El sistema endocannabinoide es un sistema de comunicación celular que regula diversas funciones fisiológicas en una amplia variedad de células del organismo. Los principales endocannabinoides son el 2-araquidonoilglicerol (2-AG) y la anandamida (AEA). Los endocannabinoides en el sistema nervioso central (SNC) funcionan como mensajeros retrógrados. Se sintetizan en los terminales postsinápticos a partir de diferentes vías, son liberados al espacio sináptico y ejercen su efecto por unión a receptores ubicados en los terminales presinápticos, principalmente inhibiendo la liberación de neurotransmisores. Entre las numerosas funciones que tienen los endocannabinoides en el SNC, se destaca su rol neuroprotector frente a enfermedades neurodegenerativas. La Enfermedad de Alzheimer (EA) es una de las mayores causas de demencias, cuya etiología se atribuye al péptido beta amiloide (βA) que se agrega formando primeramente oligómeros y luego fibrillas que se depositan en forma de placas amiloides. Las enfermedades neurodegenerativas tienen en común que se presentan generalmente en individuos de edad avanzada por lo que se las relaciona íntimamente con el envejecimiento. Se sabe que los endocannabinoides regulan numerosos procesos celulares y moleculares relacionados con el envejecimiento. En base a esto, los principales objetivos de esta Tesis Doctoral fueron: determinar el metabolismo de los endocannabinoides en el proceso de envejecimiento fisiológico y patológico, y estudiar la modulación de este metabolismo por agonistas y antagonistas de receptores cannabinoides. Los estudios fueron realizados en fracciones de membrana y soluble, y en sinaptosomas de la corteza cerebral (CC) de rata, una de las regiones del cerebro en la que el sistema endocannabinoide ejerce su rol neuroprotector. Los resultados del metabolismo del 2-AG y de la AEA en el envejecimiento mostraron que existe una desregulación del mismo tanto en los terminales sinápticos como en las membranas de las células de la CC. Con respecto al 2-AG, se observó que su síntesis en la CC de rata se produce por la acción de las enzimas lisofosfatidato fosfohidrolasa (LPAasa) y diacilglicerol lipasa (DAGL). A su vez, la hidrólisis de este endocannabinoide en los terminales sinápticos de la CC de ratas adultas y seniles así como en membranas de ratas adultas, se produce principalmente por la acción de la enzima monoacilglicerol lipasa (MAGL). Sin embargo, en membranas de la CC de ratas seniles la enzima serina hidrolasa con dominios α/β (ABHD) sería la principal involucrada. En la CC, la desregulación del metabolismo del 2-AG observada en ratas seniles se inclina hacia una disminución de este endocannabinoide, lo cual comprometería su función neuroprotectora. En cuanto a la AEA, se determinó que su hidrólisis tanto en los terminales sinápticos como en las membranas de la CC, es a través de la enzima ácido graso amidohidrolasa (FAAH). A su vez, se observó que el envejecimiento ejerce un efecto diferente en la degradación de la AEA ya sea que se tratara de las membranas celulares o de los terminales sinápticos. En las membranas de ratas seniles, se observó que la hidrólisis de este endocannabinoide aumenta, lo cual podría comprometer la función antiinflamatoria de la AEA en la CC. Por otra parte, en los terminales sinápticos se observó que la hidrólisis de la AEA disminuye en el envejecimiento, lo cual podría proteger al cerebro frente a los daños neurodegenerativos propios de la edad. Al analizar los niveles de los receptores CB1 y CB2, se observó que ambos disminuyen en los terminales sinápticos durante el envejecimiento. Teniendo en cuenta estos resultados además de la alteración observada en el metabolismo de los endocannabinoides, se propuso evaluar la regulación de la síntesis y/o de la degradación de los endocannabinoides mediante el uso de agonistas y antagonistas de los receptores cannabinoides. Estos estudios mostraron que los antagonistas de receptores CB1 y CB2 aumentarían la disponibilidad del 2-AG y de la AEA en los terminales sinápticos de la CC de ratas seniles. Por otra parte, los agonistas de estos receptores, especialmente del CB2, aumentarían la disponibilidad de la AEA tanto en terminales sinápticos como en membranas de células de la CC de ratas seniles. Para estudiar la hidrólisis de los endocannabinoides en estadios tempranos y tardíos de la EA, se preincubaron los sinaptosomas de la CC con el péptido βA1-40 en sus conformaciones de oligómeros (etapas tempranas de la EA) y fibrillas (etapas tardías de la EA). Estos estudios mostraron que la hidrólisis del 2-AG y de la AEA disminuye en el modelo de EA temprana, por lo que los niveles de ambos endocannabinoides se verían aumentados protegiendo a la célula frente a procesos neurodegenerativos. Sin embargo, en el modelo de EA tardía, se observó que la hidrólisis de la AEA disminuye mientras que la degradación del 2-AG aumenta drásticamente comprometiendo los efectos neuroprotectores atribuidos a este endocannabinoide. En conclusión, los resultados de esta Tesis Doctoral indican que el metabolismo de los endocannabinoides se encuentra desregulado tanto en el envejecimiento fisiológico como en el patológico, lo cual pone en peligro sus funciones neuroprotectoras, y que la modulación de este metabolismo en la senescencia puede darse a través de agonistas y/o antagonistas de los receptores CB1 y/o CB2. / The endocannabinoid system is a cell communication mechanism which regulates several physiological functions in a variety of cells. 