The basal ganglia (BG) is known for its function not only in motor modulation but also in action selection and reward learning. There are two major anatomical pathways through the BG, the direct and the indirect pathways. The direct pathway starts from the striatum and then directly projects to the globus pallidus, pars interna (GPi) and the substantia nigra, pars reticulata (SNr) respectively, while the indirect pathway starts from the striatum but then indirectly projects to GPi and SNr through the globus pallidus, pars externa and then to the subthalamic nucleus. In addition, the output from GPi not only projects to the thalamus where it has been proposed to function in motor control, but also to the lateral habenula (LHb) where it has been proposed to function in outcome evaluation. Previous studies have found that there are three major genetically distinct neuron groups in the entopeduncular nucleus (EP) (rodent homologue of the primate GPi): 1) purely glutamatergic neurons projecting to LHb neurons expressing parvalbumin (PVLHb); 2) purely GABAergic neurons projecting to motor thalamic neurons expressing parvalbumin (PVThal); 3) GABA/glutamate co-releasing neurons projecting to LHb neurons expressing somatostatin (SstLHb). In this study, we knocked out the vesicular glutamate transporter 2 in SstLHb neurons through an adeno-associated virus in mice to test for the impact on goal-directed behavior using a probabilistic switching, two-armed bandit task (2ABT). Results obtained from the freely moving, water-restricted somatostatin-cre mice with the vesicular glutamate transporter 2 ablated in SstLHb neurons showed that: 1) there was neither improvement nor decline in their performance on the task; 2) they might be more distracted between trials while more concentrated within a trial; 3) they had an increase in the probability of switching between ports on consecutive trials when uncertainty in the location of the highly rewarded port was maximum compared to the control animals with intact glutamate release from SstLHb neurons to LHb. The success of the viral expression was then confirmed through whole-cell voltage-clamp recordings of postsynaptic neurons of the LHb, receiving projections from SstLHb neurons. In conclusion, our study has suggested that the glutamate release from the GABA/glutamate co-releasing neurons of EP projecting to LHb may play a role in reinforcement learning and motivation to obtain rewards, and the loss of glutamate in the GABA/glutamate co-releasing vesicles results in increasing uptake of GABA into these vesicles, leading to possible rebound burst firing of SstLHb neurons that eventually increases the sensitivity towards low rate of reward-delivery dramatically. / 2026-03-13T00:00:00Z

Les récepteurs mu aux opioïdes (MORs) jouent un rôle central dans l’addiction et ce majoritairement via le contrôle qu’ils exercent sur les phénomènes de récompense. Les effets des MORs sur le système de récompense sont généralement attribués à leur expression dans le circuit mésocorticolimbique. Les MORs sont toutefois exprimés dans d’autres régions du cerveau, notamment dans une petite structure cérébrale épithalamique qui exprime la plus forte densité de MORs : l’habénula médiane (MHb). Le rôle de cette population spécifique de récepteurs n’a jamais été exploré, malgré l’implication fortement suggérée dans la littérature de l’habénula dans l’addiction et les processus de récompense/aversion sous-jacents. Nous avons donc généré un modèle de souris knockout conditionnel chez lesquelles nous supprimons les MORs uniquement dans l’habénula et nous les avons soumis à des tests comportementaux dans le but d’évaluer les fonctions de ces récepteurs et leur impact sur le comportement avec un focus sur les processus aversifs, cognitifs et récompensant. Nos résultats révèlent que les récepteurs mu que nous supprimons dans l’habénula limitent l’aversion somatique et affective mais ne modulent ni les réponses locomotrices, analgésiques et de récompense à la morphine, ni les fonctions cognitives que nous avons testées. Nous identifions donc pour la première fois une population de MORs dans l’habénula qui freinent l’aversion et nous suggérons par là-même que les récepteurs mu de l’habénula pourraient être cruciaux dans le stage «sevrage aversif» des cycles d’addiction. / Mu opioid receptors (MORs) have been extensively studied for their addictive properties that are thought to operate through the control of reward processes. While the importance of MORs in reward is generally attributed to their presence in the mesocorticolimbic circuitry, their role in the medial habenula (MHb), the structure in which MORs are most densely expressed, remains unexplored to date. This is quite surprising given the increasing literature on the habenula’s role in addiction as well as reward/aversion processes. Here we generated a conditional knockout mouse model that lacks MORs solely in the MH band we investigated the contribution of habenular MORs in brain functions and behavioural out comes with emphasis on reward, aversion and cognition. While the performance of our mutant model did not differ in locomotor, analgesic and reward responses to morphine norincognitive tasks compared to control mice, we uncovered a novel role for MORs in aversive states.This is the first report demonstrating that MORs control both somatic and affective aversion specifically at the level of the MHb. Habenular MORs could thus be crucial to the aversive with drawal stage of addiction cycles that is thought to increase craving and prevent success in quitting.

