PHARMACOLOGICAL MODULATION OF LACTATE LEVELS AFFECTS THE FORMATION OF LONG-TERM MEMORY AND THE STRUCTURE OF EXCITATORY SYNAPSES IN MICE HIPPOCAMPUS

The block of glycogenolysis in the rat hippocampus impairs memory while L-lactate (but not equicaloric concentrations of glucose) rescues amnesia, suggesting that the metabolic coupling between astrocytes and neurons via lactate is required for long-term memory formation. The inactivation of glycogen metabolism and the consequent decrease of lactate production in hippocampal astrocytes leads to an impairment of long-term memory and defects in long-term potentiation (LTP). Since it has been demonstrated that LTP defects are associated with alterations in dendritic spines morphology and density, we studied whether anatomical differences in hippocampal excitatory synapses were related to defects in long-term memory in mice treated with the inhibitor of glycogen phosphorylase, 1,4-dideoxy-1,4-imino-D-arabinitol (DAB), and if these defects were rescued by L-lactate administration. We first tested the effects associated to DAB administration on long-term memory using the passive avoidance (PA) task, a fear-aggravated test used to evaluate learning and memory in rodents. DAB was injected 15 minutes before training and long-term memory was tested 24 hours after the administration of the drug. The analysis of the PA test shows that DAB did not affect long-term memory. We then performed the novel object recognition (NOR) test to study intermediate- and long-term episodic memory after DAB injection into the hippocampus and here we showed that this treatment affected both intermediate- and long-term episodic memory, in a dose dependent manner. We performed low resolution analyses of spine density in apical dendrites from CA1 pyramidal neurons by means of the Golgi-Cox staining technique, 24 hours after training of mice treated with DAB and challenged with both PA and NOR tasks, to determine whether behavioural impairments are paralleled by defects on dendritic spine density. Our data demonstrated that 24 hours after treatment with DAB and following memory training, hippocampal neurons showed a marked reduction in dendritic spine density if compared to that of mice injected with vehicle, regardless of the memory paradigm used. To evaluate if this defect in spine density was associated to ultrastructural defects in hippocampal excitatory synapses, we performed a morphometric analysis on 2D transmission electron microscopy projections. These analyses highlighted a significant reduction of vesicle density in synapses of mice treated with DAB and an unvaried pre-synaptic surface. Interestingly, despite the effect of DAB on spine density, the post-synaptic density (PSD) architecture is unaffected as length and thickness were the same in animals treated with vehicle or DAB. We then analysed the astrocyte-neuron lactate shuttle’s role on the 3D architecture of excitatory synapses and mitochondria as a consequence of DAB administration. These data confirmed the observations obtained with the Golgi and the 2D analyses: the density of the dendritic spines was reduced and the PSD volume remained unvaried in mice treated with DAB. Altered brain metabolism is associated with changes in mitochondria dynamic cycles of fission and fusion. The remodelling of mitochondrial cristae and mitochondrial morphology are sensitive to the energetic demand of the cell, and they directly affects the development and maintenance of synapses. We analysed the mitochondrial structure in order to see whether the block of glycogen metabolism in astrocytes could affect the architecture of these organelles in neurons. We observed a significant change in dendritic mitochondria number, shape, and volume in DAB treated mice. Compared to uniformly compact and well-separated mitochondria in the hippocampi of mice treated with vehicle, DAB-treated neurons exhibited elongated mitochondria, in which two or more units were interconnected by tubular membrane extensions. Dendritic mitochondria in DAB treated mice were longer, reduced in number and bigger compared to dendritic mitochondria of mice treated with vehicle, suggesting that the alteration of glycogen metabolism induced by DAB could be the basis of mitochondria dynamics changes. L-lactate, was administered to investigate its ability to revert the DAB-induced memory impairment. Co-injection of DAB and L-lactate was able to revert the memory impairment induced by DAB, 24 hours after NOR training. Besides behavioural rescue, the Golgi-Cox staining showed that in mice treated with DAB and L-lactate there was a rescue of spine density in apical dendrites of CA1 pyramidal neurons, bringing back its value to that of mice treated with vehicle. Focusing our attention on the ultrastructure of excitatory synapses, we measured the pre-synaptic surface and vesicle density of these samples and we showed that the vesicle density was comparable to that found in the synapses of mice treated only with DAB. Our results suggest that, bilateral hippocampal injection of DAB caused an impairment of long-term memory formation 24 hours after NOR test. This behavioural alteration was accompanied by a strong morphological alteration of excitatory hippocampal synapses and dendritic mitochondria. L-lactate is able to rescue the behavioural phenotype, the reduction in spine density and the morphological alterations of dendritic mitochondria induced by this drug, but not the pre-synaptic effect induced by DAB on excitatory hippocampal synapses.