2-arachidonoylglycerol (2-AG) and anandamide (AEA) are the main endocannabinoids known to date. In the central nervous system (CNS) endocannabinoids behave as retrograde messengers. They are synthesized by different pathways in postsynaptic endings, released to the synaptic clef, and they finally exert their functions by coupling to cannabinoid receptors located in presynaptic endings, thus inhibiting neurotransmitter release. Among the several functions assigned to endocannabinoids in the CNS, is their role as neuroprotective agents against several neurodegenerative diseases. Alzheimer Disease (AD) is one of the major causes of dementia and its etiology is thought to be due to the aggregation of β-amyloid peptide (βA) which firstly generates oligomers and subsequently fibril deposits known as amyloid plaques. One of the most common factors of neurodegenerative diseases is aging due to the fact that they usually occur in old people. Several studies have in fact shown that endocannabinoids participate in the modulation of numerous cellular and molecular processes related to aging. In view of this, the main purposes of this Ph. D. Thesis were to: analyze endocannabinoid metabolism in the physiological and pathological aging process, and to study the modulation of this metabolism by cannabinoid receptor agonists and antagonists. These studies were carried out in rat cerebral cortex (CC) membrane and soluble fractions and synaptosomes, a brain area in which endocannabinoids exert their neuroprotective role. 2-AG and AEA hydrolysis and/or synthesis assays in synaptic endings and also in membranes from CC cells during aging showed a deregulation of their metabolism. As to 2-AG, it could be observed that its synthesis in rat CC was carried out by the enzymes lysophosphatidate phosphohydrolase (LPAase) and diacylglycerol lipase (DAGL). Furthermore, its hydrolysis in CC synaptic endings of adult and aged rats and in CC membranes of adult rats was exerted mainly by the enzyme monoacylglycerol lipase (MAGL). However, in CC membranes of aged rats the main enzyme involved was α/β serine hydrolase domain (ABHD). Deregulation of 2-AG metabolism in CC was observed to favor the decrease of these endocannabinoid levels, though compromising its neuroprotective role. As to AEA, these studies demonstrated that its hydrolysis either in CC synaptic endings or in CC membranes was carried out by the enzyme fatty acid amide hydrolase (FAAH). In addition, a different effect in AEA breakdown between cell membranes and synaptic endings was observed as a result of aging. In aged CC membranes, its hydrolysis was observed to increase. This could, in turn, compromise AEA antiinflamatory role in CC. On the other hand, AEA hydrolysis in synaptic endings during aging was observed to decrease, possibly protecting the brain against neurodegenerative damages inflicted by aging. Assays of CB1 and CB2 receptors levels in aged synaptic endings showed they were lower than in adults. Based on these results and on the disarrangement observed in endocannabinoid metabolism, regulation of endocannabinoid hydrolysis and/or synthesis by cannabinoid receptor agonists and antagonists were then assayed. These studies showed an increased 2-AG and AEA availability in CC synaptic endings of aged rats as a result of the presence of CB1 and CB2 receptor antagonists. On the other hand, cannabinoid receptor agonists, particularly through CB2, increased AEA availability either in synaptic endings or cell membranes from CC of aged rats. Endocannabinoid hydrolysis in early and late AD stages was assayed by preincubating CC synaptosomes with βA1-40 peptide in different conformations: oligomers (mimicking early AD stages) and fibrils (mimicking late AD stages). These studies showed that 2-AG and AEA hydrolysis in the early AD model diminished, thus indicating that both endocannabinoid levels could be increased and therefore protect the cell against neurodegenerative processes. Moreover, in the late AD model AEA hydrolysis decreased while 2-AG breakdown highly increased, thus compromising the neuroprotective role assigned to 2-AG. Summing up, results from this Ph. D. Thesis show a deregulation in endocannabinoid metabolism either in physiological or pathological aging, which could, in turn, lead to a failure of their neuroprotective functions, and a modulation of this metabolism as a result of aging, which could be exerted by CB1 and/or CB2 receptor agonists and/or antagonists.

