Analysis, modelling and classification of geospatial data using machine learning

The research considers the problem of spatial data classification using machine learning algorithms: probabilistic neural networks (PNN) and support vector machines (SVM). As a benchmark model simple k-nearest neighbor algorithm is considered. PNN is a neural network reformulation of well known nonparametric principles of probability density modeling using kernel density estimator and Bayesian optimal or maximum a posteriori decision rules. PNN is well suited to problems where not only predictions but also quantification of accuracy and integration of prior information are necessary. An important property of PNN is that they can be easily used in decision support systems dealing with problems of automatic classification. Support vector machine is an implementation of the principles of statistical learning theory for the classification tasks. Recently they were successfully applied for different environmental topics: classification of soil types and hydro-geological units, optimization of monitoring networks, susceptibility mapping of natural hazards. In the present paper both simulated and real data case studies (low and high dimensional) are considered. The main attention is paid to the detection and learning of spatial patterns by the algorithms applied.

Phosphorylation of neurofilament subunit NF-M is regulated by activation of NMDA receptors and modulates cytoskeleton stability and neuronal shape.

The cytoskeleton is essential for the structural organization of neurons and is influenced during development by excitatory stimuli such as activation of glutamate receptors. In particular, NMDA receptors are known to modulate the function of several cytoskeletal proteins and to influence cell morphology, but the underlying molecular and cellular mechanisms remain unclear. Here, we characterized the neurofilament subunit NF-M in cultures of developing mouse cortical neurons chronically exposed to NMDA receptor antagonists. Western blots analysis showed that treatment of cortical neurons with MK801 or AP5 shifted the size of NF-M towards higher molecular weights. Dephosphorylation assay revealed that this increased size of NF-M observed after chronic exposure to NMDA receptor antagonists was due to phosphorylation. Neurons treated with cyclosporin, an inhibitor of the Ca(2+)-dependent phosphatase calcineurin, also showed increased levels of phosphorylated NF-M. Moreover, analysis of neurofilament stability revealed that the phosphorylation of NF-M, resulting from NMDA receptor inhibition, enhanced the solubility of NF-M. Finally, cortical neurons cultured in the presence of the NMDA receptor antagonists MK801 and AP5 grew longer neurites. Together, these data indicate that a blockade of NMDA receptors during development of cortical neurons increases the phosphorylation state and the solubility of NF-M, thereby favoring neurite outgrowth. This also underlines that dynamics of the neurofilament and microtubule cytoskeleton is fundamental for growth processes.

Nuclear and cytoplasmic triiodothyronine-binding sites in primary sensory neurons and Schwann cells: radioautographic study during development.

The effects of the thyroid hormones on target cells are mediated through nuclear T3 receptors. In the peripheral nervous system, nuclear T3 receptors were previously detected with the monoclonal antibody 2B3 mAb in all the primary sensory neurons throughout neuronal life and in peripheral glia at the perinatal period only (Eur. J. Neurosci. 5, 319, 1993). To determine whether these nuclear T3 receptors correspond to functional ones able to bind T3, cryostat sections and in vitro cell cultures of dorsal root ganglion (DRG) or sciatic nerve were incubated with 0.1 nM [125I]-labeled T3, either alone to visualize the total T3-binding sites or added with a 10(3) fold excess of unlabeled T3 to estimate the part due to the non-specific T3-binding. After glutaraldehyde fixation, radioautography showed that the specific T3-binding sites were largely prevalent. The T3-binding capacity of peripheral glia in DRG and sciatic nerve was restricted to the perinatal period in vivo and to Schwann cells cultured in vitro. In all the primary sensory neurons, specific T3-binding sites were disclosed in foetal as well as adult rats. The detection of the T3-binding sites in the nucleus indicated that the nuclear T3 receptors are functional. Moreover the concomitant presence of both T3-binding sites and T3 receptors alpha isoforms in the perikaryon of DRG neurons infers that: 1) [125I]-labeled T3 can be retained on the T3-binding 'E' domain of nascent alpha 1 isoform molecules newly-synthesized on the perikaryal ribosomes; 2) the alpha isoforms translocated to the nucleus are modified by posttranslational changes and finally recognized by 2B3 mAb as nuclear T3 receptor. In conclusion, the radioautographic visualization of the T3-binding sites in peripheral neurons and glia confirms that the nuclear T3 receptors are functional and contributes to clarify the discordant intracellular localization provided by the immunocytochemical detection of nuclear T3 receptors and T3 receptor alpha isoforms.

