Langage et cerveau : vingt ans d'imagerie fonctionnelle [Language and the brain: 20 years of functional imagery]

La masse considérable de travaux publiés dans le domaine de la neuroimagerie fonctionnelle concernant les fonctions ou modalités du langage (compréhension et expression de la parole, lecture) ou les différents processus linguistiques qui les sous-tendent (sémantique, phonologie, syntaxe) permet de dégager de grandes tendances en termes de substrats anatomiques. Si les « fondamentaux » issus des origines aphasiologiques du domaine n'ont pas été bouleversés, certaines spécificités non explorées par l'approche lésionnelle sont identifiables. Les méta-analyses, en regroupant les résultats de la littérature, nous procurent aujourd'hui une vision globale des substrats cérébraux du langage. Cependant la variabilité inter-individuelle reste importante en raison de multiples facteurs dont certains sont mal identifiés ; cartographier exhaustivement les fonctions du langage à l'échelle individuelle reste une gageure. La quête des images du langage est sans doute aussi inachevable que celle de l'étude du langage lui-même.

Effect of chronic hypoglycaemia on glucose concentration and glycogen content in rat brain: A localized 13C NMR study.

While chronic hypoglycaemia has been reported to increase unidirectional glucose transport across the blood-brain barrier (BBB) and to increase GLUT1 expression at the endothelium, the effect on steady-state brain d-glucose and brain glycogen content is currently unknown. Brain glucose and glycogen concentrations were directly measured in vivo using localized 13C magnetic resonance spectroscopy (MRS) following 12-14 days of hypoglycaemia. Brain glucose content was significantly increased by 48%, which is consistent with an increase in the maximal glucose transport rate, Tmax, by 58% compared with the sham-treated animals. The localized 13C NMR measurements of brain glucose were directly validated by comparison with biochemically determined brain glucose content after rapid focused microwave fixation (1.4 s at 4 kW). Both in vivo MRS and biochemical measurements implied that brain glycogen content was not affected by chronic hypoglycaemia, consistent with brain glucose being a major factor controlling brain glycogen content. We conclude that the increased glucose transporter expression in chronic hypoglycaemia leads to increased brain glucose content at a given level of glycaemia. Such increased brain glucose concentrations can result in a lowered glycaemic threshold of counter-regulation observed in chronic hypoglycaemia.

Decoding sequence learning from single-trial intracranial EEG in humans.

We propose and validate a multivariate classification algorithm for characterizing changes in human intracranial electroencephalographic data (iEEG) after learning motor sequences. The algorithm is based on a Hidden Markov Model (HMM) that captures spatio-temporal properties of the iEEG at the level of single trials. Continuous intracranial iEEG was acquired during two sessions (one before and one after a night of sleep) in two patients with depth electrodes implanted in several brain areas. They performed a visuomotor sequence (serial reaction time task, SRTT) using the fingers of their non-dominant hand. Our results show that the decoding algorithm correctly classified single iEEG trials from the trained sequence as belonging to either the initial training phase (day 1, before sleep) or a later consolidated phase (day 2, after sleep), whereas it failed to do so for trials belonging to a control condition (pseudo-random sequence). Accurate single-trial classification was achieved by taking advantage of the distributed pattern of neural activity. However, across all the contacts the hippocampus contributed most significantly to the classification accuracy for both patients, and one fronto-striatal contact for one patient. Together, these human intracranial findings demonstrate that a multivariate decoding approach can detect learning-related changes at the level of single-trial iEEG. Because it allows an unbiased identification of brain sites contributing to a behavioral effect (or experimental condition) at the level of single subject, this approach could be usefully applied to assess the neural correlates of other complex cognitive functions in patients implanted with multiple electrodes.

Atlas-based segmentation of pathological brain MR images

We propose a method for brain atlas deformation inpresence of large space-occupying tumors, based on an apriori model of lesion growth that assumes radialexpansion of the lesion from its starting point. First,an affine registration brings the atlas and the patientinto global correspondence. Then, the seeding of asynthetic tumor into the brain atlas provides a templatefor the lesion. Finally, the seeded atlas is deformed,combining a method derived from optical flow principlesand a model of lesion growth (MLG). Results show that themethod can be applied to the automatic segmentation ofstructures and substructures in brains with grossdeformation, with important medical applications inneurosurgery, radiosurgery and radiotherapy.

Brain tissue properties differentiate between motor and limbic basal ganglia circuits.

