Hypothermia therapy (HT) is used clinically following global cerebral ischemia (GCI) but its therapeutic mechanisms are not completely understood. An elucidation of such mechanisms may lead to novel therapeutic approaches that improve patient outcome. Using a murine model of GCI, we determined the effect of HT on the expression of inflammatory proteins in the hippocampus and serum. We also examined its effect on microglia/macrophage activation and neurodegeneration in the brain at 72 hours following ischemia, and its effect on long-term spatial memory/learning and contextual fear response. GCI led to increased neurodegeneration and microglia/macrophage activation in the hippocampus, and increased IL-1β and KC protein expression in the hippocampus at 72 hours. Hypothermia therapy attenuated these inflammatory responses. It also improved spatial learning/memory at 7 and 21 days, and preserved contextual fear response 21 days post-ischemia. Hypothermia therapy attenuated the post-ischemic inflammatory response, protected hippocampal neurons, and preserved long-term memory and learning.

Alzheimer’s disease is an insidious neurodegenerative disease that affects millions of people worldwide. Currently, there are no determinants that can accurately predict the onset cognitive decline in AD. This thesis investigates and defines changes in the lipidome that are linked to symptomatic onset and cognitive impairment in mouse models of AD. Using a targeted lipidomic approach employing high performance liquid chromatography electrospray ionization tandom mass spectrometry, direct biochemical assessments, and behavioural evaluation, I was able to (a) profile and quantify cortical and hippocampal glycerophosphocholine and glycerophosphoethanolamine metabolites and signaling molecules in the APPSwe/PS1dE9 and the N5 TgCRND8 murine models of AD and (b) associate changes in lipid metabolism with learning and memory impairment. I demonstrate that glycerophosphocholine metabolism in the cortex but not the hippocampus is altered at symptomatic onset in both mouse models. These same metabolic changes were seen in younger animals exposed to chronic intermittent hypoxia, an environmental risk factor that accelerates their phenoconversion. In fully impaired transgenic mice, I defined metabolic changes associated with disease progression. To further assess the impact of sex, another risk factor of Alzheimer’s disease cognitive decline, I characterized an AD model of sex-specific cognitive resistance. I demonstrated that transgenic males but not females exhibit behavioural indices of cognitive reserve when tested in the Morris Water Maze. Using this mouse line, I then investigated how measures of learning and memory associated with glycerophosphocholine and glycerophosphoethanolamine metabolism. I identified increases in critical glycerophosphoethanolamine metabolites linked to spatial learning and memory impairment in the cortex of N5 TgCNRD8 mice and demonstrated that these changes could be predicted by profiling the plasma glycerophosphoethanolamine lipidome. Taken together, this thesis links glycerophospholipid metabolism to the onset and progression of learning and memory impairment in experimental models of AD and provides the first evidence that changes in cortical lipid metabolism can be predicted by changes in the plasma lipidome.

A growing body of evidence implicates the importance of the endogenous cannabinoid 2-arachidonyl glycerol (2-AG) in memory regulation. The biosynthesis of 2-AG occurs primarily through the diacylglycerol lipases (DAGL-α and -β), with 2-AG serving as a bioactive lipid to both activate cannabinoid receptors and as a rate limiting precursor for the production of arachidonic acid and subsequent pro-inflammatory mediators. Gene deletion of DAGL-α shows decrements in synaptic plasticity and hippocampal neurogenesis suggesting this biosynthetic enzyme may be important for processes of normal spatial memory. Additionally, 2-AG is elevated in response to pathogenic events such as traumatic brain injury (TBI), suggesting its regulatory role may extend to conditions of neuropathology. As such, this dissertation investigates the in vivo role of DAGL-α and -β to regulate spatial learning and memory in the healthy brain and following neuropathology (TBI). The first part of this dissertation developed a mouse model of learning and memory impairment following TBI, using hippocampal-dependent tasks of the Morris water maze (MWM). We found modest, but distinct differences in MWM performance between left and right unilateral TBI despite similar motor deficits, histological damage, and glial reactivity. These findings suggest that laterality in mouse MWM deficit might be an important consideration when modeling TBI-induced functional consequences. The second part of this dissertation work evaluated DAGL-β as a target to protect against TBI-induced learning and memory deficit given its selective expression on microglia and the role of 2-AG as a precursor for eicosanoid production. The gene deletion of DAGL-β did not protect against TBI-induced MWM or motor deficits, but unexpectedly produced a survival protective phenotype. These findings suggest that while DAGL-β does not contribute to injury-induced memory deficit, it may contribute to TBI-induced mortality. The third and final set of experiments investigated the role of DAGL-α in mouse spatial learning and memory under physiological conditions (given the predominantly neuronal expression of DAGL-α). Complementary pharmacological and genetic manipulations produced task specific impaired MWM performance, as well as impaired long-term potentiation and alterations to endocannabinoid lipid levels. These results suggest that DAGL-α may play a selective role in the integration of new spatial information in the normal mouse brain. Overall, these data point to DAGL-α, but not DAGL-β, as an important contributor to hippocampal-dependent learning and memory. In contrast, DAGL-β may contribute to TBI-induced mortality.