Comprehensive spatiotemporal transcriptomic analyses of the ganglionic eminences demonstrate the uniqueness of its caudal subdivision.

The elucidation of mechanisms underlying telencephalic neural development has been limited by the lack of knowledge regarding the molecular and cellular aspects of the ganglionic eminence (GE), an embryonic structure that supplies the brain with diverse sets of GABAergic neurons. Here, we report a comprehensive transcriptomic analysis of this structure including its medial (MGE), lateral (LGE) and caudal (CGE) subdivisions and its temporal dynamics in 12.5 to 16 day-old rat embryos. Surprisingly, comparison across subdivisions showed that CGE gene expression was the most unique providing unbiased genetic evidence for its differentiation from MGE and LGE. The molecular signature of the CGE comprised a large set of genes, including Rwdd3, Cyp26b1, Nr2f2, Egr3, Cpta1, Slit3, and Hod, of which several encode cell signaling and migration molecules such as WNT5A, DOCK9, VSNL1 and PRG1. Temporal analysis of the MGE revealed differential expression of unique sets of cell specification and migration genes, with early expression of Hes1, Lhx2, Ctgf and Mdk, and late enrichment of Olfm3, SerpinE2 and Wdr44. These GE profiles reveal new candidate regulators of spatiotemporally governed GABAergic neuronogenesis.

ROLES OF ACID-SENSING ION CHANNELS (ASICS) IN PERIPHERAL NEUROPEPTIDE SECRETION AND PAIN

Acid-sensing ion channels (ASICs) are non-voltage-gated sodium channels activated by an extracellular acidification. They are widely expressed in neurons of the central and peripheral nervous system. ASICs have a role in learning, the expression of fear, in neuronal death after cerebral ischemia, and in pain sensation. Tissue damage leads to the release of inflammatory mediators. There is a subpopulation of sensory neurons which are able to release the neuropeptides calcitonin gene-related peptide (CGRP) and substance P (SP). Neurogenic inflammation refers to the process whereby peripheral release of the neuropeptides CGRP and SP induces vasodilation and extravasation of plasma proteins, respectively. Our laboratory has previously shown that calcium-permeable homomeric ASIC1a channels are present in a majority of CGRP- or SP-expressing small diameter sensory neurons. In the first part of my thesis, we tested the hypothesis that a local acidification can produce an ASIC-mediated calcium-dependant neuropeptide secretion. We have first verified the co-expression of ASICs and CGRP/SP using immunochemistry and in-situ hybridization on dissociated rat dorsal root ganglion (DRG) neurons. We found that most CGRP/SP-positive neurons also expressed ASIC1a and ASIC3 subunits. Calcium imaging experiments with Fura-2 dye showed that an extracellular acidification can induce an increase of intracellular Ca2+ concentration, which is essential for secretion. This increase of intracellular Ca2+ concentration is, at least in some cells, ASIC-dependent, as it can be prevented by amiloride, an ASIC antagonist, and by Psalmotoxin (PcTx1), a specific ASIC1a antagonist. We identified a sub-population of neurons whose acid-induced Ca2+ entry was completely abolished by amiloride, an amiloride-resistant population which does not express ASICs, but rather another acid-sensing channel, possibly transient receptor potential vanilloïde 1 (TRPV1), and a population expressing both H+-gated channel types. Voltage-gated calcium channels (Cavs) may also mediate Ca2+ entry. Co-application of the Cavs inhibitors (ω-conotoxin MVIIC, Mibefradil and Nifedipine) reduced the Ca2+ increase in neurons expressing ASICs during an acidification to pH 6. This indicates that ASICs can depolarise the neuron and activate Cavs. Homomeric ASIC1a are Ca2+-permeable and allow a direct entry of Ca2+ into the cell; other ASICs mediate an indirect entry of Ca2+ by inducing a membrane depolarisation that activates Cavs. We showed with a secretion assay that CGRP secretion can be induced by extracellular acidification in cultured rat DRG neurons. Amiloride and PcTx1 were not able to inhibit the secretion at acidic pH, but BCTC, a TRPV1 inhibitor was able to decrease the secretion induced by an extracellular acidification in our in vitro secretion assay. In conclusion, these results show that in DRG neurons a mild extracellular acidification can induce a calcium-dependent neuropeptide secretion. Even if our data show that ASICs can mediate an increase of intracellular Ca2+ concentration, this appears not to be sufficient to trigger neuropeptide secretion. TRPV1, a calcium channel whose activation induces a sustained current - in contrary of ASICs - played in our experimental conditions a predominant role in neurosecretion. In the second part of my thesis, we focused on the role of ASICs in neuropathic pain. We used the spared nerve injury (SNI) model which consists in a nerve injury that induces symptoms of neuropathic pain such as mechanical allodynia. We have previously shown that the SNI model modifies ASIC currents in dissociated rat DRG neurons. We hypothesized that ASICs could play a role in the development of mechanical allodynia. The SNI model was performed on ASIC1a, -2, and -3 knock-out mice and wild type littermates. We measured mechanical allodynia on these mice with calibrated von Frey filaments. There were no differences between the wild-type and the ASIC1, or ASIC2 knockout mice. ASIC3 null mice were less sensitive than wild type mice at 21 day after SNI, indicating a role for ASIC3. Finally, to investigate other possible roles of ASICs in the perception of the environment, we measured the baseline heat responses. We used two different models; the tail flick model and the hot plate model. ASIC1a null mice showed increased thermal allodynia behaviour in the hot plate test at three different temperatures (49, 52, 55°C) compared to their wild type littermates. On the contrary, ASIC2 null mice showed reduced thermal allodynia behaviour in the hot plate test compared to their wild type littermates at the three same temperatures. We conclude that ASIC1a and ASIC2 in mice can play a role in temperature sensing. It is currently not understood how ASICs are involved in temperature sensing and what the reason for the opposed effects in the two knockout models is.