Despite advances in understanding basic organizational principles of the human basal ganglia, accurate in vivo assessment of their anatomical properties is essential to improve early diagnosis in disorders with corticosubcortical pathology and optimize target planning in deep brain stimulation. Main goal of this study was the detailed topological characterization of limbic, associative, and motor subdivisions of the subthalamic nucleus (STN) in relation to corresponding corticosubcortical circuits. To this aim, we used magnetic resonance imaging and investigated independently anatomical connectivity via white matter tracts next to brain tissue properties. On the basis of probabilistic diffusion tractography we identified STN subregions with predominantly motor, associative, and limbic connectivity. We then computed for each of the nonoverlapping STN subregions the covariance between local brain tissue properties and the rest of the brain using high-resolution maps of magnetization transfer (MT) saturation and longitudinal (R1) and transverse relaxation rate (R2*). The demonstrated spatial distribution pattern of covariance between brain tissue properties linked to myelin (R1 and MT) and iron (R2*) content clearly segregates between motor and limbic basal ganglia circuits. We interpret the demonstrated covariance pattern as evidence for shared tissue properties within a functional circuit, which is closely linked to its function. Our findings open new possibilities for investigation of changes in the established covariance pattern aiming at accurate diagnosis of basal ganglia disorders and prediction of treatment outcome.

Cellular and molecular mechanisms underlying the effects of sleep loss on hippocampal synaptic plasticity

