Received: September 19, 2002 Accepted: September 29, 2002

IGF-I and NT-3 Signaling Pathways in Developing Oligodendrocytes: Differential Regulation and Activation of Receptors and the Downstream Effector Akt Jennifer K. Ness Nina E. Mitchell Teresa L. Wood Department of Neuroscience and Anatomy, Penn State College of Medicine, Hershey, Pa., USA

Key Words Glutamate W Apoptosis W Akt W Phosphatidylinositol 3-kinase W Neurotrophin W Trk receptor

Abstract A previous study from our laboratory demonstrated differences in the ability of insulin-like growth factor-I (IGF-I) and neurotrophin-3 (NT-3) to promote survival of pro-oligodendroblasts (pro-OLs) against glutamate-mediated apoptosis. In the current study, we tested whether submaximal concentrations of NT-3 would maintain receptor tyrosine kinase TrkC activation and Akt phosphorylation and thus promote long-term survival of the proOL against glutamate. Our results demonstrate that NT-3 at any concentration sufficient to activate the TrkC receptor results in a transient phosphorylation of the receptor and of Akt due, in part, to downregulation of the Trk receptor. In contrast, even submaximal IGF-I concentrations maintain long-term Akt activation and prevent glutamate-mediated apoptosis in pro-OLs. In addition, we also present data showing that IGF-I and NT-3 differentially activate their receptors and Akt depending on the maturational stage of the oligodendrocyte. Copyright © 2002 S. Karger AG, Basel

ABC

© 2002 S. Karger AG, Basel 0378–5866/02/0245–0437$18.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dne

Introduction

The development of oligodendrocyte lineage cells is regulated by extracellular signals that control cell proliferation, migration, differentiation, and survival. Oligodendrocyte development has been well-characterized in vitro and involves progression through several developmental stages from the bipolar oligodendrocyte progenitor cell (OPC) to the pro-oligodendroblast (pro-OL), that is multipolar but still mitotic, to the postmitotic, immature oligodendrocyte, and finally to the myelin-producing, mature oligodendrocyte. Progression through each developmental stage in the oligodendrocyte lineage is regulated by extracellular signals including growth factors, cytokines and hormones [for a review, see ref. 1]. Extracellular factors that promote survival in the oligodendrocyte lineage are of interest in the normal generation of oligodendrocytes in vivo, but also are of significant interest for how to protect these cells from death in pathological states such as following perinatal hypoxia-ischemia and in demyelination associated with multiple sclerosis. Two factors that promote survival of cells in the oligodendrocyte lineage both in vitro and in vivo are insulin-like growth factor-I (IGF-I) [2–5] and neurotrophin-3 (NT-3) [2, 5–8]. NT-3 and IGF-I promote maximal survival of OPCs in vitro when in combination with a mem-

Teresa L. Wood Department of Neuroscience and Anatomy, H109 Penn State College of Medicine, PO Box 850, 500 University Drive Hershey, PA 17033 (USA) Tel. +1 717 531 8650, Fax +1 717 531 5184, E-Mail [email protected]

ber of the interleukin-6 family of cytokines [2]. IGF-I also protects mature oligodendrocytes from cell death mediated by tumor necrosis factor-· [3]. More recent data have demonstrated that IGF-I and NT-3 protect cells in the oligodendrocyte lineage against glutamate-mediated apoptosis [5, 8]. A previous study from our laboratory focused on the ability of IGF-I and NT-3 to protect the intermediate-stage pro-OL from glutamate toxicity [5], since this is the stage of oligodendrocyte progenitor that is preferentially affected by hypoxia-ischemia in the immature brain [9]. Specifically, we showed that while both IGF-I and NT-3 provided immediate protection of the pro-OL from glutamate-mediated death, only IGF-I could provide long-term protection of these cells from glutamate toxicity [5]. A major signal transduction pathway that mediates survival in many cell types is the phosphatidylinositol 3kinase (PI3-K)/Akt pathway [for a review, see ref. 10]. The PI3-K pathway is essential for the survival of oligodendrocyte lineage cells, and expression of a dominant/ negative Akt induces oligodendrocyte apoptosis, indicating that the PI3-K/Akt signaling pathway is a potent mediator of oligodendrocyte survival [11, 12]. Our previous studies further demonstrated that the PI3-K/Akt signaling pathway is the primary pathway by which IGF-I and NT-3 protect pro-OLs from glutamate-mediated apoptosis [5]. Both IGF-I and NT-3 utilize the PI3-K pathway in pro-OLs to prevent glutamate-mediated caspase-3 activation [5]. Importantly, we demonstrated that sustained Akt activation is a key point in the survival pathway, which correlated with the ability of IGF-I versus NT-3 to promote prolonged survival in pro-OLs. While IGF-I maintained Akt phosphorylation through 48 h and promoted long-term survival, NT-3 only transiently activated Akt phosphorylation resulting in a delay in glutamate-mediated apoptosis [5]. The transient effect of NT-3 on Akt activation was correlated with a rapid downregulation of activation and total levels of its primary receptor tyrosine kinase, TrkC [5]. The previous studies raised several important questions concerning the PI3-K/Akt signaling pathway activated by IGF-I and NT-3 in developing oligodendrocytes. While concentrations of NT-3 that maximally stimulate phosphorylation of Akt were shown to rapidly downregulate TrkC receptor activity, it is not known whether lower, submaximal, concentrations of NT-3 could maintain TrkC activation and Akt phosphorylation. Furthermore, it is possible that the combination of IGF-I and NT-3 at submaximal concentrations could have an additive or synergistic effect on Akt phosphorylation, which would

438

Dev Neurosci 2002;24:437–445

correlate with previous reports of increased survival with combinations of trophic factors [2, 7]. Finally, our previous studies focused on Akt activation in the pro-OL, but it is unknown whether IGF-I and NT-3 differentially activate Akt in other stages of the oligodendrocyte lineage. A previous study reported maturational differences in the ability of NT-3 to activate mitogen-activated protein kinase (MAPK) in oligodendrocytes [6], suggesting the possibility that a similar difference may occur in the activation of Akt. In order to fully elucidate the effects of IGF-I and NT-3 on PI3-K/Akt signaling in developing oligodendrocytes, we investigated how these factors stimulate receptor and Akt activation at submaximal concentrations, alone and in combination. In addition, we explored the possibility that submaximal concentrations of IGF-I are sufficient to provide long-term activation of Akt and protection of pro-OLs from glutamate-mediated death. To understand maturational differences in Akt activation in the oligodendrocyte lineage, we investigated how IGF-I and NT-3 activate the PI3-K/Akt pathway in early OPCs and in postmitotic, immature oligodendrocytes. The results presented here indicate that NT-3 at any concentration sufficient to activate the TrkC receptor only transiently phosphorylates its receptor and Akt in pro-OLs. In contrast, submaximal IGF-I concentrations, sufficient to induce Akt phosphorylation, maintain long-term Akt activation and prevent glutamate-mediated apoptosis in proOLs. Finally, we also demonstrate that the ability of IGF-I and NT-3 to activate Akt is different depending on the developmental stage of the oligodendrocyte lineage cell.

Materials and Methods Materials MEM cell culture media, fetal bovine serum (FBS), and trypsin were purchased from Gibco-BRL (Long Island, N.Y., USA). Recombinant rat IGF-I and NT-3 were purchased from Upstate Biochemicals (Lake Placid, N.Y., USA). Recombinant human fibroblast growth factor-2 (FGF-2) was purchased from R&D Systems (Minneapolis, Minn., USA). Recombinant rat ciliary neurotrophic factor was purchased from Alomone Labs (Jerusalem, Israel). Antibodies to activated caspase-3 (cat. No. 9661S), Akt (cat. No. 9272), phospho(Ser473)-Akt (cat. No. 9271), phospho(Tyr490)-Trk (cat. No. 9141), and phospho-IGF-IR (Tyr1131)/insulin receptor (Tyr1146) (cat. No. 3021) were purchased from Cell Signaling Technology (Beverly, Mass., USA). Antibodies to IGF-IRß (C-20) and the TrkC receptor (C-14) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, Calif., USA). Conjugated secondary antibodies were purchased from Jackson Immunoresearch Laboratories (West Grove, Pa., USA). The antibody to actin and cell culture media supplements were purchased from Sigma Chemical Company (St. Louis, Mo.,

Ness/Mitchell/Wood

USA). Standard laboratory reagents were purchased from Fisher Scientific (Pittsburgh, Pa., USA) or VWR (Bridgeport, N.J., USA). Cell Culture and Treatment Conditions Primary cultures of OPCs were isolated from newborn rat mixed glial cultures and amplified as described previously [5]. Briefly, forebrain cortices were enzymatically digested with trypsin and DNase I, then mechanically dissociated, and plated in minimal essential medium containing 10% FBS with antibiotics. OPCs were isolated from confluent cultures by a differential shake, then seeded into poly-dlysine-coated flasks in chemically defined media (N2B2) containing 1:3 B104-conditioned media and 5 ng/ml FGF-2. OPCs were amplified 7–12 days, during which time the cells were passaged once with papain. Amplified OPCs were passaged with papain and seeded onto poly-d-lysine-coated 60-mm plates or chamber slides at a density of 4 ! 104 cells/cm2 for OPCs and pro-OLs and at 5 ! 104 cells/cm2 for immature oligodendrocytes. To prepare OPCs for treatment, cells were grown in amplification media overnight. To obtain highly enriched cultures of O4+/Ranscht– pro-OLs, amplified OPCs were cultured in N2B2 with 0.5% FBS with 10 ng/ml FGF-2 for 48 h. To obtain enriched cultures of GC+ immature OLs, amplified OPCs were cultured in N2B2 with 1.95 Ìg/ml T3 hormone, 5 ng/ml CNTF, and 0.5% FBS for 5 days. For glutamate and trophic factor treatments, a hormone-supplemented medium was used that was identical to the N2B2 medium with the exception that the insulin concentration was lowered to 5 ng/ml, a concentration adequate to stimulate the insulin receptor but not the IGF-IR [13]. Treatment media did not contain growth factors or FBS that were present in the differentiation media. Cells were treated for various time points with 500 ÌM glutamate, a concentration that produces maximal cell death in oligodendrocyte lineage cells [5, 14]. Trophic factors IGF-I or NT-3 were added at concentrations that maximally stimulated Akt phosphorylation and survival (10 and 5 ng/ml, respectively) or at submaximal concentrations (5 and 1 ng/ml, respectively). Media were replaced every 16 h. For analysis of receptor and Akt phosphorylation, cultures were preincubated for 4 h with control media without trophic factors to return activation of the PI3-K pathway to basal levels [15]. Western Blotting Total protein from oligodendroglial cultures was isolated in sodium dodecyl sulfate sample buffer (62.5 mM Tris-HCl, 2% sodium dodecyl sulfate, 10% glycerol, 50 mM dithiothreitol) containing 1:100 protease inhibitor cocktail, 1 mM sodium vanadate, and 10 mM sodium fluoride (Sigma) and sonicated for 10–15 s on ice. The RC DC protein assay was used to determine protein concentrations (Biorad Laboratories, Hercules, Calif., USA). Equal amounts of protein were heated to 95–100 ° C for 5 min, cooled, and loaded onto a 4–20% Tris gradient gel (Biorad Laboratories). Proteins were then electrotransferred to a nitrocellulose membrane. Membrane blocking and primary and secondary antibody incubations were done in 5% milk in Tris-buffered saline (TBS) containing 0.1% Tween-20. Membrane blocking and secondary antibody incubations were 1 h at room temperature, and primary incubations were overnight at 4 ° C. Blots were washed for 5 min three times each with TBS-0.1% Tween-20 following each incubation in antibody. The incubation buffer for detection of the phosphorylated IGF-IR contained 5% BSA in TBS0.1% Tween-20. Detection of the horseradish peroxidase-conjugated secondary antibody was by enhanced chemiluminescence (NEN).

IGF-I and NT-3 Signaling in Developing Oligodendrocytes

Statistical Analyses Analysis of variance was used for group comparisons within an experiment. For experiments with multiple comparisons, adjusted p values associated with post hoc group comparisons were calculated using Fisher’s PLSD method. All analyses were done using the Statview statistical software program. All statistical analyses were repeated on multiple experiments.

Results

Effect of Submaximal Trophic Factor Concentrations on Akt Phosphorylation in Pro-OLs Initial studies were conducted to determine the concentration of IGF-I and NT-3 that produced approximately 50% of the level of Akt phosphorylation seen previously with higher concentrations of IGF-I (10 ng/ml) and NT3 (5 ng/ml) that provided maximal stimulation of Akt in pro-OLs [5]. In dose-response studies, we determined the concentrations of IGF-I (5 ng/ml) and NT-3 (1 ng/ml) that induced Akt phosphorylation by 15 min at approximately half the level of phosphorylation induced by the maximal doses of the factors (fig. 1). At the submaximal concentrations, the combination of IGF-I and NT-3 had an additive effect on the level of Akt phosphorylation after 15 min, significantly increasing Akt phosphorylation as compared with each factor alone (fig. 1; p ! 0.01). However, the combination of IGF-I and NT-3 at submaximal concentrations did not significantly increase Akt phosphorylation above the levels seen with either IGF-I or NT-3 at higher concentrations (10 and 5 ng/ml, respectively; p 1 0.1). In our previous studies, we found that activation of Akt in pro-OLs was maintained over time by IGF-I but was reduced to control levels by 2 h after stimulation with NT-3 due to degradation of the TrkC receptor [5]. Thus, we investigated whether concentrations of IGF-I and NT-3 that provided submaximal receptor activation could maintain Akt phosphorylation after 2 h in cultures of pro-OLs. Similar to the results seen following stimulation with NT-3 at maximal concentrations, Akt phosphorylation was lost by 2 h in pro-OLs treated with 1 ng/ml NT-3 (fig. 2a). Consistent with these data, the levels of Akt phosphorylation in 1 ng/ml NT-3-treated pro-OLs at 10 h was not significantly different from control levels (p = 0.4; fig. 2b). In contrast to NT-3, submaximal doses of IGF-I maintained Akt phosphorylation after 10 h (fig. 2b). In addition, the additive effect of IGF-I and NT3 seen at 15 min was not maintained at 10 h. The level of Akt phosphorylation after 10 h in cells treated with IGF-I (5 ng/ml) plus NT-3 (1 ng/ml) was significantly greater

Dev Neurosci 2002;24:437–445

439

15 min

a CTRL

2h

5N

1N

CTRL

1N

p-Akt Akt

b

8 7

15 min 10 h

b

b

Akt phospho

6

Fig. 1. Submaximal concentrations of IGF-I and NT-3 have an addi-

tive effect on Akt phosphorylation after 15 min. Pro-OLs were treated with IGF-I (10 ng/ml or 5 ng/ml), NT-3 (5 or 1 ng/ml) or the combination of IGF-I (5 ng/ml) and NT-3 (1 ng/ml) for 15 min (5I+1N). Cells were lysed and Akt phosphorylation was analyzed by Western immunoblotting of isolated protein. Blots were stripped and used for analysis of total Akt levels. Band density was quantified and the percentage of phosphorylated Akt to total Akt was calculated. The graph shows data from one representative experiment (n = 3). CTRL = Control; phospho = phosphorylation. a p ! 0.02 vs. CTRL; b p ! 0.01 vs. CTRL; c p ! 0.001 vs. CTRL.

than in control cells (p ! 0.05) but not significantly different than the levels of Akt phosphorylation seen in proOLs treated with IGF-I alone (fig. 2b). Submaximal Concentrations of NT-3 Downregulate the Trk Receptor To determine whether the loss of Akt activation over time with the lower doses of NT-3 involved the stability of the TrkC receptor, the phosphorylation of the receptor was investigated following different treatment conditions. After stimulation of pro-OLs for 10 h with 1 ng/ml NT-3, phosphorylation of the TrkC receptor was reduced to levels that were no longer significantly different from control levels (p = 0.27; fig. 3a, b). The decrease in TrkC receptor phosphorylation occurred as early as 2 h following the addition of NT-3 to the pro-OLs (data not shown). Moreover, the addition of 5 ng/ml IGF-I had no effect on TrkC phosphorylation by NT-3 after either 15 min or 10 h (data not shown). We have recently reported that after the initial downregulation of TrkC receptor phosphorylation in the presence of NT-3, adding fresh NT-3 to the pro-OLs after 10 h in NT-3 media had no effect on Trk receptor phosphorylation [5]. In cultured neurons, the TrkC recep-

440

Dev Neurosci 2002;24:437–445

5 a

4

c

a

3

c

2 1 0

CTRL

IGF 5

NT-31

5I+1N

Fig. 2. IGF-I (I) but not NT-3 (N) at submaximal concentrations maintains Akt activation through 10 h. Pro-OLs were treated with IGF-I or NT-3 for the indicated times and concentrations, and isolated protein was analyzed for phosphorylated and total Akt levels. a Representative Western blots of Akt phosphorylation at 15 min and 2 h. b Graph showing quantitation of Western blots for multiple samples. Band density is expressed as the percentage of Akt phosphorylation compared with 15 min control levels. The graph shows data from one representative experiment (n = 3). CTRL = Control; phospho = phosphorylation; p-Akt = phosphorylated Akt. a p ! 0.01 vs. CTRL; b p ! 0.001 vs. CTRL; c p ! 0.001 vs. 15 min treatment.

tor is also downregulated in response to NT-3 treatment but can be restimulated if NT-3 is removed for a minimum of 15 min and then reapplied to the cells [16]. In pro-OLs, even when NT-3 was removed from the media for 2 h after the initial 8-hour NT-3 treatment period, restimulation with NT-3 for 15 min did not induce phosphorylation of TrkC significantly from control cells (p = 0.19; fig. 3b). Submaximal Concentrations of IGF-I Protect Pro-OLs from Glutamate-Mediated Apoptosis Since our previous study demonstrated that the sustained activation of Akt by IGF-I was correlated with long-term protection of the pro-OLs from glutamateinduced apoptosis [5], we investigated whether submaximal concentrations of IGF-I also provided long-term protection of the pro-OLs from glutamate-induced caspase

Ness/Mitchell/Wood

15 min

a CTRL

increase in caspase-3 activation in pro-OLs from either glutamate or the absence of trophic factors.