Les endocannabinoïdes, anandamide (AEA) et 2-arachidonoylglycérol (2-AG), appartiennent au système endocannabinoïde et sont impliqués dans la régulation de nombreux systèmes biologiques. Ils ont un rôle dans la modulation du système immunitaire et la défense anti-tumorale. Dans cette étude, nous avons montré que l'AEA et le 2-AG peuvent moduler l'activation des cellules Natural Killer CD56 + CD3- (NK) et des cellules NK92. A de faibles doses, ils potentialisent la production d'IFN γ induite par des doses suboptimales d'IL-12, d'IL-15 et d'IL-2. En outre, l'AEA et le 2-AG potentialisent la cytotoxicité des cellules NK. Des expériences de quantification des transcrits codant FasL, TRAIL, perforine et granzyme et des expériences d'inhibition indiquent que cette augmentation de la cytotoxicité est corrélée à une augmentation de l'expression de FasL, tandis que la production de granzyme, de perforine et l'expression TRAIL reste inchangées. Les effets des endocannabinoides sur les cellules NK sont inhibés en présence d'un antagoniste du récepteur endocannabinoidien de type 2 (CB2). Ces travaux in vitro on été confirmés in vivo sur un model de tumeur murine. Les cellules NK murines stimulées ex-vivo par les endocannabinoides et réinjectées en intra-tumoral, induisent un ralentissement significatif de la croissance tumorale Ces résultats montrent pour la première fois que les endocannabinoïdes AEA et 2-AG utilisées à faible dose sont impliqués dans la modulation des activités régulatrices (objectivée par la production d'IFN γ) et cytotoxiques des cellules NK.

Background: Background of this work was the investigation of the endocannabinoid system (ECS) in the Fmr1 knock- out (KO) mouse. The Fmr1- KO mouse is a mouse model for fragile X syndrome (FXS). FXS is the leading monogenic cause for autism spectrum disorders (ASD) in humans. The Fmr1- KO mouse displays autistic behavior such as an impaired social interaction, repetitive behavior, cognitive deficits, increased anxiety and aggressiveness. Alterations of the ECS have been suggested to play a key role in the etiopathology of a variety of neuropsychiatric disorders. Until today, little has been described about the involvement of the ECS in ASD. Interrogation: 1. Evaluating the manifestation of typical cannabinoid- induced effects in the Fmr1- KO mouse 2. Investigating the influenceability of autistic symptoms with THC treatment in the Fmr1- KO mouse 3. Analyzing the signaling cascade of the stimulated and unstimulated ECS in different brain regions of the Fmr1- KO mouse Material and Methods: Experiments were carried out on adult (12±1 weeks old) male Fmr1- KO and Fmr1- wild- type (WT) mice from the C57BL/6J- (B6)- background. N= 15 mice received THC (10mg/kg bodyweight) and N= 16 received WIN55,212 (3mg/kg bodyweight). 30min after injection, the body temperature was measured and the distance animals moved in an open field during 15min was recorded (locomotion). Then, animals were placed with their forepaws onto a horizontally fixed bar and the time remaining in this position (catalepsy) was measured. Finally animals were placed on a preheated plate and the temperature at which a pain stimulus occurred was determined (testing analgesia). All 4 experiments are called tetrad experiment. Afterwards changes in body temperature, locomotion, catalepsy and analgesia of the animals was evaluated. To explore long-term effects of THC after the tetrad, N= 15 animals were tested in a social interaction test with a female contact mouse, 10 and 20 days after THC treatment. Therefore, the tested mouse and the contact mouse were placed together into a cage and the time mice spent in social interaction (nose, body and anogential sniffing, allogrooming