Acid-sensing ion channel 1A (ASIC1A) is abundant in the nucleus accumbens (NAc), a region known for its role in addiction. Because ASIC1A has been previously suggested to promote associative learning, we hypothesized that disrupting ASIC1A in the NAc would reduce drug-associated learning and memory. However, contrary to this hypothesis, we found that disrupting ASIC1A in the NAc increased cocaine-conditioned place preference, suggesting an unexpected role for ASIC1A in addiction-related behavior. Investigating the underlying mechanisms, we identified a novel postsynaptic current during neurotransmission mediated by ASIC1A and ASIC2 and thus well-positioned to regulate synapse structure and function. Consistent with this possibility, disrupting ASIC1A altered dendritic spine density and glutamate receptor function, and increased cocaine-evoked plasticity in AMPA-to-NMDA ratio, all resembling changes previously associated with cocaine-induced behavior. Together, these data suggest ASIC1A inhibits plasticity underlying addiction-related behavior, and raise the possibility of therapies for drug addiction by targeting ASIC-dependent neurotransmission. The amygdala plays critical roles in the learning and expression of fear-related behavior. Previous studies have implicated the amygdala in CO2-evoked fear-like behavior in mice; however, a more recent study demonstrated that humans lacking the amygdala bilaterally experience fear and panic with CO2-inhalation. Because all subjects lacking the amygdala had panic attacks after inhaling CO2 compared to only 25% of controls, this data suggests the amygdala may play an inhibitory role in CO2-evoked panic. To assess the role of the amygdala in CO2-evoked behaviors in mice, we lesioned the amygdala and optogenetically stimulated different amygdalar nuclei. We found that large unilateral and bilateral amygdala lesions caused the emergence of escape-like jumping behavior in mice exposed to CO2 and a relative deficit in CO2-evoked freezing. This jumping behavior depended on the dorsal periaqueductal gray, a brain area previously associated with panic attacks. Additionally, the putative CO2 chemosensor ASIC1A and ASIC2 are not necessary for CO2-evoked jumping, and may even play an inhibitory role in this behavior. Optogenetic manipulation of the amygdala revealed that stimulation of the basolateral amygdala enhanced jumping behavior and inhibited freezing behavior. This may be due to the basolateral amygdala's ability to inhibit the main output center of the amygdala, the central nucleus. Together, these results suggest that different amygdalar nuclei differentially modulate CO2-evoked behavior by regulating the switch between mobile and immobile defense responses. Additionally, they provide additional evidence that amygdalar dysfunction may contribute to panic disorder.

We have recently established that aberrant alkylacylglycerophosphocholine metabolism results in the increased tissue concentration of platelet activating factors (PAFs) in the temporal cortex of Alzheimer Disease (AD) patients and in TgCRND8 mice over-expressing mutant human amyloid precursor protein. PAF lipids activate a G-protein coupled receptor (PAFR) reported to be expressed by microglia and subsets of neurons in rat. It is not known whether this same expression pattern is recapitulated in mice however, as the expression has only been inferred by use of pharmacological PAFR antagonists, many of which impact on both PAFR-dependent and PAFR-independent signalling pathways. PAFR plays a role in long term potentiation (LTP) induction in rats. PAFR has also been implicated in behavioural indices of spatial learning and memory in rats. Contradictory reports using mice provide ambiguity regarding the role of PAFR in LTP induction in mice. To assess whether PAFR is expressed in murine neurons, I localized PAFR mRNA in wild-type C57BL/6 mice using PAFR KO mice as a negative control. I further showed that the loss of PAFR did not impair learning and memory although this assessment must be considered preliminary as the behavioural test employed was not optimized to detect changes in learning and memory of C57BL/6 mice over time adequately.Finally, I showed that the loss of PAFR in TgCRND8 mouse model of AD had no impact upon Aβ plaque number. My observations suggest that PAFR is restricted to microglial-like cells in mouse hippocampus and as such, it may not play a role in learning and memory.