2h

5N

1N

5N

CTRL

1N

p-Trk Trk

b

6

a

p-Trk c

5 4

a

3 b 2

b

1 0

CTRL

NT-35

NT-31

15 min Treatment

CTRL

NT-35 NT-31 (15 min)

10 h Treatment

Fig. 3. TrkC activation is permanently downregulated after initial stimulation by NT-3 (N) at both submaximal and maximal concentrations. Pro-OLs were treated with NT-3 for 15 min or 10 h. Another group of cells was treated with 5 ng/ml NT-3 for 8 h, the media was replaced with control media without NT-3 for 2 h, then the cells were restimulated with 5 ng/ml NT-3 for 15 min [labeled as ‘NT-3 5 (15 min)’ in the graph in b]. Protein was isolated and analyzed for phosphorylated and total Trk receptor. a Representative Western blots of Trk phosphorylation and total Trk receptor at 15 min and 10 h. b Graph showing quantitation of Western blots for multiple samples. Band density is expressed as the percentage of Trk phosphorylation compared with control samples (15 min). The graph shows data from one representative experiment (n = 3). CTRL = Control; p-Trk = phosphorylated Trk. a p ! 0.001 vs. CTRL; b p ! 0.004 10 h vs. 15 min of equivalent NT-3 concentrations.

activation. Cells were treated with or without glutamate in combination with 5 or 10 ng/ml IGF-I for 22 or 48 h. As shown in figure 4, 5 ng/ml IGF-I maintained Akt phosphorylation through 48 h in the presence of glutamate (fig. 4a, b; p ^ 0.02). Furthermore, 5 ng/ml IGF-I was sufficient to prevent the activation of caspase-3 through 48 h in pro-OLs cultured in the presence of glutamate (fig. 4a, c; p ! 0.01). In the absence of any exogenous trophic factors, pro-OLs have significantly increased levels of active caspase-3 indicative of apoptosis by 48 h as seen by the increase in caspase-3 activation in pro-OLs in the control cultures at 48 h compared with control cultures at 22 h (fig. 4c; p ! 0.001). IGF-I at 5 ng/ml prevented the

IGF-I and NT-3 Signaling in Developing Oligodendrocytes

Maturational Stage-Specific Effects of IGF-I and NT-3 on Akt Phosphorylation To determine the functional state of the IGF-IR and TrkC receptors on the earlier-stage OPCs and on the laterstage immature oligodendrocytes, cultures of oligodendrocyte lineage cells enriched for all three maturational stages were treated with IGF-I and NT-3 at the maximal doses previously used in studies on the pro-OLs. For these studies, we prepared cultures of cells with the following characteristics: the OPCs were 95% A2B5+/O4–, the proOLs were 90–95% O4+/O1–, and the immature OLs were 65% O1+. Total IGF-IR levels and levels of phosphorylated IGF-IR were determined in the three lineage stages after 20 h in the presence of IGF-I (fig. 5a). Only the proOLs significantly increased the ratio of phosphorylated IGF-IR to unphosphorylated IGF-IR in response to IGF-I (fig. 5a, b). To investigate the ability of IGF-IR stimulation to activate Akt at the different stages of oligodendrocyte maturation, cell lysates from cultures of all three lineage stages that were treated with or without IGF-I for 20 h were used to determine levels of phosphorylated Akt. Significant activation of Akt by IGF-I was observed in both OPCs and pro-OLs, but not in immature oligodendrocytes (fig. 5c, d). Surprisingly, while activation of the IGF-IR in the OPCs in response to IGF-I was not statistically significant, IGF-I did significantly stimulate Akt phosphorylation in these cells (compare fig. 5b, d). Similar to the studies on IGF-IR stimulation in stages of the oligodendrocyte lineage, we investigated NT-3 activation of TrkC and Akt across the three lineage stages. Similar to our previous results on the pro-OLs, NT-3 treatment of OPCs resulted in detectable phosphorylated Trk after 15 min but not after 2 h in the presence of NT-3 (fig. 6a). TrkC activation in the immature oligodendrocytes was not detectable after 15 min or 2 h of NT-3 treatment (fig. 6a). Our previous studies demonstrated that total Trk receptor levels were reduced by 10 h in pro-OLs treated with NT-3 [5]. After only 2 h of NT-3 treatment, total Trk receptor levels were significantly reduced in OPCs but not in pro-OLs or immature oligodendrocytes (fig. 6b). Interestingly, analyses of Akt activation revealed that only pro-OLs showed significant levels of phosphorylated Akt after 15 min of NT-3 stimulation (fig. 6c, d). As previously demonstrated in the pro-OLs, total Akt levels did not differ with trophic factor stimulation or between developmental stages (data not shown) [5].

Dev Neurosci 2002;24:437–445

441

48 h

22 h

a CTRL

GLU

G+I (10) G+I (5)

CTRL

GLU

G+I (10)

G+I (5)

p-Akt Active C3 Actin

b 6

c 2.5

a

b

b

b

3 2 a

c 1.5 d 1.0

f

f

G+I (10)

G+I (5)

d

0.5

1 0

e

2.0

a

4

Active C3

Akt phospho

5

CTRL GLU

G+I (10)

G+I (5)

CTRL IGF (5)

22 h

G+I (10)

G+I (5)

48 h

0

CTRL GLU

G+I (10)

G+I (5)

CTRL IGF (5)

22 h

48 h

Fig. 4. Submaximal IGF-I maintains Akt phosphorylation and prevents glutamate (GLU)-mediated caspase-3 (C3) activation through 48 h. Pro-OLs were treated with IGF-I (I) at either 5 ng/ml or 10 ng/ml in the presence of GLU for 22 or 48 h [G+I (5), G+I (10), respectively]. Protein was isolated and used for immunoblotting to detect phosphorylated Akt (p-Akt). The blots were then stripped and reanalyzed for activated C3 and ß-actin. a Representative Western blots of p-Akt and active C3 at 22 and 48 h. b, c Graphs showing

quantitation of Western blots for multiple samples analyzed for p-Akt (b) or active C3 (c) and normalized to ß-actin band density. Results are expressed as fold increase over 22 h control levels. The graphs show data from one representative experiment. CTRL = Control; phospho = phosphorylation. a p ! 0.001 vs. CTRL 22 h; b p ^ 0.02 vs. CTRL 48 h; c p ! 0.01 vs. CTRL 22 h; d p ! 0.01 vs. GLU 22 h; e p ! 0.001 vs. CTRL 22 h; f p ! 0.01 vs. CTRL 48 h.

Fig. 5. IGF-I stimulation of IGF-IR and Akt phosphorylation differs

Fig. 6. Levels of Trk receptor and NT-3 activation of TrkC and Akt differ during oligodendrocyte maturation. OPCs, pro-OLs and immature (imm.) OLs were treated with 5 ng/ml NT-3 for 15 min or 2 h. Isolated protein was used in Western immunoblotting analyses. a Representative Western blots of phosphorylated Trk (p-Trk) receptor that were subsequently stripped and analyzed for total Trk and ß-actin. b Graph showing quantitation of band density from analysis of Trk receptor 2 h after addition of NT-3. Data are expressed as fold increase or decrease over pro-OL control levels. c Representative Western blot analyzed for phosphorylated Akt (p-Akt) after 15 min NT-3 treatment, then stripped and reanalyzed for ß-actin. d Graph showing quantitation of band density from p-Akt blots after 15 min treatment with NT-3. Data shown were normalized to ß-actin levels and expressed as fold control over control levels. The graphs show data from one representative experiment (n = 3). CTRL = Control; phospho = phosphorylation. a p ! 0.02 vs. OPC CTRL, b p ! 0.001 vs. pro-OL CTRL.

during oligodendrocyte maturation. OPCs, pro-OLs, and immature (imm.) OLs were treated with 10 ng/ml IGF-I for 20 h. Protein was isolated and analyzed by Western immunoblotting. a Representative Western blots of phosphorylated IGF-IR levels (p-IGFR). Blots were stripped and then used to analyze total receptor levels and ß-actin. b Graph showing quantitation of band density from Western analysis of multiple samples and expressed as the ratio of phosphorylated/ unphosphorylated IGF-I receptor as compared with OPC control levels. c Representative Western blot analyzed for phosphorylated Akt (p-Akt), then stripped and analyzed for ß-actin. d Graph showing quantitation of band density from Western analysis of multiple samples after normalization to ß-actin band density, and expressed as fold increase over OPC control levels. The graphs show data from one representative experiment (n = 3). CTRL = Control; phospho = phosphorylation; p-IGF-I receptor = phosphorylated IGF-I receptor. a p ! 0.001 vs. CTRL pro-OL, b p ! 0.01 vs. CTRL OPC, c p ! 0.05 vs. CTRL pro-OL.

442

Dev Neurosci 2002;24:437–445

Ness/Mitchell/Wood

b 4.5 OPC

a

imm. OL

pro-OL CTRL

IGF

IGF

CTRL

CTRL IGF

a

3.5

IGF p-IGF-I receptor

CTRL

4.0

p-IGFR IGF-IR b-Actin

3.0 2.5 2.0 1.5 1.0 0.5 0

8

d

CTRL

imm. OL

pro-OL CTRL

IGF

IGF

CTRL

c

6 Akt phospho

OPC CTRL

imm. OL

b

7

c

pro-OL

IGF

p-Akt

5 4 3

b-Actin

2 1 0

CTRL

pro-OL

imm. OL

5

b 3 OPC 15 CTRL

15 NT-3

2h CTRL

imm. OL 2h NT-3

15 CTRL

15 IGF 2 h NT-3 CTRL

2h NT-3

p-Trk

Trk receptor

a

2

a

1

Trk Actin

0

d 10 9 OPC CTRL p-Akt Actin

pro-OL NT-3

CTRL

NT-3

CTRL

NT-3

CTRL NT-3

pro-OL

imm. OL

c

8

imm. OL Akt phospho

c

CTRL

7 6 5 4 3 2 1 0

CTRL

pro-OL

imm. OL

6

IGF-I and NT-3 Signaling in Developing Oligodendrocytes

Dev Neurosci 2002;24:437–445

443

Discussion

The results of this study demonstrate that submaximal concentrations of IGF-I maintain Akt phosphorylation and can prevent glutamate-mediated apoptosis through 48 h. Conversely, submaximal concentrations of NT-3 do not maintain receptor or Akt phosphorylation beyond 2 h. Moreover, while submaximal IGF-I and NT-3 are additive in activating Akt after 15 min, this combined trophic factor stimulation of Akt is lost by 10 h. Finally, we also demonstrated stage-specific differences in IGF-IR and TrkC receptor activation and in the ability of IGF-I or NT-3 to activate Akt in the oligodendrocyte lineage. The inability of NT-3 to sustain receptor and Akt activation in pro-OLs is due, in part, to receptor downregulation. Our previous studies demonstrated that 5 ng/ml of NT-3 caused downregulation of TrkC protein levels in pro-OLs [5], and the current study shows similar results even with concentrations of NT-3 that provide 50% maximal receptor phosphorylation. The response of the TrkC receptor to sustained stimulation by NT-3 in the pro-OLs is in contrast to ligand regulation of TrkC in neurons. TrkC downregulation in the pro-OLs showed no recovery even after removal of the ligand for 2 h. In neurons, the TrkC receptor was restimulated by addition of NT-3 after only a 15-min removal of the ligand [16]. However, similar to our results, brain-derived neurotrophic factor (BDNF) induced a rapid ligand-induced downregulation of the TrkB receptor in neurons, which was refractory through 6 h [16]. Reduced BDNF phosphorylation of the TrkB receptor persisted even after a 24-hour withdrawal of BDNF [16]. Similar to our results in pro-OLs, downregulation of TrkB and TrkC in neurons is reflected in a reduction in total receptor protein levels as well as a reduction in messenger RNA levels [16, 17]. Importantly, chronic infusion of BDNF into the hippocampus of adult rats also decreased the total levels of TrkB receptor protein, suggesting that the ligand-induced downregulation of TrkB receptors described in vitro is relevant to in vivo regulatory mechanisms [17]. Taken together, these results suggest that NT-3 would only be effective as an acute therapy to prevent death of oligodendrocyte lineage cells in vivo following insult or in disease states. The differences we observed in the levels of IGF-IR activation across maturational stages of the oligodendrocyte lineage suggest stage-specific differences in receptor levels and, specifically, in the levels of cell surface IGF receptors. Indeed, our data support the hypothesis that levels of available cell surface receptors vary across the developmental stages since the total levels of the IGF-IR

444

Dev Neurosci 2002;24:437–445

were significantly reduced in the IGF-treated pro-OLs compared with either the IGF-treated OPCs or immature oligodendrocytes (data not shown); however, the pro-OLs were the only stage that showed a significant induction of IGF-IR activation in response to IGF-I. Similarly, the different maturational stages also showed different levels of Trk phosphorylation in response to NT-3. Downregulation of the TrkC receptor in O1+ oligodendrocytes has been previously reported, supporting the decreased responsiveness of immature oligodendrocytes to NT-3 in our studies [8]. Similarly, we demonstrated a reduction in total Trk receptor levels in both the pro-OL and immature oligodendrocyte control cultures compared with the OPC control cultures. It should be noted that the analysis of Trk receptors in our studies was with an antibody that detects all Trk receptor isoforms, as we were not able to obtain reliable results for immunoblot analyses using a TrkC-specific antibody. Thus, the exact levels of total TrkC in our cultures could not be assessed, since oligodendrocyte lineage cells also express TrkA that was recognized by the antibody we utilized for these studies. However, our analysis of the TrkC response to NT-3 in the pro-OLs was specific since the concentration of NT-3 that we used does not activate TrkA and, moreover, the NT-3 responsive splice variant of the TrkA receptor is not expressed in the progenitor stages of the oligodendroglial lineage [6, 18]. The differential ability of both IGF-I and NT-3 to activate Akt in the different stages of the oligodendrocyte lineage may involve differential regulation of receptor recycling and maintenance of active signaling complexes. Surprisingly, although NT-3 treatment of OPCs induced TrkC phosphorylation by 15 min, no detectable Akt phosphorylation was seen at that time (fig. 6). This result could indicate differences between OPCs and pro-OLs in the rate of Trk internalization similar to a report that cell surface TrkA receptors mediate prolonged activation of the PI3-K/Akt pathway, while internalization terminates Akt activation in PC12 cells [19]. Similarly, internalization of TrkC receptors and termination of the PI3-K/Akt pathway could explain the transient activation of Akt and the limited survival effect against glutamate-mediated apoptosis in pro-OLs. The IGF-IR also activates different signaling pathways through receptor recycling. While the MAPK pathway requires receptor internalization, IRS-1 activation and association with PI3-K occurs from both cell surface and endosomal receptors [20]. Although the IGF-IR internalizes in response to IGF-I binding, evidence also suggests that IGF-I does not necessarily dissociate from the IGF-

Ness/Mitchell/Wood

IR after internalization and thus may be recycled back to the cell surface with the receptor [21]. The results presented in this study showing that IGF-I sustains Akt phosphorylation through 20 h in both stages of oligodendrocyte progenitors supports the hypothesis that IGF-I-mediated activation of the PI3-K/Akt pathway is essential to survival of these cells. Finally, the dissociation between the ratio of phosphorylated to nonphosphorylated IGF-IR and the levels of Akt activation suggests additional mechanisms of regulation of the IGF-I signaling pathway. Even though the ratio of phosphorylated to nonphosphorylated IGF-IR was decreased in OPCs compared with pro-OLs,

IGF-I treatment of both OPCs and pro-OLs resulted in significant stimulation of Akt. Thus, the activation of the IGF-IR is proportional to Akt activation in the pro-OLs but not in the OPCs. Taken together, with the data showing significant Akt activation only in the pro-OL lineage stage in response to NT-3, our results suggest that the PI3-K/Akt pathway is regulated distinctly in response to several trophic factors in pro-OLs. Moreover, these data suggest a possible mechanistic basis for the selective sensitivity of the pro-OLs to insult and loss of trophic factor signaling.

References 1 Grinspan J: Cells and signaling in oligodendrocyte development. J Neuropathol Exp Neurol 2002;61:297–306. 2 Barres BA, Schmid R, Sendnter M, Raff MC: Multiple extracellular signals are required for long-term oligodendrocyte survival. Development 1993;118:283–295. 3 Ye P, D’Ercole AJ: Insulin-like growth factor I protects oligodendrocytes from tumor necrosis factor-·-induced injury. Endocrinology 1999; 140:3063–3072. 4 Mason J, Ye P, Suzuki K, D’Ercole A, Matsushima G: Insulin-like growth factor-1 inhibits mature oligodendrocyte apoptosis during primary demyelination. J Neurosci 2000;20: 5703–5708. 5 Ness J, Wood T: Insulin-like growth factor I, but not neurotrophin-3, sustains Akt activation and provides long-term protection of immature oligodendrocytes from glutamate-mediated apoptosis. Mol Cell Neurosci 2002;20:476– 488. 6 Cohen RI, Marmur R, Norton WT, Mehler MF, Kessler JA: Nerve growth factor and neurotrophin-3 differentially regulate the proliferation and survival of developing rat brain oligodendrocytes. J Neurosci 1996;16:6433– 6442. 7 Kumar S, Kahn MA, Dinh L, de Vellis J: NT3-mediated TrkC receptor activation promotes proliferation and cell survival of rodent progenitor oligodendrocyte cells in vitro and in vivo. J Neurosci Res 1998;54:754–765.

IGF-I and NT-3 Signaling in Developing Oligodendrocytes

8 Kavanaugh B, Beesley J, Itoh T, Itoh A, Grinspan J, Pleasure D: Neurotrophin-3 (NT-3) diminishes susceptibility of the oligodendroglial lineage to AMPA glutamate receptor-mediated excitoxicity. J Neurosci Res 2000;60:725–732. 9 Back SA, Han BH, Luo NL, Chricton CA, Xanthoudakis S, Tam J, Arvin KL, Holtzman DM: Selective vulnerability of late oligodendrocyte progenitors to hypoxia-ischemia. J Neurosci 2002;22:455–463. 10 Datta SR, Brunet A, Greenberg ME: Cellular survival: A play in three Akts. Genes Dev 1999; 13:2905–2927. 11 Vemuri GS, McMorris FA: Oligodendrocytes and their precursors require phosphatidylinositol 3-kinase signaling for survival. Development 1996;122:2529–2537. 12 Flores AI, Mallon BS, Matsui T, Ogawa W, Rosenzweig A, Okamoto T, Macklin WB: Aktmediated survival of oligodendrocytes induced by neuregulins. J Neurosci 2000;20:7622– 7630. 13 Rechler M, Nissley S: The nature and regulation of the receptors for insulin-like growth factors. Ann Rev Physiol 1985;47:425–442. 14 McDonald J, Althomsons S, Hyrc K, Choi D, Goldberg M: Oligodendrocytes from forebrain are highly vulnerable to AKPA/kainate receptor-mediated excitotoxicity. Nat Med 1998;4: 291–297. 15 Ebner S, Dunbar M, McKinnon RD: Distinct roles for PI3K in proliferation and survival of oligodendrocyte progenitor cells. J Neurosci Res 2000;62:336–345.

