Rapamycin Treatment Reduces Brain Pericyte Constriction in Ischemic Stroke

Abstract

The contraction and subsequent death of brain pericytes may play a role in microvascular no-reflow following the reopening of an occluded artery during ischemic stroke. Mammalian target of rapamycin (mTOR) inhibition has been shown to reduce motility/contractility of various cancer cell lines and reduce neuronal cell death in stroke. However, the effects of mTOR inhibition on brain pericyte contraction and death during ischemia have not yet been investigated. Cultured pericytes exposed to simulated ischemia for 12 h in vitro contracted after less than 1 h, which was about 7 h prior to cell death. Rapamycin significantly reduced the rate of pericyte contraction during ischemia; however, it did not have a significant effect on pericyte viability at any time point. Rapamycin appeared to reduce pericyte contraction through a mechanism that is independent of changes in intracellular calcium. Using a mouse model of middle cerebral artery occlusion, we showed that rapamycin significantly increased the diameter of capillaries underneath pericytes and increased the number of open capillaries 30 min following recanalisation. Our findings suggest that rapamycin may be a useful adjuvant therapeutic to reduce pericyte contraction and improve cerebral reperfusion post-stroke.

Keywords: Brain pericytes; Calcium; Experimental stroke; Rapamycin; mTOR.

Fig. 1

Rapamycin reduces pericyte contractility during OGD. A Average normalised cell index for vehicle (blue), 10 nM rapamycin– (red) and 100 nM rapamycin– (green) treated cells during 12 h of OGD. B Slope of cell index 0–1 h post-OGD, vehicle (blue), 10 nM rapamycin (red), and 100 nM rapamycin (green); circle, square, and triangles signify 3 independent cultures, 4 wells per culture. A two-way ANOVA was conducted with culture and treatment as variables. Treatment, F (2, 27) = 9.716, p = 00007; culture, F (2, 27) = 19.04, p < 0.0001; interaction, F (4, 27) = 7.275, p = 0.0004, with Tukey’s multiple comparisons tests to assess the main treatment effects. *p < 0.05, ***p <0.001. C Slope 1–2 h post-OGD. Two-way ANOVA, treatment, F (2, 27) = 20.38, p < 00001; culture, F (2, 27) = 119.3, p < 0.0001; interaction, F (4, 27) = 3.985, p = 0.0114, with Tukey’s multiple comparisons tests to assess the main treatment effects. ****p < 0.0001. D Slope 2–3 h post-OGD. Two-way ANOVA, treatment, F (2, 27) = 28.13, p < 00001; culture, F (2, 27) = 209.5, p < 0.0001; interaction, F (4, 27) = 9.048, p < 0.0001, with Tukey’s multiple comparisons tests to assess the main treatment effects. ****p < 0.0001. E Slope 3–4 h post-OGD. Two-way ANOVA, treatment, F (2, 27) = 4.526, p = 0.0202; culture, F (2, 27) = 68.8, p < 0.0001; interaction, F (4, 27) = 0.7043, p = 0.5959, with Tukey’s multiple comparisons tests to assess the main treatment effects. *p < 0.05

Fig. 2

Rapamycin did not affect pericyte cell death during OGD. Vehicle (blue), 10 nM rapamycin (red), and 100 nM rapamycin (green); circle, square, and triangles signify 3 independent cultures, 2–5 wells per culture. A Pericyte viability after 2-h OGD (% of cells negative for annexin V (AV) and propidium iodide (PI) staining). Two-way ANOVA, treatment, F (5, 18) = 5.717, p = 0.0025; culture, F (2, 18) = 69.72, p < 0.0001; interaction, F (10, 18) = 0.7912, p = 0.6382, with Sidak’s multiple comparisons tests to assess the main treatment effects. *p < 0.05. B Pericyte viability after 8-h OGD. Two-way ANOVA, treatment, F (5, 18) = 49.59, p < 0.0001; culture, F (2, 18) = 125.8, p < 0.0001; interaction, F (10, 18) = 5.105, p = 0.0014, with Sidak’s multiple comparisons tests to assess the main treatment effects. *p < 0.0001. C Pericyte viability after 12-h OGD. Two-way ANOVA, treatment, F (10, 62) = 280.5, p < 0.0001; culture, F (2, 62) = 257, p < 0.0001; interaction, F (10, 62) = 6.682, p < 0.0001, with Sidak’s multiple comparisons tests to assess the main treatment effects. ****p < 0.0001. D Apoptosis of pericytes after 12-h OGD (% cells positive for AV but negative for PI). Two-way ANOVA, treatment, F (5, 62) = 4.604, p = 0.0012; culture, F (2, 62) = 113.9, p < 0.0001; interaction, F (10, 62) = 2.797, p = 0.0064, with Sidak’s multiple comparisons tests to assess the main treatment effects. *p < 0.05, **p < 0.01. E Pericyte cell death 12 h after OGD (% cells positive for PI +/− annexin V). Two-way ANOVA, treatment, F (5, 68) = 215.1, p < 0.0001; culture, F (2, 58) = 90.85, p < 0.0001; interaction, F (10, 58) = 5, p < 0.0001, with Sidak’s multiple comparisons tests to assess the main treatment effects. ****p < 0.01

