Ultrastructural and physiological studies on the cannabinoid CB1 receptor localized in astroglia

Résumé

The cannabinoid CB1 receptor-mediated functions in astrocytes are highly dependent on the CB1 receptordistribution in these glial cells relative to neuronal sites, particularly at the nearby synapses under normalor pathological conditions. However, the whole picture of the subcellular CB1 receptor distribution inastroglial compartments remains uncompleted due to the scattered CB1 receptor expression, and thereforedifficult to detect, in astrocytes. Our laboratory has in previous studies estimated that about 5-6 % of thetotal CB1 receptors in the hippocampal CA1 stratum radiatum are localized in astrocytes identified by themarker glial fibrillary acidic protein (GFAP). However, GFAP is a cytoskeleton protein mostly restrictedto the astroglial cell bodies and their main branches. This might be distorting the actual proportion andtotal amount of CB1 receptors in astrocytes. Therefore, the search for alternative astroglial markers todecipher the precise mapping of CB1 receptors in astrocytes is a timely goal in the cannabinoid field. Theglutamate aspartate transporter (GLAST) is used as astroglial marker and raises as a good astroglialmarker candidate to study in detail the CB1 receptor distribution in astrocytes.To prove this hypothesis, I have used a pre-embedding immunogold method for electron microscopy tocompare first the astroglial distribution of GLAST versus GFAP. GLAST labeling was along the plasmamembrane of astrocytes, including cell bodies and the smallest astrocytic projections in close contact withneurons, capillaries and other glial cells, covering altogether a much broader labeled area than GFAP.Furthermore, the use of a pre-embedding immunoperoxidase method for electron microscopy served meto assess that almost three times more astroglial area is visualized with GLAST than GFAP, and thatGLAST detects four times as much astroglial membranes as GFAP. Finally, a double pre-embeddingimmunogold/immunoperoxidase method allowed to estimate that about 12 % of the total CB1 receptorparticles are localized in GLAST-positive astrocytes, but the value drops to 5-6 % in GFAP-positiveastrocytes, as published previously by our laboratory.Once these findings were obtained, I studied in more detail the CB1 receptor localization in astroglialmitochondria. We discovered the presence of functional CB1 receptors in mitochondrial membranes ofhippocampal neurons and astrocytes. Accordingly, I used double GLAST-CB1 immunolabeling toanalyze in the electron microscope the density of mitochondrial CB1 receptors in neurons and astrocytesin four brain regions: CA1 hippocampus, prefrontal cortex, piriform cortex and nucleus accumbens. Theresults showed that the CB1 receptor density in astrocytic mitochondria is higher than in neuronalmitochondria. Altogether, despite the lower absolute levels of CB1 receptors in astrocytes than inneurons, the density of mitochondrial CB1 receptors in astrocytes is higher than in neurons in the fourbrain regions studied. Namely, CB1 receptors are more expressed in astroglial than neuronalmitochondria. Activation of mitochondrial CB1 receptors alters energy production in neurons and cancause memory impairment. Likewise, activation of mitochondrial CB1 receptors in astrocytes interfereswith glucose metabolism and lactate production, disrupting neuronal functions and social behavior.In conclusion, the findings that astrocytes of the four brain regions studied contain more CB1 receptors intheir mitochondria than the neuronal mitochondria, and that the cannabinoid-induced reduction of oxygenconsumption is absent in mitochondria isolated from the forebrain of GFAP-CB1-KO mice, suggest thatmitochondrial CB1 receptors in astrocytes play a crucial role in the global effects of cannabinoids onbrain mitochondrial respiration. // The cannabinoid CB1 receptor-mediated functions in astrocytes are highly dependent on the CB1 receptordistribution in these glial cells relative to neuronal sites, particularly at the nearby synapses under normalor pathological conditions. However, the whole picture of the subcellular CB1 receptor distribution inastroglial compartments remains uncompleted due to the scattered CB1 receptor expression, and thereforedifficult to detect, in astrocytes. Our laboratory has in previous studies estimated that about 5-6 % of thetotal CB1 receptors in the hippocampal CA1 stratum radiatum are localized in astrocytes identified by themarker glial fibrillary acidic protein (GFAP). However, GFAP is a cytoskeleton protein mostly restrictedto the astroglial cell bodies and their main branches. This might be distorting the actual proportion andtotal amount of CB1 receptors in astrocytes. Therefore, the search for alternative astroglial markers todecipher the precise mapping of CB1 receptors in astrocytes is a timely goal in the cannabinoid field. Theglutamate aspartate transporter (GLAST) is used as astroglial marker and raises as a good astroglialmarker candidate to study in detail the CB1 receptor distribution in astrocytes.To prove this hypothesis, I have used a pre-embedding immunogold method for electron microscopy tocompare first the astroglial distribution of GLAST versus GFAP. GLAST labeling was along the plasmamembrane of astrocytes, including cell bodies and the smallest astrocytic projections in close contact withneurons, capillaries and other glial cells, covering altogether a much broader labeled area than GFAP.Furthermore, the use of a pre-embedding immunoperoxidase method for electron microscopy served meto assess that almost three times more astroglial area is visualized with GLAST than GFAP, and thatGLAST detects four times as much astroglial membranes as GFAP. Finally, a double pre-embeddingimmunogold/immunoperoxidase method allowed to estimate that about 12 % of the total CB1 receptorparticles are localized in GLAST-positive astrocytes, but the value drops to 5-6 % in GFAP-positiveastrocytes, as published previously by our laboratory.Once these findings were obtained, I studied in more detail the CB1 receptor localization in astroglialmitochondria. We discovered the presence of functional CB1 receptors in mitochondrial membranes ofhippocampal neurons and astrocytes. Accordingly, I used double GLAST-CB1 immunolabeling toanalyze in the electron microscope the density of mitochondrial CB1 receptors in neurons and astrocytesin four brain regions: CA1 hippocampus, prefrontal cortex, piriform cortex and nucleus accumbens. Theresults showed that the CB1 receptor density in astrocytic mitochondria is higher than in neuronalmitochondria. Altogether, despite the lower absolute levels of CB1 receptors in astrocytes than inneurons, the density of mitochondrial CB1 receptors in astrocytes is higher than in neurons in the fourbrain regions studied. Namely, CB1 receptors are more expressed in astroglial than neuronalmitochondria. Activation of mitochondrial CB1 receptors alters energy production in neurons and cancause memory impairment. Likewise, activation of mitochondrial CB1 receptors in astrocytes interfereswith glucose metabolism and lactate production, disrupting neuronal functions and social behavior.In conclusion, the findings that astrocytes of the four brain regions studied contain more CB1 receptors intheir mitochondria than the neuronal mitochondria, and that the cannabinoid-induced reduction of oxygenconsumption is absent in mitochondria isolated from the forebrain of GFAP-CB1-KO mice, suggest thatmitochondrial CB1 receptors in astrocytes play a crucial role in the global effects of cannabinoids onbrain mitochondrial respiration.