Ce travail de thèse avait pour objectif d’étudier le rôle de l’habénula latérale (HbL) dans les processus mnésiques chez le Rat en utilisant une approche par inactivation réversible grâce à l’administration de muscimol ou de CNQX. Nous avons ainsi montré l’implication de l’HbL dans : i) les processus d’encodage et de rappel d’une mémoire spatiale en piscine de Morris ; ii) la mémoire de travail, comme relais potentiel d’informations en provenance du cortex préfrontal médian, dans un paradigme de non-appariement différé à la position en boites de conditionnement opérant; iii) la réponse émotionnelle, aux niveaux comportemental et physiologique, à une situation stressante. L’ensemble de ces résultats suggèrent que l’HbL est impliquée dans les processus « online » de gestion des informations sensorielles, et qu’elle participe à la prise en compte de l’aspect émotionnel d’une situation. Ces particularités en font un lien important potentiel entre gestion des émotions et cognition. / The main objective of this thesis was to investigate the role of the lateral habenula (LHb) in mnemonic processes in rats using reversible inactivations with muscimol or CNQX. We have shown the involvement of the LHb in : i) encoding and retrieval of spatial reference memory in the Morris water maze ; ii) working memory, as a potential relay of top-down information coming from the medial prefrontal cortex, in a delayed non-matching to position paradigm using operant chambers ; iii) the behavioral and physiological responses to stressful situations. Altogether, those results suggest that the LHb is involved in the « online » process of sensory information. They also suggest that it is involved in coping with particularly stressful situations, and further position the LHb as an interface between emotions and cognition.

Nicotine dependence is responsible for perpetuating the adverse health effects due to tobacco use, the leading cause of preventable death worldwide. Nicotine is an agonist for nicotinic acetylcholine receptors, which are enriched in the mesocorticolimbic and habenulo-interpeduncular circuitries, underlying nicotine reward and withdrawal, respectively. Drugs of abuse, including nicotine, induce stable neuroadaptations, requiring protein synthesis through regulation of transcription factors, epigenetic mechanisms, and non-coding RNAs. It also been shown that miRNAs in brain are regulated by nicotine and that miRNA dysregulation contributes to brain dysfunction, including drug addiction. While much is known about the neurocircuitry responsible for the behaviors associated with nicotine reward or withdrawal, the underlying molecular mechanisms of how these changes in behavior are induced are less clear. Using miRNA-/mRNA-Seq, we demonstrate that there are widespread changes in both miRNA and mRNA expression in brain regions comprising the mesocorticolimbic circuit after chronic nicotine treatment, and the habenulo-interpeduncular circuit during acute nicotine withdrawal. Conserved, differentially expressed miRNAs were predicted to target inversely regulated mRNAs. We determined that expression of miR-106b-5p is up-regulated and Profilin 2 (Pfn2), an actin-binding protein enriched in the brain, is down-regulated in the interpeduncular nucleus (IPN) during acute nicotine withdrawal. Further we show that miR-106b-5p represses Pfn2 expression. We demonstrate that knockdown of Pfn2 in the IPN is sufficient to induce anxiety, a symptom of withdrawal. This novel role of Pfn2 in nicotine withdrawal-associated anxiety is a prime example of this dataset’s utility, allowing for the identification of a multitude of miRNAs/mRNA which may participate in the molecular mechanisms underlying the neuroadaptations of nicotine dependence.

