Accelerated cerebromicrovascular senescence contributes to cognitive decline in a mouse model of paclitaxel (Taxol)-induced chemobrain

Abstract

Chemotherapy-induced cognitive impairment ("chemobrain") is a frequent side-effect in cancer survivors treated with paclitaxel (PTX). The mechanisms responsible for PTX-induced cognitive impairment remain obscure, and there are no effective treatments or prevention strategies. Here, we test the hypothesis that PTX induces endothelial senescence, which impairs microvascular function and contributes to the genesis of cognitive decline. We treated transgenic p16-3MR mice, which allows the detection and selective elimination of senescent cells, with PTX (5 mg/kg/day, 2 cycles; 5 days/cycle). PTX-treated and control mice were tested for spatial memory performance, neurovascular coupling (NVC) responses (whisker-stimulation-induced increases in cerebral blood flow), microvascular density, blood-brain barrier (BBB) permeability and the presence of senescent endothelial cells (by flow cytometry and single-cell transcriptomics) at 6 months post-treatment. PTX induced senescence in endothelial cells, which associated with microvascular rarefaction, NVC dysfunction, BBB disruption, neuroinflammation, and impaired performance on cognitive tasks. To establish a causal relationship between PTX-induced senescence and impaired microvascular functions, senescent cells were depleted from PTX-treated animals (at 3 months post-treatment) by genetic (ganciclovir) or pharmacological (treatment with the senolytic drug ABT263/Navitoclax) means. In PTX treated mice, both treatments effectively eliminated senescent endothelial cells, rescued endothelium-mediated NVC responses and BBB integrity, increased capillarization and improved cognitive performance. Our findings suggest that senolytic treatments can be a promising strategy for preventing chemotherapy-induced cognitive impairment.

Keywords: aging; chemotherapy; chemotherapy-induced cognitive impairment; dementia; functional hyperemia; senescence; vascular cognitive impairment.

FIGURE 1

GCV and ABT263 treatment successfully deplete senescent endothelial cells from the brains of PTX‐treated p16‐3MR mice. (a) Schematic representation of the experimental design. PTX: paclitaxel, GCV: ganciclovir, NVC: neurovascular coupling. (b–f) Flow cytometric detection of RFP⁺/CD‐31⁺ senescent endothelial cells, RFP⁻/CD‐31⁺ non‐senescent endothelial cells and RFP⁺/CD‐31⁻ senescent non‐endothelial cells in single‐cell suspensions obtained from the brains of control and PTX treated p16‐3MR mice that received vehicle, GCV or the senolytic drug ABT263. Brains were analyzed 6 months after PTX. (b) shows a representative dot plot of RFP‐Booster Atto647N fluorescence (which correlates with p16‐3MR expression) versus CD‐31 staining (583/26 nm), depicting the percentage of senescent endothelial cells with bright fluorescence in the brain of a PTX treated mouse. Summary data for all senescent cells are shown in (e). Data are mean ± SEM (n = 4 for each data point). ***p < 0.001 versus control. (d) Representative density plots of RFP‐Booster Atto647N fluorescence versus autofluorescence obtained from the FACS sorting of single cell brain suspension of a PTX treated mouse. Note the distinct population of RFP⁺ senescent cells (purple arrow). (e) Flow cytometric detection of RFP⁺/CD‐31⁺ senescent endothelial cells and RFP⁺/CD‐31⁻ senescent non‐endothelial cells in single‐cell suspension obtained from the brain of a PTX treated p16‐3MR mouse, which was enriched for RFP+ cells by FACS. Dot plot of RFP‐Booster Atto647N fluorescence versus CD‐31 staining depicts the percentages of senescent endothelial cells with bright fluorescence in both channels and that of RFP⁺/CD‐31⁻ senescent non‐endothelial cells. (f) Pie charts in the upper row show the ratio of senescent endothelial cells as percentage of all endothelial cells. Note that PTX treatment significantly increases the presence of senescent endothelial cells in the mouse brain, which is reversed by both GCV and ABT263 treatment. Pie charts in the bottom row show the ratio of senescent endothelial cells as percentage of all senescent cells. (g) Identification of cerebromicrovascular endothelial cells based on differentially expressed marker genes (scRNAseq data). Shown is two‐dimensional UMAP plot based on differentially expressed marker genes, colored by cluster. Cluster identity was assigned based on previously reported differentially expressed genes listed in Table S1. (h) Cells with high expression of senescence markers overlaid on UMAP plots for brains of PTX treated mice. (i) Bar charts showing percentage of senescent endothelial cells, microglia, pericytes, vascular smooth muscle cells (VSMCs) and oligodendrocytes in brains of PTX treated mice. (j) Principal component analysis of RNA‐Seq data generated from brain endothelial cells derived from PTX treated mice. Capillary‐, venous‐, and arterial endothelial cells were identified based on their specific gene expression signatures. Shown is visualization of endothelial cell subclusters, color coded for the identified endothelial phenotypes. Marker genes of each cluster are provided in Table S3. Senescent endothelial cells identified based on high expression of senescence markers are highlighted (red) in (k). (l) Spatially‐resolved mRNA expression of the senescence marker gene Cdkn2a (green ST spots) in brains of control and PTX treated p16‐3MR mice that received vehicle, GCV or ABT263. Representative H&E stained coronal sections of brains are shown. (m) Spatial distribution of Cdkn2a positive spots, expressed as a percentage of all spots in the different anatomical regions in brains of each group of mice. GCV and ABT263 treatment successfully depleted senescent cells from the isocortex of PTX treated p16‐3MR mice.