Studying protein-protein interactions has been vital for understanding how proteins function within the cell, how biological processes are strictly regulated by these interactions, and what molecular mechanisms underlie cellular functions and diseases. Recent biochemical and biophysical studies have provided evidence supporting that G protein-coupled receptors (GPCRs) can and do interact with one another to form dimers or larger oligomeric complexes, which may determine the structure and function of GPCRs, including receptor trafficking, scaffolding and signaling. This may help to understand the physiological roles of GPCRs and mechanisms underlying certain disease pathologies and to provide an alternative approach for drug intervention.<p>Cannabinoid CB1 and dopamine D2 receptors are the most common GPCRs in the brain and exert a mutual regulation in brain functions involved in learning, memory and drug addiction. There is structural and functional evidence supporting the idea that CB1 and D2 receptors physically interact with each other in hippocampal and striatal neurons to modulate their functions. Direct evidence supporting a physical interaction between the CB1 and D2 receptors was obtained from cultured HEK293 cells stably coexpressed with both receptors.<p> This research project was designed to critically test the hypothesis that a specific protein sequence (i.e. motif) in the D2 receptor is responsible for in vitro protein-protein interactions between the CB1 and D2 receptors. To reach this goal, fusion proteins containing various domains and motifs of the CB1 and D2 receptors were prepared and then used first to determine the domains of the CB1 and D2 receptors responsible for in vitro protein-protein interactions between CB1 and D2 receptors, and then to identify the specific motifs in the D2 receptor responsible for in vitro CB1 coupling with the D2 receptors. The major method used in this study is in vitro pull-down assay, which uses a purified and tagged bait protein to generate a specific affinity support that is able to bind and purify a prey protein from a lysate sample. The present study provides the first evidence that CB1 intracellular C-terminal (CB1-CT) and D2 intracellular loop 3 (D2-IL3) can directly interact with each other, and that the specific motifs D2-IL3(Ⅳ1) and D2-IL3(Ⅳ3) in the D2 receptor are likely responsible for their in vitro coupling with the CB1 receptors. <p>The results of the present study are invaluable for future research exploring in vivo protein-protein interaction between the CB1 and D2 receptors in the rat striatum by co-immunoprecipitation. Specifically, future studies will determine whether the identified specific motifs D2-IL3(Ⅳ1) and D2-IL3(Ⅳ3) in the D2 receptor are indeed critical for their in vivo coupling with the CB1 receptors.

We have recently established that aberrant alkylacylglycerophosphocholine metabolism results in the increased tissue concentration of platelet activating factors (PAFs) in the temporal cortex of Alzheimer Disease (AD) patients and in TgCRND8 mice over-expressing mutant human amyloid precursor protein. PAF lipids activate a G-protein coupled receptor (PAFR) reported to be expressed by microglia and subsets of neurons in rat. It is not known whether this same expression pattern is recapitulated in mice however, as the expression has only been inferred by use of pharmacological PAFR antagonists, many of which impact on both PAFR-dependent and PAFR-independent signalling pathways. PAFR plays a role in long term potentiation (LTP) induction in rats. PAFR has also been implicated in behavioural indices of spatial learning and memory in rats. Contradictory reports using mice provide ambiguity regarding the role of PAFR in LTP induction in mice. To assess whether PAFR is expressed in murine neurons, I localized PAFR mRNA in wild-type C57BL/6 mice using PAFR KO mice as a negative control. I further showed that the loss of PAFR did not impair learning and memory although this assessment must be considered preliminary as the behavioural test employed was not optimized to detect changes in learning and memory of C57BL/6 mice over time adequately.Finally, I showed that the loss of PAFR in TgCRND8 mouse model of AD had no impact upon Aβ plaque number. My observations suggest that PAFR is restricted to microglial-like cells in mouse hippocampus and as such, it may not play a role in learning and memory.