16 Knusel B, Gao H, Okazaki T, Yoshida T, Mori N, Hefti F, Kaplan DR: Ligand-induced downregulation of TRK messenger RNA, protein, and tyrosine phosphorylation in rat cortical neurons. Neuroscience 1997;78:851–862. 17 Frank L, Ventimiglia R, Anderson K, Linsay RM, Rudge JS: BDNF down-regulates neurotrophin responsiveness, TrkB protein, and TrkB mRNA levels in cultures rat hippocampal neurons. Eur J Neurosci 1996;8:1220–1230. 18 Clary DO, Reichardt LF: An alternatively spliced form of the nerve growth factor receptor TrkA confers an enhanced response to neurotrophin-3. Proc Natl Acad Sci USA 1994;91: 11133–11137. 19 Zhang Y, Moheban DB, Conway BR, Bhattacharyya A, Segal RA: Cell surface Trk receptors mediate NGF-induced survival while internalized receptors regulate NGF-induced differentiation. J Neurosci 2000;20:5671–5678. 20 Chow J, Condorelli G, Smith R: Insulin-like growth factor-I receptor internalization regulates signaling via the Shc/mitogen-activated protein kinase pathway, but not the insulin receptor substrate-1 pathway. J Biol Chem 1998;273:4672–4680. 21 Zapf A, Hsu D, Olefsky JM: Comparison of the intracellular itineraries of insulin-like growth factor-I and insulin and their receptors in Rat-1 fibroblasts. Endocrinology 1994;134:2445– 2452.

Dev Neurosci 2002;24:437–445

445

Dev Neurosci 2002;24:446–464 DOI: 10.1159/000069041

Abstracts

Cerebrovascular Effects of Cocaine in the Fetus, Newborn and Adult Richard J. Traystman, Christine A. Gleason Johns Hopkins School of Medicine, Baltimore, Md., USA

Cocaine abuse continues to be a major health problem in the United States. Maternal abuse of cocaine has been associated with neonatal stroke, seizures, intracranial hemorrhages, microcephaly and neurobehavioral abnormalities, but the pathogenesis of these effects remains unknown. Newborns may be exposed to cocaine passively via smoke, or actively via breast milk. In the adult, cocaine abuse can lead to cerebral hemorrhagic or vaso-occlusive injury, resulting in seizures, ischemic attacks, or stroke. In addition, the effects (direct or indirect) of cocaine on cerebral vessels is controversial. Conflicting results from animal studies have not yielded a consistent explanation regarding the mechanisms for these cocaineinduced cerebrovascular effects. Whereas species and methodological differences may explain some of the conflicting results, we question whether developmental differences may play a role. Therefore, we compared the cerebrovascular responses of cocaine in unanesthetized fetal, newborn and adult sheep to determine whether the same dose of cocaine, without the confounding variable of the utero plancental circulation, would cause cerebral vasodilation at all developmental stages, and to determine what, if any, differences there were in response to cocaine. We compared the cerebrovascular and metabolic responses to a 2 mg/kg i.v. cocaine dose in unanesthetized fetal (n = 8), newborn (n = 6), and adult (n = 12) sheep. We measured cerebral blood flow (CBF), mean arterial blood pressure, and arterial and venous O2 content, and we calculated cerebral O2 consumption and cerebrovascular resistance at baseline and at 30 s and 5, 15 and 60 min after cocaine injection. CBF increased 5 min after injection in the fetus and newborn, but not until 15 min in the adult. In the fetus, cocaine caused a transient cerebral vasoconstriction at 30 s. In all groups, cocaine caused cerebral vasodilation which was delayed in the adult. Cerebral metabolic O2 consumption increased 5 min after the injection in the fetus and newborn, but not until 15 min after injection in the adult. Arterial O2 content decreased 5 min after injection in the fetus, and 15 min after injection in the adult. Overall, the results of this unique, comparative sheep study confirmed that cocaine causes cerebral vasodilation and increased O2 consumption in sheep of all ages, but we have observed developmental differences in the timing and extent of these responses. Our study results support the hypothesis that clinical, and perhaps species variability in cocaine cerebrovascular effects may be due in part to developmental responses. The mechanism for these developmental differences in

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2002 S. Karger AG, Basel

Accessible online at: www.karger.com/dne

the cerebrovascular responses to cocaine is not known. However, one possible mechanism could be developmental differences in cocaine metabolism. Another may involve differences in cerebral O2 consumption, either intrinsically, or in response to cocaine. Our previous studies demonstrate cerebral vasodilation with cocaine; however, other investigators have observed vasoconstriction. We speculate that species and/or methodological differences (route of injection, anesthesia technique, or timing of measurements), and developmental differences in cerebrovascular or metabolic development could account for these conflicting results in laboratory animals.

Acute and Chronic Behavioral Outcome after Unilateral Brain Injury in Neonatal or Adult Rats: Use Dependent Neural Events and Plasticity Tim Schallert, John Barks, Barbara Felt Center for Human Growth and Development, and Departments of Pediatrics and Neurosurgery, University of Michigan, Ann Arbor, Mich., USA

Behavioral outcome after unilateral neonatal and adult ischemic damage was examined, acutely and chronically. Testing in the neonates occurred early after the injury, through development and into adulthood. Key deficits observed in adult operated animals were never observed in the neonatally injured rats, yet the neonatal rats ‘developed into’ other deficits which had not been detected early in life and recovered as they matured. Brain injury in adult animals appears to create conditions that in many ways resemble neural events that occur during brain development. Neurotrophic factor expression in astrocytes, exuberant neural growth, synaptogenesis, and dendritic pruning may contribute to restoration of function. As in development, these events are influenced markedly by experience. Brain damage can lead to self-regulated functional impoverishment in both adult and neonatally injured animals, which may reduce mechanisms of neural plasticity associated with under-used functions. Motor enrichment procedures aimed specifically at overcoming behavioral deficits in adult operated animals have been found to restore plasticity and improve functional and neurochemical outcome, whereas restricting motor behavior can interfere with plasticity. The essential role of motor behavior in structural plasticity and recovery of function will be discussed.

An Early T2-Weighted MRI Signal in Ventral Cortices Is Predictive for the Development of Temporal Lobe Epilepsy in Immature Rats Catherine Roch, Claire Leroy, Izzie J. Namer, Astrid Nehlig

leading to the development of epilepsy while the hippocampus seems to be involved only at later times. The measurement of T2 relaxation times at 24 h after SE in 21-day-old rats will allow to discriminate the population of rats that will be become epileptic and hence to study further the critical steps leading to epilepsy in this model.

CNRS UMR 7004 and INSERM U398, Faculty of Medicine, Strasbourg, France Rationale: Patients with temporal lobe epilepsy (TLE) usually presented in early childhood an initial insult such as febrile seizures, status epilepticus (SE), trauma or encephalitis. However, not all children who have undergone an early insult become epileptic. As in patients, the consequences of the lithium-pilocarpine-induced SE are age-dependent and only a subset of 21-day-old rats subjected to SE will become epileptic. Thus, with magnetic resonance imaging (MRI), we explored the differences in the evolution of lesions in these two populations of rats. Methods: SE was induced in 61 21-day-old rats by the injection of lithium (3 mEq/kg) followed 20 h later by pilocarpine (30 mg/kg) and 35 rats survived the SE. T2-weighted images and T2 relaxation time measurements were performed for the detection of lesions at various times ranging from 6 h to 4 months after SE. The occurrence of spontaneous seizures was surveyed using a video recording device for 10 h per day. Results: Three populations of rats could be distinguished (fig. 1). The first one (n = 6) presented no visible MRI anomalies nor modification of the T2-relaxation time in any brain region and at any time and these rats did not become epileptic. In the second group (n = 13), a hypersignal appeared in the piriform and entorhinal cortices 24 h after SE (increase of 49% of the T2 relaxation time in the piriform cortex), that disappeared 48–72 h after SE; these animals did all become epileptic. The third group (n = 16) showed a more moderate increase of the T2 relaxation time in cortices (14% in the piriform cortex) that was not visible on T2-weighted images. Rats of this group did also all become epileptic. A hippocampal hypersignal indicative of hippocampal sclerosis appeared after a mean delay of one month in 69% and 25% of the rats of the second and third groups, respectively. The other rats of these two groups developed epilepsy without clear signs of hippocampal sclerosis. Conclusions: These results suggest that the piriform and entorhinal cortices represent key structures in the initial step

Effects of Cerebral Maturation on Regional nNOS Expression, Cerebrovascular Reactivity to NO, and Responses to Ischemia William J. Pearce Center for Perinatal Physiology, Department of Physiology, Loma Linda University School of Medicine, Loma Linda, Calif., USA

Fetal development and postnatal maturation are dynamic processes that dramatically influence the structure and function of all intracranial tissues. One major consequence of these influences is that vulnerability to hypoxic and ischemic insults is generally greater in the mature than in the immature brain. Whereas it is clear that cerebral responses to ischemia involve multiple mechanisms, one important molecule in these responses appears to be NO, which has both cytotoxic and vasodilator effects. In view of the broad variety of evidence indicating that cerebral synthesis and metabolism of NO change significantly during development and maturation, the present studies explore the hypothesis that postnatal maturation selectively enhances the capacity for neuronal NO synthesis while simultaneously depressing vascular reactivity to the NO-cGMP vasodilator pathway, such that vulnerability to ischemic insults is enhanced. To evaluate this hypothesis, we have conducted experiments designed to determine the effects of postnatal maturation upon: (1) the regional vulnerability to ischemia; (2) the regional distribution of nNOS enzyme and activity; (3) cerebrovascular reactivity to NO; (4) cerebrovascular calcium stores that are targets of NO action; and (5) thick- and thin-filament contractile proteins that are candidates for

Fig. 1. Changes in T2 relaxation times in the 3 groups of rats, from left to right, (1) rats with no measurable lesion and no subsequent epilepsy, (2) rats with a visible and measurable signal and subsequent epilepsy and (3) rats with a measurable but not visible signal and subsequent epilepsy. The dotted lines limit the range of T2 relaxation values in control rats.

Abstracts

Dev Neurosci 2002;24:446–464

447

NO action. Together, these experiments reveal that postnatal increases in the neuronal capacity for NO synthesis are highly region specific, involve increases in the ratio of soluble to particulate nNOS, and that soluble nNOS expression is highly correlated with regional vulnerability to ischemic insult. At the level of the cerebral arteries, postnatal maturation dramatically increases the intracellular pool size of agonist-releasable calcium from 28% in the fetus to 68% in the adult, while also decreasing thick-filament regulation and increasing thin-filament reactivity in a NO/cGMP-sensitive manner. Overall, these findings are highly consistent with our core hypothesis and strengthen the view that the greater vulnerability of the adult brain to ischemic insults involves both increased ischemic neuronal synthesis of cytotoxic NO leading to increased protein nitrosylation and freeradical damage, but decreased vasodilator responses to NO.

Excessive GABAA Receptor Activation Damages the Hippocampus of Neonatal Male and Female Rats: A New Model for Prenatal Brain Damage? Joseph L. Nuñez, Jesse J. Alt, Margaret M. McCarthy Physiology Department, University of Maryland School of Medicine, Baltimore, Md., USA

Premature infants are at exceptionally high risk for seizures, hypoxia-ischemic insults and other traumatic events that result in permanent brain damage. Most models of pediatric brain damage have focused on the damaging effects of glutamate in the postnatal day 7 rat, which is considered analogous to the newborn human. Exposure of newborn rats to hypoxia increases extracellular GABA levels up to 15 fold [Andine et al, Dev Brain Res 64 (1991) 115–120]. Importantly, the major excitatory drive in immature neurons is derived from depolarizing responses following activation of the Á-aminobutyric acid (GABA)A receptor. GABAA receptor mediated membrane depolarization is capable of opening L-type voltage sensitive calcium channels and removing the endogenous block on the NMDA receptor [Leinekugel et al, Adv Neurol 79 (1999) 189–201]. Excessive calcium entry through the NMDA channel and voltage sensitive calcium channels is a major mediator of excitotoxicity [Choi, TINS 11 (1988) 465–469]. We report that activation of GABAA receptors by muscimol in newborn rats, a time when GABAA receptor activation leads to depolarization, increases cell death in the developing hippocampus. The effects are region specific, persistent and greater in males. This damage is attenuated by pretreatment with diltiazem, an L-type voltage sensitive calcium channel blocker. Results using hippocampal cultures parallel those observed in vivo, indicating the effects are mediated directly in the hippocampus. When tested at postnatal day 21, animals treated with muscimol as newborns display deficits on hippocampal dependent learning tasks such as a preweanling version of the Morris water maze task and the open field task. Therefore, exposure to exogenous GABAA receptor activation over the first two days of postnatal life produces anatomical and behavioral deficits observed into adolescence. We propose that treatment of newborn rats with muscimol mimics the initiation of a cell death cascade induced by hypoxia or injury in premature infants, and is analogous to the accepted method of glutamate administration to the 1-weekold rat pup to model the newborn human.

448

Dev Neurosci 2002;24:446–464

Serial Quantitative Diffusion Tensor MRI of the Premature Brain Reveals Injury S.P. Miller, D.B. Vigneron, R.G. Henry, M.A. Bohland, C. Ceppi-Cozzio, C. Hoffman, N. Newton, J.C. Partridge, D.M. Ferriero, A.J. Barkovich Departments of Neurology, Radiology, and Pediatrics, University of California, San Francisco, Calif., USA Objective: To determine the change over time of the directionally averaged apparent diffusion coefficient of water (ADC) and relative anisotropy in a cohort of serially studied premature newborns with and without white matter injury. Methods: 23 newborns, median post-conceptual age 29.8 weeks (range 25–33.8 weeks), were studied at median 32.8 weeks (27.6–38.5 weeks) and again at median 37.3 weeks (33.2–43 weeks). Newborns were classified as normal (n = 11), minimal white matter injury (n = 7) and moderate white matter injury (n = 5). Diffusion tensor imaging was performed using a locally developed sequence. Mean ADC and anisotropy were calculated from 0.5-cm3 regions of interest bilaterally: 1) frontal white matter, 2) posterior white matter, 3) optic radiations, 4) visual association, 5) cortical spinal tracts, 6) thalamus 7) basal ganglia, 8) calcarine gray, 9) hippocampus (0.15 cm3). Results: Unlike the adult brain, ADC values were substantially higher in white matter regions compared with gray matter regions. ADC decreased significantly with age in all brain regions in normals and those with minimal white matter injury. ADC increased with age or failed to decline in widespread areas of white matter in newborns with moderate white matter injury. Anisotropy increased with age in all white matter regions in normals. Anisotropy did not increase in frontal white matter in those with minimal white matter injury and in widespread white matter areas in those with moderate white matter injury (see figure). Conclusions: We found a lack of normal developmental changes in water diffusion in the brains of newborns with early white matter injury. The serial measurement of ADC and anisotropy appears to be a promising technique for the early detection of impaired brain development in premature newborns with this type of injury. Supported by NIH NS35902, M01-RR01271, RO1 NS40117 (DMF, DBV, CCC and AJB), CIHR and Glaser Foundation (SPM).

Abstracts

Desflurane Confers Neuroprotection for Hypothermic Low-Flow Cardio-Pulmonary Bypass in Neonatal Pigs A.W. Loepke, M.A. Priestley, J. Golden, S.E. Schultz, J. McCann, C.D. Kurth Brain Research Laboratory, Department of Anesthesiology and Critical Care Medicine, Joseph Stokes Jr. Research Institute, The Children’s Hospital of Philadelphia and the Departments of Anesthesia, Pediatrics, and Pathology, University of Pennsylvania School of Medicine, Philadelphia, Pa., USA Introduction: Despite improvements in surgical management for congenital heart disease, neurologic and cardiac complications continue to occur related to ischemia during low-flow cardiopulmonary bypass (LF-CPB) and deep hypothermic circulatory arrest (DHCA). Desflurane has been shown to be both neuro- and cardioprotective against ischemia at normothermia and for DHCA [1–3]. This study investigates the neuroprotective effects of Desflurane during hypothermic LF-CPB. Methods: After IACUC approval, 30 piglets aged 1 week were randomized to receive Fentanyl/Droperidol (F/D), Desflurane 4.5% (Des4.5), or Desflurane 9% (Des9) during surgical preparation and CPB. Physiologic variables (arterial blood gases, glucose, heart rate, arterial pressure, brain temperature, and cortical blood flow) were recorded. After piglets were CPB-cooled to 22 ° C (brain) using pH stat management, LF-CPB was performed for 150 min (MAP 9 B 2 mm Hg, pump flow 10 B 6 ml/kg/min). Piglets were then CPB rewarmed, separated from CPB, extubated, and survived for 2 days. After sacrifice, the brain was processed as previously described and slices were put on slides. (3) Both behavioral and neuropathologic criteria were used to assess neurologic outcome. Outcome was scored by physical exam from no disability to death (0–4) by a blinded observer; brain slides were ranked from no damage to very severe injury (0–5) by a blinded neuropathologist. Results: Physiologic variables were similar between groups before, during, and after LF-CPB except for lower arterial pO2 in the Desflurane groups during LF-CPB. Cerebral cortical blood flow during LFCPB compared to baseline did not differ between the groups (11 B 5% in F/D vs. 12 B 4% in Des4.5 vs. 16 B 5% in Des9). Functional and histopathologic outcome was better in the Desflurane groups than in the F/D group. Disability scores were 1.4 B 1.6 in F/D vs. 0.3 B 0.5 in Des4.5 (p = 0.1) vs. 0.2 B 1 in Des9 (p = 0.04). Neuropathologic damage scores in neocortex were 2.6 B 1.1 in F/D vs. 1.5 B 0.8 in Des4.5 (p = 0.03) vs. 1.3 B 0.8 in Des9 (p = 0.009) and in the hippocampus 2.4 B 1.6 in F/D vs. 1.3 B 1.1 in Des4.5 (p = 0.08) vs. 0.8 B 0.7 in Des9 (p = 0.01). The incidence of ventricular fibrillation (VF) during LF-CPB was 90%, 60%, and 10% for F/D, Des4.5 (p = 0.06) and Des9 (p = 0.0002), respectively. Discussion: Desflurane improved neurologic outcome for LF-CPB compared with F/D in neonatal piglets, as indicated by less functional disability and less histologic damage. Desflurane may have produced cardiac protection, as suggested by a lower incidence of VF.

References 1 2 3

Br J Anaesth 1999;83:415. Anesthesiology 2000;92:1731. Anesthesiology 2001;95:959.

