Estradiol augments peri-infarct cerebral vascular density in experimental stroke

Abstract

Peri-infarct increase of vascular density has been observed in animals and in humans with ischemic stroke. Increased peri-infarct vascular density correlates with improved functional outcome after stroke. We hypothesized that pre-treatment with estradiol will increase post-ischemic peri-infarct capillary density in a rat model of transient ischemic stroke. Estradiol, compared to placebo, augmented post-ischemic peri-infarct vascular density by 22% 10 days after stroke. Recovery of forelimb function was not improved with estradiol treatment on day three and nine post-stroke. Loss of estradiol may limit repair in the peri-infarct region by limiting angiogenesis, but functional significance in stroke recovery requires further investigation.

Figure 1

A and B. Eosin-stained coronal rat brain section photographed at 1.25X illustrating the approximate peri-infarct region analyzed (arrows) in rats with the largest (A) and smallest (B) infarcts on post stroke day three. Infarct dimensions and sub-regional localization of the cortical portion of the lesion were heterogeneous within and between cohorts. Scale bar = 5 mm.

Figure 1

A and B. Eosin-stained coronal rat brain section photographed at 1.25X illustrating the approximate peri-infarct region analyzed (arrows) in rats with the largest (A) and smallest (B) infarcts on post stroke day three. Infarct dimensions and sub-regional localization of the cortical portion of the lesion were heterogeneous within and between cohorts. Scale bar = 5 mm.

Figure 2

A and B. Low magnification views of CD31-labeled structures in the non-ischemic cortex (A) and peri-infarct region (B) photographed 10 days after experimental ischemic stroke. Counts of these structures in the peri-infarct and contralateral mirror-image regions were used to derive ischemic/nonischemic vascular indices. The structures located in the peri-ischemic region were more numerous, larger, and torturous. Scale bar = 200μm. C and D. High magnification views of CD31-labeled structures (arrows) in intact brain (C) and in the peri-infarct region (D) 10 days after experimental stroke. In intact brain, the morphology is typical of a normal parenchymal capillary: the endothelial cells form a narrow lumen, with relatively few nuclei oriented along the axis of the vessel. In the injured area, the morphology somewhat resembles that of angiogenic blood vessels: the lumen is wide, and there are numerous endothelial cell nuclei clustered around the lumen. CD31 outlines the surface membranes. Nuclei are labeled with hematoxylin. Scale bar = 50μm.

Figure 2

A and B. Low magnification views of CD31-labeled structures in the non-ischemic cortex (A) and peri-infarct region (B) photographed 10 days after experimental ischemic stroke. Counts of these structures in the peri-infarct and contralateral mirror-image regions were used to derive ischemic/nonischemic vascular indices. The structures located in the peri-ischemic region were more numerous, larger, and torturous. Scale bar = 200μm. C and D. High magnification views of CD31-labeled structures (arrows) in intact brain (C) and in the peri-infarct region (D) 10 days after experimental stroke. In intact brain, the morphology is typical of a normal parenchymal capillary: the endothelial cells form a narrow lumen, with relatively few nuclei oriented along the axis of the vessel. In the injured area, the morphology somewhat resembles that of angiogenic blood vessels: the lumen is wide, and there are numerous endothelial cell nuclei clustered around the lumen. CD31 outlines the surface membranes. Nuclei are labeled with hematoxylin. Scale bar = 50μm.

Figure 2

A and B. Low magnification views of CD31-labeled structures in the non-ischemic cortex (A) and peri-infarct region (B) photographed 10 days after experimental ischemic stroke. Counts of these structures in the peri-infarct and contralateral mirror-image regions were used to derive ischemic/nonischemic vascular indices. The structures located in the peri-ischemic region were more numerous, larger, and torturous. Scale bar = 200μm. C and D. High magnification views of CD31-labeled structures (arrows) in intact brain (C) and in the peri-infarct region (D) 10 days after experimental stroke. In intact brain, the morphology is typical of a normal parenchymal capillary: the endothelial cells form a narrow lumen, with relatively few nuclei oriented along the axis of the vessel. In the injured area, the morphology somewhat resembles that of angiogenic blood vessels: the lumen is wide, and there are numerous endothelial cell nuclei clustered around the lumen. CD31 outlines the surface membranes. Nuclei are labeled with hematoxylin. Scale bar = 50μm.