FIGURE 2

Elimination of senescent cells increases capillary density and improves endothelium‐mediated neurovascular coupling responses in PTX‐treated mice. (a) Segmentation of blood vessels in brain parenchyma on OCT images. Original z‐stack images acquired in brains of control and PTX treated p16‐3MR mice that received vehicle, ganciclovir (GCV) or ABT263 were maximum projected. Thresholded binary images and skeletons were used to calculate indices (vascular density index, vessel skeleton density index) corresponding to microvascular density. (b,c) Microvascular density was significantly decreased by PTX treatment and rescued by treatment with GCV or ABT263. Data are mean ± SEM. *p < 0.05. (d) Representative pseudocolour laser speckle flowmetry maps of baseline CBF (upper row; shown for orientation purposes) and CBF changes in the whisker barrel field relative to baseline during contralateral whisker stimulation (bottom row, right oval, 30 s, 5 Hz) in control and PTX treated p16‐3MR mice that received vehicle, GCV or ABT263. Color bar represents CBF as percent change from baseline. The NO synthase inhibitor L‐NAME was administered to test NO mediation of functional hyperemia. Bottom: time‐course of CBF changes after the start of contralateral whisker stimulation (horizontal bars). Summary data are shown in (e). Data are mean ± SEM (n = 8–9 in each group), *p < 0.05 versus control; ^# p < 0.05 versus before L‐NAME. (one‐way ANOVA with post‐hoc Tukey's tests). n.s., not significant. Capillary density and NVC responses were assessed 6 month post‐PTX treatment (see Materials and methods).

FIGURE 3

Elimination of senescent cells rescues blood brain barrier integrity and attenuates neuroinflammation in PTX‐treated mice. (a) Two‐photon‐imaging‐based measurement of microvascular permeability to fluorescent tracers in brains of control and PTX treated p16‐3MR mice that received vehicle, ganciclovir (GCV) or ABT263. Left: Z‐stack projection of cerebral vasculature. Red fluorescence: WGA‐Alexa594 staining of the glycocalyx of the endothelial cells. Green autofluoresent background in the tracer channel is shown. Right: Changes in tracer fluorescence intensity in the extravascular space and brain parenchyma in each group of animals upon injection of the fluorescent tracers of different molecular weight. Images captured subsequent to tracer administrations were maximum projected (z‐stack) and subtraction of the images of WGA‐Alexa594 stained vasculature was performed. Intensity plots derived from median projection of time‐stacks of the parenchymal recordings show increased extravasation of tracers of different molecular weights in brains of PTX‐treated mice as compared to those of control mice. Note that treatment with GCV or ABT263 attenuates BBB disruption. Relative fluorescent intensity scale is shown at top right. Scale: 100 μm. (b) Summary data for microvascular permeability to fluorescent tracers with different molecular weights in brains of each group of animals. Note the increased permeability for the different sized tracers in PTX‐treated animals and reversal by senolytics. Data are mean ± SEM. *p < 0.05. (n = 5–10 for each datapoint). (c) Confocal images showing perivascular IBA1 positive microglia (red fluorescent cells located adjacent to green fluorescent endomucin positive capillaries; blue fluorescence: nuclear staining with DAPI) in brains of control and PTX treated p16‐3MR mice that received vehicle, GCV or ABT263. (d) Bar graphs are summary data of relative changes in activated microglia. Data are mean ± SEM (n = 4 for each datapoint). ***p < 0.001.

FIGURE 4

Rescue of microvascular function by elimination of senescent cells improves performance of PTX treated p16‐3MR mice in the radial‐arm water maze (RAWM). Control p16‐3MR mice and PTX treated p16‐3MR mice that received vehicle, ganciclovir (GCV) or ABT263 were tested in the RAWM. (a) Representative probe test search path of a randomly selected animal from each group is shown with the target position highlighted in green. Note the PTX treated mice that received vehicle required more time and a longer path length in order to find the hidden escape platform than both control animals and GCV‐ or ABT263 treated PTX mice. The PTX treated mouse also entered multiple incorrect arms. (b) PTX treated animals have longer path lengths throughout day 2 and 3 of the learning phase, and the retrieval day 10 (“P”) as compared to control animals. Both treatment with GCV and ABT263 improved learning and memory performance. (c) PTX treated animals also made significantly more errors during probe trial and probe reversal than control animals. In contrast, PTX treated mice treated with GCV or ABT263 perform this task significantly better than vehicle treated PTX mice. Data are mean ± SEM (n = 10–15 for each data point). *p < 0.05 versus control, ^# p < 0.05 versus vehicle treated PTX. (d) Scheme depicting the microvascular mechanisms (microvascular rarefaction, neurovascular dysfunction and BBB disruption) by which PTX‐induced endothelial senescence may contribute to the genesis of chemotherapy‐induced cognitive impairment.