Surprising up-regulation of P-selectin glycoprotein ligand-1 (PSGL-1) in endotoxin-induced uveitis

Abstract

P-selectin glycoprotein ligand-1 (PSGL-1) is constitutively expressed on leukocytes and was thought to be down-regulated with cell activation. However, this work shows the surprising finding of functional PSGL-1 up-regulation during acute inflammation. PSGL-1 function was studied in our autoperfusion assay, in which blood from a mouse carotid flows through a microchamber coated with a fixed density of P-selectin. Under the inflammatory conditions--uveitis induced by systemic lipopolysaccharide injection--we recorded significantly reduced leukocyte rolling velocity, which suggests PSGL-1 up-regulation; however, flow cytometry showed reduced PSGL-1. When bound leukocytes were released from the vasculature by PSGL-1 blockade, a large peripheral blood leukocyte (PBL) population showed elevated PSGL-1, which could account for the reduced PSGL-1 in the remaining unbound population. In the eye, systemic blockade of PSGL-1 with a monoclonal antibody or recombinant soluble PSGL-1 drastically reduced the severe manifestations of uveitis. Furthermore, PSGL-1 blockade was significantly more effective in reducing retinal leukostasis than was P-selectin blockade. Our results provide surprising evidence for functional PSGL-1 up-regulation in PBLs during acute inflammation. The temporal overlap between PSGL-1 and P-selectin up-regulation reveals an as yet unrecognized collaboration between this receptor-ligand pair, increasing efficiency of the first steps of the leukocyte recruitment cascade.

Figure 1.

Schematic of the autoperfused micro-flow chamber assay, originally introduced by Hafezi-Moghadam et al. . The micro-flow chamber assembly, microsurgically connected to the carotid artery and jugular vein of a live mouse under anesthesia, creates a closed circuit with the animal’s vascular system. The native blood cells flow from the carotid artery, pass through the translucent area of the chamber for microscopy, and subsequently reenter the animal’s body through the jugular vein. The pressure gradient between the artery (high pressure) and the vein (low pressure) provides a continuous flow of cells through the chamber. Circulating PBLs interact with immobilized adhesion molecules (i.e., P-selectin at 5 μg/ml coating concentration) on translucent microslides. Regulation of the entry pressure (i.e., 40 mmHg) before the microslide provides controlled flow conditions. Interacting leukocytes are observed and recorded by using transluminescence video microscopy.

Figure 2.

Visualization of PSGL-1-mediated leukocyte rolling during EIU. A) Cumulative histograms of the velocities of rolling leukocytes on immobilized P-selectin in micro-flow chambers connected to control or EIU mice at various time points after LPS stimulation. The average rolling velocity was decreased in EIU animals 6 h after LPS injection (n=4) compared with control animals (n=6, P<0.05). However, 24 h after LPS injection, the average rolling velocity in EIU animals was similar to that of control animals (n=4, P=0.4). Each curve represents rolling velocities of 50 leukocytes. B) Composite micrographs of representative rolling leukocytes on immobilized P-selectin in the autoperfused micro-flow chamber: control mouse (a), EIU mouse 6 h after LPS injection (b), EIU mouse 24 h after LPS injection (c), and EIU mouse treated with anti-PSGL-1 mAb injected through the side port into the chamber, 6 h after LPS injection (d). The displacement of a representative rolling leukocyte from each group from the left border (0 μm) to the red dotted line within 15 s illustrates the differences in leukocyte rolling velocity.

Figure 3.

PSGL-1 expression on PBLs during EIU. A) Representative histograms of PSGL-1 expression on PBLs from control (CTR) or EIU mice at 6 and 24 h after LPS injection. Leukocytes were stained with anti-PSGL-1 (2PH1) or IgG1 control antibodies conjugated with PE. PBLs from EIU animals 6 h after LPS injection showed decreased PSGL-1 mean fluorescence intensity values compared with normal controls. In contrast, leukocytes from EIU animals at 24 h after LPS injection showed PSGL-1 mean fluorescence values similar to control animals. Each curve is representative of 4 independent experiments. B) Quantification of PSGL-1 mean fluorescence intensity in control and EIU animals at 6 and 24 h after LPS injection. N.S., not significant. **P < 0.01.

