P-Selectin glycoprotein ligand 1 (PSGL-1) is a physiological ligand for E-selectin in mediating T helper 1 lymphocyte migration

Abstract

P-selectin glycoprotein ligand 1 (PSGL-1) is a sialomucin expressed on leukocytes that mediates neutrophil rolling on the vascular endothelium. Here, the role of PSGL-1 in mediating lymphocyte migration was studied using mice lacking PSGL-1. In a contact hypersensitivity model, the infiltration of CD4(+) T lymphocytes into the inflamed skin was reduced in PSGL-1-deficient mice. In vitro-generated T helper (Th)1 cells from PSGL-1-deficient mice did not bind to P-selectin and migrated less efficiently into the inflamed skin than wild-type Th1 cells. To assess the role of PSGL-1 in P- or E-selectin-mediated migration of Th1 cells, the cells were injected into E- or P-selectin-deficient mice. PSGL-1-deficient Th1 cells did not migrate into the inflamed skin of E-selectin-deficient mice, indicating that PSGL-1 on Th1 cells is the sole ligand for P-selectin in vivo. In contrast, PSGL-1-deficient Th1 cells migrated into the inflamed skin of P-selectin-deficient mice, although less efficiently than wild-type Th1 cells. This E-selectin-mediated migration of PSGL-1-deficient or wild-type Th1 cells was not altered by injecting a blocking antibody to L-selectin. These data provide evidence that PSGL-1 on Th1 cells functions as one of the E-selectin ligands in vivo.

Figure 1

Histological analysis of contact hypersensitivity response. (A–C) Hematoxylin and eosin–stained sections from control (A) or challenged ears (B) and anti-CD4–stained frozen ear sections from the challenged ears (C) from sensitized PSGL-1^+/+ (left panel) or PSGL-1^−/− mice (right panel). Ear specimens were taken 24 h after challenge. The scale bar indicates 100 μm for A, B top panel, and C; 25 μm for B bottom panel. (D) The number of infiltrating cells in the dermis of the control and challenged ears was quantitated. Data are means ± SEM from six mice. (E) The number of infiltrating CD4⁺ T cells was quantitated on frozen sections stained with anti-CD4. Data are means ± SEM from four mice.

Figure 1

Histological analysis of contact hypersensitivity response. (A–C) Hematoxylin and eosin–stained sections from control (A) or challenged ears (B) and anti-CD4–stained frozen ear sections from the challenged ears (C) from sensitized PSGL-1^+/+ (left panel) or PSGL-1^−/− mice (right panel). Ear specimens were taken 24 h after challenge. The scale bar indicates 100 μm for A, B top panel, and C; 25 μm for B bottom panel. (D) The number of infiltrating cells in the dermis of the control and challenged ears was quantitated. Data are means ± SEM from six mice. (E) The number of infiltrating CD4⁺ T cells was quantitated on frozen sections stained with anti-CD4. Data are means ± SEM from four mice.

Figure 2

Migration of PSGL-1–deficient Th1 cells into the inflamed skin is impaired. ⁵¹Cr-labeled PSGL-1^+/+ or PSGL-1^−/− Th1 cells were injected into the tail veins of PSGL-1^+/+ mice. The mice had been sensitized 6 d before with oxazolone and challenged 24 h before on the left ear. The mice were killed 3 h after injection, and the radioactivity in the control and challenged ears was counted. Values are means ± SEM from four mice.

Figure 3

Selectin-binding activities of PSGL-1–deficient Th1 cells. In vitro–generated Th1 cells from PSGL-1^+/+ or PSGL-1^−/− mice were labeled with BCECF and added to 96-well plates coated with human IgG, P-selectin–IgG chimera (P-IgG), or E-selectin–IgG chimera (E-IgG). The plates were incubated for 20 min at 4°C under rotating conditions, unbound cells were removed, and the fluorescence per well was determined. Percent adhesion equals 100 × bound cells/total cells added. (A) Adhesion of PSGL-1^+/+ or PSGL-1^−/− unstimulated CD4⁺ cells and Th1 cells to selectin–IgG chimeras. (B) Adhesion of PSGL-1^+/+ Th1 cells treated with anti–PSGL-1 antibodies to selectin–IgG chimeras. PSGL-1^+/+ Th1 cells were incubated with either rabbit anti–PSGL-1 or nonimmune IgG for 30 min on ice and washed before addition to the wells. Values are means ± SEM from triplicate wells.

Figure 4

Migration of PSGL-1–deficient Th1 cells into the inflamed skin of E-selectin– or P-selectin–deficient mice. ⁵¹Cr-labeled PSGL-1^+/+ or PSGL-1^−/− Th1 cells were injected into the tail veins of E-selectin^−/− mice and wild-type B6/129S F2 mice (A) or P-selectin^−/− mice and wild-type B6 mice (B). The mice were killed 3 h after injection, and the radioactivity in the control and challenged ears was assayed. Values are means ± SEM from four mice.

Figure 4

Migration of PSGL-1–deficient Th1 cells into the inflamed skin of E-selectin– or P-selectin–deficient mice. ⁵¹Cr-labeled PSGL-1^+/+ or PSGL-1^−/− Th1 cells were injected into the tail veins of E-selectin^−/− mice and wild-type B6/129S F2 mice (A) or P-selectin^−/− mice and wild-type B6 mice (B). The mice were killed 3 h after injection, and the radioactivity in the control and challenged ears was assayed. Values are means ± SEM from four mice.

Figure 5

Effect of an anti–L-selectin antibody on Th1 migration in vivo and adhesion in vitro. (A) ⁵¹Cr-labeled PSGL-1^+/+ or PSGL-1^−/− Th1 cells were injected together with anti–L-selectin mAb MEL-14, anti–E-selectin mAb 9A9, or its isotype control. The mice had been sensitized 6 d before with oxazolone and challenged 24 h before on the left ear. Values are means ± SEM from three mice. (B) Adhesion of PSGL-1^+/+ or PSGL-1^−/− Th1 cells treated with MEL-14 to selectin–IgG chimeras. PSGL-1^+/+ or PSGL-1^−/− Th1 cells were incubated with either MEL-14 or its isotype control for 30 min on ice and washed before addition to wells. Values are means ± SEM from triplicate wells.

Figure 5

Effect of an anti–L-selectin antibody on Th1 migration in vivo and adhesion in vitro. (A) ⁵¹Cr-labeled PSGL-1^+/+ or PSGL-1^−/− Th1 cells were injected together with anti–L-selectin mAb MEL-14, anti–E-selectin mAb 9A9, or its isotype control. The mice had been sensitized 6 d before with oxazolone and challenged 24 h before on the left ear. Values are means ± SEM from three mice. (B) Adhesion of PSGL-1^+/+ or PSGL-1^−/− Th1 cells treated with MEL-14 to selectin–IgG chimeras. PSGL-1^+/+ or PSGL-1^−/− Th1 cells were incubated with either MEL-14 or its isotype control for 30 min on ice and washed before addition to wells. Values are means ± SEM from triplicate wells.