P-selectin glycoprotein ligand-1 mediates L-selectin-dependent leukocyte rolling in venules

Abstract

Leukocyte rolling in postcapillary venules of inflamed tissues is reduced in L-selectin-deficient mice and mice treated with L-selectin blocking antibodies, but the glycoprotein ligand for L-selectin in inflamed venules is unknown. Here, we show that L-selectin-dependent rolling after P-selectin blockade is completely absent in P-selectin glycoprotein ligand-1 (PSGL-1)-/- mice or wild-type mice treated with a PSGL-1 blocking monoclonal antibody. Immunohistochemistry and flow cytometry failed to show PSGL-1 expression on resting or inflamed endothelium or on platelets. To investigate whether leukocyte-expressed PSGL-1 is mediating L-selectin-dependent rolling, we reconstituted lethally irradiated wild-type mice with PSGL-1-/- bone marrow cells. These chimeric mice showed no L-selectin-dependent rolling, suggesting that leukocyte-expressed PSGL-1 mediates L-selectin-dependent rolling. Frame-to-frame video analysis of L-selectin-dependent rolling in wild-type mice showed that the majority of observed L-selectin-dependent leukocyte rolling was between free flowing leukocytes and already adherent leukocytes or possibly leukocyte fragments, followed by E-selectin-dependent leukocyte rolling along the endothelium. Leukocyte rolling was significantly slower for leukocyte-endothelial than leukocyte-leukocyte interactions. We conclude that leukocyte-expressed PSGL-1 serves as the main L-selectin ligand in inflamed postcapillary venules. L-selectin binding to PSGL-1 initiates tethering events that enable L-selectin-independent leukocyte-endothelial interactions. These findings provide a molecular mechanism for the inflammatory defects seen in L-selectin-deficient mice.

Figure 1.

Leukocyte rolling flux fraction (mean ± SEM) in (A) untreated cremaster muscle venules of PSGL-1 deficient mice (black bars, n = 7) and wild-type mice (gray bars, n = 9), (C) in 4–6 h TNF-α treated cremaster muscle venules of PSGL-1–deficient (n = 4) and wild-type mice (n = 4), or in (B) untreated or (D) TNF-α–treated wild-type mice reconstituted with bone marrow from PSGL-1–deficient mice (black bars, n = 3 and n = 2, respectively). anti-P, P-selectin blocking mAb RB40.34; anti-PSGL-1, PSGL-1 blocking mAb 4RA10; anti-L, L-selectin blocking F(ab′)₂ of mAb MEL-14, anti-E, E-selectin blocking mAb 9A9. <1 h means within 1 h of exteriorization, all other data at >1 h after exteriorization. Significant differences (P < 0.05) in leukocyte rolling flux fraction between PSGL-1^−/− and wild-type group are indicated by *.

Figure 2.

L-selectin–dependent leukocyte rolling flux fraction (mean ± SEM) at >60 min after exteriorization of (A) the untreated (trauma) or (B) TNF-α–treated cremaster muscle in wild-type mice in which P-selectin (anti-P, black bars, n = 3), both P- and E-selectin (anti-P/E, black bars, n = 4) or CD18 (dark gray bars, n = 3 for trauma, n = 2 for TNF-α) was blocked before exteriorization of the cremaster muscle (pretreated). For comparison, L-selectin–dependent rolling in both models is shown (light gray bars, data from Fig. 1). Significant differences between different pretreatments (P < 0.05) indicated by *. To demonstrate that rolling is L-selectin–dependent in all cases, anti-L, L-selectin blocking F(ab′)₂ of mAb MEL-14 was added (lower group of bars in each panel), resulting in significantly reduced rolling flux fractions (indicated by #) compared with mice in which only P-selectin (A) or P- and E-selectin were blocked (B).

Figure 3.

Leukocyte rolling velocity distribution in untreated cremaster muscle venules of (A) wild-type mice treated with P-selectin blocking mAb RB40.34 (anti-P), (B) PSGL-1^−/− mice treated with anti-P, and (C) wild-type mice treated with anti-P and anti-L (L-selectin blocking F(ab′)₂ of mAb MEL-14) >60 min after exteriorization. Rolling was completely blocked by adding L- and E-selectin mAb in panel A, or E-selectin mAb in panels B and C. Absolute cell numbers are shown for leukocyte velocity groups with bin sizes of 10 μm/s.

Figure 4.

Some leukocytes rolling slowly along the endothelium (A) showed transient adhesion, during which time (here, 13 s) they pulled visible tethers of considerable length (B, arrow). Scale bar 5 μm. Supplemental video clip at http://www.jem.org/cgi/content/full/jem.20021854/DC1.