Abstracts

Ventilation with Oxygen in Fetal Sheep Produces Pial Arteriolar Constriction That Is Blocked by an Inhibitor of Cytochrome P450 ˆ-Hydroxylase Activity Raymond C. Koehler, Hiroto Ohata, Richard J. Traystman Department of Anesthesiology/Critical Care Medicine, The Johns Hopkins University, Baltimore, Md., USA

With the onset of ventilation at birth, cerebral blood flow decreases as oxygenation increases, but the mechanism of cerebral vasoconstriction is unknown. Cytochrome P450 ˆ-hydroxylase activity metabolizes arachidonic acid to 20-HETE, a potent vasoconstrictor, with a Km for O2 in the physiological range. This pathway has been shown to contribute to autoregulation in kidney and to be involved in vasoconstriction in skeletal muscle during increased oxygenation. In mature brain, 20-HETE produced by cytochrome P450 4A in cerebral arterial smooth muscle contributes to myogenic constriction by inhibiting Kca channels and thereby limiting overoxygenation during arterial hypertension (Circ Res 87:60–65, 2000). We tested the hypothesis that during increased oxygenation associated with in utero O2 ventilation in fetal sheep, the ˆ-hydroxylase inhibitor, 17-octadecynoic acid (17-ODYA), reduces pial arteriolar vasocontriction. In halothane-anesthetized pregnant sheep near term (F140 days gestation), the axillary artery of the fetus was catheterized and an endotracheal tube was inserted into the trachea. The fetal head was exposed with the rest of the body remaining in utero so as not to disrupt the umbilical circulation. The exteriorized head was supported on a head holder and a closed cranial window was constructed over parietal cortex with the dura removed and temperature maintained with a heating lamp. The window was superfused with vehicle (n = 6) or 10 ÌM 17-ODYA (n = 6). Pial arteriolar diameter responses were measured during in utero ventilation, first with a low O2 mixture to maintain normal fetal PaO2 and to control for mechanical ventilation, and then with a high O2 mixture. Cerebral arteries removed from the Circle of Willis were found to express cytochrome P450 4A. However, expression in near-term fetuses was less than in adult sheep, as might be expected from the lower arterial blood pressure and PO2 in the fetus resulting in less myogenic tone. Ventilation with the low O2 mixture did not change pial arteriolar diameter in the vehicle (–4 B 6%) or 17-ODYA (–4 B 6%) groups (B SE). Ventilation with a high O2 mixture in the vehicle group increased PaO2 from 18 B 1 to 88 B 17 mm Hg and decreased arteriolar diameter by 24 B 3% without a change in arterial pressure. In contrast, increasing PaO2 from 20 B 1 to 91 B 8 mm Hg in the 17-ODYA group resulted in no change in diameter (–2 B 3%) or arterial pressure. Increasing arterial pressure from about 55 to 88 mm Hg with phenylephrine resulted in little additional vasoconstriction (8 B 6%) in the vehicle group, indicating that constriction was already near-maximal at high O2. In the 17-ODYA group, there was also little additional vasoconstiction (5 B 4%), presumably because inhibition of ˆ-hydroxylase is known to impair autoregulation. After returning to a low O2 mixture, fetuses were hyperventilated to decrease PaCO2 to 19 mm Hg. Arteriolar diameter decreased similarly in the vehicle (32 B 7%) and 17ODYA (20 B 4%) groups, thereby indicating that the effect of 17ODYA was selective for increased oxygenation. We conclude that P450 ˆ-hydroxylase activity represents an important homeostatic regulatory mechanism for protecting the brain from over-oxygenation at birth. If this meachanism is not fully developed in the prema-

Dev Neurosci 2002;24:446–464

449

ture fetus with lower arterial pressure and lower myogenic tone, then the premature brain might become over-oxygenated at birth and possibly lead to oxidative vascular damage and hemorrhage. Supported by NIH NS20020.

Amelioration of Hypoxia-Ischemia Induced Brain Damage by a Traditional Chinese Medicine in Newborn SD Rats Mingyan Hei, Inderjeet Bhatia, Helen K.W. Law, Nikolaus Sucher, Pik To Cheung Department of Paediatrics and Neuroscience Research Centre, the University of Hong Kong Department of Biology, Hong Kong University of Science and Technology, Hong Kong SAR, China

Neonatal hypoxic-ischemic (HI) encephalopathy is a disease with long-term neurological deficits and high mortality. Activation of glutamate receptors (such as NMDA receptors) is an important event in the pathogenesis cascade of HIE suggesting that glutamate receptor antagonists might be a valuable protective agent to HI-induced brain damage. Mitochondrial membrane potential (MMP) is an indicator of mitochondrial function which can be altered by NMDA receptor activation by agonists and measured by flow cytometry. Compound T (TCM-T) is a traditional chinese medicine recently identified by one of us (N.S.) to be a potent non-competitive NMDA receptor antagonist. The objective of this study was to assess the effect of TCM-T on HI-induced brain infarction by TTC staining, and the mitochondrial membrane potential (MMP) of dissociated brain cells by studying rhodamine 123 retention using flow cytometry. Postnatal day 7 SD rats with unilateral carotid artery ligation followed by 2.5 h of hypoxia in 8% O2/92% N2 were used. TCM-T was injected intraperitoneally at 0.5 Ìg/g/dose, 3 doses/day, 1 Ìg/g/dose, 2 doses/ day, and 1 Ìg/g/dose, 2 doses/day (1st dose given just before induction of hypoxia) for 24 h (n = 12), 3 days (n = 20), and 7 days (n = 25), respectively. Equal amount of vehicle (DMF) was given to control animals. TTC straining of 2-mm thick brain slices were traced with Stereo-Investigator system. Comparison was made between the ipsilateral and contralateral side. At 24 h, 3 days, and 7 days of treatment, the damages in treated animals were reduced by 19% (p = 0.232), 26% (p = 0.374), and 68% (p = 0.035), respectively. Regional tracing showed that the reductions were most prominent at thalamus and hippocampus after 3 days and 7 days of treatment. Preliminary experiments indicated that MMP was reduced as soon as the HI experiment was concluded (ligation plus 2.5 h of hypoxia – designated time 0 h), the 0 h and 1 h time points post HI insult were therefore chosen for the current study. Animals were randomly divided into ‘HI’ group (n = 18), ‘HI + T’ group (n = 11), and ‘HI + vehicle’ group (n = 11). 1 Ìg/g of TCM-T or vehicle was given just before induction of hypoxia. Ipsilateral or contralateral hemispheres were dissociated into single cell suspension in a concentration of 6–8 ! 106/ml. Rhodamine 123 was loaded to the cell suspension at final concentration of 1 ÌM for 45 min at 37 ° C. The Gmean value of rhodamine fluorescence was determined to indicate the MMP of the hemisphere. Analysis was focused on the ipsi-/contralateral ratio in order to minimize experimental variance. The results showed that, at 0 h and 1 h time

450

Dev Neurosci 2002;24:446–464

points, the ratio was 0.8 and 0.91 respectively in ‘HI’ group, 1.09 and 1.142 respectively in ‘HI + T’ group (both were significantly increased with p ! 0.05), and 0.82 and 0.914 respectively in ‘HI + vehicle’ group. Thus TCM-T can reduce HI-induced brain cell MMP depolarization 0 h (which was equivalent to 2.5 h after TCM-T was administered) and 1h after HI insult. We conclude that 1 Ìg/g TCMT protects neonatal SD rat from HI-induced brain damage in a timedependent pattern, with preferential protection for thalamus and hippocampus. This protective effect may be related to its ability to block NMDAR activity with consequent reduction of HI-induced MMP depolarization. Further work will be performed to document the therapeutic window offered by TCM-T, with or without other therapeutic agents, and the drug effect on MMP at later time points. This work is supported by Hong Kong Research Grant Council Fund HKU7249/99M (PTC, HKFY, MHS) and Sir Edward HoTung Paediatric Education and Research Fund (PTC).

Treatment of Hypoxic-Ischemic Brain Injury to Newborn Rats with TPCK 3 Hours after Hypoxia Decreases Caspase-9 Activation and Improves Outcome Yangzheng Feng, Michael H. LeBlanc Department of Pediatrics, University of Mississippi Medical School, Jackson, Miss., USA Background and Purpose: N-tosyl-L-phenylalanyl-chloromethyl ketone (TPCK) reduced apoptosis in vitro. Pretreatment with TPCK reduced brain injury in the neonatal rat. Would treatment after injury reduce brain damage? Methods: Seven-day-old rats had the right carotid artery ligated and were subjected to 2.5 h of 8% oxygen. Results: Treatment with 10 mg/kg of TPCK i.p. 3 h after hypoxia reduced the decrease in right hemisphere weight caused by injury when measured 22 days after injury from 27.6 B 2.8% SEM (vehicle, n = 56) to 19.8 B 2.8% (TPCK, n = 61, p ! 0.05). Posttreatment increased the percentage of brains scored as normal by a blinded observer from 14/58 (24%) to 30/61 (49%), p ! 0.01. Posttreatment reduced bcl-2 in the cortex from 130 B 7% to 106 B 6% (p ! 0.05 vs. vehicle) without effecting mitochondrial bax. Post-treatment reduced caspase-9 activity measured 24 h after hypoxia from 176 B 27% to 116 B 8% (p ! 0.05 vs. vehicle) in the cortex without effecting cytoplasmic cytochrome c. Post-treatment reduced caspase3 activity measured 24 h after hypoxia from 262 B 50% to 165 B 34% (p ! 0.05 vs. vehicle). Conclusions: Treatment with TPCK 3 h after hypoxia reduced brain infarct size. The neuroprotection seen may have been caused by a reduction in caspase-9 activation caused by TPCK [DeFreitas, MF, Hamrick, SEG, Ferriero, DM and PS McQuillen: Subplate Neuron Cell Death and mRNA Expression Profiling Following Oxygen Glucose Deprivation. Developmental Cerebral Blood Flow and Metabolism. June 2002. Hershey, Pa.]. Periventricular leukomalacia (PVL) occurs following hypoxic-ischemic cerebral injury in the preterm human and involves excitotoxic cell death of developing preoligodendrocytes (Back SA, et al, J Neurosci 22:455–463). Observed clinical deficits include spastic diplegia, developmental delay and cortical visual impairment. We have charac-

Abstracts

terized a rodent model of PVL to demonstrate that subplate neurons also die following hypoxia-ischemia at P1. To evaluate the mechanism of subplate sensitivity to hypoxia-ischemia, they were immunopurified and cultured in vitro (DeFreitas MF, et al, J Neurosci 21:5121–5129). Subplate neurons were compared to immunopurified cortical plate and hippocampal neurons. In agreement with observations in vivo, subplate neurons were twice as sensitive as hippocampal neurons to oxygen-glucose deprivation (OGD). In order to identify signaling pathways that account for the differential susceptibility to OGD, we have applied mRNA expression profiling to the purified neuronal populations using Affymetrix rat genome DNA microarrays. Differentially expressed genes will also serve as cell type specific markers for identifying subplate neurons in vivo. We have identified 46 genes, and 24 ESTs that are expressed at least 4-fold higher in subplate neurons than in the other neuron types. Most of these genes are completely absent in the other cell types. We have similarly identified genes that may serve as markers for cortical plate and hippocampal neurons. We are applying mRNA expression profiling to individual cultures of purified neurons using amplified mRNA before and after OGD to identify the cellular signaling pathways modulated by OGD in each cell type.

PET for Localization of Epileptogenic Lesions in Children Harry T. Chugani Children’s Hospital of Michigan, Wayne State University, Detroit, Mich., USA

lepsy syndromes characterized by generalized seizures, such as infantile spasms and Lennox-Gastaut syndrome, resection of focal PET abnormalities corresponding to focal ictal and interictal EEG abnormalities is associated with improved seizure and cognitive outcome. In PET studies on patients with infantile spasms, there is evidence of complex cortico-subcortical interactions believed to be important in an age-dependent secondary generalization of focal cortical discharges to result in the typical spasms. Only about 20% of patients with ‘cryptogenic infantile spasms’ show a single PET focus amenable to resection. The other 80% of patients show more than one focus, and most of these will not be optimal surgical candidates. Some of these infants will benefit from the use of newer PET tracers (see below), which may reveal the primary seizure focus. The finding of bilateral symmetric hypometabolism suggests a nonlesional etiology for the spasms and is a useful indication to pursue further neurogenetic/neurometabolic evaluation rather than a surgical approach. Recent studies using other PET probes have attempted to provide a more specific and accurate assessment of seizure foci. 11C-Flumazenil (FMZ): this labels central benzodiazepine receptors and is particularly useful in showing decreased receptor binding in medial temporal sclerosis. FMZ-PET is also useful in: (i) showing decreased FMZ binding in medial temporal lobe structures associated with a brain lesion elsewhere (dual pathology), (ii) depicting the perilesional epileptogenic zone, (iii) depicting the seizure onset zone, (iv) depicting potential secondary epileptic foci. 11C-a-methyl-L-tryptophan (AMT): this analogue of tryptophan traces serotonin synthesis and kynurenine pathways in the brain. AMT-PET in patients with epilepsy demonstrate focally increased uptake in cortical regions of epileptogenesis as correlated with scalp and intracranial EEG. AMT-PET in children with tuberous sclerosis and intractable epilepsy demonstrate focal increases in AMT uptake in the region of epileptogenic tubers, but not in the apparently nonepileptogenic tubers. Thus, the development of newer more specific PET probes for epilepsy has led to improved and more accurate localization of seizure foci that should ultimately improve outcome of extratemporal epilepsy surgery.

The role of positron emission tomography (PET) with 2-deoxy2(18F)fluoro-D-glucose (FDG) in seizure focus localization is dependent upon the epilepsy syndrome being considered. Temporal lobe epilepsy: In temporal lobe epilepsy (TLE), interictal FDG-PET identifies areas of decreased glucose utilization that correspond to epileptogenic lesions (sensitivity: 80–90%). Occasionally, focal ‘interictal’ hypermetabolism may be seen corresponding to an active focal epileptiform EEG discharge potentially resulting in false lateralization. Therefore, EEG monitoring is important during PET. With newer MR techniques (MRS and volumetry), FDG-PET is now used mainly in MR negative cases or in localization discordant cases. Nonlesional extratemporal lobe epilepsy: These are the most difficult cases in epilepsy surgery and current success rates for seizure-free outcome are 50–55% without the use of PET. FDG-PET provides useful lateralization and localization data in these cases to guide the placement of intracranial electrodes and improve outcome. Regions of interictal cortical hypometabolism should be covered with subdural grids to the extent that is technically practical. Neurophysiological correlations based on coregistration of MRI, PET and subdural electrodes indicate that the seizure onset zone typically lies in the periphery or boundary of the cortical hypometabolism rather than within the hypometabolic zone. Our data also show that the size of the hypometabolic zone increases along the major propagation pathways as a function of duration of intractable epilepsy, thus supporting the concept of network growth and secondary epileptogenesis proposed by Morrell. Surgery for infantile spasms: In a number of childhood epi-

Behavioural/sleep state activity may impact on synthetic processes within the brain, thus accounting for the developmental change in such activity and suggesting a role in the brain’s growth and development. We have therefore determined both the relative leucine uptake (as a measure of protein synthesis and oxidation) by the brain in relation to spontaneous changes in behavioural state, as well

Abstracts

Dev Neurosci 2002;24:446–464

Regional Protein Synthesis and Leucine Metabolism in the Near-Term Ovine Fetal Brain: Relationship to Behavioural State Activity M.J. Czikk, J.C. Sweeley, J.H. Homan, J.R. Milley, B.S. Richardson CIHR Group in Fetal and Neonatal Health and Development, Departments of OB/GYN and Physiology, Lawson Health Research Institute, University of Western Ontario, London, Canada, and Department of Pediatrics, Division of Neonatology, University of Utah, Salt Lake City, Utah, USA

451

as the regional flux of leucine into brain tissue (as a measure of protein synthesis) in the chronically catheterized ovine fetus near term. Nine animals were studied during a 6-hour infusion of L-[1-14C]leucine, with blood sampling for cerebral arterial-venous (A-V) differences during both low-voltage (LV)/rapid eye movement (REM) and high-voltage (HV)/NREM epochs. A flow probe was used on the superior sagittal sinus as a continuous measure of cerebral blood flow (representing flow to F35% of the brain). Following the infusion, animals were sacrificed and the fetal brain dissected and frozen in liquid nitrogen. All plasma and tissue aliquots were analyzed for leucine concentration and [14C]leucine specific activity. Results are presented as grouped means B SEM. Relative leucine uptake was greater during HV/NREM than during LV/REM (77 B 25 vs 44 B 30 nmol/min, p ! 0.02), as was the A-V difference for leucine across the brain (4.7 B 1.0 vs 2.0 B 0.8 mmol/l, p ! 0.01). Relative [14C]leucine uptake was also greater during HV/NREM than LV/REM (971 B 199 vs 341 B 218 DPM/min, p = 0.05), with the change in uptake for [14C]leucine F2 fold that for unlabelled leucine. Regional tissue protein synthetic data is as follows: (FSR = fractional synthetic rate, Flux = unidirectional flux of leucine into the brain. Letter differences, p ! 0.05). Relative leucine uptake by the ovine fetal brain is thus increased during HV/NREM indicating that protein turnover must also be increased at this time and supporting a restorative role whereby the lower metabolic activity during this state favours biosynthetic processes and the anabolic restoration of tissues. The greater increase in the uptake of the labeled versus unlabelled leucine would also indicate a significant level of protein degradation within the ovine fetal brain near term. Regional protein synthesis was highest in the pituitary gland, likely reflecting the active synthesis and secretion of neuropeptides within this area, whereas the lower synthetic rates in the spinal cord may reflect slower turnover of structural components within the nervous system compared to those involved in functional attributes.

FSR, %/day Flux, mmol/ 100g/min

Cortex

Cerebellum

Brain Stem

Spinal Cord

Pituitary

19.2B1.1a

18.1B0.8a

17.3B1.1a

16.1B0.9a

29.3B2.7a, b

0.58B0.05a, b 0.63B0.04b

0.53B0.03a, b 0.41B0.03c

1.33B0.11d

Perinatal Hypoxia-Ischemia Induces Apoptotic and Excitotoxic Death of Progenitors in the Subventricular Zone and White Matter Michael J. Romanko, Raymond P. Rothstein, Jennifer K. Ness, Roland Meyers, Teresa L. Wood, Steven W. Levison Laboratory for Regenerative Medicine, Department of Neuroscience and Anatomy, Pennsylvania State University College of Medicine, Hershey, Pa., USA

Perinatal hypoxia/ischemia (H/I) occurs during development when the cells of the subventricular zone (SVZ) are actively proliferating, migrating and differentiating. Therefore, we hypothesized that progenitors are damaged as a consequence of H/I. To determine the type and time course of cell death after H/I, P7 rats were subjected to

452

Dev Neurosci 2002;24:446–464

unilateral common carotid artery ligation followed by 90 min of H/I (8% O2). Brains were perfusion-fixed at recovery intervals ranging from 2 to 48 h and sections were examined using hematoxylin and eosin, in situ end labeling, TUNEL, electron microscopy and immunofluorescence. By electron microscopy, necrotic, apoptotic and cells showing the features of both necrotic and apoptotic cells were observed. Necrosis was apparent as early as 2 h of recovery while hybrid cells predominated at 12 h in both the SVZ and white matter. Few ISEL+ cells were observed at 4 h, whereas ISEL+ cells became prevalent at distal time points of recovery, peaking at 12 h with 1.3% of the cells staining at this time and persisting to 48 h. In addition 3.2% of cells in the SVZ were positive for active caspase-3 at 12 h. Interestingly, more TUNEL and pyknotic cells were visualized (9.4% and 10.5% respectively) at 12 h than active caspase-3+ cells, suggesting that a subset of cells are dying via a caspase-3 mechanism. In the white matter active caspase-3+ cells were also observed. We conclude that progenitors in the SVZ and white matter are vulnerable to H/I which will result in the production of fewer neurons and glial cells. The temporal evolution from necrotic to hybrid to apoptotic deaths suggests that as the energy supply is restored with reperfusion, damaged cells initiate the cell death program. Supported by MH59950, HD30705, NS37560 and an AHA fellowship award to JKN.

