Abstract :
[en] Nowadays, there is evidence that brain glucose metabolism and Alzheimer's disease (AD) are linked (1). Patients suffering from type II diabetes present a higher risk to develop AD while in AD patients but also in preclinical stage (MCI) the brain glucose metabolism is reduced, leading to a general hypometabolism (2). It is therefore very important to better understand the link between brain utilization of glucose and AD. On the other hand, while beta-amyloid aggregates are one of the principal hallmarks of the disease, all strategies targeting these aggregates have failed until now to prove their efficiency. Targeting the amyloid precursor protein (APP) itself and its role in brain metabolism could bring some new insights and lead to novel therapeutic strategies. Our hypothesis is that APP is involved in energy flux between the body and the brain. During ageing or in case of pathology such as AD (2, 3, 4), Down Syndrome (5) and insulin resistance (6), glucose availability can be reduced in the brain leading to a compensatory increase in the expression of APP. This compensatory increase could be the starting point of disruption of metabolic and neurotransmitter homeostasis leading to cognitive deficit (3, 4). The aim of this project is to better understand the link between APP expression and brain glucose metabolism and its impact on neuronal activity and synaptic connections.
Three levels of APP expression are investigated thanks to APP WT, HT and KO mice. APP roles in metabolic pathways in the hippocampus is evaluated by 1H-NMR spectroscopy. The neurophysiological impact of the genotype, of the glucose restriction and of the interaction between these two parameters is studied by extracellular electrophysiological recordings of cell excitability and synaptic activity in acute hippocampal slices incubated in control condition (10mM), mild (5mM) and severe glucose restriction (2.5mM). Because APP KO mice are known to be at risk of developing seizures (7), we also studied the effect of disinhibition on electrical activity by adding the GABA-A receptor antagonist picrotoxin to the aCSF.
1HNMR spectroscopy showed a strong increase in glutamate abundance while GABA decreased in APP KO mice. These metabolic modifications are consistent with the literature and could explain the hyperexcitability reported in APP KO mice (7). This overage in glutamate seems to be converted in glutamine whose abundance was also increased. Cholinergic metabolism is also modified as there was less choline and phosphocholine in KO mice but more of their precursor: glycerophosphocholine. This dysregulation is important in AD (8). Moreover, APP role in ATP production at the level of the mitochondria (9, 10) seems consistent with our observations as ADP/ATP ratio and AMP level were modified according to APP expression. Interestingly, HT mice presented an intermediate level of expression for every metabolite characterised, confirming the importance of the level of APP expression in metabolism regulation. In vivo hypoglycemia are currently carried out and 1HNMR spectroscopy will be performed to determine if the differences in metabolic profiles observed can be intensified when glucose is restricted.
Electrophysiological recordings also highlighted electrical differences in sensitivity to glucose restriction between WT, HT and KO mice and susceptibility to hyperexcitability in KO ones. Here again, HT mice presented an intermediate phenotype, strengthening the working hypothesis. Firstly, glucose restriction reduced synaptic activity and excitability of a neuronal network in a concentration dependent way. This indicates that a more pronounced glucose hypometabolism has more deleterious consequences on neural viability. Then, we observed that fEPSP and fiber volleys were not different according to the genotype in the basal condition. However, when glucose supply was impoverished to 5mM or 2.5mM, differences in synaptic activities appeared. Indeed, glucose restriction induced a large decrease of fEPSP in WT mice while this decrease in synaptic activity was considerably attenuated in KO mice. On the opposite, neuronal excitability was not modified as action potentials propagation measured by the fiber volley were not different. Also, ageing had an effect on the hippocampus functioning as 6 month-old mice showed smaller synaptic activity and excitability compared to 6 week-old mice. This reduction was observed for slices perfused in the three conditions. The lack of APP could have an influence on ageing as fEPSP of 6 months-old KO mice were smaller in the glucose restriction conditions than the other genotypes. As 1HNMR spectroscopy and literature showed that GABA is modified in APP KO mice, we studied the epileptiform activity of polyphasic fEPSP induced by a GABA-A receptor antagonist: picrotoxin. We observed that the intensity of the epileptiform activity was higher in KO mice (10mM and 5mM) and when the glucose was reduced to 2.5mM, we observed an extinction of fEPSP in most of the WT slices but not in the KO slices, where epileptiform activity was still high. Moreover, the drug decreased synaptic activity in WT mice but increased it in KO mice and had no significant effect on HT fEPSP.
The next step is to confirm that the reduction in glucose supply causes an increase in APP expression as described in the literature (11). If this hypothesis is validated, it could allow us to have a new level of APP expression: the overexpression one. This hypothesis is critical to determine if modifications observed in ex vivo glucose restrictions can be related to molecular changes found in AD and Down syndrome. Nevertheless, we can already conclude that APP expression and glucose metabolism are indeed linked in the hippocampus and that further investigations need to be conducted in the future to better understand this relationship.
