The Neuron-Astrocyte-Microglia Triad in a Model of Chronic Cerebral Ischemia in the Rat Hippocampus: Effect of Dipyridamole

D Lana1, A Melani2, D Nosi3, F Pedata2, MG Giovannini1. 1University of Florence, Dept. of Health Sciences, Florence, Tuscany, Italy, 2University of Florence, Dept. of NEUROFARBA, Florence, Tuscany, Italy, 3University of Florence, Dept. of Experimental and Clinical Medicine, Florence, Tuscany, Italy

We recently characterized morphological and functional alterations in the neuron-astrocyte-microglia triad in the rat hippocampus in two different models of neurodegeneration: normal brain aging and LPS-induced acute inflammation. In particular we found that astrocytes and microglia actively cooperate in the clearance of apoptotic neurons and neuronal debris in the hippocampus (1). Moving from the results obtained in our previous paper here we wanted to characterize a different model of neurodegeneration: chronic cerebral ischemia (CI) induced by bilateral common carotid artery occlusion (bCCAo) in the rat. Aim of our research was to elucidate the pathophysiological mechanisms that link CI to the cognitive impairment previously found in this model of cerebral hypoxia (2) and the effect of dipyridamole as an anti-inflammatory drug (3).

A group of adult male Wistar rats (n=11) was subjected to permanent bCCAo according to Sarti et al. (4). A different group of rats, operated for bCCAo, was infused into the jugular vein, using an osmotic minipump, with dipyridamole (Persantin, Boehringer Ingelheim, 5 mg/mL) for 7 days, 4 mg/kg/day (n=11). Sham-operated rats were used as controls (n=12). Immunohistochemical staining and analysis of neurons, astrocytes and microglia was performed according to our previous paper (1), 3 months after bCCAo. Statistical analysis: One-way ANOVA and Newman-Keuls post test.

No differences in astrogliosis, in the number of pyramidal CA1 neurons or in the thickness of the CA1 Str. Pyramidalis were found. Pyramidal ectopic neurons and neuronal debris density in CA1 Str. Radiatum were significantly higher in bCCAo rats vs shams (76.22±2.63 vs 51.10±6.07 neurons/mm2 and 294.5±27.71 vs 215.2±29.5 debris/mm2, p<0.05) and dipyridamole reverted this effect (41.67±9.89 neurons/mm2 and 165.3±20.21 debris/mm2, p<0.01 dipyridamole vs bCCAo). In bCCAo rats we found a significant decrease in total microglia vs shams (166.0±9.41 vs 201.1±11.87 microglia/mm2, p<0.05) and this effect was reverted by dipyridamole (203.6±9.05 microglia/mm2, p<0.05 dipyridamole vs bCCAo). These results demostrate the presence of a neuronal damage in CA1 of bCCAo rats and the protective role of microglial cells. Qualitative triple fluorescent immunostaining of neurons-astrocytes-microglia revealed that in the CA1 Str. Radiatum of bCCAo rats microglia cells were actively phagocytosing pyramidal ectopic neurons and were in close interplay with the nearer astrocytes that appeared to compartmentalize the neurons engulfed; furthermore we observed that some astrocytes branches were closely apposed to ectopic neurons and intermingled the neuronal body, cooperating in their fragmentation. We did not observe any of these events in sham or in dipyridamole treated rats.

In conclusion, we hypothesize that the signs of alteration we found in the CA1 hippocampal region could be the pathophysiological mechanisms linking CI to the cognitive impairment observed in this model of neurodegeneration (2). Dipyridamole administration could revert the progression of this pathophysiological mechanism because of its role as an anti-inflammatory drug during the acute phase of CI.

References

(1) Cerbai F et al, PLoS One 7:e45250, 2012

(2) Melani A et al, Ann NY Acad Sci 1207:89, 2010

(3) Weyrich AS et al, Circulation 111: 633, 2005

(4) Sarti C et al, Behav Brain Res 136:13, 2002