Hexagonal Coil: A Single Channel Multi-Coil Design for Small Animal Imaging Jelena Lazovic-Stojkovic a, Qing X. Yang a, Dragan Stojkovic b, Wanzhan Liu a, J. Thomas Vaughan c, Michael B. Smith a a Center

for NMR Research, Department of Radiology, The Pennsylvania State University College of Medicine, Hershey, Pa., b Department of Physics, The Pennsylvania State University, University Park, Pa., c Center for Magnetic Resonance Research, Department of Radiology, School of Medicine, University of Minnesota, Minneapolis, Minn., USA

Introduction: There is a growing interest in imaging multiple animals simultaneously to increase throughput. Fitting multiple animals into a single volume coil may not provide sufficient SNR for the voxel size required for micro-imaging. We propose a single channel multi-coil design based on a coaxial cavity with high sensitivity. This hardware development is aimed toward studying the changes in the brain during hypoxia-ischemia in neonatal (P7) rat. Material and Methods: The coil consists of 6 pairs of parallel copper plates (5 ! 2 cm, 2 cm distance between two plates) conformed into a hexagonal coaxial cavity. The inner conductor is a continuous copper sheet, while the outer conductor plates are connected with wires at each end forming a complete ‘endring’. Each inner plate is connected to the corresponding outer plate with 4 capacitors at the 4 corners. Capacitor values are chosen so that coil is tuned to resonate at 125 MHz. The coil is connected to the tuning and matching circuit through any of these capacitors. Thus, each pair of the parallel plates can be regarded as an individual coil similar to a slotted tube resonator and the space between the plates forms an imaging cell with high filling factor. Results: The unloaded quality factor (Q-value), measured on

Abstracts

Structural Proteins during Brain Development in the Preterm (PT) and Near-Term (NT) Ovine Fetus and the Effect of Intermittent Umbilical Cord Occlusion (UCO) E. Rocha a, S. Totten a, R. Hammond b, B. Richardson a a Departments

of Physiology and Ob/Gyn, CIHR Group in Fetal and Neonatal Health and Development, Child Health Research Institute; b Department of Pathology, University of Western Ontario, London, Ont., Canada

Fig. 1. An axial T2-weighted image of anesthetized six 7-day old rat pup brains and enlarged image of one rat pup brain, acquired with a 2D spin echo method.

Fig. 2. An axial T1-weighted image of anesthetized six 7-day old rat pup brains and enlarged image of one rat pup brain, acquired with a 3D gradient echo method.

Objective: High rates of cerebral protein synthesis are evident during early life in support of the brain’s growth and development, which may be disrupted by intermittent hypoxic insults with UCO antenatally, thus contributing to aberrant neurological development. We have therefore determined the changes in immunoreactivity (IR) of selected proteins during brain development in the ovine fetus and the response to severe, but limited hypoxic insults with intermittent UCO: vimentin and glial fibrillary acidic protein (GFAP), as markers for astroglial maturation and astrogliosis, and myelin basic protein (MBP), as a marker for oligodendrocytes and myelin formation. Methods: Fourteen PT (0.75 gestation; control group n = 7 and UCO group n = 7) and 15 NT (0.9 gestation; control group n = 7 and UCO group n = 8) animals were studied over 4 successive days with UCO of 90 s duration every 30 min for 3 to 5 h daily in the UCO animals. Animals were sacrificed and the fetal brain dissected and processed for histologic analysis of the white and gray matter. IR was quantified with an image analysis system (Northern Eclipse) and expressed as the fractional area positively stained for each protein. Results are presented as grouped means B SEM. Results: Intermittent UCO in both the PT and NT animals produced a severe but limited hypoxic insult (fetal PaO2 F22–7 mm Hg) with a modest fall in pHa (F7.46–7.30), but no cumulative acidosis. p ! 0.05.

Group

the HP spectrum analyzer (HP 4195A) was 530, while the loaded Q-value with six rat pups was 290. A T2-weighted spin echo (TE = 19.4–213.51 ms, TR = 3,000 ms, 512 ! 256 matrix, 8.22 cm2 FOV, 1 mm slice thickness and 160 ! 320 Ìm voxel size, 1 average) and T1-weighted 3D gradient echo image (TE = 10 ms, TR = 40 ms, 512 ! 512 ! 32 matrix, 8 ! 8 ! 2 cm3 FOV, and 157 ! 157 ! 625 Ìm voxel size, 1 average) of the six rat pups is shown in figures 1 and 2, respectively. Discussion: The advantage of the multi-animal probe is that six animals will be imaged at the same time, thus excluding any ambiguity due to different experimental conditions, which have proved to have significant influence on the outcome of the experiments. Also, since six animals can be studied simulataneously, statistical significance can be achieved more rapidly.

Abstracts

PT control PT UCO NT Control NT UCO

Vimentin

GFAP

MBP

white matter

gray matter

white matter

gray matter

white matter

30.6B5.9 20.8B4.7 15.3B2.5* 15.9B2.8

22.6B4.6 14.4B3.3 12.4B3.0 12.6B3.1

3.8B1.6 0.7B0.1+ 5.3B0.4* 5.1B1.5

0.08B0.02 0.02B0.01+ 0.35B0.10** 0.18B0.09

14.8B2.5 13.5B2.0 52.3B3.6** 51.5B5.8

Control PT vs. NT, * p ! 0.05, ** p ! 0.01; Control vs. UCO, + p ! 0.01.

Vimentin IR was decreased in the white matter (F50%) with advancing gestation between the PT and NT control group animals. Conversely, GFAP IR was increased in both the white and gray matter (F1- and 4-fold, respectively) with advancing gestation, while MBP IR was increased in the white matter (F3-fold). Intermittent UCO in the PT animals resulted in a modest decrease in vimentin IR (F35%) and a marked decrease in GFAP IR (F80%) in both the white and gray matter, but with no change in MBP IR. However, these proteins were little changed in the NT animals with UCO. Conclusions: The changes in vimentin, GFAP and MBP from PT to NT animals are consistent with the normal developmental pattern of these proteins during astrocyte and neuronal maturation. The selective decrease in both vimentin and GFAP in the PT group animals in response to UCO may reflect their participation in the high rates of

Dev Neurosci 2002;24:446–464

453

protein turnover evident at this stage of brain development, and thus a gestational age dependent vulnerability to cord related hypoxic insults.

Neuroprotection of Creatine Supplementation in Neonatal Rats with Transient Hemispheric Cerebral Hypoxia-Ischemia B.P. Wagner a, J. Nedelcu, K. Adcock, T. Loenneker, E. Martin aPediatric

Effects of Fasting or Dexamethasone Pretreatment on Oxygen Consumption during Hypoxia-Ischemia in the Immature Rat Jianli Wang, Shannon Beabes, Michael B. Smith Center for NMR Research, The Pennsylvania State University College of Medicine, Hershey, Pa., USA

Twenty-four hours of fasting or a single dose of dexamethasone (DEX) pretreatment have been shown to be neuroprotective in a 7day-old neonatal rat model of cerebral hypoxia-ischemia (HI) [1, 2]. They can make immature animals lose bodyweight and increase their blood ketone body levels [3, 4]. The purpose of this study was to investigate the metabolic characteristics of this animal model before, during, and after cerebral HI under these two pretreatments. Methods: Sixty-two Wistar rat pups were randomly divided into 3 groups: control, DEX pretreatment, and fasted group. The control and DEX groups received i.p. injection with either saline or 0.1 mg/kg DEX 24 h prior to the onset of HI. On day 7 they were subjected to a combined insult of right common carotid artery ligation followed by 3 h of hypoxia in 8% O2 [5]. HI was induced inside an oxygen monitoring chamber (n = 54) at 36.1 ° C. The oxygen consumption rate was measured with a Columbus Instruments Oxymax system. The pups were sacrificed 42 h post HI to assess brain damage and edema rate. Core temperature was recorded with a multi-channel OMEGA microprocessor thermometer (n = 12). Results: Compared to the control group which all had brain damage after HI (% edema = 4.70 B 1.57), only 2 fasted pups and 2 of the DEX group had slight brain damages (% edema = 0.57 B 1.60 and 0.12 B 1.06). In 21% O2, VO2 of the fasted group was much less than the other two groups (p ! 0.01). In 8% O2, hypoxia lowered VO2 by 13.5%, 24.4%, and 21.6% in DEX, fasted, and control groups, respectively. Animals pretreated with DEX had higher VO2 during the HI period than the saline group (p ! 0.001), while the fasted group had lower (p ! 0.001). During the 42hour recovery period, the DEX group gained more bodyweight than the control and fasted groups (p ! 0.001). Under the same environment and pretreatment, gender difference did not affect pathologic outcome, edema rate, or bodyweight changes during the whole timecourse (p 1 0.05). There is no significant difference of the rat pup core temperature between or within the three groups before, during, and after 3 h of HI (p 1 0.05). Conclusion: These results suggest that while both fasting and DEX pretreatment increase the availability of energy to the brain, only DEX pretreatment improves the efficiency of neonatal oxygen utilization under hypoxic-ischemic conditions. Thus, there may be a different mechanism between fasting and DEX pretreatment in neonatal neuroprotection.

Chumas P.D., et al, J. Neurosurg., 1993;79:414. Tuor U.I., Annals NY Acad. Sci, 1995;765:179. Enders B., Smith M.B., SMRM Abstract, 1993;1489. Dardzinski B.J., Smith M.B. et al, Pediatr. Res., 2000;48:248. Vannucci R.C., Mujsco D., Biol. Neonate, 1992;62:215.

454

Aim: A recent study showed age-dependent protection of creatine supplementation against hypoxia-induced seizures [1]. We hypothesized that creatine supplementation would minimize the secondary energy failure and the extent of brain edema seen after severe but transient cerebral hypoxia-ischemia in the neonatal rat model. Methods: In the first set of 6-day-old rats (n = 16), 3 g/kg bodyweight/d creatine-monohydrate was injected subcutaneously for 3 consecutive days, followed by 31P-MRS at P9. Noninvasive 31P-MRS (during halothane anesthesia) was carried out with a home-built surface coil to fit one hemisphere as previously described [2]. In a 2nd set of 4-day-old rats (n = 16), again after creatine substitution, the classical cerebral hypoxic-ischemic insult was induced during halothane anesthesia on P7. The transient, mostly hemispheric insult consisted of right common carotic artery ligation followed 1 h later by 100 min of hypoxia (8% O2). Rats were maintained at 37 ° C rectal temperature until MRI were performed 24 h after insult. Fractional brain edema was calculated from DWI and T2-weighted imaging. A 2-T whole-body MR system (Bruker-Medical, Fällanden, Switzerland) with a thermoregulated animal holder was used for all MR measurements. Results (see table): 3 days of creatine substitution significantly increased the ratio of PCr/ATP as measured by 31PMRS. In rats with hemispheric cerebral hypoxic-ischemic insult, creatine-substituted rats showed a significant reduction of brain edema by 25% when compared with nonsubstituted rats.

Creatine-substituted rats Control (no creatine)

Dev Neurosci 2002;24:446–464

PCr/ATP ratio in normal rats

Brain edema (% of total brain) in hypoxic-ischemic rats

1.53 (n = 8)* 0.96 (n = 8)

32.0 (n = 8)* 42.5 (n = 8)

* p ! 0.01, creatine-substituted versus control.

Conclusions: We confirmed the results of previous studies showing that creatine substitution increases cerebral energy storage in neonatal rats. In addition, prophylactic creatine substitution demonstrated a significant neuroprotective effect 24 h after transient cerebral hypoxia-ischemia. Neuroprotection is probably due to reduction of the secondary energy failure, because DWI were reported to correlate with PCr/ATP ratio in the acute phase of injury (2, and ongoing study). However, longterm benefits of creatine substitution need to be examined before starting evaluation in the clinical settings of perinatal cerebral hypoxia-ischemia.

References 1

References 1 2 3 4 5

Intensive Care, University Childrens Hospital, Berne; Neuroradiology and Magnetic Resonance Research, University Childrens Hospital, Zurich, Switzerland

2

Holtzman D, Togliatti A, Khait I, Jensen F: Creatine increases survival and suppresses seizures in the hypoxic immature rat. Pediatr Res 1998; 44:410–414. Nedelcu J, Klein MA, Aguzzi A, Boesiger P, Martin E: Resuscitative hypothermia protects the neonatal brain from hypoxic-ischemic injury. Brain Pathol 2000;10:61–71.

Abstracts

Inhibition of nNOS and iNOS upon Hypoxia-Ischemia Improves Long-Term Outcome but Does Not Influence the Inflammatory Response in the Brain Evelyn R.W. van den Tweel a, Frank van Bel a, Klaas Nicolay b, Cobi J. Heijnen c, Floris Groenendaal a, a Neonatology, b Experimental c Psycho-Neuro-Immunology,

In Vivo NMR; University Medical Center,

Utrecht, The Netherlands Background: Neuronal (nNOS) and inducible NOS (iNOS) contribute to neuronal damage following perinatal hypoxia-ischemia. Inhibition of these enzymes may be neuroprotective. The inflammatory response has been indicated to contribute to the neuronal damage and recently increased levels of IL-1ß, IL-6 and TNF· have been reported in the CSF of asphyxiated fullterm neonates. Upregulation of the heat shock protein 70 (HSP70) can be provoked by hypoxicischemic stress. Objective: We hypothesized that by inhibition of nNOS with 7-nitroindazole (7NI) and iNOS with aminoguanidine (AG) the neuronal damage and the inflammatory response following hypoxia-ischemia will be reduced. Design/Methods: The Vannucci-Rice model was adapted to 12-day-old rats. Hypoxia-ischemia was induced in 48 rat pups by right common carotid artery ligation followed by 90 min of hypoxia (FiO2 0.08) in an incubator with a constant temperature (37 ° C) and humidity. In this model unilateral damage to the right hemisphere is induced. Immediately after hypoxia-ischemia 24 rats received placebo treatment (PLAC) and 24 rats AG (100 mg/kg i.p. twice a day for two days) and 7-NI (50 mg/kg i.p. once) (AG +7NI). Left and right hemispheres were removed and snap frozen in liquid nitrogen 12 h post HI, n = 12 per group. From the individual hemispheres total RNA and protein were isolated. mRNA for the cytokines IL-1·, IL-1ß, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, TNF·, TNFß, and IFN-Á was examined. By means of a Western blot assay the presence of specific HSP70 protein was assessed. In the other 24 animals T2-weighted MRI was performed 36 h post hypoxia-ischemia on a 4.7-Tesla SISCO/Varian spectrometer, with TR 2 s and TEs of 17, 54, 91, and 128 ms. Six weeks post hypoxiaischemia MRI measurements were repeated and brains were removed for histology. From T2 maps T2 values were calculated in the region of interest (about 2 mm3 volume) indicated by histological landmarks. Histopathology was scored on a 4-point scale: 1 = normal, 2 = few neurons damaged mainly in hippocampus, 3 = moderate number of neurons damaged mainly in hippocampus and cortex, 4 =

cystic infarction. Results: Data are represented as mean B SEM, cytokines and HSP70 in arbitrary density units, T2 in ms (table 1). mRNA for the cytokines IL-1·, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, and IFN-Á was negative at 12 h post HI. Conclusions: Long-term neuroprotection in 12-day-old rats with hypoxia-ischemia is obtained with specific, simultaneous inhibition of nNOS and iNOS, but not eNOS. The amount of HSP70 protein is significantly increased in the ipsilateral site of both groups. Moreover HSP70 tended to be lower in the ipsilateral side of the group with NOS inhibition. The inflammatory response at 12 h post HI was similar in both hemispheres and both groups. We speculate that this early inflammatory response did not contribute to the brain injury, induced by hypoxiaischemia and nNOS and iNOS activation, in the present model.

Cytokines in the Brain following Hypoxia-Ischemia in the 12-Day-Old Rat Evelyn R.W. van den Tweel a, Floris Groenendaal a, Joost M. Bakker b, Frank van Bel a, Cobi J. Heijnen b a Neonatology

and b Psycho-Neuro-Immunology, Wilhelmina Children’s Hospital, University Medical Center, Utrecht, The Netherlands

Background: In recent studies increases of IL-1ß, IL-6 and TNF· in CSF have been demonstrated in fullterm neonates with perinatal hypoxia-ischemia on the first day of life. Cytokines and chemokines in the brain will recruit leukocytes, thereby causing increased cerebral damage. The 70-kD heat shock protein (HSP70) is induced following hypoxic-ischemic stress and is reported to have neuro-protective effects. More insight in the production of cytokines and HSP70 in the neonatal brain following hypoxia-ischemia can provide insight necessary for novel intervention strategies. In a pilot study a peak of cytokine production was demonstrated 12 h post hypoxia-ischemia, therefore this time point was chosen in the present experiment. Objective: To study cytokine and HSP70 production in the brain following hypoxia-ischemia and compare this to long-term histology in the 12-day-old rat as a model of perinatal hypoxia-ischemia in the human fullterm neonate. Design/Methods: The Vannucci-Rice model was adapted to 12-day-old rats. Rat pups (n = 24)

Table 1. Group

IL-1ß

TNF·

TNFß

HSP70

Histology score

T2 36 h

T2 6 weeks

PLAC contralateral PLAC ipsilateral AG + 7NI contralateral AG + 7NI ipsilateral

2,006B725 1,743B600 1,515B698 1,920B598

949B248 1,610B550 627B89 856B191

1,582B540 1,746B329 1,012B390 1,454B320

382B177 2,922B937* 285B167 1,922B322*

1.2B0.1 2.6B0.4* 1.1B0.1 1.6B0.4*, #

74.3B1.6 77.3B2.7 76.4B0.8 76.9B0.9

50.1B1.3 58.1B7.1 51.5B1.3 51.5B1.0

* p ! 0.05 ipsilateral vs contralateral, Wilcoxon signed ranks test. # p ! 0.05 AG + 7NI ipsilateral vs PLAC ipsilateral, Mann-Whitney U test.