and body contact) was manually quantified during 6min of recorded testing time. Another group of N= 19 received a premedication of rimonabant (Cannabinoid- receptor 1 (CB1) antagonist, 3mg/kg bodyweight) 30min prior to THC treatment. Rimonabant prevents THC from binding to CB1 and therefore allows the assessment of the involvement of CB1 in mediating social behavior. Furthermore the suggestibility of context-dependent fear conditioning with THC treatment has been tested on N= 13 mice. Animals were placed into a conditioning chamber that delivered 6 short electric shocks with a 30sec pause to their paws (conditioning phase). Immediately afterwards mice received THC or placebo. 24h later contextdependent fear was evaluated by quantification of the time mice spent freezing in the conditioning-chamber (fear) without receiving foot shocks. Intraneuronal signaling of the ECS was analyzed with N= 29 animals using western blots. Quantities of phosphorylated (“activated”) protein kinases (ERK, AKT and S6) from different brain homogenates (hippocampus, striatum, cortex and cerebellum) were therefore measured after THC or placebo injection (30 minutes prior to sacrificing). Results: Cannabinoids induced hypothermia, hypolocomotion, analgesia and catalepsy in WTmice. These effects were significantly less detectable in Fmr1- KO mice. Effects of both cannabinoids, THC and WIN55,212, were comparable with a slightly greater but not significant efficiency of THC. THC treated WT- mice exhibited further reduced social interaction 10 days after treatment, an effect that was partially prevented by premedication with rimonabant. THC increased social interaction in Fmr1- KO mice comparable to the level of untreated WT- mice. THC had no effect on behavior of WT- mice in context-dependent fear conditioning. Fmr1- KO mice showed significant less contextdependent fear conditioning compared to WT- mice. THC facilitated the recognition of an anxiety-correlated context in Fmr1- KO mice comparable to untreated WT- mice. In western blots significant changes in the THC- induced signaling cascade were detectable and depending on genotype, brain-region and analyzed protein-kinase. In the hippocampus there were no changes in untreated Fmr1- KO mice compared to WT- mice. THC had no effect on activation of protein-kinases in WT- and Fmr1- KO mice. In the striatum there were no changes in untreated Fmr1- KO mice compared to WTmice. THC significantly increased activity of ERK, AKT and S6 in WT-mice and not in Fmr1- KO mice. In the cortex of untreated Fmr1- KO mice AKT showed a significantly increased activity compared to WT- mice. THC significantly increased AKT activity in WT- mice without having an effect on KO- mice. In the cerebellum there were no changes in untreated Fmr1- KO mice compared to WT- mice. THC significantly increased ERK- activity in Fmr1- KO mice but had no effect on protein kinase activity in WT- mice. Conclusion: We observed physiological cannabinoid effects in WT- mice after treatment with THC and WIN55,212. These effects are significantly attenuated in Fmr1- KO mice. This may be interpreted as a desensitization of the ECS in the Fmr1- KO mouse. At the same time it was demonstrated that THC has the potential to improve context dependent memory consolidation and to increase social interaction in the Fmr1- KO mouse. In particular the influence of THC on impaired social interaction should be a target of further investigations to find possible therapeutic options for this typical symptom of Autism. Underlying molecular mechanisms remain unclear and the analysis of THC stimulated intraneuronal signaling gave no clear indication of possible molecular alterations in the Fmr1- KO mouse.

