Selectin-mediated recruitment of bone marrow stromal cells in the postischemic cerebral microvasculature

Abstract

Background and purpose: The therapeutic potential of bone marrow stromal cells (BMSCs) has been demonstrated in different models of stroke. Although it is well established that BMSCs selectively migrate to the site of brain injury, the mechanisms underlying this process are poorly understood. This study addresses the hypothesis that selectins mediate the recruitment of BMSCs into the postischemic cerebral microvasculature.

Methods: Focal ischemic stroke was induced by middle cerebral artery occlusion and reperfusion. Cell recruitment was monitored using either fluorescent- or radiolabeled BMSCs detected by intravital microscopy or tissue radioactivity. Mice were treated with either a blocking antibody against P- or E-selectin or with the nonselective selectin antagonist, fucoidin. The role of CD44 in cell recruitment was evaluated using BMSCs from CD44 knockout mice.

Results: Middle cerebral artery occlusion and reperfusion was associated with a significantly increased adhesion of BMSCs in cerebral venules compared with sham mice. Immunoneutralization of either E- or P-selectin blocked the middle cerebral artery occlusion and reperfusion-induced recruitment of adherent BMSCs. An attenuated recruitment response in the postischemic hemisphere was also noted after fucoidin treatment or administration of CD44-deficient BMSCs.

Conclusions: Cerebral vascular endothelium assume a proadhesive phenotype after ischemic stroke that favors the recruitment of BMSCs, which use both P- and E-selectin to home into the infarct site. CD44 may serve as the critical ligand for selectin-mediated BMSC recruitment.

Figure 1. Adhesion of wild type (WT) and large-T bone marrow stromal cells (BMSC) in cerebral venules of sham (Sham) and postischemic (I/R) mice

Quantification of adherent cells (Panel A). * denotes P<0.05 relative to Sham mice; N.S. denotes no statistical significance. (WT Sham and I/R n=3 and 4, Large-T Sham and I/R n=6 and 8 respectively). Images demonstrating the adhesion of Large T BMSC in cerebral venules of Sham (Panel B) and I/R (Panel C) mice. Bar=50 μm.

Figure 2. Effects of immunoneutralization of either E- or P-selectin on the number of adherent large-T BMSC in postischemic (I/R) cerebral venules

Data are also shown for the isotype-matched (control) antibodies for E-selectin (E-Sel iso) and P-selectin (P-Sel iso). * denotes p<0.05 compared to the corresponding sham value, ^ denotes p<0.05 relative to corresponding I/R group. (in groups Sham I/R, Anti-E-Sel, E-Sel Iso, Anti-P-Sel and P-Sel Iso n= 8, 6, 8, 6, 4 and 6 respectively.)

Figure 3. Number of radiolabelled large-T BMSC recruited into the postischemic brain 3 days after an ischemic stroke (I/R) or sham operation (Sham), with or without treatment with the pan-selectin antagonist fucoidin

* denotes p<0.05 compared to the contralateral (non-ischemic) hemisphere, ^ denotes p<0.05 relative to the corresponding Fucoidin group. # denotes p<0.05 compared to the corresponding Sham value. Mean and SEM values are derived from three independent experiments.

Figure 4. Surface expression of selectin ligands CD24 and CD44 on BMSC under basal conditions and following incubation with ischemic brain extract (I/R brain)

Values represent duplicate measurements from the same sample.

Figure 5. Comparison of BMSC recruitment in wild type (WT) and CD44 knockout (CD44KO) mice following focal ischemia-reperfusion (I/R)

* denotes p<0.05 compared to contralateral (non-ischemic) hemisphere, ^ denotes p<0.05 relative to corresponding CD44KO group. (n=5 CD44KO, n=3 WT+I/R)