Abstracts

Dev Neurosci 2002;24:446–464

455

were anesthetized (O2:N2O 1:1, halothane 2%) and the right carotid artery was ligated. After at least 60 min recovery, the animals were placed in an incubator for 90 min with a constant temperature (37 ° C) and humidity and a FiO2 of 0.08. In this model unilateral damage to the right hemisphere is induced. In 12 rat pups the left and right hemispheres were removed and snap frozen in liquid nitrogen 12 h post HI. From the individual hemispheres total RNA and protein were isolated. mRNA for the cytokines IL-1· , IL-1ß, IL-3, IL-4, IL-5, IL-6, IL-10, TNF·, TNFß, IL-2 and IFN-Á were examined with a RNase protection assay. By means of a Western blot assay the presence of specific HSP70 protein was assessed. Histological studies were performed in 12 rat pups 6 weeks post HI. The histopathology was scored on a 4-point scale: 1 = normal, 2 = few neurons damaged mainly in hippocampus, 3 = moderate number of neurons damaged mainly in hippocampus and cortex, 4 = cystic infarction. Results: Data are represented as mean B SEM, cytokines and HSP70 in arbitrary density units. mRNA for the cytokines IL-1·, IL-3, IL-4, IL-5, IL-6, IL-10, IL-2 and IFN-Á was negative at 12 h post HI. IL-1ß

Contralateral 2,006B725 Ipsilateral 1,743B600

TNF·

TNFß

HSP70

Histological score

949B248 1,610B550

1,582B540 1,745B329

382B178 2,922B937*

1.2B0.1 2.6B0.4*

* p ! 0.05 ipsilateral vs contralateral, Wilcoxon signed ranks test.

Conclusions: The HSP70 protein was significantly increased in the ipsilateral hemisphere, while the contralateral hemisphere showed basal levels. mRNA for IL-1ß, TNF· and TNFß were not different between both hemispheres at 12 h post HI. Histological examination 6 weeks post HI shows damage on the ipsilateral side of the brain only. We speculate that elevated levels of mRNA for cytokines at 12 h post HI are not indicative of irreversible brain damage.

MRI, DWI, Perfusion, MRS and 14C-2-DG-Autoradiography of the Brain in Nearly-Term Fetal Lambs Subjected to Fetal Cord Occlusion K. Thorngren-Jerneck a, D. Ley a, I. Burtscher d, R. D'Argy e, B. Geijer d, l. Hellström-Westas a, E. Hernandez b, G. Lingman b, T. Ohlsson e, S.-E. Strand e, O. Werner c, S. Holtås d, K. Marsal b Departments of a Pediatrics, b Obstetrics and Gynecology, c Anesthetics, d Neuroradiology and e Radiation Physics, Lund University Hospital, Lund, Sweden

Aim: To investigate the consequences of fetal systemic asphyxia on brain morphology, function and metabolism, by sequentially, during postnatal recovery, applying MR (T1 and T2), DWI (diffusion weighted imaging), 1H-MRS and perfusion imaging; to compare the results with cerebral glucose metabolism, measured by 14C-DG-autoradiography at the end of the study period, 5 h and 27 h respectively; and to compare with sham controls. Methods: Three lambs were subjected to systemic fetal asphyxia by ligation of the umbilical cord. Two lambs served as controls. All lambs (gestational age 134–137

456

Dev Neurosci 2002;24:446–464

days) were examined with MR at 2.5 h postnatal age. In two lambs, one asphyxiated and one control, the MR investigation was extended, with five serial MR examinations at 4, 5.5, 7, 8.5 h and finally at 25 h postnatal age (serial experiment). CMRgl measured by 14CDG-autoradiography was performed at the end of the study, at 5 h and at 27 h respectively. Results: In the asphyxiated animal, increased signal in the parasagittal region was seen at 4 h (examination no. 2) and onwards on T2-weighted images. The T1-weighted examination after contrast medium injection at 4 h showed bloodbrain-barrier damage in the parasagittal area and in the basal ganglia. These changes increased with time and at 8.5 h. Similar changes were seen in the central gray matter. At 25 h these changes had become widespread and diffuse. Diffusion Measurements: The ADC-values were always lower in the asphyxiated than in the control animals, and the changes were most pronounced in the central gray matter and parasagittal cortex. The variation of ADC-measurements in the control animals was very small. Perfusion measurements: A marked reduction in the parasagittal cortex and a central hyperperfusion were noted in the asphyxiated animal at 4 h. At later measurements this distribution pattern could no longer be observed, and there was a general reduction of the perfusion. In the control animal there was a normal circulation pattern, which was consistent at all three measurements. Proton MR spectroscopy (1H-MRS): A marked difference was seen in the Lac/Cr ratio between the asphyxiated (high Lac/Cr) and the control animals (low Lac/Cr). In the serial experiments an increase in Cr/NAA was seen over time in the asphyxiated animal compared to stable Cr/NAA ratios in the control animal. 14C-deoxyautoradiography: At 5 h the asphyxiated lamb had a total CMRglc of 34.9 Ìmol/min/100 g as compared with the control lamb 51.1 Ìmol/ min/100 g. At 27 h there were no detectable CMRglc in the asphyxiated lamb, to compare with 9.2 Ìmol/min/100 g in the control lamb.

A Cerebral Palsy Phenotype in Newborn Rabbits after Antenatal Hypoxia-Ischemia in Preterm Fetuses Sidhartha Tan, Matthew Derrick, Joanne C. Bregman, Tamas Jilling Department of Pediatrics, Northwestern University and Evanston Northwestern Healthcare, Evanston, Ill., USA

Hypoxia-ischemia to the fetus at a preterm gestation has been implicated in periventricular leukomalacia and later cerebral palsy. The objective was to investigate the effect of hypoxia-ischemia to the fetus at a preterm gestation to develop a model for cerebral palsy, as there is a paucity of animal models with hypertonicity. Initially nearterm animals were subjected to repetitive uterine ischemia at E28 (89% term gestation) and then allowed to deliver spontaneously at E32. A panel of neurobehavioral tests were developed for New Zealand white rabbits (4 pups/litter, n = 12 pups/group) on postnatal day 1 (postconception day 32, E32). The examination was videotaped and scored by two blinded observers (0–3 scale). Significant impairment of the olfactory and tactile responses, movements of head and legs, movement patterns, activity, righting reflex, and coordinated sucking and swallowing were observed in the hypoxic animals com-

Abstracts

pared to controls (Wilcoxon signed rank test) but there was no difference in muscle tone in the extremities. A second group of animals were subjected to sustained hypoxia at E22 (70% term gestation) and then allowed to deliver spontaneously at E32. In addition to changes similar to that described above, there was significant increase in hypertonicity in the hypoxic animals compared to controls. The percentage of newborn deaths was 23% and of those tested (n = 32), 33% had spasticity. Similar results were obtained following sustained hypoxia at E21. Antenatal hypoxia-ischemia at preterm gestation results in motor, sensory and reflex deficits in the newborn rabbit, which are consistent with deficits observed in cerebral palsy.

Time-Course of Cellular and Axonal Damage after Mild Acute Hypoxia-Ischemia in P3 Rat Brain Stéphane Sizonenko a, b, Chris Williams a, Terrie Inder c, Peter Gluckman a a Liggins

Institute, University of Auckland, Auckland, New Zealand; b Department of Pediatrics, School of Medicine, University of Geneva, Geneva, Switzerland; c Murdoch and Children’s Research Institute, Melbourne, Australia

Background: White matter (WM) injury is a major cause of neurodevelopmental disabilities in preterm infants however its pathophysiology is not fully understood. Objective: To study the time course of cellular and axonal damage after a focal acute HI injury in the developing P3 rat brain. Design and Methods: We previously developed a model of cerebral HI injury in the P3 rat that leads to loss of cortical volume and alterations in cortical myelination at P21 [1]. Using this unilateral HI model we have determined cellular degeneration with Fluoro-Jade B (FJ-B) and axonal disruption with ß-APP immunostaining. After right carotid ligation and 30 min hyp-

Abstracts

oxia at 6%, animals from each litter were collected at set time-points and their brains were stained with FJ-B B/DAPI and for ß-APP. A score of positive staining was used to assess the injury: 0 = no positive cells, 1 = scattered positive cells, 2 = moderate amount of positive cells, 3 = large amount of positive cells. Results: Positive fluorescent cells for FJ-B and neuronal ß-APP accumulation were present in the ipsilateral cortex with no involvement of other cerebral regions. This is consistent with the reduction of the cortical size and altered pattern of myelination seen at P21 in this model. The time course of cellular degeneration in the right lateral cortex stained with FJ-B is shown below (left graph). It was maximum 24 h after the HI injury. A clear columnar pattern of cellular degeneration was seen extending from the deep cortical layer directly above the WM tracts of the corpus callosum (corresponding to the subplate neurons layer at this stage of development) to layers 6 and 5 of the developing cortex. Similarly, axonal disruption in the lateral cortex immunostained with ß-APP was maximum at 24 h and delayed up to 72 h after the injury (right graph). The damaged axons were organized in the same columnar pattern and showed the same localisation as with the FJ-B staining. As shown by the level of scoring the overall amount of damage was low in this model suggesting that the damage is very mild. Further the DAPI stain revealed highly condensed and fragmented nucleus in positive FJ-B cells suggestive of delayed cell death. Conclusion: Delayed cell death and axonal degeneration in the deeper subplate and cortical layers 5 and 6 occurred after a mild transient hypoxic ischemic injury in the P3 rat brain. This unique pattern of cellular and axonal degeneration precedes the development of diffuse cortical WM lesions. These data suggest that a window exists when it may be possible to treat after an injury to prevent delayed cell death and subsequent development of diffuse WM lesions. References 1 2

Sizonenko, S.V., et al., Hypoxia-ischaemia in P3 rats leads to an alteration in subcortical myelination at 21 days. Pediatr Res, 2001;49:441A. Sizonenko, S.V., et al., Selective alteration in myelination after hypoxicischemic injury in the very immature rat brain. Pediatr Res, 2002. Submitted.

Dev Neurosci 2002;24:446–464

457

Under What Circumstances Can Seizures Produce Hippocampal Injury: Evidence for Age-Specific Effects

Hypoxic-Ischemic Injury in the Immature Rat Brain Induces a Biphasic Molecular Response

S.L. Moshé, A.S. Galanopoulou

Departments of a Pathology, b Human Biological Chemistry and Genetics, University of Texas Medical Branch, Galveston, Tex., USA

Departments of Neurology, Neuroscience and Pediatrics, Monefiore/Einstein Epilepsy Management Center, Albert Einstein College of Medicine and Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, N.Y., USA

Clinical and research evidence indicates that the immature brain is highly susceptible to the development of seizures in response to environmental stimuli such as fever, infection or head trauma. In addition, epileptic seizures can occur early in life in association with brain anomalies. In either case, the seizures can be prolonged and often the first presentation may be status epilepticus. Because of the increased frequency of seizures, particularly complex febrile seizures during infancy and early childhood, it has often been assumed that these early seizures are the cause of mesial temporal sclerosis. Mesial temporal sclerosis is the most common pathological substrate of temporal lobe epilepsy and is often intractable to medical treatment. Retrospective clinical data support the notion that complex febrile seizures may be related to mesial temporal sclerosis. However, this link is not supported by prospective studies. Experimental studies in normally developing rats indicate that seizures early in life do not produce hippocampal damage similar to that observed in adult rats under the same conditions. However, under certain circumstances these seizures may produce subtle hippocampal changes including alterations in hippocampal physiology or dendritic morphology, granule cell neoneurogenesis, or age-specific patterns of synaptic reorganization as well interfere with learning although to a lesser degree than that observed after severe seizures in adults. Some of these changes may be associated with a decrease in provoked seizure threshold in adulthood although only rarely spontaneous seizures have been reported. On the other hand, provoked seizures in neurologically abnormal rats may produce hippocampal damage that resembles the damage observed in adult rats with normal brains. These experimental data appears to be supported by recent clinical studies utilizing MRI imaging in patients with complex febrile seizures soon after the seizure occurred. At this juncture, the data suggest that there is a gamut of changes that can occur following seizures early in life. Factors such as the presence of prior neurological abnormalities, age, repetitive seizures, genetic predisposition, as well as etiology of the provoked seizures may affect this range of hippocampal changes. The key point is to identify the significance of these changes and design preventive treatments. Previous strategies dealt with the continuous administration of antiepileptic drugs with or without loading doses and maintenance for a long time. New neuroprotective treatments may include a rational combination of various agents including currently available AEDs, hormonotherapy, stimulation paradigms, exposure to enriched environment, vaccinations or other immune treatments and prevention of secondary ‘hits’ that may aggravate an already compromised brain. Studies will need to be designed to determine timing and type of intervention, dosing, follow-up and most importantly what should be the desired endpoint.

458

Dev Neurosci 2002;24:446–464

S.S. Martin a, J.R. Perez-Polo b, M.R. Grafe a

Background: DNA damage, a well-documented consequence of hypoxic-ischemic brain injury (HII), activates Poly (ADP-Ribose) Polymerase-1 (PARP-1). Activated PARP-1 utilizes cellular NAD+ to ribosylate nuclear proteins in an effort to maintain the integrity of the DNA repair process. Severe DNA damage (like that seen in the ischemic core of an infarct) following HII, may over-activate PARP1, deplete cellular energy stores and exacerbate tissue destruction. On the other hand, moderate levels of DNA damage (often observed in the penumbra of the infarct) may result in PARP-1 activation followed by caspase 3-mediated PARP-1 inactivation. Active caspase 3 cleaves the 116-kDa PARP-1 protein into the 89-kDa and 25-kDa fragments. PARP-1 cleavage is considered a hallmark of apoptosis. Hypotheses: 1) PARP-1 activity is increased following HII in the P7 rat and 2) cell death following HII occurs via both apoptosis and necrosis. Experimental Studies: Using the Rice-Vannucci model, we examined bilateral changes in the levels of PARP-1, cleaved PARP-1, caspase 3 activity and cell death between 0.5 and 48 h posthypoxia. Results: We observed biphasic changes in the levels of PARP-1 and cleaved PARP-1 in the cortex ipsilateral to the ligated carotid artery, which peaked at 0.5 and 12 h post-hypoxia. Mild cell death, evident at 0.5 h, progressed to moderate levels of cell death in some animals by 2 h. Elevated levels of caspase 3 activity were also found at 0.5 h; however, caspase 3 activity remained elevated for the remainder of the study. In contrast to many studies, we also detected a peak in caspase 3 activity levels in the contralateral cortex immediately following hypoxia; caspase 3 activity was also increased in sham-operated rats. This transient rise in caspase 3 activity was associated with elevated levels of PARP-1, cleaved PARP-1 and cell death at 1 h post-hypoxia. Conclusions: We demonstrated biphasic changes in the ipsilateral cortex, which supports previous studies describing biphasic responses to brain injury in the immature rat. We also showed contralateral changes, which are most likely induced by hypoxia/reoxygenation as similar changes were found in the shamoperated animals. Finally, these studies suggest bilateral cell death occurring after hypoxia can be partially attributed to apoptosis.

Abstracts

The Mitochondrially Targeted Antioxidant TPPB (2-[2-(Triphenylphosphonio)Ethyl]-3,4-Dihydro2,5,7,8-Tetramethyl-2h-1-Benzo Pyran-6-ol Bromide) Protects Rat Primary Oligodendrocyte Precursors but Not Neurons from Oxidative Stress E.A. Malecki a, B. Todorich b, M.P. Haaf b, J.R. Connor a a Department

of Neuroscience and Anatomy, College of Medicine, The Pennsylvania State University, Hershey, Pa.; b Department of Chemistry, Elizabethtown University, Elizabethtown, Pa., USA

Mitochondria play a key role in modulating a variety of toxic insults, including oxidative stress. Immature oligodendrocytes are particularly vulnerable to hypoxic insults; hypoxia/ischemia in premature infants results in death of oligodendrocytes and the development of periventricular leukomalacia and cerebral palsy. Smith et al. (Eur J Biochem 1999;263:709–716) have described an antioxidant compound, TPPB, which accumulates in isolated mitochondria and in mitochondria of living cells. We hypothesized that such a compound would be protective against oxidative stress in rat primary oligodendrocyte precursor and neuronal cultures. Cells were pretreated with 0 or 10 ÌM TMH-ferrocene for 24 h. Thirty minutes before oxidative insult, vehicle or TPPB was added to a final concentration of 5 ÌM. Medium was replaced with medium containing 0.5 mg/ml MTT and 0 or 1.5 mM t-bOOH; MTT uptake was quantitated 2 h later. Iron/peroxide treatment resulted in the death of 48% of oligodendrocyte precursors and 19% of neurons (as measured by MTT uptake). Pre-treatment with TPPB attenuated iron/peroxideinduced cytotoxicity in oligodendrocyte precursors, but not in neurons. Mitochondrial targeting of antioxidants remains a compelling strategy for neural cell protection. An immediate goal will be to assess whether this compound protects oligodendrocyte precursors from hypoxic insult.

lowed by 90 min of H/I (8% O2). Within 4 h after this insult, approximately 20% of the total cells are deleted from the SVZ. These early damaged cells appear necrotic. Dying cells continue to accumulate within the next 48 h of recovery. At 12 h, dying cells with features shared by both apoptotic and necrotic cells prevail. At 24 h and later, deaths are predominantly apoptotic. By 48 h, the SVZ sustains a 25% reduction in cellularity. This temporal evolution from necrotic to hybrid to apoptotic deaths suggests that as the energy supply is restored with reperfusion, damaged SVZ cells initiate the cell death program. To better understand the biochemical cascades leading to the deaths of these cells, we analyzed the proportion of cells positive for active caspase 3 relative to the numbers of apoptotic cells. Approximately 3% of the cells in the SVZ stain for active caspase 3, whereas approx 9% are TUNEL+ at 12 h, suggesting that a subset of the apoptotic deaths are mediated by caspase 3. The locations of the dying cells within the SVZ suggest that progenitors rather than stem cells are dying. Confirming this interpretation, an assay for neural stem cells reveals that the number of neural stem cells within the SVZ increases approximately 2-fold by 48 h. However, at 3 weeks survival, the SVZ is smaller, less cellular, and it contains less than 1/4 the normal complement of neural stem cells. We conclude that neural stem cells and progenitors in the SVZ are differentially affected by H/I. Progenitors are eliminated, whereas neural stem cells are resistant, and in fact, the stem cells transiently increase in abundance. These data provide new insights into the mechanisms responsible for the failed generation and regeneration of CNS myelin in periventricular loci. However, they also demonstrate that compensatory responses are initiated in response to brain damage, which if appropriately harnessed and extended could provide the means to regenerate the brain after perinatal injury. Supported by MH 59950 and HD 30705 awarded to SWL.

