Pericyte Remodeling in Health and Alzheimer's Disease

Abstract

Pericyte loss is known to occur in aging and disease and has been linked to vessel regression through changes in capillary flow patterns. However, pericytes possess an intrinsic repair strategy to compensate for pericyte loss, as the surviving pericytes can remodel their processes to restore coverage of the exposed endothelium. Pericyte remodeling is impaired in the aged mouse brain, but we still have limited knowledge of the mechanisms that drive pericyte growth and how pericyte remodeling is affected by the amyloid-beta (Aβ) pathology characteristic of Alzheimer’s disease (AD). The platelet-derived growth factor (PDGF-B/PDGFRβ) signaling axis is an important pathway for pericyte recruitment to the vessel wall during development, and changes in PDGFRβ have been implicated in AD. As such, we hypothesized that PDGF-B/PDGFRβ signaling drives pericyte growth and that pericyte remodeling is diminished in mouse models of AD. To test these hypotheses, we performed longitudinal in vivo two-photon imaging in both 6 – 12-month PDGFRβCre;Ai14 mice and 12 – 18-month TgSwDI;PDGFRβCre;Ai14 mice with the Swedish, Dutch, and Iowan mutations of the APP gene. Optical ablation of brain pericytes was performed, and pericyte remodeling was measured over the span of one week while the potent and selective PDGFRβ inhibitor, SU16f, was delivered at 10mg/kg/day to the healthy adult mice. Interestingly, SU16f did not change the overall rate of pericyte growth compared to vehicle controls. However, processes with shorter baseline lengths grew faster in drug-treated animals, suggesting involvement of PDGFRβ in the remodeling process. In TgSwDI mice, we found that spontaneous pericyte loss occurred closer to the venules than the arterioles. Interestingly, this loss did not positively correlate with Aβ burden. Further, pericyte remodeling was significantly slower in TgSwDI mice, with the largest deficits found in cells closest to the venules. Spontaneous pericyte loss was associated with longer and more tortuous vessels indicative of vessel rarefaction. Mimicking capillary regression in 4 – 8-month wild-type mice reduced blood flow in venous outputs by ~50%, increased upstream blood flow heterogeneity, and induced chronic constriction of upstream transitional vessels, reducing flow to regions distant from the regression site. These results suggest a selective vulnerability of venous-associated pericytes in disease, further connect pericyte loss to cerebral hypoperfusion, and pinpoint PDGFRβ as a potential therapeutic target to promote pericyte coverage. Future studies will examine potential causes of pericyte loss in TgSwDI mice and investigate pericyte remodeling capabilities in conditional CNS pericyte-specific PDGFRβ mutant mouse lines.