Role of salt-inducible kinase 1 in the activation of MEF2-dependent transcription by BDNF.

Substantial evidence supports a role for myocyte enhancer factor 2 (MEF2)-mediated transcription in neuronal survival, differentiation and synaptic function. In developing neurons, it has been shown that MEF2-dependent transcription is regulated by neurotrophins. Despite these observations, little is known about the cellular mechanisms by which neurotrophins activate MEF2 transcriptional activity. In this study, we examined the role of salt-inducible kinase 1 (SIK1), a member of the AMP-activated protein kinase (AMPK) family, in the regulation of MEF2-mediated transcription by the neurotrophin brain-derived neurotrophic factor (BDNF). We show that BDNF increases the expression of SIK1 in primary cultures of rat cortical neurons through the extracellular signal-regulated kinase 1/2 (ERK1/2)-signaling pathway. In addition to inducing SIK1 expression, BDNF triggers the phosphorylation of SIK1 at Thr182 and its translocation from the cytoplasm to the nucleus of cortical neurons. The effects of BDNF on the expression, phosphorylation and, translocation of SIK1 are followed by the phosphorylation and nuclear export of histone deacetylase 5 (HDAC5). Blockade of SIK activity with a low concentration of staurosporine abolished BDNF-induced phosphorylation and nuclear export of HDAC5 in cortical neurons. Importantly, stimulation of HDAC5 phosphorylation and nuclear export by BDNF is accompanied by the activation of MEF2-mediated transcription, an effect that is suppressed by staurosporine. Consistent with these data, BDNF induces the expression of the MEF2 target genes Arc and Nur77, in a staurosporine-sensitive manner. In further support of the role of SIK1 in the regulation of MEF2-dependent transcription by BDNF, we found that expression of wild-type SIK1 or S577A SIK1, a mutated form of SIK1 which is retained in the nucleus of transfected cells, is sufficient to enhance MEF2 transcriptional activity in cortical neurons. Together, these data identify a previously unrecognized mechanism by which SIK1 mediates the activation of MEF2-dependent transcription by BDNF.

Stable alpha-synuclein oligomers strongly inhibit chaperone activity of the Hsp70 system by weak interactions with J-domain co-chaperones.