The Lateral Habenula (LHb) have been implicated in both reward-seeking behavior and in depressive disorders due to its modulatory effects on dopamine rich areas. Excitatory projections from LHb target GABAergic interneurons of both ventral tegmental area (VTA) and rostromedial tegmental nucleus (RMTg) and consequently provide strong inhibition on VTA‟s dopaminergic neurons. These reward related signals are provided to LHb from distinct neuronal populations in internal Globus Pallidus (GPi). Here by using a dual viral combination of an adeno-associated helper virus (AAV) and a genetically modified rabies virus that displays specific transsynaptic retrograde spread we are providing anatomical evidence for a strong innervations of the LHb by VGLUT2+ glutaminergic and SOM+ GABAergic GPi neurons. Our results provide the first direct evidence for both an excitatory and an inhibitory projection m, from GPi to the LHb. Given the importance of the LHb as a modulatory nucleus of the dopaminergic system, the definition of its connectivity and function will give valuable insights in the understanding of both reward-seeking behavior and depressive disorders.

Recently, the medial habenula-interpeduncular (MHb-IPN) axis has been hypothesized to modulate anxiety although neuronal populations and molecular mechanisms regulating affective behaviors in this circuit are unknown. Here we show that MHb cholinergic neuron activity directly regulates anxiety-like behavior. Optogenetic silencing of MHb cholinergic IPN inputs reduced anxiety-like behavior in mice. MHb cholinergic neurons are unique in that they robustly express neuronal nicotinic acetylcholine receptors (nAChRs), although their role as autoreceptors in these neurons has not been described. nAChRs are ligand-gated cation channels that are activated by the excitatory neurotransmitter, acetylcholine (ACh), as well as nicotine, the addictive component of tobacco smoke. We expressed novel nAChR subunits that render nAChRs hypersensitive to ACh, ACh detectors, selectively in MHb cholinergic neurons of adult mice. Mice expressing these ACh detectors exhibited increased baseline anxiety-like behavior that was alleviated by blocking the mutant receptors. Under stressful conditions, such as during nicotine withdrawal, nAChRs were functionally upregulated in MHb cholinergic neurons mediating an increase in anxiety-like behavior. Together, these data indicate that MHb cholinergic neurons regulate anxiety via signaling through nicotinic autoreceptors and point toward nAChRs in MHb as molecular targets for novel anxiolytic therapeutics.

The classic surfactant proteins (SPs) A, B, C, and D were discovered in the lungs, where they contribute to host defense and regulate the alveolar surface tension during breathing. Their additional importance for brain physiology was discovered decades later. SP-G, a novel amphiphilic SP, was then identified in the lungs and is mostly linked to inflammation. In the brain, it is also present and significantly elevated after hemorrhage in premature infants and in distinct conditions affecting the cerebrospinal fluid circulation of adults. However, current knowledge on SP-G-expression is limited to ependymal cells and some neurons in the subventricular and superficial cortex. Therefore, we primarily focused on the distribution of SP-G-immunoreactivity (ir) and its spatial relationships with components of the neurovascular unit in murine forebrains. Triple fluorescence labeling elucidated SP-G-co-expressing neurons in the habenula, infundibulum, and hypothalamus. Exploring whether SP-G might play a role in Alzheimer’s disease (AD), 3xTg-AD mice were investigated and displayed age-dependent hippocampal deposits of β-amyloid and hyperphosphorylated tau separately from clustered, SP-G-containing dots with additional Reelin-ir—which was used as established marker for disease progression in this specific context. Semi-quantification of those dots, together with immunoassay-based quantification of intra- and extracellular SP-G, revealed a significant elevation in old 3xTg mice when compared to age-matched wildtype animals. This suggests a role of SP-G for the pathophysiology of AD, but a confirmation with human samples is required.

