E-selectin engages PSGL-1 and CD44 through a common signaling pathway to induce integrin alphaLbeta2-mediated slow leukocyte rolling

Abstract

In inflamed venules, neutrophils rolling on E-selectin induce integrin alpha(L)beta(2)-dependent slow rolling on intercellular adhesion molecule-1 by activating Src family kinases (SFKs), DAP12 and Fc receptor-gamma (FcRgamma), spleen tyrosine kinase (Syk), and p38. E-selectin signaling cooperates with chemokine signaling to recruit neutrophils into tissues. Previous studies identified P-selectin glycoprotein ligand-1 (PSGL-1) as the essential E-selectin ligand and Fgr as the only SFK that initiate signaling to slow rolling. In contrast, we found that E-selectin engagement of PSGL-1 or CD44 triggered slow rolling through a common, lipid raft-dependent pathway that used the SFKs Hck and Lyn as well as Fgr. We identified the Tec kinase Bruton tyrosine kinase as a key signaling intermediate between Syk and p38. E-selectin engagement of PSGL-1 was dependent on its cytoplasmic domain to activate SFKs and slow rolling. Although recruiting phosphoinositide-3-kinase to the PSGL-1 cytoplasmic domain was reported to activate integrins, E-selectin-mediated slow rolling did not require phosphoinositide-3-kinase. Studies in mice confirmed the physiologic significance of these events for neutrophil slow rolling and recruitment during inflammation. Thus, E-selectin triggers common signals through distinct neutrophil glycoproteins to induce alpha(L)beta(2)-dependent slow rolling.

Figure 1

Slow rolling of neutrophils on E-selectin and ICAM-1 requires Hck and Lyn as well as Fgr, lipid rafts, and Btk, but not PI3K. (A) Velocities of WT or SFK-deficient neutrophils rolling on E-selectin with or without coimmobilized ICAM-1. (B) Velocities of WT neutrophils rolling on E-selectin with or without coimmobilized ICAM-1 in the presence or absence of the vehicle control DMSO, MβCD or its inactive analog α-cyclodextrin (αCD), MβCD plus 15% FBS, or filipin III. (C) Velocities of WT neutrophils rolling on E-selectin with or without coimmobilized ICAM-1 in the presence or absence of the vehicle control DMSO, the PI3K inhibitor LY294002, or the Tec kinase inhibitor LFM-A13. (D) Velocities of WT, Btk^−/−, SHIP1^−/−, PI3Kγ^−/−, or PI3Kδ^−/− neutrophils rolling on E-selectin with or without coimmobilized ICAM-1. (E) Numbers of WT, SHIP1^−/−, PI3Kγ^−/−, or PI3Kδ^−/− neutrophils rolling or firmly adherent (arrest) on coimmobilized E-selectin, ICAM-1, and CXCL1 in the presence or absence of the PI3K inhibitor LY294002. In all experiments, the E-selectin density was 200 sites/μm² and the ICAM-1 density was 240 sites/μm². The wall shear stress was 1 dyne/cm². The data represent the mean ± SEM from 5 experiments. *P < .01.

Figure 2

E-selectin engages PSGL-1 and CD44 through a common signaling pathway to slow neutrophil rolling. (A) Velocities of neutrophils from the indicated genotype rolling on E-selectin at the indicated site density with or without coimmobilized ICAM-1. (B) CD44^−/− neutrophils or (C) PSGL-1^−/− neutrophils rolling on E-selectin with or without coimmobilized ICAM-1 in the presence or absence of vehicle control DMSO, α-cyclodextrin (αCD), MβCD, or MβCD with serum. (D) CD44^−/− neutrophils or (E) PSGL-1^−/− neutrophils rolling on E-selectin with or without coimmobilized ICAM-1 in the presence or absence of DMSO, the Syk inhibitor piceatannol, the SFK inhibitor PP2 or its inactive analog PP3, the PI3K inhibitor LY294002, the Tec kinase inhibitor LFM-A13, or the p38 inhibitor SB203580. Except where indicated in panel A, the E-selectin density was 200 sites/μm². The ICAM-1 density was 240 sites/μm². The wall shear stress was 1 dyne/cm². The data represent the mean ± SEM from 5 experiments. *P < .01.