α-Synuclein aggregation and accumulation in Lewy bodies are implicated in progressive loss of dopaminergic neurons in Parkinson disease and related disorders. In neurons, the Hsp70s and their Hsp40-like J-domain co-chaperones are the only known components of chaperone network that can use ATP to convert cytotoxic protein aggregates into harmless natively refolded polypeptides. Here we developed a protocol for preparing a homogeneous population of highly stable β-sheet enriched toroid-shaped α-Syn oligomers with a diameter typical of toxic pore-forming oligomers. These oligomers were partially resistant to in vitro unfolding by the bacterial Hsp70 chaperone system (DnaK, DnaJ, GrpE). Moreover, both bacterial and human Hsp70/Hsp40 unfolding/refolding activities of model chaperone substrates were strongly inhibited by the oligomers but, remarkably, not by unstructured α-Syn monomers even in large excess. The oligomers acted as a specific competitive inhibitor of the J-domain co-chaperones, indicating that J-domain co-chaperones may preferably bind to exposed bulky misfolded structures in misfolded proteins and, thus, complement Hsp70s that bind to extended segments. Together, our findings suggest that inhibition of the Hsp70/Hsp40 chaperone system by α-Syn oligomers may contribute to the disruption of protein homeostasis in dopaminergic neurons, leading to apoptosis and tissue loss in Parkinson disease and related neurodegenerative diseases.

Regional and cellular patterns of reelin mRNA expression in the forebrain of the developing and adult mouse

The reelin gene encodes an extracellular protein that is crucial for neuronal migration in laminated brain regions. To gain insights into the functions of Reelin, we performed high-resolution in situ hybridization analyses to determine the pattern of reelin expression in the developing forebrain of the mouse. We also performed double-labeling studies with several markers, including calcium-binding proteins, GAD65/67, and neuropeptides, to characterize the neuronal subsets that express reelin transcripts. reelinexpression was detected at embryonic day 10 and later in the forebrain, with a distribution that is consistent with the prosomeric model of forebrain regionalization. In the diencephalon, expression was restricted to transverse and longitudinal domains that delineated boundaries between neuromeres. During embryogenesis,reelin was detected in the cerebral cortex in Cajal-Retzius cells but not in the GABAergic neurons of layer I. At prenatal stages, reelin was also expressed in the olfactory bulb, and striatum and in restricted nuclei in the ventral telencephalon, hypothalamus, thalamus, and pretectum. At postnatal stages, reelin transcripts gradually disappeared from Cajal-Retzius cells, at the same time as they appeared in subsets of GABAergic neurons distributed throughout neocortical and hippocampal layers. In other telencephalic and diencephalic regions,reelin expression decreased steadily during the postnatal period. In the adult, there was prominent expression in the olfactory bulb and cerebral cortex, where it was restricted to subsets of GABAergic interneurons that co-expressed calbindin, calretinin, neuropeptide Y, and somatostatin. This complex pattern of cellular and regional expression is consistent with Reelin having multiple roles in brain development and adult brain function.

Alpha1b-adrenergic receptors control locomotor and rewarding effects of psychostimulants and opiates.

Drugs of abuse, such as psychostimulants and opiates, are generally considered as exerting their locomotor and rewarding effects through an increased dopaminergic transmission in the nucleus accumbens. Noradrenergic transmission may also be implicated because most psychostimulants increase norepinephrine (NE) release, and numerous studies have indicated interactions between noradrenergic and dopaminergic neurons through alpha1-adrenergic receptors. However, analysis of the effects of psychostimulants after either destruction of noradrenergic neurons or pharmacological blockade of alpha1-adrenergic receptors led to conflicting results. Here we show that the locomotor hyperactivities induced by d-amphetamine (1-3 mg/kg), cocaine (5-20 mg/kg), or morphine (5-10 mg/kg) in mice lacking the alpha1b subtype of adrenergic receptors were dramatically decreased when compared with wild-type littermates. Moreover, behavioral sensitizations induced by d-amphetamine (1-2 mg/kg), cocaine (5-15 mg/kg), or morphine (7.5 mg/kg) were also decreased in knock-out mice when compared with wild-type. Ruling out a neurological deficit in knock-out mice, both strains reacted similarly to novelty, to intraperitoneal saline, or to the administration of scopolamine (1 mg/kg), an anti-muscarinic agent. Finally, rewarding properties could not be observed in knock-out mice in an oral preference test (cocaine and morphine) and conditioned place preference (morphine) paradigm. Because catecholamine tissue levels, autoradiography of D1 and D2 dopaminergic receptors, and of dopamine reuptake sites and locomotor response to a D1 agonist showed that basal dopaminergic transmission was similar in knock-out and wild-type mice, our data indicate a critical role of alpha1b-adrenergic receptors and noradrenergic transmission in the vulnerability to addiction.