Figure 4.

PBL counts and PSGL-1 expression in EIU animals with anti-PSGL-1-neutralizing mAb. A) At 6 h after LPS injection, mice were treated with PSGL-1-neutralizing mAb (4RA10) or isotype control. The animal blood was harvested at 30, 120, and 180 min after antibody treatment, and the PBL counts were obtained. In parallel, leukocytes were evaluated for their PSGL-1 surface expression through indirect immunofluorescence. B) PSGL-1 surface expression on PBLs from EIU mice 6 h after LPS injection at 30, 120, and 180 min after antibody treatment. PBLs from animals 120 and 180 min after PSGL-1 blockade showed higher expression of PSGL-1 compared with the cells obtained 30 min after mAb treatment. C) Increased PSGL-1 surface expression on PBLs from EIU mice 6 h after simultaneous administration of PSGL-1-neutralizing mAb (4RA10) and LPS, compared with 4RA10-treated controls. Each curve is representative of 3 independent experiments.

Figure 5.

PSGL-1 blockade diminishes inflammatory cell infiltration into the aqueous humor during EIU. The number of infiltrated cells counted in 1 μl of aqueous humor of normal and EIU mice 24 h after LPS injection, treated with either anti-PSGL-1 mAb (n=9) or control IgG (n=7). Values are means ± se. *P < 0.05.

Figure 6.

Effect of PSGL-1 blockade on retinal leukostasis. A) Representative micrographs of flatmounted retinas from EIU mice at 24 h treated with isotype control mAb show a large number of firmly adhering leukocytes on the retinal vasculature (a, c; arrows). In contrast, anti-PSGL-1-treated EIU animals show very few firmly adhering leukocytes (b, d). B) Quantification of firmly adhering leukocytes in the retina of normal and EIU animals treated with IgG control or mAb against P-selectin or PSGL-1. EIU mice treated with anti-PSGL-1 mAb (n=12) showed significantly fewer firmly adhering leukocytes than those treated with either anti-P-selectin mAb (n=8) or IgG control (n=8). Results represent means ± se; *P < 0.05; **P < 0.01.

Figure 7.

Effect of PSGL-1 blockade on retinal expression of inflammatory mediators. A) TNF-α and MCP-1 expression in retinal tissue of EIU mice. Retinal mRNA expression of TNF-α and MCP-1 in control (CTR) and EIU mice, treated with PSGL-1-neutralizing mAb (Ab) or isotype-matched control IgG (Iso), 6 and 24 h after LPS injection. B, C) Semiquantitative analysis showing the relative amount of TNF-α (B) and MCP-1 mRNA expression (C) as a ratio of the gene product relative to the expression of the GAPDH (arbitrary units). Each data point represents an average of 6 to 10 experiments. *P < 0.05; **P < 0.01.

Figure 8.

Effect of competitive blockade of PSGL-1 on inflammatory outcome during EIU. A) Leukocyte infiltration into the aqueous humor. Number of infiltrated cells was counted in 1 μl of aqueous humor of EIU mice at 24 h after LPS injection and treatment with rPSGL-Ig (n=10) or as a control with human IgG (n=12). Values are means ± se. **P < 0.01. B) Representative micrographs of flatmounted retinas from human IgG-treated (a) and rPSGL-Ig-treated EIU mice (b). rPSGL-Ig suppressed firm leukocyte adhesion to the retinal vasculature. Arrows indicate firmly adhering leukocytes. C) Quantification of adherent retinal leukocytes. EIU mice treated with rPSGL-Ig (n=9) showed significant reduction of adherent leukocytes to the retinal vessels compared with IgG-treated EIU mice (n=6). **P < 0.01.