Fig. 3

Rapamycin prevents pericyte contraction in response to chemical ischemia. Representative DIC and fluorescent Fura-2 AM Ca²⁺ flux images of pericytes treated with a chemical ischemia + vehicle (DMSO) or b rapamycin, at baseline (0 min) and 20-min time points. The white dotted outline indicates cell membrane boundaries for each representative cell at baseline. Colorimetric scale represents intracellular Ca²⁺ levels with the ratiometric Fura-2 AM Ca²⁺ indicator. Lower right scale bar = 10 μm. For each cell: c normalised cell area (A/A₀) and e normalised Fura-2 AM calcium flux (R/R₀), compared to baseline, were calculated over time following each treatment. The dotted line represents baseline cell area or calcium flux. The dark lines represent the mean and the shaded area represents the SEM for each treatment group. Number of individual cells analysed for each group: vehicle (DMSO) (N = 47); ischemia (N = 36); ischemia + Rap (rapamycin) (N = 37); over 6 replicate cultures. d Net area under the curve (nAUC) of A/A₀ and f R/R₀ was calculated for each treatment group. Bars represent mean ± SD. One-way ANOVA (treatment, F (2,117) = 8.785, p = 0.0003) with Dunnett’s multiple comparisons test was used to compare A/A₀ groups. The Kruskal-Wallis test (treatment, H (2) = 15.28, p = 0.0005) with Dunn’s multiple comparisons test was used to compare R/R₀ groups. *p < 0.05, ***p < 0.001

Fig. 4

Rapamycin increases luminal width and the proportion of open vessels in the striatum following MCAO. a Schematic of mouse brain coronal section under bregma demonstrating contralateral and ipsilateral hemispheres, relative location of laser doppler (LD) probe, the area of vascular perfusion deficit in the ipsilateral hemisphere, and regions of interest sampled. b Example fluorescent image of whole coronal brain section labelled with FITC-albumin (green), NG2DsRed (Red), and isolectin B4 (ILB4, white), with the area of vascular perfusion deficit annotated with white dotted line. c Example images of capillaries with luminal FITC-albumin labelling (left) and reduced luminal FITC-albumin labelling (right) under pericyte soma. d, e Quantification of mean perfused vessel luminal width under pericytes in the striatum and cortex in both the contralateral and ipsilateral hemispheres following 60-min MCAO and 30-min reperfusion. Individual data points are presented as mean ± SD and were determined from 1086 individual vessel width measurements. Striatum (ipsilateral N = 1–18, contralateral N = 24–-50) and cortex (ipsilateral N = 10–49, contralateral N = 25–50) vessels per animal. Repeated measures two-way ANOVA (d hemisphere, F (1,8) = 162.9, p < 0.0001; treatment, F (1,8) = 4.831, p = 0.0592, and e hemisphere, F (1,5) = 38.93, p < 0.0015; treatment, F (1,9) = 2.961, p = 0.1194) with Sidak’s multiple comparisons test was used to compare treatment and hemisphere groups. When values were missing, a mixed-effects model with Sidak’s multiple comparisons test was used. f, g Vessel data (from d, e) were binned based on luminal width being 0 (closed) or > 0 (open) in the striatum and cortex. Data are presented as a percentage of open or closed vessels. Fisher’s exact test was used to compare groups. *p < 0.05, **p < 0.01, ****p < 0.0001

Fig. 5

Rapamycin had no effect on the reduction in pericyte number in the striatum following MCAO. a Example images of capillaries (ILB4, white) with pericytes (NG2, red) in contralateral (left) and ipsilateral (right) striatum of mouse subject to 60-min MCAO followed by 30-min reperfusion. b, c Quantification of pericyte number in the striatum and cortex. Individual data points represent individual mice overlayed with mean ± SD. Vehicle treatment N = 7, rapamycin treatment N = 8. Two-way ANOVA (b hemisphere, F (1,13) = 70.43, p < 0.0001; treatment, F (1,8) = 0.5430, p = 0.4743, and c hemisphere, F (1,13) = 4.579, p = 0.0519; treatment, F (1,13) = 0.2980, p = 0.5944) with Tukey’s multiple comparisons test was used to compare groups. ***p<0.001, ****p < 0.0001