SUMMARY : The function of sleep for the organism is one of the most persistent and perplexing questions in biology. Current findings lead to the conclusion that sleep is primarily for the brain. In particular, a role for sleep in cognitive aspects of brain function is supported by behavioral evidence both in humans and animals. However, in spite of remarkable advancement in the understanding of the mechanisms underlying sleep generation and regulation, it has been proven difficult to determine the neurobiological mechanisms underlying the beneficial effect of sleep, and the detrimental impact of sleep loss, on learning and memory processes. In my thesis, I present results that lead to several critical steps forward in the link between sleep and cognitive function. My major result is the molecular identification and physiological analysis of a protein, the NR2A subunit of NMDA receptor (NMDAR), that confers sensitivity to sleep loss to the hippocampus, a brain structure classically involved in mnemonic processes. Specifically, I used a novel behavioral approach to achieve sleep deprivation in adult C57BL6/J mice, yet minimizing the impact of secondary factors associated with the procedure,.such as stress. By using in vitro electrophysiological analysis, I show, for the first time, that sleep loss dramatically affects bidirectional plasticity at CA3 to CA1 synapses in the hippocampus, a well established cellular model of learning and memory. 4-6 hours of sleep loss elevate the modification threshold for bidirectional synaptic plasticity (MT), thereby promoting long-term depression of CA3 to CA 1 synaptic strength after stimulation in the theta frequency range (5 Hz), and rendering long-term potentiation induction.more difficult. Remarkably, 3 hours of recovery sleep, after the deprivation, reset the MT at control values, thus re-establishing the normal proneness of synapses to undergo long-term plastic changes. At the molecular level, these functional changes are paralleled by a change in the NMDAR subunit composition. In particular, the expression of the NR2A subunit protein of NMDAR at CA3 to CA1 synapses is selectively and rapidly increased by sleep deprivation, whereas recovery sleep reset NR2A synaptic content to control levels. By using an array of genetic, pharmacological and computational approaches, I demonstrate here an obligatory role for NR2A-containing NMDARs in conveying the effect of sleep loss on CA3 to CAl MT. Moreover, I show that a genetic deletion of the NR2A subunit fully preserves hippocampal plasticity from the impact of sleep loss, whereas it does not alter sleepwake behavior and homeostatic response to sleep deprivation. As to the mechanism underlying the effects of the NR2A subunit on hippocampal synaptic plasticity, I show that the increased NR2A expression after sleep loss distinctly affects the contribution of synaptic and more slowly recruited NMDAR pools activated during plasticity-induction protocols. This study represents a major step forward in understanding the mechanistic basis underlying sleep's role for the brain. By showing that sleep and sleep loss affect neuronal plasticity by regulating the expression and function of a synaptic neurotransmitter receptor, I propose that an important aspect of sleep function could consist in maintaining and regulating protein redistribution and ion channel trafficking at central synapses. These findings provide a novel starting point for investigations into the connections between sleep and learning, and they may open novel ways for pharmacological control over hippocampal .function during periods of sleep restriction. RÉSUMÉ DU PROJET La fonction du sommeil pour l'organisme est une des questions les plus persistantes et difficiles dans la biologie. Les découvertes actuelles mènent à la conclusion que le sommeil est essentiel pour le cerveau. En particulier, le rôle du sommeil dans les aspects cognitifs est soutenu par des études comportementales tant chez les humains que chez les animaux. Cependant, malgré l'avancement remarquable dans la compréhension des mécanismes sous-tendant la génération et la régulation du sommeil, les mécanismes neurobiologiques qui pourraient expliquer l'effet favorable du sommeil sur l'apprentissage et la mémoire ne sont pas encore clairs. Dans ma thèse, je présente des résultats qui aident à clarifier le lien entre le sommeil et la fonction cognitive. Mon résultat le plus significatif est l'identification moléculaire et l'analyse physiologique d'une protéine, la sous-unité NR2A du récepteur NMDA, qui rend l'hippocampe sensible à la perte de sommeil. Dans cette étude, nous avons utilisé une nouvelle approche expérimentale qui nous a permis d'induire une privation de sommeil chez les souris C57BL6/J adultes, en minimisant l'impact de facteurs confondants comme, par exemple, le stress. En utilisant les techniques de l'électrophysiologie in vitro, j'ai démontré, pour la première fois, que la perte de sommeil est responsable d'affecter radicalement la plasticité bidirectionnelle au niveau des synapses CA3-CA1 de l'hippocampe. Cela correspond à un mécanisme cellulaire de l'apprentissage et de la mémoire bien établi. En particulier, 4-6 heures de privation de sommeil élèvent le seuil de modification pour la plasticité synaptique bidirectionnelle (SM). Comme conséquence, la dépression à long terme de la transmission synaptique est induite par la stimulation des fibres afférentes dans la bande de fréquences thêta (5 Hz), alors que la potentialisation à long terme devient plus difficile. D'autre part, 3 heures de sommeil de récupération sont suffisant pour rétablir le SM aux valeurs contrôles. Au niveau moléculaire, les changements de la plasticité synaptiques sont associés à une altération de la composition du récepteur NMDA. En particulier, l'expression synaptique de la protéine NR2A du récepteur NMDA est rapidement augmentée de manière sélective par la privation de sommeil, alors que le sommeil de récupération rétablit l'expression de la protéine au niveau contrôle. En utilisant des approches génétiques, pharmacologiques et computationnelles, j'ai démontré que les récepteurs NMDA qui expriment la sous-unité NR2A sont responsables de l'effet de la privation de sommeil sur le SM. De plus, nous avons prouvé qu'une délétion génétique de la sous-unité NR2A préserve complètement la plasticité synaptique hippocampale de l'impact de la perte de sommeil, alors que cette manipulation ne change pas les mécanismes de régulation homéostatique du sommeil. En ce qui concerne les mécanismes, j'ai .découvert que l'augmentation de l'expression de la sous-unité NR2A au niveau synaptique modifie les propriétés de la réponse du récepteur NMDA aux protocoles de stimulations utilisés pour induire la plasticité. Cette étude représente un pas en avant important dans la compréhension de la base mécaniste sous-tendant le rôle du sommeil pour le cerveau. En montrant que le sommeil et la perte de sommeil affectent la plasticité neuronale en régulant l'expression et la fonction d'un récepteur de la neurotransmission, je propose qu'un aspect important de la fonction du sommeil puisse être finalisé au règlement de la redistribution des protéines et du tracking des récepteurs aux synapses centraux. Ces découvertes fournissent un point de départ pour mieux comprendre les liens entre le sommeil et l'apprentissage, et d'ailleurs, ils peuvent ouvrir des voies pour des traitements pharmacologiques dans le .but de préserver la fonction hippocampale pendant les périodes de restriction de sommeil.

Canine distemper virus infection of primary hippocampal cells induces increase in extracellular glutamate and neurodegeneration.