La facilitación del aprendizaje y la memoria por la autoestimulación eléctrica intracraneal (AEIC) se ha observado principalmente en tareas de memoria implícita, como la evitación activa de dos sentidos, en ratas. En este trabajo pretendemos investigar si la AEIC post-entrenamiento, además de facilitar el aprendizaje y la formación de la memoria en paradigmas de memoria implícita, puede facilitar también el aprendizaje y la formación de la memoria en un paradigma de memoria espacial explícita o relacional, hipocampo-dependiente, en el Laberinto Acuático de Morris (LAM), en ratas Wistar. Dado que nuestra investigación previa ha puesto de manifiesto que el efecto de la AEIC para facilitar el aprendizaje y la memoria podría ser sensible a la capacidad básica o inicial de los sujetos para aprender las tareas, en el presente trabajo hemos realizado 4 experimentos consecutivos con ratas Wistar que han sido diseñados para dificultar progresivamente el aprendizaje en el LAM reduciendo, en cada siguiente experimento, la cantidad de ensayos de entrenamiento administrados. En cada experimento, los sujetos se dividieron en dos grupos, grupo AEIC y grupo Control y realizaron, en función del experimento, 8, 5, 3 o 1 ensayo diario. Inmediatamente después de cada sesión de entrenamiento los sujetos del grupo AEIC recibieron el tratamiento de AEIC (2000 trenes de corriente reforzante y estimulante a través de un electrodo crónicamente implantado en el hipotálamo lateral derecho de los sujetos) mientras que los sujetos Control no recibieron tratamiento alguno. Los ensayos tenían una duración de 2 minutos y se iniciaban desde una posición que variaba de uno a otro. Los animales debían nadar en la piscina hasta localizar una plataforma escondida que estaba señalizada por una pelota de playa colocada en la periferia de la piscina en el cuadrante derecho adyacente a la plataforma (experimento 1) o en el cuadrante opuesto al de la plataforma (experimentos 2, 3 y 4). Durante el entrenamiento en esta tarea de plataforma-escondida, registramos la latencia de nado y otras variables como la velocidad media de nado, el tiempo en paredes y la distancia total nadada de cada sujeto. Tres días después de la última sesión de entrenamiento, los sujetos realizaron el ensayo de prueba que consistió en dejar nadar al animal durante 60 segundos con la pelota presente pero sin la plataforma. En este ensayo de prueba registramos la media de tiempo que cada sujeto pasó en cada cuadrante así como las trayectorias de nado. A lo largo de las sesiones de entrenamiento observamos una fuerte y consistente facilitación de la ejecución en el LAM en aquellas ratas que habían recibido el tratamiento de autoestimulación tras un único ensayo de adquisición, es decir, cuando el aprendizaje fue más difícil (experimento 4). En este mismo experimento, durante el ensayo de prueba, los sujetos AEIC nadaron significativamente más tiempo en el cuadrante donde debería haberse encontrado la plataforma que en el resto de zonas de la piscina. Esta facilitación también se observó en la última sesión y en el ensayo de prueba de aquellas ratas que habían realizado tres ensayos por sesión (experimento 3). No obstante, no observamos diferencias entre los sujetos tratados y los controles cuando realizaron 8 o 5 ensayos diarios por sesión (experimentos 1 y 2 respectivamente). Estos hallazgos apoyan y confirman resultados previos de nuestro laboratorio en los que, como hemos comentado anteriormente, observamos que la capacidad de la AEIC post-entrenamiento para facilitar la consolidación de la memoria ha sido y es mayor en sujetos con bajos niveles iniciales de condicionamiento y claramente prueban que este tratamiento es capaz de facilitar la memoria explícita, relacional o hipocampo dependiente. / Learning and memory improvement by post-training intracranial self-stimulation has been observed mostly in implicit tasks, such as active avoidance, in rats. Here we wanted to know whether post-training self-stimulation is also able to facilitate a spatial hippocampus-dependent task in the Morris water maze. Four experiments were run with Wistar rats. In each of them subjects were given at least 5 acquisition sessions, one daily, consisting of two-minute trials. Starting from a variable position, rats had to swim in a pool until they located a hidden platform signalled by a cue located on its opposite site. Each daily session was followed by an immediate treatment of intracranial self-stimulation. Control subjects did not receive the self-stimulation treatment and were placed in their home cage. In the three successive experiments, independent groups of rats were given five, three and one trial per session, respectively. Temporal latencies and trajectories to locate the platform were measured for each subject. Three days after the last acquisition session, the animals were placed again in the pool, for 60 sec, but without the platform and the time spent in each quadrant and the swim trajectories were registered for each subject. A strong and consistent improvement of performance was observed in the self-stimulated rats when they were given only one trial per session, i.e. when learning was more difficult. These findings agree with our previous data showing the capacity of post-training self-stimulation to improve memory especially in rats with low conditioning levels (little training), and clearly prove that post-training self-stimulation can also improve explicit or relational memory.