Differential Expression of Chemokines and Chemokine Receptors during Microglial Activation Hypoxia-Ischemia Eliminates Progenitors, but Not Stem Cells from the Perinatal Subventricular Zone: Consequences for Brain Development Steven W. Levison, Michael J. Romanko, Raymond P. Rothstein, Matthew J. Snyder

Sergey G. Kremlev, Rebecca L. Roberts, Charles Palmer Department of Pediatrics, Pennsylvania State University, Hershey, Pa., USA

Birth asphyxia occurs during development when the brain’s stem cells in the subventricular zone are actively producing progenitors, which then migrate into the developing brain to produce new neurons and glia. Therefore, if these precursors are affected by perinatal hypoxia/ischemia (H/I), then the formation of the brain may be irreparably altered. In the late trimester fetus the germinal matrix, the principle source of new brain cells, consists predominantly of subventricular zone (SVZ) cells. Therefore, the goal of our studies has been to evaluate the vulnerability of neural stem cells and progenitors in the subventricular zone to hypoxic/ischemic insults. Postnatal day 7 Wistar rats were subjected to unilateral carotid artery ligation fol-

Intrauterine infection is associated with a fetal inflammatory response characterized by increased inflammatory cytokines in the fetal brain and activation of brain microglial cells. Intrauterine infection can release bacterial cell wall products into the fetal circulation like lipopolysaccharide (LPS). LPS can activate brain microglial cells to produce inflammatory chemokines and chemokine receptors. We speculate that the degree of injury in the brain is associated with the degree of microglial cell activation. Minocycline (MN) is a drug that can modulate microglial cell activity. To determine if MN can reduce the production of chemokines and chemokine receptors in response to LPS microglial-like BV-2 cells were cultured in the presence or absence of 100 ng/ml of LPS. ELISA and semi-quantitative RT-PCR were used to examine changes in inflammatory chemokines (MCP-1, MIP-1·, and RANTES) and chemokine receptors (CXCR3 and CCR5) production. Cell supernatants were collected at intervals following exposure to LPS, frozen, and analyzed together by ELISA.

Abstracts

Dev Neurosci 2002;24:446–464

Department of Neuroscience and Anatomy, Pennsylvania State University College of Medicine, Hershey, Pa., USA

459

Total RNA was extracted, reverse transcribed and amplified with gene-specific primers. Chemokine release after 8 h exposure to LPS was significantly higher compared to non-exposed cells for all the chemokines measured (P ! 0.001). The inhibitory effect of MN on LPS-stimulated BV-2 cell chemokine release is shown in the table (a).

No LPS (a) Chemokine (pg/ml B SD)* MIP-1· 10,352B989 RANTES 128B30

LPS

LPS B MN

74,235B1311 1,421B196

65,352B2,242 12% 1,200B149 16%

(b) Chemokine receptor/chemokine (O.D. ratio B SD)** CXCR3 0.9B0.34 9.24B1.14 3.88B1.51 MCP-1 1.11B0.66 10.2B2.23 3.97B1.12

Inhibition

58% 61%

* Chemokine release after 16 h exposure to 100 ng/ml of LPS and 1 ÌM of MN (p ! 0.001). ** CXCR3 expression after 1 h, MCP-1 expression after 2 h exposure to 100 ng/ml of LSP and 1 ÌM of MN (p ! 0.001).

Similar inhibition was found for chemokine/chemokine receptors mRNA expression after 1 or 2 h exposure to LPS (p ! 0.001) (table, (b)). LPS directly stimulates microglial cells to produce inflammatory chemokines and chemokine receptors. MN inhibited the release of inflammatoary chemokines by an average of 14% and reduced chemokine/chemokine receptors mRNA expression by an average of 60%. Modulation of inflammatory response may be a valuable therapeutic strategy in the reduction of brain injury associated with fetal exposure to bacterial infection.

Role of MAP Kinase Signaling Pathways in Neonatal Hypoxic-Ischemic Brain Injury David M. Holtzman, Kara Arvin, Byung Hee Han Washington University School of Medicine, Department of Neurology, Molecular Biology and Pharmacology, Center for the Study of Nervous System Injury, St. Louis, Mo., USA

Hypoxic-ischemic (H-I) injury to the developing brain is a common cause of morbidity and mortality. Rodent models of neonatal H-I brain injury have been utilized to study the pathogenesis of cell death and to develop potential treatments. Post-natal day 7 rats and mice undergoing unilateral carotid ligation and exposure to hypoxia (Rice/Vannucci model) develop many of the pathological features seen in the human neonatal brain following H-I. In this model, prominent features of both early excitotoxic/necrotic as well as delayed apoptotic cell death are seen. We have found that H-I turns on specific MAP kinase intracellular signaling pathways that appear to play a role in brain injury following H-I, and neuroprotective treatments that influence these pathways can protect the brain from injury. The neurotrophic factor BDNF given ICV prior to H-I is markedly protective against H-I. BDNF activates ERK1/2 MAP kinase and PI3 kinase in neurons and inhibition of the BDNF’s ability to activate ERK1/2 blocks its protective effects. Following neonatal H-I, we have found that there is activation of both ERK1/2 and p38 MAP

460

Dev Neurosci 2002;24:446–464

kinases. Preliminary evidence suggests that this activation is not in neurons but in glia. While pharmacological inhibition of ERK1/2 had no significant effect on H-I induced brain injury, inhibition of the p38 MAP kinase significantly blocked H-I induced damage. Our recent experiments with the semi-synthetic tetracycline, minocycline, demonstrate that its systemic administration is profoundly neuroprotective against neonatal H-I injury. While more protective than p38 inhibitors, minocycline completely blocks early activation of the p38 pathway suggesting that part of its neuroprotective effects is via suppression of this H-I induced stress MAP kinase. Overall, these results suggest that modulation of MAP kinase signaling may provide new insights into pathogenesis and treatment of neonatal H-I induced brain injury.

Hypoxia-Ischemia in Immature Mice: Disruption of the Poly(ADP-Ribose)Polymerase (PARP)-1 Gene Modifies the Response to Injury H. Hagberg a, M.-A. Wilson b, H. Matsushita b, M. Lange b, M. Poitras c, T. Dawson c, V. Dawson c, F. Northington d, M. Johnston b a Perinatal Center, Sahlgrenska University Hospital, Göteborg, Sweden; b Neuroscience Lab, Kennedy Krieger Institute; Departments of c Neurology and d Pediatrics, Johns Hopkins School of Medicine, Baltimore, Md., USA

PARP-1 is a nuclear enzyme activated by DNA strand breaks to participate in DNA repair. Previous studies in adult animals indicate that PARP-1 is overactivated in response to various insults including ischemia, which depletes its substrate NAD+ and may lead to energy failure and cell death. PARP inhibitors are neuroprotective in models of focal ischemia and PARP-1 deficiency confers neuroprotection [1]. Recently, PARP-1 has been implicated to modify inflammatory [2] and apoptotic [3] responses which also may be relevant to explain its protective effects. The aim of the present study was to evaluate the role of PARP-1 for development of hypoxic-ischemic injury in the immature brain. Methods: 129SV PARP knock-out mice were back-crossed with CD-1 mice and experiments were performed on F3 CD-1 (87.5%)/SV-129 (12.5%) mouse pup (7-day-old) litters of mixed genotype: homozygous KO (KO), heterozygous KO (HET) and wild-type (WT). The right common carotid artery was ligated under isoflurane inhalant anesthesia and litters were exposed to hypoxia (10% oxygen in nitrogen) for 50 min at 36 ° C. The animals were allowed to recover and were re-anesthetized 10 days later. The tail was sampled for genotyping using PCR , and mice were perfused transcardially and fixed with paraformaldehyde. Brain injury was evaluated in cresyl violet/hematoxyline sections (10/animal) and each region was scored with regard to injury and degree of hypotrophy (cerebral cortex 0–4, striatum 0–6, hippocampus 0–6, thalamus 0–6 and total injury score 0–22) without knowing group belonging. Results: The mortality was not significantly different between the the three groups of mice (KO 6.5%; HET 6.4% and WT 13%) leaving 43, 72 and 33 animals for histological evalution in the KO, HET and WT groups, respectively. The total brain injury was reduced (p = 0.016, ANOVA) from 12.8 B 1.4 in WT to 8.5 B 1.1 (–34%) in KO mice. Attenuation of brain injury was more pronounced in the thala-

Abstracts

mus (–50% p = 0.005, KO vs. WT) than in the cerebral cortex (–28% p = 0.04), hippocampus (–24% p = 0.04) or striatum (–39%, p = 0.03). The mean injury score in HET mice was intermediate to that of KO and WT mice, being significantly lower than WT only in the thalamus (–34%, p = 0.03). Subanalysis according to gender, revealed that the reduction of injury in KO vs. WT was more pronounced in males (total score –43%, cerebral cortex –31%, hippocampus –26%, striatum –51% thalamus –71%) than in females (total score –18%, cerebral cortex –16%, hippocampus –16%, striatum –22% thalamus –20%). Brain injury in females tended to smaller in HETs than in KOs, whereas in males the level of injury in HET mice was always between that of WT and KO genotypes. Conclusion: PARP-deficiency rendered immature mice resistant to hypoxia-ischemia to a somewhat lesser degree than in adult animals. However, the response in immature males was comparable to previous studies on ischemia in adult mice executed exclusively in males. On the contrary, complete PARP disruption in female mice did not confer protection, whereas partial deficiency was associated with slightly reduced brain injuries. References 1 Eliasson et al, Nat Med 1997;3:1089–1095. 2 Ullrich et al, Nat Cell Biol 2001;3:1035–1042. 3 Wang et al, Soc Neuroscience, 2001;96;16.

Mitogen-Activated Protein (MAP) Kinase Activation after Transient Focal Cerebral Ischemia in the Neonatal Rat C.K. Fox, N. Derugin, M. Wendland, D.M. Ferriero, Z.S. Vexler Department of Child Neurology, University of California, San Francisco, Calif., USA Objective: To examine the role of MAP kinase signaling in neonatal stroke. Background: JNK, ERK1/2, and p38 MAP kinases are

reactive cells, a marker for activated microglia, were determined after 24 h of reperfusion. The number of ED-1 immunoreactive cells were counted in two random fields of view in the core injured cortex in two sections per brain at the level of the caudate. Results are mean B SD. Results: Increases in p38 phosphorylation were observed within 1 h after reperfusion. The increase was transient, peaking at 8 h of reperfusion, and returning to baseline by 24 h. No changes were observed in total protein expression. Preliminary data with p38 inhibitor studies showed attenuation of microglial activation in the drug-treated group. The number of ED-1 immunoreactive cells decreased by 50%, from 51.3 B 12.9 in placebo to 28.8 B 16.8 in the inhibitor-treated group. Neuronal injury as measured by cresyl violet staining was also decreased by administration of p38 inhibitor. In the P38 inhibitor treated group, the size of infarct measured as a percentage of injured ipsilateral hemisphere was 34.1 B 13% while the placebo treated group was 45.9 B 6.5%. Interestingly, ERK1/2 phosphorylation was also observed after MCA occlusion and reperfusion, following a similar time course as p38 phosphorylation, while no changes were seen in the phosphorylation levels of JNK. Conclusions: Early activation of p38 and ERK 1/2 suggest a role for these signaling cascades in evolving injury after focal cerebral ischemia and reperfusion in neonates. Our preliminary results show that administration of a p38 inhibitor causes attenuation of microglial activation and a tendency towards injury reduction, suggesting a pathologic role of p38 in neonatal stroke. Supported by a Genentech Foundation Fellowship and the American Heart Association Grant-in-Aid, Western Affiliates.

Decreased Expression of Type II Thyroxine Deiodinase and Neurogranin in Sturge-Weber Syndrome Surgical Brain Tissue Anne M. Comi Johns Hopkins University, Baltimore, Md., USA

a family of signaling molecules that mediate various outcomes on the cellular level including growth, differentiation, release of inflammatory cytokines, and apoptosis. Recent models of adult stroke have suggested MAP kinases signaling may be involved in evolving injury. Phosphorylation of p38 has been associated with activation of microglia, which are important mediators of ischemic and reperfusion injury in developing brain. However, little is known about the role of MAP kinases, and p38 in particular, in neonatal stroke. Methods: Neonatal P7 rats were subjected to 3 h of middle cerebral artery (MCA) occlusion or sham surgery followed by reperfusion. At 1 h of occlusion, diffusion-weighted MRI was used to identify the region of cytotoxic edema and evolving infarction. At 30 min, 1, 4, 8, 18 and 24 h after reperfusion, cortical tissue from ischemic core and the contralateral hemisphere was collected. Protein expression and phosphorylation of p38, ERK1/2 and JNK were determined by Western blot using antibodies specific for phosphorylated and total forms of these protein kinases. To further determine the role of p38 signaling after stroke and reperfusion, a p38 specific inhibitor, SB203580 (67 nmol/kg in 1% DMSO, n = 6) or vehicle (1% DMSO, n = 4) was administered by intraventricular injection during MCA occlusion. Histological outcomes (cresyl violet) and number of ED-1 immuno-

Objective: To study the molecular basis of Sturge-Weber syndrome (SWS) by analyzing gene expression profiles of brain tissue from subjects with SWS. Background: Sturge-Weber syndrome is a neurocutaneous disorder characterized by a facial port-wine stain, pial angioma and ocular abnormalities. Individuals with SWS frequently develop epilepsy, migraines, stroke-like episodes, hemiparesis and learning disabilities. Imaging studies of the CNS in subjects with SWS have demonstrated atrophy, cortical calcification, regional hypoperfusion, ischemia-related white matter abnormalities, abnormal glucose metabolism, and decreased N-acetyl aspartate on magnetic resonance spectroscopy, affecting cortical regions underlying the leptomeningeal angioma. Design/Methods: Differentially expressed genes were screened for in surgical SWS cortical brain tissue (n = 2) using microarray analysis of 9,183 genes. The SWS brain tissue (one male and one female) was compared to age and region matched surgical epilepsy cortical tissue (n = 2; both male) which were pathologically normal except for small areas of cortical dysgenesis. SNOMAD normalization (http://pevsnerlab.kennedykrieger. org/snomadinputCarlo.html) was performed upon the microarray data. Genes were identified which had Z scores 1/! 2 in both comparisons indicating their differential regulation within the context of

Abstracts

Dev Neurosci 2002;24:446–464

461

the data set. Rt-PCR confirmation of regulated genes was performed with 4 SWS brain samples compared to 5 control samples (2 epilepsy and 3 postmortem) and relative amounts of cDNA measured by LightCycler. Results: In addition to the increased expression of glial acidic fibrillary protein, fibronectin, and immediate early genes previously reported, a decrease in the expression of type II thyroxine deiodinase (Z scores = –5.6 and –3.3) and neurogranin (Z scores = –7.7 and –5.2) was found in the SWS brain tissue. Rt-PCR studies found a 1.7-fold decrease in type II thyroxine deiodinase expression (p = 0.046) and 3.0-fold decrease in neurogranin expression (p = 0.002) in the SWS brain tissue compared to controls. Conclusions: Type II thyroxine deiodinase is expressed in astrocytes and converts T4 to the active T3 in brain tissue. Thyroid hormone is known to regulate neurogranin expression by neurons. In SWS brain tissue, these results suggest an alteration in the expression of thyroid hormone-related genes in the SWS cortex. These findings also suggest a role for the neuro-endocrine system in the brain’s response to chronic hypoxic injury and in the molecular basis of neurodegeneration in SWS. Further studies are on going to pursue these results with in situ hybridization and Western analysis. Study supported by grants from the Sturge-Weber Foundation and the NINDS.

Regulation of Phosphorylation of c-Jun N-Terminal Kinases (JNKs) in Neonatal Mouse Hypoxic Ischemic Encephalopathy Model H.W.T. Cheung a, T.W.C. Leung a, H.K.F. Yip b, M.H. Sham c, P.T. Cheung a a Department

of Paediatrics and Neuroscience Research Centre; Departments of b Anatomy and c Biochemistry, University of Hong Kong, Hong Kong SAR, China

Jun N-terminal kinases (JNKs) can be phosphorylated and activated in response to diverse stress signals in vitro, such as UV irradiation, hyperosmotic shock, and have been implicated in the pathogenesis of glutamate excitotoxicity and cerebral ischemia in adult animal models. The three isoforms (JNK1, 2 and 3) also have differential roles in regulating developmental neuronal apoptosis. However, their role(s) in postnatal brain injury has not been addressed. We adopted neonatal mouse hypoxic ischemic encephalopathy model to study their role in mediating delayed cell death after injury. Western blot analysis with phospho-specific JNK antibody (recognizes all isoforms of JNK) revealed high basal level of phosphorylated JNK (p-JNK) in the mouse brain hippocampal lysates on postnatal day 7. At the conclusion of the HI experiments (permanent unilateral common carotid artery ligation plus 30 min of hypoxia – designated time 0 h), the ipsilateral p-JNK level was reduced to 1/10 that of contralateral (which was reduced to about 80% of control animals). Such reduction persisted for 30 min., then dramatically rose from 45 min to a peak at 3 h and a secondary fall from 6 to 18 h after HI. Given that there could be differential changes of individual pJNK isoforms, further work is in progress to define the relative proportion of these isoforms throughout the course. The spatial and cellular changes in pJNK were investigated with immunohistochemical study. At 1 h post HI, p-JNK staining was not detectable in the majority of cells, likely reflecting that the high basal pJNK on West-

462

Dev Neurosci 2002;24:446–464

ern blot was contributed by many neural cells. However, positive cells were evident in the hippocampus with the temporal change compatible with that revealed by Western blot. Specifically, in the ipsilateral hippocampus at 1 h after HI, cytoplasmic p-JNK was detected in neurons in the pyramidal cell layer, and the granule cell layer of dentate gyrus. At 3 and 6 h after HI, the number of positive neurons increased, some with condensed nuclei and strong p-JNK immunoreactivity throughout the whole cell. At 24 h after HI, the staining intensity and cell number reduced, except for some hippocampal neurons retaining the strong cytosolic p-JNK immunoreactivity. Parallel immunostaining with anti-phosphorylated-c-Jun antibody (phospho-Serine-73) showed a similar pattern of changes after HI. In the ipsilateral hippocampus at 3 h after HI, increased nuclear immunoreactivity in the pyramidal cell layer of hippocampus and the granule cell layer of dentate gyrus, with much lower intensity over the contralateral side. At 6 and 24 h after HI, the staining intensity decreased. To investigate if JNK phosphorylation may be associated with caspase activation, double immunostaining for phospho-JNK and active caspase-3 was performed. At 6 h post HI, the positive active caspase-3 immunoreactivity present in some ipsilateral hippocampal pyramidal neurons and the granule neuons of dentate gyrus, with condensed and/or apoptotic nuclei, did not colocalise with pJNK staining. By 24 h after HI, the lack of colocalization between p-JNK and active caspase-3 persisted in the pyramidal cell layer of hippocampus. However, some of these pJNK positive cells had condensed/apoptotic nuclei. In contrast, almost all neurons in granule cell layer of dentate gyrus were double stained for p-JNK and active caspase-3 staining. Thus, while our data support that JNK phosphorylation (activation) may lead to caspase-3 activation in dentate gyrus, but not in pyramidal cell layer, it remains possible that other caspasedependent or even caspase-independent apoptotic death pathway could be differentially recruited by the various pJNK isoforms. Complicating such theory is the fact that some strongly p-JNK positive neurons maintained healthy nuclear morphologies, casting an additional possibility of specific JNK isoform phosphorylation being prosurvival for selective neurons after HI insult. Along this line JNK1 and 2 have been shown to be pro-survival during neuronal development while JNK-3 being pro-apoptotic. We therefore conclude that further detailed dissection of the specific involvement of different JNK isoforms is crucial in defining the roles of JNK pathways in neonatal HI insult. This work is supported by Hong Kong Research Grant Council Fund HKU7249/99M (PTC, HKFY, MHS) and Sir Edward HoTung Paediatric Education and Research Fund (PTC).