The canine distemper virus (CDV) belongs to the Morbillivirus genus which includes important human pathogens like the closely related measles virus. CDV infection can reach the nervous system where it causes serious malfunctions. Although this pathology is well described, the molecular events in brain infection are still poorly understood. Here we studied infection in vitro by CDV using a model of dissociated cell cultures from newborn rat hippocampus. We used a recombinant CDV closely related to the neurovirulent A75/17 which also expresses the enhanced green fluorescent protein. We found that infected neurons and astrocytes could be clearly detected, and that infection spreads only slowly to neighboring cells. Interestingly, this infection causes a massive cell death of neurons, which includes also non-infected neurons. Antagonists of NMDA-type or alpha-amino-3-hydroxy-5-methylisoxazole-4-propinate (AMPA)-type glutamate receptors could slow down this neuron loss, indicating an involvement of the glutamatergic system in the induction of cell death in infected and non-infected cells. Finally, we show that, following CDV infection, there is a steady increase in extracellular glutamate in infected cultures. These results indicate that CDV infection induces excitotoxic insults on neurons via glutamatergic signaling.

Biochemical, morphological and content characterization of synaptobrevin 2-expressing vesicles in astrocytes

Neuron-astrocyte reciprocal communication at synapses has emerged as a novel signalling pathway in brain function. Astrocytes sense the level of synaptic activity and, in turn, influence its efficacy through the regulated release of ''glio- transmitters'' such as glutamate, ATP or D-serine. A calcium- dependent exocytosis is proposed to drive the release of gliotransmitters but its existence is still debated. To shed light onto the mechanisms controlling the storage and the release of gliotransmitters and namely D-serine, we have developed a new method for the immunoisolation of synaptobrevin 2-positive vesicles from rat cortical astrocytes in culture. The purified organelles are clear round shape vesicles of excellent purity as judged by electron microscopy. Immunoblotting analysis revealed that isolated vesicles contain most of the major proteins already described for neuron-derived vesicles. In addition, we have analyzed the content for various amino acids of these vesicles by means of chiral capillary electro- phoresis coupled to laser-induced fluorescence detection and liquid chromatography coupled to mass spectrometry. Post- embedding immunogold labelling of the rat neocortex and hippocampus further revealed the expression of D-serine and glutamate in astrocyte processes contacting excitatory sy- napses. Our results provide significant support for the existence of secretory glial vesicles storing chemical substances like D- serine and glutamate and thus point to the co-release of amino acids by exocytosis in astrocytes.

Correlation of diffusion tensor tractography and intraoperative macrostimulation during deep brain stimulation for Parkinson disease.

Object The purpose of this study was to investigate whether diffusion tensor imaging (DTI) of the corticospinal tract (CST) is a reliable surrogate for intraoperative macrostimulation through the deep brain stimulation (DBS) leads. The authors hypothesized that the distance on MRI from the DBS lead to the CST as determined by DTI would correlate with intraoperative motor thresholds from macrostimulations through the same DBS lead. Methods The authors retrospectively reviewed pre- and postoperative MRI studies and intraoperative macrostimulation recordings in 17 patients with Parkinson disease (PD) treated by DBS stimulation. Preoperative DTI tractography of the CST was coregistered with postoperative MRI studies showing the position of the DBS leads. The shortest distance and the angle from each contact of each DBS lead to the CST was automatically calculated using software-based analysis. The distance measurements calculated for each contact were evaluated with respect to the intraoperative voltage thresholds that elicited a motor response at each contact. Results There was a nonsignificant trend for voltage thresholds to increase when the distances between the DBS leads and the CST increased. There was a significant correlation between the angle and the voltage, but the correlation was weak (coefficient of correlation [R] = 0.36). Conclusions Caution needs to be exercised when using DTI tractography information to guide DBS lead placement in patients with PD. Further studies are needed to compare DTI tractography measurements with other approaches such as microelectrode recordings and conventional intraoperative MRI-guided placement of DBS leads.

Peroxisome proliferator-activated receptor beta regulates acyl-CoA synthetase 2 in reaggregated rat brain cell cultures.

Peroxisome proliferator-activated receptors (PPARs) are nuclear hormone receptors that regulate the expression of many genes involved in lipid metabolism. The biological roles of PPARalpha and PPARgamma are relatively well understood, but little is known about the function of PPARbeta. To address this question, and because PPARbeta is expressed to a high level in the developing brain, we used reaggregated brain cell cultures prepared from dissociated fetal rat telencephalon as experimental model. In these primary cultures, the fetal cells initially form random aggregates, which progressively acquire a tissue-specific pattern resembling that of the brain. PPARs are differentially expressed in these aggregates, with PPARbeta being the prevalent isotype. PPARalpha is present at a very low level, and PPARgamma is absent. Cell type-specific expression analyses revealed that PPARbeta is ubiquitous and most abundant in some neurons, whereas PPARalpha is predominantly astrocytic. We chose acyl-CoA synthetases (ACSs) 1, 2, and 3 as potential target genes of PPARbeta and first analyzed their temporal and cell type-specific pattern. This analysis indicated that ACS2 and PPARbeta mRNAs have overlapping expression patterns, thus designating the ACS2 gene as a putative target of PPARbeta. Using a selective PPARbeta activator, we found that the ACS2 gene is transcriptionally regulated by PPARbeta, demonstrating a role for PPARbeta in brain lipid metabolism.