Nous avons étudié les relations anatomiques entre les systèmes de neurotransmission à substance P (SP) et à sérotonine (5-hydroxytryptamine, 5-HT) dans le noyau du raphé dorsal (NRD) du rongeur, afin de mieux comprendre les interactions entre ces systèmes durant la régulation de l’humeur. Le NRD reçoit une innervation SP provenant de l’habenula, et le blocage pharmacologique des récepteurs neurokinine-1 (rNK1) de la SP aurait des effets antidépresseurs. Chez le rongeur, le traitement par les antagonistes des rNK1 s’accompagne d’une désensibilisation des autorécepteurs 5-HT1A de la 5-HT et d’une hausse de l’activité des neurones 5-HT dans le NRD, suggérant des interactions locales entre ces deux systèmes. Dans un premier temps, nous avons démontré par doubles marquages immunocytochimiques en microscopies optique, confocale et électronique, la présence du rNK1 dans une sous-population de neurones 5-HT du NRD caudal. Lors de l’analyse en microscopie électronique, nous avons pu constater que les rNK1 étaient principalement cytoplasmiques dans les neurones 5-HT et membranaires sur les neurones non 5-HT du noyau. Grâce à d’autres doubles marquages, nous avons aussi pu identifier les neurones non-5-HT porteurs de rNK1 comme étant GABAergiques. Nous avons ensuite combiné l’immunomarquage de la SP avec celui du rNK1, dans le but d’examiner les relations entre les terminaisons (varicosités *) axonales SP et les neurones 5-HT (pourvus de rNK1 cytoplasmiques du NRD caudal. En simple marquage de la SP, nous avons pu estimer à 41% la fréquence avec laquelle les terminaisons SP font synapse. Dans le matériel doublement marqué pour la SP et son récepteur, les terminaisons SP ont été fréquemment retrouvées en contact direct ou à proximité des dendrites munies de rNK1 cytoplasmiques, mais toujours éloignées des dendrites à rNK1 membranaires. Pour tester l’hypothèse d’une internalisation soutenue des rNK1 par la SP dans les neurones 5-HT, nous avons ensuite examiné la localisation subcellulaire du récepteur chez le rat traité avec un antagoniste du rNK1, le RP67580. La densité du marquage des rNK1 a été mesurée dans le cytoplasme et sur la membrane des deux types de dendrites (5-HT: rNK1 cytoplasmiques; non 5-HT: rNK1 membranaires). Une heure après une injection unique de l’antagoniste, la distribution du rNK1 est apparue inchangée dans les deux types de neurones (5-HT et non 5-HT). Par contre, après un traitement quotidien de 7 ou 21 jours avec l’antagoniste, nous avons mesuré une augmentation significative des densités cytoplasmique et membranaire du rNK1 dans les neurones 5-HT, sans aucun changement dans les neurones non 5-HT. Ces traitements ont aussi augmenté l’expression du gène rNK1 dans le NRD. Enfin, nous avons mesuré une hausse de la densité membranaire du rNK1 dans les neurones 5-HT, sans hausse de densité cytoplasmique, par suite d’une lésion bilatérale de l’habenula. Ces résultats confortent l’hypothèse d’une activation et d’une internalisation soutenues des rNK1 par la SP dans les neurones 5-HT du NRD caudal. Ils suggèrent aussi que le trafic des rNK1 dans les neurones 5-HT du NRD représente un mécanisme cellulaire en contrôle de l’activation du système 5-HT par les afférences SP en provenance de l’habenula. / We have studied in detail the relationships between substance P (SP) and serotonin (5-hydroxytryptamine, 5-HT) neurotransmission systems in the dorsal raphe nucleus (DRN) of rodents, in order to further our understanding of their interaction during mood regulation. The DRN receives a SP innervation arising from the habenula and, in human, it is known that blockade of the neurokinin-1 receptor (NK1r) of SP by antagonists may have antidepressant effects. In rodents, treatment with NK1r antagonists is known to increase the firing of DRN 5-HT neurons and to induce a desensitization of their 5-HT1A autoreceptors, suggesting local interactions between the SP and 5-HT systems. In a first step, we were able to demonstrate by means of light, confocal, and electron microscopic immunocytochemistry, including double immunolabelings of NK1r and of the biosynthetic enzyme of 5-HT, tryptophane hydroxylase, the presence of NK1r in a subpopulation of 5-HT neurons in the caudal DRN of rat