Ce travail s'inscrit dans le cadre général de l'étude des mécanismes d'intégration des informations viscérales. Nous avons étudié la dépression synaptique à long terme (DLT) dans le noyau du tractus solitaire et sa modulation par le statut nutritionnel Dans la première étude, nous avons caractérisé une DLT au niveau du NTS. Cette DLT, déclenchée par la stimulation des afférences viscérales à basse fréquence, est exprimée au niveau de l’élément présynaptique. Elle met en jeu une libération d'endocannabinoïdes qui en agissant au niveau de l’élément présynaptique réduisent la probabilité de libération de glutamate. De manière surprenante l’élément postsynaptique ne semble jouer aucun rôle dans cette DLT. Elle nécessite une activation séquentielle des récepteurs NMDA, la libération d'anandamide et l'activation des récepteurs aux cannabinoïdes de type 1 (CB1) et l’activité de l’élément présynaptique. Nos résultats suggèrent que cette DLT pourrait être entièrement organisée dans le compartiment présynaptique des afférences viscérales. Dans une deuxième partie du travail, nous nous sommes intéressés à la modulation de cette DLT dépendante des endocannabinoïdes (DLT-eCBs) par le statut nutritionnel. La privation de nourriture pendant 24 h empêche l'induction de la DLT-eCBs par la stimulation des afférences viscérales. Ces effets sont mimés par l'activation des récepteurs à la ghréline au niveau du NTS. Une re-nutrition pendant 3h restaure la DLT-eCBs via l'action périphérique de la Cholécystokinine (CCK) et l'activation de la voie ERK. Au total ces travaux soulignent la forte plasticité des afférences viscérales en fonction du statut nutritionnel. / This work joins within the framework of studies about the mechanisms of integration of the visceral informations. We studied long term synaptic depression in the nucleus of tractus solitarii (NTS) and it's modulation during changes in the nutritional status. In the first study, we characterized a long-term synaptic depression (LTD) in the NTS. This LTD, triggered by low frequency stimulation of visceral afferents is expressed at the presynaptic level. It involves release of endocannabinoids that would eventually reduce glutamate release probability. Surprisingly the postsynaptic element seems to play no role in this LTD. It requires sequential activation of NMDA receptors, the release of anandamide and activation of the cannabinoids type 1 receptors (CB1) and presynaptic activation. Our results suggest that this LTD could be entirely organized at the presynaptic compartment of visceral afferents. In the second part of this work, we were interested on the modulation of this endocannabinoïds dependent long-term depression (eCBs-LTD) by the nutritional status. Food deprivation during 24 h prevents the induction of eCBs-LTD by the stimulation of visceral afferents. These effects are mimicked by the activation of ghrelin receptors in the NTS. 3 h refeeding restores the eCBs-LTD via peripheral action of cholecystokinin (CCK) and the activation of the ERK pathway. Altogether, this work emphasizes the high plasticity of visceral afferents and its regulation by the nutritional status.

Background: Background of this work was the investigation of the endocannabinoid system (ECS) in the Fmr1 knock- out (KO) mouse. The Fmr1- KO mouse is a mouse model for fragile X syndrome (FXS). FXS is the leading monogenic cause for autism spectrum disorders (ASD) in humans. The Fmr1- KO mouse displays autistic behavior such as an impaired social interaction, repetitive behavior, cognitive deficits, increased anxiety and aggressiveness. Alterations of the ECS have been suggested to play a key role in the etiopathology of a variety of neuropsychiatric disorders. Until today, little has been described about the involvement of the ECS in ASD. Interrogation: 1. Evaluating the manifestation of typical cannabinoid- induced effects in the Fmr1- KO mouse 2. Investigating the influenceability of autistic symptoms with THC treatment in the Fmr1- KO mouse 3. Analyzing the signaling cascade of the stimulated and unstimulated ECS in different brain regions of the Fmr1- KO mouse Material and Methods: Experiments were carried out on adult (12±1 weeks old) male Fmr1- KO and Fmr1- wild- type (WT) mice from the C57BL/6J- (B6)- background. N= 15 mice received THC (10mg/kg bodyweight) and N= 16 received WIN55,212 (3mg/kg bodyweight). 