Apolipoprotein E protects cultured pericytes and astrocytes from D-Abeta(1-40)-mediated cell death

Abstract

Cerebral amyloid angiopathy (CAA) is a common pathological finding in Alzheimer's disease and hereditary cerebral hemorrhage with amyloidosis of the Dutch type; in this latter condition it is caused by deposition of mutated amyloid beta protein (Abeta Glu22Gln; D-Abeta(1-40)). Previously, we found a dependence of the Abeta-mediated toxicity and apolipoprotein E (apoE) production by cultured pericytes on apoE genotype. Given their close association with the cerebrovascular wall both astrocytes and pericytes may be involved in CAA development, a process that includes Abeta deposition and clearance and that may be affected by interaction with locally produced apolipoprotein E (apoE). Although astrocytes are regarded as the major source of apolipoprotein E (apoE) in the brain, also pericytes produce apoE. In this study we compared the apoE production capacity, the effects of apoE on D-Abeta(1-40) internalization, D-Abeta(1-40) cell surface accumulation and the vulnerability for D-Abeta(1-40)-induced toxicity of either cell type in order to quantify the relative contributions of astrocytes and pericytes in the various processes that contribute to CAA formation. Strikingly, cultured astrocytes produced only 3-10% of the apoE amounts produced by pericytes. Furthermore, pericytes with the apoE epsilon4 allele produced three times less apoE and were more vulnerable to D-Abeta(1-40) treatment than pericytes without an epsilon4 allele. Such relations were not observed with astrocytes in vitro. Both pericytes and astrocytes, however, were protected from Abeta-induced cytotoxicity by high levels of pericyte-derived apoE, but not recombinant apoE. In addition, pericyte-derived apoE dose-dependently decreased both internalization of Abeta and Abeta accumulation at the cell surface in either cell type. The present data suggest that apoE produced by pericytes, rather than astrocyte-produced apoE, modulates Abeta cytotoxicity and Abeta removal near the vasculature in the brain. Furthermore, since apoE production in pericytes is genotype dependent, this may contribute to the apoE genotype-dependent development of CAA in vivo.