The BH4 domain of Bcl-X(L) rescues astrocyte degeneration in amyotrophic lateral sclerosis by modulating intracellular calcium signals.

Collective evidence indicates that motor neuron degeneration in amyotrophic lateral sclerosis (ALS) is non-cell-autonomous and requires the interaction with the neighboring astrocytes. Recently, we reported that a subpopulation of spinal cord astrocytes degenerates in the microenvironment of motor neurons in the hSOD1(G93A) mouse model of ALS. Mechanistic studies in vitro identified a role for the excitatory amino acid glutamate in the gliodegenerative process via the activation of its inositol 1,4,5-triphosphate (IP(3))-generating metabotropic receptor 5 (mGluR5). Since non-physiological formation of IP(3) can prompt IP(3) receptor (IP(3)R)-mediated Ca(2+) release from the intracellular stores and trigger various forms of cell death, here we investigated the intracellular Ca(2+) signaling that occurs downstream of mGluR5 in hSOD1(G93A)-expressing astrocytes. Contrary to wild-type cells, stimulation of mGluR5 causes aberrant and persistent elevations of intracellular Ca(2+) concentrations ([Ca(2+)](i)) in the absence of spontaneous oscillations. The interaction of IP(3)Rs with the anti-apoptotic protein Bcl-X(L) was previously described to prevent cell death by modulating intracellular Ca(2+) signals. In mutant SOD1-expressing astrocytes, we found that the sole BH4 domain of Bcl-X(L), fused to the protein transduction domain of the HIV-1 TAT protein (TAT-BH4), is sufficient to restore sustained Ca(2+) oscillations and cell death resistance. Furthermore, chronic treatment of hSOD1(G93A) mice with the TAT-BH4 peptide reduces focal degeneration of astrocytes, slightly delays the onset of the disease and improves both motor performance and animal lifespan. Our results point at TAT-BH4 as a novel glioprotective agent with a therapeutic potential for ALS.

Does the physical disector method provide an accurate estimation of sensory neuron number in rat dorsal root ganglia?

The physical disector is a method of choice for estimating unbiased neuron numbers; nevertheless, calibration is needed to evaluate each counting method. The validity of this method can be assessed by comparing the estimated cell number with the true number determined by a direct counting method in serial sections. We reconstructed a 1/5 of rat lumbar dorsal root ganglia taken from two experimental conditions. From each ganglion, images of 200 adjacent semi-thin sections were used to reconstruct a volumetric dataset (stack of voxels). On these stacks the number of sensory neurons was estimated and counted respectively by physical disector and direct counting methods. Also, using the coordinates of nuclei from the direct counting, we simulate, by a Matlab program, disector pairs separated by increasing distances in a ganglion model. The comparison between the results of these approaches clearly demonstrates that the physical disector method provides a valid and reliable estimate of the number of sensory neurons only when the distance between the consecutive disector pairs is 60 microm or smaller. In these conditions the size of error between the results of physical disector and direct counting does not exceed 6%. In contrast when the distance between two pairs is larger than 60 microm (70-200 microm) the size of error increases rapidly to 27%. We conclude that the physical dissector method provides a reliable estimate of the number of rat sensory neurons only when the separating distance between the consecutive dissector pairs is no larger than 60 microm.

Astrocytic αVβ3 integrin inhibits neurite outgrowth and promotes retraction of neuronal processes by clustering Thy-1.