Figure 3

E-selectin engagement of PSGL-1 or CD44 activates SFKs in a lipid raft–dependent manner and then sequentially activates Syk, Btk, and p38. (A) WT leukocytes were rotated on E-selectin–IgM in the presence or absence of EDTA (ethylenediaminetetraacetic acid) or on control CD45-IgM for 5 minutes. Lysates were immunoprecipitated (IP) with antibody to the indicated SFK and then Western blotted (WB) with the same antibody or with anti–phospho-SFK (Y416). (B) Leukocytes of the indicated genotype were incubated as in panel A. Lysates were probed with antibody that recognizes all SFKs or with anti–phospho-SFK (Y416). (C) WT leukocytes treated with the indicated agent were incubated as in panel A. Lysates were probed with antibody that recognizes all SFKs or with anti–phospho-SFK (Y416). (D) WT leukocytes treated with the indicated agent or leukocytes of the indicated genotype were incubated as in panel A. Lysates were probed with antibody that recognizes all SFKs or with anti–phospho-SFK (Y416). (E) WT leukocytes treated with the indicated agent were incubated as in panel A. Lysates were probed with antibody to Syk or to phospho-Syk. (F) Lysates were Western blotted with antibody to Syk or to phospho-Syk. (G) Lysates were immunoprecipitated with antibody to Btk and then Western blotted with the same antibody or with antiphosphotyrosine antibody. (H) Lysates were probed with antibody to p38 or to phospho-p38. The E-selectin density was 400 sites/μm² for PSGL-1^−/−/CD44^−/− leukocytes and 200 sites/μm² for leukocytes of all other genotypes. The data are representative of at least 3 independent experiments.

Figure 4

E-selectin engagement of PSGL-1 requires its cytoplasmic domain to trigger slow rolling and to activate SFKs. (A) Velocities of CD44^−/− neutrophils, OSGE-treated CD44^−/− neutrophils, or ΔCD PSGL-1/CD44^−/− neutrophils rolling on E-selectin at the indicated density with or without coimmobilized ICAM-1 (240 sites/μm²). The inset shows the expression levels of PSGL-1 on cells from the indicated genotype as measured by flow cytometry. (B) Velocities of mock- or OSGE-treated PSGL-1^−/− neutrophils rolling on E-selectin (200 sites/μm²) with or without coimmobilized ICAM-1 (240 sites/μm²). (C) OSGE-treated CD44^−/− neutrophils or ΔCD PSGL-1/CD44^−/− neutrophils were rotated on E-selectin–IgM (400 sites/μm²) in the presence or absence of EDTA or on control CD45-IgM for 5 minutes. Lysates were Western blotted with antibody that recognizes all SFKs or with anti–phospho-SFK (Y416). The data are representative of at least 3 independent experiments. *P < .01.

Figure 5

E-selectin–mediated slow rolling in vivo requires PSGL-1 and CD44, all 3 SFKs, and Btk. (A) Velocities of leukocytes rolling in TNF-α–stimulated venules of mice of the indicated genotypes, measured before and after injecting a blocking mAb to P-selectin and then a blocking mAb to β2-integrins. (B-D) Velocities of PKH26- and PKH67-labeled leukocytes of the indicated genotypes in the same TNF-α–stimulated venules. The labeled cells were pretreated with PTx, and the mice were pretreated with anti–P-selectin mAb. (E) Lethally irradiated WT mice were injected with WT LysM-GFP⁺ bone marrow cells mixed with an equal number of GFP-negative WT or Btk^−/− cells. After 8 weeks, the mice were treated with PTx and anti–P-selectin mAb, and rolling velocities of GFP-positive and GFP-negative leukocytes were measured in the same TNF-α–stimulated venules. The data represent the mean ± SEM from at least 3 experiments. *P < .01.

Figure 6

Gα_i-independent neutrophil recruitment in vivo requires PSGL-1 and CD44, all 3 SFKs, and Btk. (A) WT, PSGL-1^−/−, CD44^−/−, or PSGL-1^−/−/CD44^−/− mice received 4 μg of PTx intravenously 2 hours before injection with 1 mL of 4% thioglycollate intraperitoneally. After 4 hours, peritoneal cells were collected, and the number of neutrophils was measured by flow cytometry. (B) WT leukocytes were labeled with PKH67 or PKH26. SFK-deficient leukocytes were labeled with PKH26. Untreated or PTx-treated WT mice were injected intraperitoneally with thioglycollate. After 2 hours, they were injected intravenously with a 1:1 mixture of PKH26- and PKH67-labeled leukocytes. After another 2 hours, blood and peritoneal cells were collected, and the number of neutrophils labeled with each dye was measured by flow cytometry. Results are plotted as the ratio of PKH26-labeled neutrophils from the indicated genotype to PKH67-labeled WT neutrophils. (C) Untreated or PTx-pretreated WT or Btk^−/− mice were injected with thioglycollate intraperitoneally. After 4 hours, peritoneal cells were collected, and the number of neutrophils was measured by flow cytometry. (D) Competitive homing was measured as in panel B, except that Btk^−/− leukocytes were used instead of SFK-deficient leukocytes. The data represent the mean ± SEM from at least 3 experiments. *P < .01.

Figure 7

Model for E-selectin–mediated slow rolling. The circled numbers represent new signaling components identified in this paper. Neutrophils rolling on E-selectin engage both CD44 and PSGL-1 to initiate signaling through a common pathway that requires lipid rafts, the cytoplasmic domain of PSGL-1, all 3 SFKs, the ITAM adaptors DAP12 and FcRγ, the Tec kinase Btk, and p38. This signaling cascade activates integrin leukocyte function-associated antigen 1 (LFA-1) to a conformation that enables slow rolling but not arrest on ICAM-1. PSGL-1 and CD44 may not be located in the same raft domains as depicted in the figure.