Figure 2

A and B. Low magnification views of CD31-labeled structures in the non-ischemic cortex (A) and peri-infarct region (B) photographed 10 days after experimental ischemic stroke. Counts of these structures in the peri-infarct and contralateral mirror-image regions were used to derive ischemic/nonischemic vascular indices. The structures located in the peri-ischemic region were more numerous, larger, and torturous. Scale bar = 200μm. C and D. High magnification views of CD31-labeled structures (arrows) in intact brain (C) and in the peri-infarct region (D) 10 days after experimental stroke. In intact brain, the morphology is typical of a normal parenchymal capillary: the endothelial cells form a narrow lumen, with relatively few nuclei oriented along the axis of the vessel. In the injured area, the morphology somewhat resembles that of angiogenic blood vessels: the lumen is wide, and there are numerous endothelial cell nuclei clustered around the lumen. CD31 outlines the surface membranes. Nuclei are labeled with hematoxylin. Scale bar = 50μm.

Figure 3

A. Peri-infarct vascular density indices were determined in brain sections labeled with anti-CD31 antibody from placebo (E0) and estradiol-treated (E25) animals on post stroke day three and 10 (PSD 3 and PSD 10, respectively). *p=0.011 for E0 vs. E25 on PSD 10; **p=0.010 for E25 on PSD 3 vs. E25 on PSD 10. One-way ANOVA. Sample size: E0, PSD 3, n=4; E25, PSD 3, n=4; E0, PSD 10, n=6; E25, PSD 10, n=5. B. Ischemic/nonischemic difference in the proportions of CD31-positive structures of different sizes was derived 10 days post stroke. CD31-positive structures were counted in placebo (E0) and estradiol (E25)-treated rats and sorted into five bins by pixel area. The percent difference in proportion of structures within each bin between the peri-infarct area and the contralateral matched region was determined. *p<0.017 for E0 vs. E25 in >1000 pixel bin; **p<0.005 for >1000 versus 50-250 pixel bin for estradiol-treated animals. Two-way ANOVA.

Figure 3

A. Peri-infarct vascular density indices were determined in brain sections labeled with anti-CD31 antibody from placebo (E0) and estradiol-treated (E25) animals on post stroke day three and 10 (PSD 3 and PSD 10, respectively). *p=0.011 for E0 vs. E25 on PSD 10; **p=0.010 for E25 on PSD 3 vs. E25 on PSD 10. One-way ANOVA. Sample size: E0, PSD 3, n=4; E25, PSD 3, n=4; E0, PSD 10, n=6; E25, PSD 10, n=5. B. Ischemic/nonischemic difference in the proportions of CD31-positive structures of different sizes was derived 10 days post stroke. CD31-positive structures were counted in placebo (E0) and estradiol (E25)-treated rats and sorted into five bins by pixel area. The percent difference in proportion of structures within each bin between the peri-infarct area and the contralateral matched region was determined. *p<0.017 for E0 vs. E25 in >1000 pixel bin; **p<0.005 for >1000 versus 50-250 pixel bin for estradiol-treated animals. Two-way ANOVA.

Figure 4

Forelimb use asymmetry scores were derived for estradiol (E2) and placebo (E0)-treated animals using the cylinder test pre stroke and on post stroke day three (PSD 3) and 10 (PSD 10). Limb asymmetry score = (right - left) / (right + left + both); thus, positive scores reflect decreased left forelimb use (as expected with a right hemispheric stroke).