Thy-1 is a membrane glycoprotein suggested to stabilize or inhibit growth of neuronal processes. However, its precise function has remained obscure, because its endogenous ligand is unknown. We previously showed that Thy-1 binds directly to α(V)β(3) integrin in trans eliciting responses in astrocytes. Nonetheless, whether α(V)β(3) integrin might also serve as a Thy-1-ligand triggering a neuronal response has not been explored. Thus, utilizing primary neurons and a neuron-derived cell line CAD, Thy-1-mediated effects of α(V)β(3) integrin on growth and retraction of neuronal processes were tested. In astrocyte-neuron co-cultures, endogenous α(V)β(3) integrin restricted neurite outgrowth. Likewise, α(V)β(3)-Fc was sufficient to suppress neurite extension in Thy-1(+), but not in Thy-1(-) CAD cells. In differentiating primary neurons exposed to α(V)β(3)-Fc, fewer and shorter dendrites were detected. This effect was abolished by cleavage of Thy-1 from the neuronal surface using phosphoinositide-specific phospholipase C (PI-PLC). Moreover, α(V)β(3)-Fc also induced retraction of already extended Thy-1(+)-axon-like neurites in differentiated CAD cells as well as of axonal terminals in differentiated primary neurons. Axonal retraction occurred when redistribution and clustering of Thy-1 molecules in the plasma membrane was induced by α(V)β(3) integrin. Binding of α(V)β(3)-Fc was detected in Thy-1 clusters during axon retraction of primary neurons. Moreover, α(V)β(3)-Fc-induced Thy-1 clustering correlated in time and space with redistribution and inactivation of Src kinase. Thus, our data indicates that α(V)β(3) integrin is a ligand for Thy-1 that upon binding not only restricts the growth of neurites, but also induces retraction of already existing processes by inducing Thy-1 clustering. We propose that these events participate in bi-directional astrocyte-neuron communication relevant to axonal repair after neuronal damage.

Hypocretin (orexin) biology and the pathophysiology of narcolepsy with cataplexy.

The discovery of hypocretins (orexins) and their causal implication in narcolepsy is the most important advance in sleep research and sleep medicine since the discovery of rapid eye movement sleep. Narcolepsy with cataplexy is caused by hypocretin deficiency owing to destruction of most of the hypocretin-producing neurons in the hypothalamus. Ablation of hypocretin or hypocretin receptors also leads to narcolepsy phenotypes in animal models. Although the exact mechanism of hypocretin deficiency is unknown, evidence from the past 20 years strongly favours an immune-mediated or autoimmune attack, targeting specifically hypocretin neurons in genetically predisposed individuals. These neurons form an extensive network of projections throughout the brain and show activity linked to motivational behaviours. The hypothesis that a targeted immune-mediated or autoimmune attack causes the specific degeneration of hypocretin neurons arose mainly through the discovery of genetic associations, first with the HLA-DQB1*06:02 allele and then with the T-cell receptor α locus. Guided by these genetic findings and now awaiting experimental testing are models of the possible immune mechanisms by which a specific and localised brain cell population could become targeted by T-cell subsets. Great hopes for the identification of new targets for therapeutic intervention in narcolepsy also reside in the development of patient-derived induced pluripotent stem cell systems.

Neuronal expression of the ubiquitin ligase Nedd4-2 in rat dorsal root ganglia: modulation in the SNI model of neuropathic pain

