Use of 3D Distance Mapping to examine Neuro‐Glia‐Vascular Plasticity in Rodent Models of Cardiovascular Disease

Abstract

Regulating cerebral blood flow to supply oxygen and nutrients and remove waste products is of utmost importance for neuronal health. The brain’s capillary network responds to acute local increases of neuronal activity with vasodilation of upstream arterioles (neurovascular coupling, NVC). NVC mechanisms are impaired in a number of neurodegenerative and cardiovascular diseases. A compounding factor in these diseases is the remodelling of capillary networks. A reduced density of capillaries and/or changes in the arrangement of capillaries so that the parenchyma is not adequately supplied can exacerbate these conditions. We set out to develop novel tools to analyse the architecture of cerebral microvascular networks and assess neuronal‐glial‐vascular plasticity in rodent models of hypertension, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) and impaired endothelial Ca2+ signalling (TRPV4‐KO).Animals (mice or rats) were transcardially perfused with a liposome‐encapsulated fluorescent dye (DiI) and brains were harvested and fixed in 4 % paraformaldehyde. Transparency of the tissue was increased using the SeeDB protocol and the microvasculature was imaged in 2 mm thick coronal brain slabs using 2‐photon microscopy. Subsequently, slabs were cryopreserved, re‐sectioned at 80 μm thickness and processed for immunofluorescence (glial fibrillary astrocytic protein, Gfap). Slices were imaged using scanning confocal microscopy. Imaging volumes were taken in the rostro‐caudal axis to a depth of ~1 mm. They were deconvolved, attenuation‐corrected and scaled to equal length voxels. Then, fluorescent structures were converted into 3D objects (Volumetry G9 ‐ GWH) and non‐vessel particulates were filtered. Then, expanding boundary shells were used at every voxel in parenchymal space. Once the expanding boundary shells collided with a vessel, the distance (radius) was converted to a color and a new imaging volume containing the distance mapping results was created. The proportion of parenchymal voxels located at different distances from their closest blood vessel was calculated.We found that in both mice and rats, overall vessel density was the highest in the paraventricular nucleus of the hypothalamus (10–20 % of voxels), followed by cortex (~ 10 %) and hippocampus (~ 5 %). 90 % of the parenchymal volume in the cortex was within 26 μm of a blood vessel compared to > 34 μm in the hippocampus. Interestingly, Gfap staining was intense in the hippocampus but not in the cortex. In disease models, we found that aged mice (CADASIL, CADASIL control and TRPV4‐KO) had a lower vessel density than younger animals especially in cortex and PVN. In contrast, it was hippocampal vessel density in the dentate gyrus that was lower in spontaneously hypertensive rats than in controls.In summary, we describe a simple, yet powerful method to assess the relationship of capillary networks and parenchymal volume and show that rodent models of cardiovascular diseases are associated with plasticity of the neuro‐glia‐vascular unit.Support or Funding InformationR01 HL133211