and mouse. After the dual immunolabelings for electron microscopy, we also found that NK1r was mostly cytoplasmic in 5-HT neurons while predominating on the plasma membrane of TPH negative (non 5-HT) neurons. Subsequently, in additionnal double labeling experiments, we were able to identify most if not all non 5-HT dendrites bearing membranous NK1r as GABAergic. In a second step, we combined the immunolabeling of SP with that of NK1r, in order to examine the relationships between SP axon terminals (varicosities *) and the two categories of DRN neurons (5-HT: cytoplasmic NK1r; non 5-HT: membranous NK1r). After single SP labeling, we could estimate the frequency with which SP terminals made synapse at 41%, at least. In the material doubly labeled for SP and NK1r, the SP terminals were often found in close contact or in the immediate proximity of dendrites endowed with cytoplasmic receptor, but never near non 5-HT dendrites bearing membrane bound receptors. To test the hypothesis of a sustained internalization of NK1r in 5-HT neurons, we then tested the effects of RP67580, a selective NK1r antagonist, on the subcellular localization of the receptor. One hour after administration of a single dose, the NK1r distribution was unchanged in both types of dendrites (5-HT and non 5-HT). However, after administration for 7 (subchronic) or 21 (chronic) days, the cytoplasmic and the membrane densities of NK1r were significantly increased in 5-HT dendrites, without any change in non 5-HT dendrites. These treatments also increased NK1r gene expression in the caudal DRN. Lastly, a significant increase in the membrane density of NK1r was measured in the 5-HT neurons, without any increase of the cytoplamic density, following bilateral electrolytic lesioning of the habenula. These results strenghtened the hypothesis of a sustained activation and internalization of NK1r by SP in 5-HT neurons of the caudal DRN. They also suggested that trafficking of NK1r in these cells might represent a cellular mechanism in control of the activation of the 5-HT system by SP afferents from the habebula.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

Nous avons étudié les relations anatomiques entre les systèmes de neurotransmission à substance P (SP) et à sérotonine (5-hydroxytryptamine, 5-HT) dans le noyau du raphé dorsal (NRD) du rongeur, afin de mieux comprendre les interactions entre ces systèmes durant la régulation de l’humeur. Le NRD reçoit une innervation SP provenant de l’habenula, et le blocage pharmacologique des récepteurs neurokinine-1 (rNK1) de la SP aurait des effets antidépresseurs. Chez le rongeur, le traitement par les antagonistes des rNK1 s’accompagne d’une désensibilisation des autorécepteurs 5-HT1A de la 5-HT et d’une hausse de l’activité des neurones 5-HT dans le NRD, suggérant des interactions locales entre ces deux systèmes. Dans un premier temps, nous avons démontré par doubles marquages immunocytochimiques en microscopies optique, confocale et électronique, la présence du rNK1 dans une sous-population de neurones 5-HT du NRD caudal. Lors de l’analyse en microscopie électronique, nous avons pu constater que les rNK1 étaient principalement cytoplasmiques dans les neurones 5-HT et membranaires sur les neurones non 5-HT du noyau. Grâce à d’autres doubles marquages, nous avons aussi pu identifier les neurones non-5-HT porteurs de rNK1 comme étant GABAergiques. Nous avons ensuite combiné l’immunomarquage de la SP avec celui du rNK1, dans le but d’examiner les relations entre les terminaisons (varicosités *) axonales SP et les neurones 5-HT (pourvus de rNK1 cytoplasmiques du NRD caudal. En simple marquage de la SP, nous avons pu estimer à 41% la fréquence avec laquelle les terminaisons SP font synapse. Dans le matériel doublement marqué pour la SP et son récepteur, les terminaisons SP ont été fréquemment retrouvées en contact direct ou à proximité des dendrites munies de rNK1 cytoplasmiques, mais toujours éloignées des dendrites à rNK1 membranaires. Pour tester l’hypothèse d’une internalisation soutenue des rNK1 par la SP dans les neurones 5-HT, nous avons ensuite examiné la localisation subcellulaire du récepteur chez le rat traité avec un antagoniste du rNK1, le RP67580. La densité du