Multiple interacting cell death mechanisms in the mediation of excitotoxicity and ischemic brain damage: A challenge for neuroprotection.

There is currently no approved neuroprotective pharmacotherapy for acute conditions such as stroke and cerebral asphyxia. One of the reasons for this may be the multiplicity of cell death mechanisms, because inhibition of a particular mechanism leaves the brain vulnerable to alternative ones. It is therefore essential to understand the different cell death mechanisms and their interactions. We here review the multiple signaling pathways underlying each of the three main morphological types of cell death - apoptosis, autophagic cell death and necrosis - emphasizing their importance in the neuronal death that occurs during cerebral ischemia and hypoxia-ischemia, and we analyze the interactions between the different mechanisms. Finally, we discuss the implications of the multiplicity of cell death mechanisms for the design of neuroprotective strategies.

Intracisternal delivery of NFκB-inducible scAAV2/9 reveals locoregional neuroinflammation induced by systemic kainic acid treatment.

We have previously demonstrated disease-dependent gene delivery in the brain using an AAV vector responding to NFκB activation as a probe for inflammatory responses. This vector, injected focally in the parenchyma prior to a systemic kainic acid (KA) injection mediated inducible transgene expression in the hippocampus but not in the cerebellum, regions, respectively, known to be affected or not by the pathology. However, such a focal approach relies on previous knowledge of the model parameters and does not allow to predict the whole brain response to the disease. Global brain gene delivery would allow to predict the regional distribution of the pathology as well as to deliver therapeutic factors in all affected brain regions. We show that self-complementary AAV2/9 (scAAV2/9) delivery in the adult rat cisterna magna allows a widespread but not homogenous transduction of the brain. Indeed, superficial regions, i.e., cortex, hippocampus, and cerebellum were more efficiently transduced than deeper regions, such as striatum, and substantia nigra. These data suggest that viral particles penetration from the cerebrospinal fluid (CSF) into the brain is a limiting factor. Interestingly, AAV2/9-2YF a rationally designed capsid mutant (affecting surface tyrosines) increased gene transfer efficiency approximately fivefold. Neurons, astrocytes, and oligodendrocytes, but not microglia, were transduced in varying proportions depending on the brain region and the type of capsid. Finally, after a single intracisternal injection of scAAV2/9-2YF using the NFκB-inducible promoter, KA treatment induced transgene expression in the hippocampus and cortex but not in the cerebellum, corresponding to the expression of the CD11b marker of microglial activation. These data support the use of disease-inducible vectors administered in the cisterna magna as a tool to characterize the brain pathology in systemic drug-induced or transgenic disease models. However, further improvements are required to enhance viral particles penetration into the brain.

Visual functions in the healthy and schizophrenic brain : from stimulus control to illusory perceptions using electrical neuroimaging

