Distinct phenotype of E-selectin–deficient mice: E-selectin is required for slow leukocyte rolling in vivo

EJ Kunkel, K Ley - Circulation research, 1996 - Am Heart Assoc
EJ Kunkel, K Ley
Circulation research, 1996Am Heart Assoc
Leukocyte capture and rolling are mediated by calcium-dependent lectins expressed on
most leukocytes (L-selectin) and the vascular endothelium (P-and E-selectin). To study the
role of the selectins during inflammation, we have investigated leukocyte rolling in venules
of tumor necrosis factor-α (TNF-α)–treated mouse cremaster muscles in wild-type mice and
gene-targeted mice with homozygous deficiency for L-, P-, or E-selectin (L−/−, P−/−, or E−/−,
respectively). TNF-α treatment induces expression of E-selectin and increases expression of …
Leukocyte capture and rolling are mediated by calcium-dependent lectins expressed on most leukocytes (L-selectin) and the vascular endothelium (P- and E-selectin). To study the role of the selectins during inflammation, we have investigated leukocyte rolling in venules of tumor necrosis factor-α (TNF-α)–treated mouse cremaster muscles in wild-type mice and gene-targeted mice with homozygous deficiency for L-, P-, or E-selectin (L−/−, P−/−, or E−/−, respectively). TNF-α treatment induces expression of E-selectin and increases expression of P-selectin on endothelial cells. Consistent with previous reports of redundant P- and E-selectin function, a combination of monoclonal antibodies (mAbs) against P- and E-selectin (RB40.34 and 9A9, respectively) was necessary to block rolling in wild-type mice. The rolling leukocyte flux fraction (percent rolling cells) in L−/− mice was similar to that in wild-type mice, but rolling in these mice was blocked by a P-selectin mAb. The velocity of rolling leukocytes in TNF-α–treated wild-type, P−/−, or L−/− mice was 5 to 10 times slower (3 to 7 μm/s) than during trauma-induced rolling (20 to 50 μm/s). In contrast, leukocytes in venules of TNF-α–treated E−/− mice rolled significantly faster (12 to 20 μm/s); the rolling leukocyte flux fraction was more than doubled compared with wild-type, L−/−, or P−/− mice; and the number of adherent leukocytes was reduced. Addition of an E-selectin mAb, but not a P-selectin mAb, increased rolling flux fraction and rolling velocity in wild-type mice. Histological analysis revealed that 90% to 95% of all leukocytes interacting (rolling and adherent) with the venular endothelium in TNF-α–treated wild-type, L−/−, P−/−, and E−/− mice were granulocytes. These results identify a previously unrecognized phenotype of E−/− mice by establishing that at the site densities prevailing in vivo, E-selectin is responsible for slow (≈5 μm/s) granulocyte rolling. E-selectin–dependent slow rolling drastically increases the transit time of leukocytes rolling through an inflamed tissue and thus aids in targeting leukocytes activated by chemoattractants to the inflammatory microenvironment.
Am Heart Assoc