marquage des rNK1 a été mesurée dans le cytoplasme et sur la membrane des deux types de dendrites (5-HT: rNK1 cytoplasmiques; non 5-HT: rNK1 membranaires). Une heure après une injection unique de l’antagoniste, la distribution du rNK1 est apparue inchangée dans les deux types de neurones (5-HT et non 5-HT). Par contre, après un traitement quotidien de 7 ou 21 jours avec l’antagoniste, nous avons mesuré une augmentation significative des densités cytoplasmique et membranaire du rNK1 dans les neurones 5-HT, sans aucun changement dans les neurones non 5-HT. Ces traitements ont aussi augmenté l’expression du gène rNK1 dans le NRD. Enfin, nous avons mesuré une hausse de la densité membranaire du rNK1 dans les neurones 5-HT, sans hausse de densité cytoplasmique, par suite d’une lésion bilatérale de l’habenula. Ces résultats confortent l’hypothèse d’une activation et d’une internalisation soutenues des rNK1 par la SP dans les neurones 5-HT du NRD caudal. Ils suggèrent aussi que le trafic des rNK1 dans les neurones 5-HT du NRD représente un mécanisme cellulaire en contrôle de l’activation du système 5-HT par les afférences SP en provenance de l’habenula. / We have studied in detail the relationships between substance P (SP) and serotonin (5-hydroxytryptamine, 5-HT) neurotransmission systems in the dorsal raphe nucleus (DRN) of rodents, in order to further our understanding of their interaction during mood regulation. The DRN receives a SP innervation arising from the habenula and, in human, it is known that blockade of the neurokinin-1 receptor (NK1r) of SP by antagonists may have antidepressant effects. In rodents, treatment with NK1r antagonists is known to increase the firing of DRN 5-HT neurons and to induce a desensitization of their 5-HT1A autoreceptors, suggesting local interactions between the SP and 5-HT systems. In a first step, we were able to demonstrate by means of light, confocal, and electron microscopic immunocytochemistry, including double immunolabelings of NK1r and of the biosynthetic enzyme of 5-HT, tryptophane hydroxylase, the presence of NK1r in a subpopulation of 5-HT neurons in the caudal DRN of rat and mouse. After the dual immunolabelings for electron microscopy, we also found that NK1r was mostly cytoplasmic in 5-HT neurons while predominating on the plasma membrane of TPH negative (non 5-HT) neurons. Subsequently, in additionnal double labeling experiments, we were able to identify most if not all non 5-HT dendrites bearing membranous NK1r as GABAergic. In a second step, we combined the immunolabeling of SP with that of NK1r, in order to examine the relationships between SP axon terminals (varicosities *) and the two categories of DRN neurons (5-HT: cytoplasmic NK1r; non 5-HT: membranous NK1r). After single SP labeling, we could estimate the frequency with which SP terminals made synapse at 41%, at least. In the material doubly labeled for SP and NK1r, the SP terminals were often found in close contact or in the immediate proximity of dendrites endowed with cytoplasmic receptor, but never near non 5-HT dendrites bearing membrane bound receptors. To test the hypothesis of a sustained internalization of NK1r in 5-HT neurons, we then tested the effects of RP67580, a selective NK1r antagonist, on the subcellular localization of the receptor. One hour after administration of a single dose, the NK1r distribution was unchanged in both types of dendrites (5-HT and non 5-HT). However, after administration for 7 (subchronic) or 21 (chronic) days, the cytoplasmic and the membrane densities of NK1r were significantly increased in 5-HT dendrites, without any change in non 5-HT dendrites. These treatments also increased NK1r gene expression in the caudal DRN. Lastly, a significant increase in the membrane density of NK1r was measured in the 5-HT neurons, without any increase of the cytoplamic density, following bilateral electrolytic lesioning of the habenula. These results strenghtened the hypothesis of a sustained activation and internalization of NK1r by SP in 5-HT neurons of the caudal DRN. They also suggested that trafficking of NK1r in these cells might represent a cellular mechanism in control of the activation of the 5-HT system by SP afferents from the habebula.