ABSTRACT (FRENCH)Ce travail de thèse basé sur le système visuel chez les sujets sains et chez les patients schizophrènes, s'articule autour de trois articles scientifiques publiés ou en cours de publication. Ces articles traitent des sujets suivants : le premier article présente une nouvelle méthode de traitement des composantes physiques des stimuli (luminance et fréquence spatiale). Le second article montre, à l'aide d'analyses de données EEG, un déficit de la voie magnocellulaire dans le traitement visuel des illusions chez les patients schizophrènes. Ceci est démontré par l'absence de modulation de la composante PI chez les patients schizophrènes contrairement aux sujets sains. Cette absence est induite par des stimuli de type illusion Kanizsa de différentes excentricités. Finalement, le troisième article, également à l'aide de méthodes de neuroimagerie électrique (EEG), montre que le traitement des contours illusoires se trouve dans le complexe latéro-occipital (LOC), à l'aide d'illusion « misaligned gratings ». De plus il révèle que les activités démontrées précédemment dans les aires visuelles primaires sont dues à des inférences « top- down ».Afin de permettre la compréhension de ces trois articles, l'introduction de ce manuscrit présente les concepts essentiels. De plus des méthodes d'analyses de temps-fréquence sont présentées. L'introduction est divisée en quatre parties : la première présente le système visuel depuis les cellules retino-corticales aux deux voix du traitement de l'information en passant par les régions composant le système visuel. La deuxième partie présente la schizophrénie par son diagnostic, ces déficits de bas niveau de traitement des stimuli visuel et ces déficits cognitifs. La troisième partie présente le traitement des contours illusoires et les trois modèles utilisés dans le dernier article. Finalement, les méthodes de traitement des données EEG seront explicitées, y compris les méthodes de temps-fréquences.Les résultats des trois articles sont présentés dans le chapitre éponyme (du même nom). De plus ce chapitre comprendra les résultats obtenus à l'aide des méthodes de temps-fréquenceFinalement, la discussion sera orientée selon trois axes : les méthodes de temps-fréquence ainsi qu'une proposition de traitement de ces données par une méthode statistique indépendante de la référence. La discussion du premier article en montrera la qualité du traitement de ces stimuli. La discussion des deux articles neurophysiologiques, proposera de nouvelles d'expériences afin d'affiner les résultats actuels sur les déficits des schizophrènes. Ceci pourrait permettre d'établir un marqueur biologique fiable de la schizophrénie.ABSTRACT (ENGLISH)This thesis focuses on the visual system in healthy subjects and schizophrenic patients. To address this research, advanced methods of analysis of electroencephalographic (EEG) data were used and developed. This manuscript is comprised of three scientific articles. The first article showed a novel method to control the physical features of visual stimuli (luminance and spatial frequencies). The second article showed, using electrical neuroimaging of EEG, a deficit in spatial processing associated with the dorsal pathway in chronic schizophrenic patients. This deficit was elicited by an absent modulation of the PI component in terms of response strength and topography as well as source estimations. This deficit was orthogonal to the preserved ability to process Kanizsa-type illusory contours. Finally, the third article resolved ongoing debates concerning the neural mechanism mediating illusory contour sensitivity by using electrical neuroimaging to show that the first differentiation of illusory contour presence vs. absence is localized within the lateral occipital complex. This effect was subsequent to modulations due to the orientation of misaligned grating stimuli. Collectively, these results support a model where effects in V1/V2 are mediated by "top-down" modulation from the LOC.To understand these three articles, the Introduction of this thesis presents the major concepts used in these articles. Additionally, a section is devoted to time-frequency analysis methods not presented in the articles themselves. The introduction is divided in four parts. The first part presents three aspects of the visual system: cellular, regional, and its functional interactions. The second part presents an overview of schizophrenia and its sensoiy-cognitive deficits. The third part presents an overview of illusory contour processing and the three models examined in the third article. Finally, advanced analysis methods for EEG are presented, including time- frequency methodology.The Introduction is followed by a synopsis of the main results in the articles as well as those obtained from the time-frequency analyses.Finally, the Discussion chapter is divided along three axes. The first axis discusses the time frequency analysis and proposes a novel statistical approach that is independent of the reference. The second axis contextualizes the first article and discusses the quality of the stimulus control and direction for further improvements. Finally, both neurophysiologic articles are contextualized by proposing future experiments and hypotheses that may serve to improve our understanding of schizophrenia on the one hand and visual functions more generally.

Epidermal growth factor and bovine growth hormone stimulate differentiation and myelination of brain cell aggregates in culture.

Bovine growth hormone (bGH) and epidermal growth factor (EGF) increased the activity of ornithine decarboxylase (ODC) in brain cell aggregates cultured in a serum-free chemically defined medium. ODC is considered as a marker of cell growth and differentiation. The effect of bGH and EGF on myelination was investigated by measuring two myelin markers, 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNP) and myelin basic protein (MBP). EGF treatment at days 2 and 5 caused a dose-dependent increase of both myelin markers at culture day 12. This increase could still be observed at culture day 19, indicating a prolonged action of EGF. The continual presence of bGH in the culture medium produced a large accumulation of MBP at day 19. This effect was dose-dependent and required the presence of triiodothyronine (T3). In contrast, the effect of bGH on CNP activity did not require the presence of T3. This is the first report showing a direct effect of bGH on CNS myelination in vitro and of EGF on both MBP accumulation and ODC activity.