Hypoxia-Ischemia and Oligodendrocyte Death in the Developing Brain Stephen A. Back Oregon Health Sciences University, Departments of Pediatrics and Neurology, Portland, Oreg., USA

Periventricular leukomalacia (PVL) is the predominant form of brain injury in the premature infant and underlies the development of the chronic spastic motor deficits of cerebral palsy as well as cognitive impairment. In the premature human brain there is a window of

Abstracts

vulnerability in which hypoxia-ischemia (H-I) and other insults damage cerebral white matter producing PVL and related disturbances of myelination. Although PVL is the leading cause of neurological disability in survivors of premature birth, there is no animal model that reproduces the myelination disturbances that give rise to cerebral palsy in humans. Since the death of oligodendrocyte (OL) progenitors could explain these myelination disturbances, we have taken three complementary approaches to test the hypothesis that targeted death of OL progenitors from H-I occurs during acute perinatal white matter injury. First, we showed in vitro that late OL progenitors differ from mature OLs in that they are highly susceptibility to death triggered by multiple sources of oxidative stress. We identified an oxidative stress pathway in which glutathione depletion triggers an intracellular rise in reactive oxygen species (ROS) in late OL progenitors but not mature OLs (Back et al, J Neurosci 1998;18(16):6241– 6253). Second, we found that during human brain development, late OL progenitors are maximally abundant during the high-risk period for PVL (Back et al, J Neurosci 2001;21(4):1302–1312). Third, we developed a variant of the Vannucci model in the P2 rat to demonstrate that late OL progenitors were the major OL stage killed by apoptosis, whereas earlier and later OL stages were highly resistant (Back et al, J Neurosci 2002;22(2):455–463). Thus, the timing of appearance of late OL progenitors during development was closely associated with the severity of perinatal white matter injury. Recently, we found that in a case of acute PVL, human late OL progenitors were also selectively vulnerable to death associated with oxidative injury to the periventricular white matter from reactive oxygen and nitrogen species. Future studies are needed to relate the maturationdependent susceptibility of late OL progenitors to H-I to the myelination disturbances of PVL.

Hyperbaric Oxygen as a Treatment for Neonatal Hypoxia-Ischemia John W. Calvert a, b, Anil Nanda b, John H. Zhang a, b a Department

of Neurosurgery, University of Mississippi Medical Center, Jackson, Mississippi, b Department of Neurosurgery, Louisiana Health Sciences Center in Shreveport, Shreveport, Louisiana, USA

The occurrence of hypoxia-ischemia (HI) during the early fetal or neonatal stages of an individual leads to the damaging of immature neurons. This can result in behavioral and psychological dysfunctions, such as motor or learning disabilities, cerebral palsy, epilepsy, or even death [Cai et al, 1999; Charriaut-Marlangue et al, 1999; Cheng et al, 1998; Gustafson et al, 1999]. Various drugs [Bagenholm et al, 1996; Viswanath et al, 2000] and hypothermia [Bona et al, 1998] have been used to treat HI with some degree of success, but nothing has emerged as an effective clinical treatment. The purpose of this study was to determine the effects of hyperbaric oxygen (HBO) on the neonatal rat brain after an HI insult and to determine if HBO can act as a suitable neuroprotective treatment against the events that follow an HI attack. Seven-day-old rat pups were subjected to unilateral carotid artery ligation followed by 2.5 h of hypoxia (8% O2 at 37 ° C). HBO treatment was administered by placing pups in a chamber (3 ATA for 1 h) 1 h after hypoxia exposure. Brain injury was assessed based on ipsilateral hemispheric weight divided by contralateral hemispheric weight, light microscopy, and EM. Western blot analysis was conducted to determine PARP activity. A sensorimotor functional test was administered 5 weeks after hypoxia exposure. After HI, the ipsilateral hemisphere was 52.65% and 57.64% (p ! 0.001) of the contralateral hemisphere at 2 weeks and 6 weeks, respectively (table 1). In HBO-treated groups, the ipsilateral hemisphere was 77.77% and 84.19% (p ! 0.001) at 2 weeks and 6 weeks (table 1). There was much less atrophy and apoptosis in HBO treated animals under light or electron microscopy. Western blot analysis showed that PARP cleavage was reduced in HBO-treated groups compared to HI groups. Sensorimotor function was also improved by HBO at 5 weeks after hypoxia exposure (¯2, p ! 0.050). In summary, the results of the present study suggest that HBO is able to attenuate the effects of HI on the neonatal brain by reducing the progression of neuronal injury and increasing sensorimotor function. Since HBO has been used in the past, with some success, to treat humans [Nighoghossian et al, 1995] and because it is currently being used as an effective treatment in infants with various disorders such

Table 1. Brain weight

Ipsilateral hemisphere, g Contralateral hemisphere, g Percentage (ips/contra)

Control 2 weeks

HI 2 weeks

HI+HBO 2 weeks

Control 6 weeks

HI 6 weeks

HI+HBO 6 weeks

0.521B0.0045 0.520B0.0083 100.396B1.275

0.276B0.0184a 0.495B0.0059 52.649B2.643a

0.3820B0.0125b 0.499B0.0047 77.774B2.456b

0.600B0.0068 0.606B0.0068 98.998B0.866

0.354B0.0220a 0.561B0.0220c 57.640B3.295a

0.493B0.0133b 0.575B0.0075d 84.1950B1.637b

Values are mean B S.E. a p ! 0.001 compared to control, b p ! 0.001 compared to control, c p ! 0.010 compared to control, d p ! 0.050 compared to control and HI by ANOVA. HI stands for hypoxia-ischemia and HBO stands for hyperbaric oxygen.

Abstracts

Dev Neurosci 2002;24:446–464

463

as radiation-induced bone and soft tissue complications, cyanotic congenital heart disease, and CO poisoning [Ashamalla et al, 1996], it may prove to be an effective strategy for preventing the numerous neurological handicaps that plague many children. However, concerns regarding the toxicity of oxygen that can occur when HBO is administered at high doses or over long periods of time needs to be addressed before HBO can be used in the treatment of newborns.

Subplate Neuron Cell Death and mRNA Expression Profiling following Oxygen Glucose Deprivation M.F. DeFreitas, S.E.G. Hamrick, D.M. Ferriero, P.S. McQuillen Division of Critical Care Medicine, University of California San Francisco, San Francisco, Calif., USA

References Ashamalla HL, Thom SR, Goldwein JW (1996): Hyperbaric oxygen therapy for the treatment of radiation-induced sequelae in children. The University of Pennsylvania experience. Cancer 77:2407–2412. Bagenholm R, Andine P, Hagberg H (1996): Effects of the 21-amino steroid tirilazad mesylate (U-74006F) on brain damage and edema after perinatal hypoxia-ischemia in the rat. Pediatr Res 40:399–403. Bona E, Hagberg H, Loberg EM, Bagenholm R, Thoresen M (1998): Protective effects of moderate hypothermia after neonatal hypoxia-ischemia: shortand long-term outcome. Pediatr Res 43:738–745. Cai Z, Xiao F, Fratkin JD, Rhodes PG (1999): Protection of neonatal rat brain from hypoxic-ischemic injury by LY379268, a Group II metabotropic glutamate receptor agonist. Neuroreport 10:3927–3931. Charriaut-Marlangue C, Richard E, Ben Ari Y (1999): DNA damage and DNA damage-inducible protein Gadd45 following ischemia in the P7 neonatal rat. Brain Res Dev Brain Res 116:133–140. Cheng Y, Deshmukh M, D’Costa A, Demaro JA, Gidday JM, Shah A, Sun Y, Jacquin MF, Johnson EM, Holtzman DM (1998): Caspase inhibitor affords neuroprotection with delayed administration in a rat model of neonatal hypoxic-ischemic brain injury. J Clin Invest 101:1992–1999. Gustafson K, Hagberg H, Bengtsson BA, Brantsing C, Isgaard J (1999): Possible protective role of growth hormone in hypoxia-ischemia in neonatal rats. Pediatr Res 45:318–323. Nighoghossian N, Trouillas P, Adeleine P, Salord F (1995): Hyperbaric oxygen in the treatment of acute ischemic stroke. A double-blind pilot study. Stroke 26:1369–1372. Viswanath M, Palmer C, Roberts RL (2000): Reduction of hypoxic-ischemic brain swelling in the neonatal rat with PAF antagonist WEB 2170: lack of long-term protection. Pediatr Res 48:109–113.

464

Dev Neurosci 2002;24:446–464

Periventricular leukomalacia (PVL) occurs following hypoxicischemic cerebral injury in the preterm human and involves excitotoxic cell death of developing preoligodendrocytes (Back SA et al., J Neurosci 2002;22:455–463). Observed clinical deficits include spastic diplegia, developmental delay and cortical visual impairment. We have characterized a rodent model of PVL to demonstrate that subplate neurons also die following hypoxia-ischemia at P1. To evaluate the mechanism of subplate sensitivity to hypoxia-ischemia, they were immunopurified and cultured in vitro (DeFreitas MF et al., J Neurosci 2001;21:5121–5129). Subplate neurons were compared to immunopurified cortical plate and hippocampal neurons. In agreement with observations in vivo, subplate neurons were twice as sensitive as hippocampal neurons to oxygen-glucose deprivation (OGD). In order to identify signaling pathways that account for the differential susceptibility to OGD, we have applied mRNA expression profiling to the purified neuronal populations using Affymetrix rat genome DNA microarrays. Differentially expressed genes will also serve as cell type specific markers for identifying subplate neurons in vivo. We have identified 46 genes, and 24 ESTs that are expressed at least 4-fold higher in subplate neurons than in the other neuron types. Most of these genes are completely absent in the other cell types. We have similarly identified genes that may serve as markers for cortical plate and hippocampal neurons. We are applying mRNA expression profiling to individual cultures of purified neurons using amplified mRNA before and after OGD to identify the cellular signaling pathways modulated by OGD in each cell type.

Abstracts

Author Index Vol. 24, No. 5, 2002 (A) = Abstract

Adcock, K.H. 382, 454 (A) Alt, J.J. 448 (A) Armstrong, E.A. 367 Arvin, K. 460 (A) Back, S.A. 462 (A) Bakker, J.M. 455 (A) Barkovich, A.J. 448 (A) Barks, J.D.E 418, 446 (A) Beabes, S. 454 (A) Bel, F. van 389, 396, 455 (A) Bhatia, I. 450 (A) Blomgren, K. 396 Bohland, M.A. 448 (A) Bregman, J.C. 456 (A) Brucklacher, R.M. 411 Burtscher, I. 456 (A) Buul-Offers, S. van 396 Calvert, J.W. 463 (A) Ceppi-Cozzio, C. 448 (A) Cheung, H.W.T. 462 (A) Cheung, P.T. 450 (A), 462 (A) Choi, J. 405 Chugani, H.T. 451 (A) Comi, A.M. 461 (A) Connor, J.R. 459 (A) Czikk, M.J. 451 (A) D’Argy, R. 456 (A) Dawson, T. 460 (A) Dawson, V. 460 (A) DeFreitas, M.F. 464 (A) Derrick, M. 456 (A) Derugin, N. 461 (A) Edwards, D. 352 Felt, B.T. 418, 446 (A) Feng, Y. 450 (A) Ferriero, D.M. 349, 448 (A), 461 (A), 464 (A) Fox, C.K. 461 (A) Galanopoulou, A.S. 355, 458 (A) Geijer, B. 456 (A) Gleason, C.A. 446 (A)

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Gluckman, P. 457 (A) Golden, J. 449 (A) Grafe, M.R. 458 (A) Groenendaal, F. 389, 396, 455 (A) Haaf, M.P. 459 (A) Hagberg, H. 364, 460 (A) Hamers, N. 396 Hammond, R. 453 (A) Hamrick, S.E.G. 464 (A) Han, B.H. 405, 460 (A) Hei, M. 450 (A) Heijnen, C.J. 389, 396, 455 (A) Hellström-Westas, I. 456 (A) Henry, R.G. 448 (A) Hernandez, E. 456 (A) Hoffman, C. 448 (A) Holtås, S. 456 (A) Holtzman, D.M. 405, 460 (A) Homan, J.H. 451 (A)

McCann, J. 449 (A) McCarthy, M.M. 448 (A) McQuillen, P.S. 464 (A) Malecki, E.A. 459 (A) Marsal, K. 456 (A) Martin, E. 382, 454 (A) Martin, S.S. 458 (A) Matsushita, H. 460 (A) Meyers, R. 452 (A) Miller, S.P. 448 (A) Milley, J.R. 451 (A) Mitchell, N.E. 437 Miyashita, H. 367 Moshé, S.L. 355, 458 (A)

Ichord, R. 364 Inder, T. 457 (A)

Namer, I.J. 447 (A) Nanda, A. 463 (A) Nedelcu, J. 382, 454 (A) Nehlig, A. 447 (A) Ness, J.K. 437, 452 (A) Newton, N. 448 (A) Nicolay, K. 455 (A) Northington, F. 460 (A) Nuñez, J.L. 448 (A)

Jilling, T. 456 (A) Johnston, M. 460 (A)

Ohata, H. 449 (A) Ohlsson, T. 456 (A)

Koehler, R.C. 449 (A) Koster, J. 396 Kremlev, S.G. 459 (A) Kurth, C.D. 449 (A)

Palmer, C. 364, 459 (A) Partridge, J.C. 448 (A) Pearce, W.J. 447 (A) Peeters-Scholte, C.M.P.C.D. 389, 396 Perez-Polo, J.R. 458 (A) Poitras, M. 460 (A) Priestley, M.A. 449 (A)

Lange, M. 460 (A) Law, H.K.W. 450 (A) Lazovic-Stojkovic, J. 452 (A) LeBlanc, M.H. 450 (A) Leroy, C. 447 (A) Leung, T.W.C. 462 (A) Levison, S.W. 347, 426, 452 (A), 459 (A) Ley, D. 456 (A) Li, X. 418 Lingman, G. 456 (A) Liu, W. 452 (A) Liu, Y. 418 Loenneker, T. 382, 454 (A) Loepke, A.W. 449 (A)

© 2002 S. Karger AG, Basel

Accessible online at: www.karger.com/dne

Richardson, B.S. 451 (A), 453 (A) Roberts, R.L. 459 (A) Roch, C. 447 (A) Rocha, E. 453 (A) Romanko, M.J. 452 (A), 459 (A) Rothstein, R.P. 426, 452 (A), 459 (A)

Schallert, T. 418, 446 (A) Schultz, S.E. 449 (A) Sham, M.H. 462 (A) Shao, J. 418 Sizonenko, S. 457 (A) Smith, M.B. 452 (A), 454 (A) Snyder, M.J. 459 (A) Stojkovic, D. 452 (A) Strand, S.-E. 456 (A) Sucher, N. 450 (A) Sweeley, J.C. 451 (A) Tan, S. 456 (A) Thorngren-Jerneck, K. 456 (A) Todorich, B. 459 (A) Totten, S. 453 (A) Traystman, R.J. 446 (A), 449 (A) Tweel, E.R.W. van den 389, 396, 455 (A) Vannucci, R.C. 347, 411 Vannucci, S.J. 347, 364, 411 Vaughan, J.T. 452 (A) Vexler, Z.S. 461 (A) Vidaurre, J. 355 Vigneron, D.B. 448 (A) Wagner, B.P. 382, 454 (A) Wallimann, T. 382 Wang, J. 454 (A) Wendland, M. 461 (A) Werner, O. 456 (A) Williams, C. 457 (A) Wilson, M.-A. 460 (A) Wirrell, E.C. 367 Wood, T.L. 437, 452 (A) Yager, J.Y. 364, 367 Yang, Q.X. 452 (A) Yip, H.K.F. 462 (A) Zhang, J.H. 463 (A) Zhu, C. 396

465

Subject Index Vol. 24, No. 5, 2002

Adult 355 Akt 437 Apoptosis 405, 426, 437

Hypoxia 352 Hypoxia-isch(a)emia 382, 389, 405, 411 –, perinatal 396 Hypoxic-ischemic brain damage 367

Brain 349, 382, 411 Caspase-3 405 Caspases 396 Cell death 349, 426 Cerebral energy metabolism 367 – hypoxia-ischemia 418 – ischemia 418 – palsy 426 Creatine 382 Cytokine(s) 352, 389, 396 Energy reserves 411 Epilepsy 355 Excitotoxicity 418 Febrile seizure 355 Glutamate 437 Glycogen 411 Heat shock protein 70 389 Hippocampus 355

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2002 S. Karger AG, Basel

Accessible online at: www.karger.com/dne

IGF-1 396 Immature brain 355 Immaturity 411 In situ nick end labelling 396 Inhibition 389 iNOS 389 Intrauterine infection 352 Ischemia 352 Mesial temporal sclerosis 355 N-Methyl-D-aspartate 418 Mitogen-activated protein kinase 405 MK-801 418 Necrosis 426 Neonatal brain injury 418 – encephalopathy 352 – rat 382 – – brain 389 – seizures 367 Neonate 349 Neurodegeneration 405

Neuroprotection 355, 382, 405, 411 Neurotrophin 437 nNOS 389 p38 405 Phosphatidylinositol 3-kinase 437 Pig 396 Plasticity 418 Rat 355, 382, 389 Reperfusion injury 396 Sensorimotor deficit 418 Stroke 418, 426 Subventricular zone 426 Supplementation 382 Therapy 349 Trk receptor 437