La douleur neuropathique est une forme de douleur chronique apparaissant suite à des lésions du système nerveux somato-sensoriel. Caractérisée par une plasticité neuronale inadapté, elle est très souvent intense, invalidante, associe des symptômes comme l'allodynie ou l' hyperalgésie et reste difficile à traiter avec les agents thérapeutiques actuels. Le thème de mon travail de thèse se concentre sur des mécanismes moléculaires de modulation des canaux sodiques voltage-dépendants suite à une lésion du nerf périphérique. Dans l'article présenté en annexe, j'ai focalisé mon travail sur une protéine, Nedd4-2, qui est une ligase ubiquitine. Elle a pour rôle de réguler et d'internaliser dans la cellule des protéines membranaires dont les canaux sodiques. Suite aux lésions du système nerveux périphérique, il existe une hyperexcitabilité neuronale engendrée notamment par un surplus et une dysrégulation des canaux sodiques à la membrane cellulaire. Dans 1 'hypothèse que l'ubiquitine ligase Nedd4-2 soit présente dans les neurones sensitifs primaires et ait un rôle dans la régulation des canaux sodiques, nous avons identifié cette protéine dans les neurones nociceptifs primaires du rat. En utilisant des techniques de Western Blot et d'immunohistochimie, j'ai trouvé que Nedd4-2 est présente dans presque 50% des neurones du ganglion spinal et ces neurones sont principalement des neurones nociceptifs. Dans un modèle expérimental de douleur neuropathique (SN I, pour spared nerve injury), Nedd4-2 se retrouve significativement diminuée dans le tissu du ganglion spinal. J'ai également investigué 1' expression de 2 isoformes des canaux sodiques connues pour leur implication dans la douleur, Navl.7 et Navl.8, et ces 2 isoformes se retrouvent dans les mêmes neurones que Nedd4-2. La caractérisation détaillée est décrite dans le manuscrit: «Neuronal expression of the ubiquitin ligase Nedd4-2 in rat dorsal root ganglia: modulation in the SNI model of neuropathic pain; Cachemaille M, Laedermann CJ, Pertin M, Abriel H, Gasselin RD, Decosterd 1.» Les résultats obtenus indiquent que Nedd4-2, en étant downrégulé après une lésion nerveuse, pourrait ainsi contribuer à une augmentation des canaux sodiques fonctionnels à la membrane. Ainsi Nedd4-2 pourrait être proposée comme cible thérapeutique de manière alternative aux bloqueurs de canaux sodiques. Ce travail a permis l'initiation d'autres expériences. J'ai contribué activement à la construction de vecteurs viraux type adéno-associé recombinant (rAA V2/6) et surexprimé la protéine in vivo dans les ganglions spinaux. Cette partie de mon travail se trouve intégrée dans d'autres travaux de mon laboratoire d'accueil qui a pu démontrer les effets fonctionnels de cette approche sur les courants sodiques enregistrés par électrophysiologie et une diminution de la douleur neuropathique chez la souris. - Abstract-Neuronal hyperexcitability following peripheral nerve lesions may stem from altered activity of voltagegated sodium channels (VGSCs), which gives rise toallodynia or hyperalgesia. In vitro, the ubiquitin ligase Nedd4-2 is a negative regulator of VGSC a-subunits (Nav), in particular Nav1.7, a key actor in nociceptor excitability. We therefore studied Nedd4-2 in rat nociceptors, its co-expression with Nav1.7 and Nav1.8, and its regulation in pathology. Adult rats were submitted to the spared nerve injury (SNI) model of neuropathic pain or injected with complete Freund's adjuvant (CFA), a model of inflammatory pain. L4 dorsal root ganglia (DRG) were analyzed in shamoperated animals, seven days after SNI and 48 h after CFA with immunofluorescence and Western blot. We observed Nedd4-2 expression in almost 50% of DRG neurons, mostly small and medium-sized. A preponderant localization is found in the non-peptidergic sub-population. Additionally, 55.7± 2.7% and 55.0 ±3.6% of Nedd4-2-positive cells are co-labeled with Nav1.7 and Nav1.8 respectively. SNI significantly decreases the proportion of Nedd4-2-positive neurons from 45.9± 1.9% to 33.5± 0.7% (p < 0.01) and the total Nedd4-2 protein to 44%± 0.13% of its basal level (p <0.01, n = 4 animals in each group, mean± SEM). In contrast, no change in Nedd4-2 was found after peripheral inflammation induced by CFA. These results indicate that Nedd4-2 is present in nociceptive neurons, is downregulated after peripheral nerve injury, and might therefore contribute to the dysregulation of Navs involved in the hyperexcitability associated with peripheral nerve injuries.

Development of projections from auditory to visual areas in the cat.

In newborn kittens, cortical auditory areas (including AI and AII) send transitory projections to ipsi- and contralateral visual areas 17 and 18. These projections originate mainly from neurons in supragranular layers but also from a few in infragranular layers (Innocenti and Clarke: Dev. Brain Res. 14:143-148, '84; Clarke and Innocenti: J. Comp. Neurol. 251:1-22, '86). The postnatal development of these projections was studied with injections of anterograde tracers (wheat germ agglutinin-horseradish peroxidase [WGA-HRP]) in AI and AII and of retrograde tracers (WGA-HRP, fast blue, diamidino yellow, rhodamine-labeled latex beads) in areas 17 and 18. It was found that the projections are nearly completely eliminated in development, this, by the end of the first postnatal month. Until then, most of the transitory axons seem to remain confined to the white matter and the depth of layer VI; a few enter it further but do not appear to form terminal arbors. As for other transitory cortical projections the disappearance of the transitory axons seems not to involve death of their neurons of origin. In kittens older than 1 month and in normal adult cats, retrograde tracer injections restricted to, or including, areas 17 and 18 label only a few neurons in areas AI and AII. Unlike the situation in the kitten, nearly all of these are restricted to layers V and VI. A similar distribution of neurons projecting from auditory to visual areas is found in adult cats bilaterally enucleated at birth, which suggests that the postnatal elimination of the auditory-to-visual projection is independent of visual experience and more generally of information coming from the retina.