Crucial role of the protein C pathway in governing microvascular inflammation in inflammatory bowel disease

Abstract

Endothelial protein C receptor (EPCR) and thrombomodulin (TM) are expressed at high levels in the resting microvasculature and convert protein C (PC) into its activated form, which is a potent anticoagulant and antiinflammatory molecule. Here we provide evidence that in Crohn disease (CD) and ulcerative colitis (UC), the 2 major forms of inflammatory bowel disease (IBD), there was loss of expression of endothelial EPCR and TM, which in turns caused impairment of PC activation by the inflamed mucosal microvasculature. In isolated human intestinal endothelial cells, administration of recombinant activated PC had a potent antiinflammatory effect, as demonstrated by downregulated cytokine-dependent cell adhesion molecule expression and chemokine production as well as inhibited leukocyte adhesion. In vivo, administration of activated PC was therapeutically effective in ameliorating experimental colitis as evidenced by reduced weight loss, disease activity index, and histological colitis scores as well as inhibited leukocyte adhesion to the inflamed intestinal vessels. The results suggest that the PC pathway represents a new system crucially involved in governing intestinal homeostasis mediated by the mucosal microvasculature. Restoring the PC pathway may represent a new therapeutic approach to suppress intestinal inflammation in IBD.

Figure 1. Immunohistochemical staining of EPCR and TM in colons of normal and actively inflamed IBD patients.

The panels show brown immunohistochemical staining for EPCR and TM in the microvasculature of colonic mucosa and submucosa from histologically normal control (A and B), active IBD (C and D), and noninflamed IBD (E and F) tissues. Images are representative of 12 control, 14 UC, and 15 CD samples (original magnification, ×10). Red arrows indicate intestinal microvessels.

Figure 2. Localization of EPCR in IBD mucosa.

Fluorescence micrographs of colonic mucosa microvessels in normal mucosa. (A–C) Staining for EPCR (green, A), von Willebrand factor (red, B), and nuclei (blue, C). (D) Merge (yellow) of panels A–C. (E and F) Binary mask of colocalization (white, E) and colocalization of panels B and E (F). Images are representative of 6 control, 7 CD, and 8 UC samples. Scale bars: 60 μm.

Figure 3. Effect of different cytokines on EPCR and TM expression by HIMECs.

(A) HIMEC monolayers were left untreated (Baseline) or stimulated with TNF-α or IL-10 for 24 hours, after which cells were suspended and EPCR and TM expression measured by flow cytometry. Filled curve represents background signal from the isotype control. Values are representative of 6 separate experiments. Numbers represent the net percentage of positive cells. (B) HIMEC monolayers were left untreated or stimulated with TNF-α for 24 hours, after which supernatants were collected for ELISA measurement of soluble EPCR and TM. Cellular mRNA was extracted for real-time PCR of EPCR and TM message levels. Values are representative of 3 separate experiments. *P < 0.05, **P < 0.01 versus baseline.

Figure 4. TNF-α–inflamed HIMECs exhibit a reduced capacity for PC activation.

HIMEC monolayers were incubated 24 hours with or without TNF-α and assayed for their ability to support activation of PC by thrombin. Generation of aPC was measured using the chromogenic tripeptidyl-pNA substrate S-2366; relative capacity for PC activation is presented as change in absorbance at 405 nm over time (mOD/min). Values are representative of 3 separate experiments. *P < 0.05 versus baseline.

Figure 5. Downregulation of VCAM-1 and ICAM-1 by recombinant aPC in TNF-α–inflamed HIMECs.

HIMEC monolayers were left untreated or stimulated with TNF-α in the presence or absence of recombinant human aPC. After 24 hours, cells were suspended and VCAM-1 and ICAM-1 expression was measured by flow cytometry. Filled curve represents background signal from the isotype control. Values are representative of 5 separate experiments. Numbers represent the net percentage of positive cells.

Figure 6. Downregulation of chemokine production by recombinant aPC in TNF-α–inflamed HIMECs.

HIMEC monolayers were left untreated or stimulated with TNF-α in the presence or absence of recombinant human aPC. After 24 hours, supernatants were collected, and ENA-78, IL-8, and MCP-1 were measured by ELISA. Values are representative of 5 separate experiments. *P < 0.05 versus TNF-α without aPC.

Figure 7. Inhibition of T cell adhesion to TNF-α–inflamed HIMECs by recombinant aPC.

HIMEC monolayers were left untreated or stimulated with TNF-α with or without recombinant human aPC. Calcein-labeled MOLT4 T cells were added to the HIMEC monolayers. Values are mean ± SEM of 4 separate experiments. *P < 0.05 versus TNF-α without aPC.

Figure 8. Dextran sodium sulfate (DSS) colitic mice exhibit a reduced capacity for PC activation.

Healthy and colitic mice (n = 3 per group) were injected i.v. with exogenous human PC. After 30 minutes, blood was collected into sodium citrate anticoagulant supplemented with benzamidine to protect aPC from inhibition by endogenous irreversible inhibitors. Using the immunocapture assay described in Methods, aPC was measured using the chromogenic tripeptidyl-pNA substrate S-2366; relative capacity for PC activation is presented as the change in absorbance at 405 nm over time. *P < 0.05 versus healthy mice.

Figure 9. Therapeutic effects of murine recombinant aPC administration.

Animals undergoing DSS treatment were given daily i.v. injections of recombinant murine aPC (1 mg/kg; n = 10) or placebo (n = 12). Weight was monitored (A) as well as disease activity index (DAI) (B). After 10 days, mice were sacrificed, and colons were assessed for histological colitis (C and D). *P < 0.05, aPC versus placebo.

Figure 10. aPC inhibits CAM expression, IL-8 production, and leukocyte adhesion to the endothelium in the intestine.

Animals undergoing DSS treatment were given daily i.v. injections of recombinant murine aPC or placebo. After 10 days mice were sacrificed, and biopsies were collected for Western blot analysis for VCAM-1 and ICAM-1 (A) or organ culture for IL-8 measurement by ELISA (B). (C) Leukocyte–endothelial cell interactions were assessed by intravital microscopy in 3 groups of animals: healthy mice (n = 5); DSS colitic mice treated for 10 days with placebo (n = 8); and DSS colitic mice treated for 10 days with daily i.v. injections of recombinant aPC (n = 6). For each animal, 3–6 unbranched venules were studied and the mean value was calculated. Leukocyte adhesion is expressed as the number of firmly adherent leukocytes to the endothelium per 100 μm of venule. *P < 0.05 versus placebo; **P < 0.001 versus healthy mice.