30min after injection, the body temperature was measured and the distance animals moved in an open field during 15min was recorded (locomotion). Then, animals were placed with their forepaws onto a horizontally fixed bar and the time remaining in this position (catalepsy) was measured. Finally animals were placed on a preheated plate and the temperature at which a pain stimulus occurred was determined (testing analgesia). All 4 experiments are called tetrad experiment. Afterwards changes in body temperature, locomotion, catalepsy and analgesia of the animals was evaluated. To explore long-term effects of THC after the tetrad, N= 15 animals were tested in a social interaction test with a female contact mouse, 10 and 20 days after THC treatment. Therefore, the tested mouse and the contact mouse were placed together into a cage and the time mice spent in social interaction (nose, body and anogential sniffing, allogrooming and body contact) was manually quantified during 6min of recorded testing time. Another group of N= 19 received a premedication of rimonabant (Cannabinoid- receptor 1 (CB1) antagonist, 3mg/kg bodyweight) 30min prior to THC treatment. Rimonabant prevents THC from binding to CB1 and therefore allows the assessment of the involvement of CB1 in mediating social behavior. Furthermore the suggestibility of context-dependent fear conditioning with THC treatment has been tested on N= 13 mice. Animals were placed into a conditioning chamber that delivered 6 short electric shocks with a 30sec pause to their paws (conditioning phase). Immediately afterwards mice received THC or placebo. 24h later contextdependent fear was evaluated by quantification of the time mice spent freezing in the conditioning-chamber (fear) without receiving foot shocks. Intraneuronal signaling of the ECS was analyzed with N= 29 animals using western blots. Quantities of phosphorylated (“activated”) protein kinases (ERK, AKT and S6) from different brain homogenates (hippocampus, striatum, cortex and cerebellum) were therefore measured after THC or placebo injection (30 minutes prior to sacrificing). Results: Cannabinoids induced hypothermia, hypolocomotion, analgesia and catalepsy in WTmice. These effects were significantly less detectable in Fmr1- KO mice. Effects of both cannabinoids, THC and WIN55,212, were comparable with a slightly greater but not significant efficiency of THC. THC treated WT- mice exhibited further reduced social interaction 10 days after treatment, an effect that was partially prevented by premedication with rimonabant. THC increased social interaction in Fmr1- KO mice comparable to the level of untreated WT- mice. THC had no effect on behavior of WT- mice in context-dependent fear conditioning. Fmr1- KO mice showed significant less contextdependent fear conditioning compared to WT- mice. THC facilitated the recognition of an anxiety-correlated context in Fmr1- KO mice comparable to untreated WT- mice. In western blots significant changes in the THC- induced signaling cascade were detectable and depending on genotype, brain-region and analyzed protein-kinase. In the hippocampus there were no changes in untreated Fmr1- KO mice compared to WT- mice. THC had no effect on activation of protein-kinases in WT- and Fmr1- KO mice. In the striatum there were no changes in untreated Fmr1- KO mice compared to WTmice. THC significantly increased activity of ERK, AKT and S6 in WT-mice and not in Fmr1- KO mice. In the cortex of untreated Fmr1- KO mice AKT showed a significantly increased activity compared to WT- mice. THC significantly increased AKT activity in WT- mice without having an effect on KO- mice. In the cerebellum there were no changes in untreated Fmr1- KO mice compared to WT- mice. THC significantly increased ERK- activity in Fmr1- KO mice but had no effect on protein kinase activity in WT- mice. Conclusion: We observed physiological cannabinoid effects in WT- mice after treatment with THC and WIN55,212. These effects are significantly attenuated in Fmr1- KO mice. This may be interpreted as a desensitization of the ECS in the Fmr1- KO mouse. At the same time it was demonstrated that THC has the potential to improve context dependent memory consolidation and to increase social interaction in the Fmr1- KO mouse. In particular the influence of THC on impaired social interaction should be a target of further investigations to find possible therapeutic options for this typical symptom of Autism. Underlying molecular mechanisms remain unclear and the analysis of THC stimulated intraneuronal signaling gave no clear indication of possible molecular alterations in the Fmr1- KO mouse.