Stress affects virtually all organisms and can result in both physiological and behavioral changes. Conditioned defeat in Syrian hamsters is a model of stress-induced behavioral plasticity that occurs in a social context. In this model, hamsters are defeated by a larger, more aggressive counterpart. Defeated hamsters subsequently fail to defend their own territory and show striking and long-lasting increases in submissive behavior even when paired with a non-threatening counterpart. The present series of experiments seeks to identify the brain regions and molecular mediators that contribute to this behavioral plasticity. One brain region that has been overlooked by our laboratory is the hippocampus. The results of the first study suggested that the ventral, but not dorsal, hippocampus is important for the acquisition of conditioned defeat as temporary inactivation of the ventral hippocampus prior to defeat training significantly reduced submissive and defensive behaviors when hamsters were tested with a non-aggressive intruder. Next, we sought to identify a potential molecular mediator of social stress-induced behavioral plasticity in hamsters identified as winners or losers after a fight. Using in situ hybridization for brain-derived neurotrophic factor (BDNF) mRNA, we showed that winning and losing hamsters exhibited differences in BDNF mRNA in several regions including the basolateral and medial amygdala as well as the dentate gyrus of the dorsal hippocampus and CA1 of the ventral hippocampus. We next showed that neurotrophic activity in the basolateral amygdala is important for the acquisition of conditioned defeat because K252a infused into the basolateral amygdala prior to defeat training by an aggressive counterpart, significantly decreased submissive and defensive behavior during subsequent testing. Finally, existing data suggest that the amygdala and hippocampus interact to modulate the formation of emotional memories. To test the hypothesis that the basolateral amygdala and ventral hippocampus interact to mediate the behavioral plasticity observed in conditioned defeat, we simultaneously inactivated these regions either contralaterally or ipsilaterally prior to social defeat. Our results suggest that BLA and VHPC interact to mediate the acquisition of conditioned defeat, however, the nature of this interaction remains to be determined.

Studying protein-protein interactions has been vital for understanding how proteins function within the cell, how biological processes are strictly regulated by these interactions, and what molecular mechanisms underlie cellular functions and diseases. Recent biochemical and biophysical studies have provided evidence supporting that G protein-coupled receptors (GPCRs) can and do interact with one another to form dimers or larger oligomeric complexes, which may determine the structure and function of GPCRs, including receptor trafficking, scaffolding and signaling. This may help to understand the physiological roles of GPCRs and mechanisms underlying certain disease pathologies and to provide an alternative approach for drug intervention.<p>Cannabinoid CB1 and dopamine D2 receptors are the most common GPCRs in the brain and exert a mutual regulation in brain functions involved in learning, memory and drug addiction. There is structural and functional evidence supporting the idea that CB1 and D2 receptors physically interact with each other in hippocampal and striatal neurons to modulate their functions. Direct evidence supporting a physical interaction between the CB1 and D2 receptors was obtained from cultured HEK293 cells stably coexpressed with both receptors.<p> This research project was designed to critically test the hypothesis that a specific protein sequence (i.e. motif) in the D2 receptor is responsible for in vitro protein-protein interactions between the CB1 and D2 receptors. To reach this goal, fusion proteins containing various domains and motifs of the CB1 and D2 receptors were prepared and then used first to determine the domains of the CB1 and D2 receptors responsible for in vitro protein-protein interactions between CB1 and D2 receptors, and then to identify the specific motifs in the D2 receptor responsible for in vitro CB1 coupling with the D2 receptors. The major method used in this study is in vitro pull-down assay, which uses a purified and tagged bait protein to generate a specific affinity support that is able to bind and purify a prey protein from a lysate sample. The present study provides the first evidence that CB1 intracellular C-terminal (CB1-CT) and D2 intracellular loop 3 (D2-IL3) can directly interact with each other, and that the specific motifs D2-IL3(Ⅳ1) and D2-IL3(Ⅳ3) in the D2 receptor are likely responsible for their in vitro coupling with the CB1 receptors. <p>The results of the present study are invaluable for future research exploring in vivo protein-protein interaction between the CB1 and D2 receptors in the rat striatum by co-immunoprecipitation. Specifically, future studies will determine whether the identified specific motifs D2-IL3(Ⅳ1) and D2-IL3(Ⅳ3) in the D2 receptor are indeed critical for their in vivo coupling with the CB1 receptors.