Organ-specific protection against LPS-induced vascular leak is dependent on the endothelial protein C receptor (EPCR)

Abstract

Objective

To study the role of the endothelial protein C receptor (EPCR) in the modulation of susceptibility to inflammation-induced vascular leak in vivo.

Approach/Results

Genetically modified mice with low, <10%-EPCR expression (EPCR^low) and control mice were challenged with lipopolysaccharides (LPS) in a mouse model of endotoxemia. Infrared fluorescence and quantification of albumin-bound Evans Blue (EB) in tissues and intravascular plasma volumes were used to assess plasma extravasation. Pair wise analysis of EPCR^low and control mice matched for gender, age, and weight allowed determination of EPCR-dependent vascular leak. Kidney, lung, and brain were the organs with highest discriminative increased EB accumulation in EPCR^low versus control mice in response to LPS. Histology of kidney and lung confirmed the EPCR-specific pathology. In addition to severe kidney injury in response to LPS, EPCR^low and anti-EPCR-treated wild-type mice suffered from enhanced albuminuria and profound renal hemorrhage versus controls. Intravascular volume loss at the same extent of weight loss in EPCR^low mice compared to control mice provided proof that plasma leak was the predominant cause of EB tissue accumulation

Conclusions

This study demonstrates an important protective role for EPCR in vivo against vascular leakage during inflammation and suggests that EPCR-dependent vascular protection is organ specific.

Introduction

The endothelial protein C receptor (EPCR)is a predominantly vascular cell receptor. EPCR binds both protein C (PC) zymogen circulating and the serine protease activated protein C (APC) with similar affinity, thereby localizing PC and APC to the cell surface.^{1, 2} Although EPCR is considered a non-signaling receptor due to its short (Arg-Arg-Cys) intracellular tail, it is increasingly recognized as key receptor in the anticoagulant and cytoprotective protein C pathways by facilitating (A)PC's interactions. Binding of PC to EPCR facilitates activation of PC to APC by the thrombin-thrombomodulin complex.³ Activated protein C (APC) acts as anticoagulant, and it also conveys cytoprotective functions that require EPCR and protease activated receptor 1 (PAR1) and include anti-inflammatory and anti-apoptotic activities and endothelial barrier stabilizing effects.^4-7 The important contributions of the protein C pathway are evident from the severe thrombotic and inflammatory complications in newborns with homozygous protein C deficiency that typically present as neonatal purpura fulminans with microvascular thromboembolism and necroses.^{8, 9}

In vivo murine endotoxemia and sepsis studies showed that a cytoprotective-selective APC mutant reduced mortality equivalent to wild-type APC, whereas an anticoagulant-selective APC mutant did not.^10-12 Since the mortality reducing activity of APC in murine sepsis models required EPCR and PAR1, it suggested that the EPCR-mediated PAR1-cleavage by APC is likely a major contributing factor to mortality reduction in sepsis.⁶

In addition to therapeutic cytoprotection when APC is administered exogenously at high doses, there is mounting evidence that the protein C pathway involving EPCR plays an important role for endogenous cytoprotection during inflammation. Endogenous depletion of PC or APC by inhibitory antibodies blocking selectively APC's proteolytic or cytoprotective activities increased susceptibility to death in animal models of endoxemia and sepsis (mice and baboons).^{13, 14} Susceptibility to death during LPS-induced mortality in mice with genetically low PC expression (10-50% of normal) or low EPCR expression (<10% of normal; EPCR^low) is significantly increased,^{11, 15, 16}while mice over-expressing EPCR are resistant to LPS-induced mortality.¹⁷ In humans, acquired PC deficiency during sepsis is associated with higher organ failure rates and higher mortality. PC depletion is thought to be caused by liver failure, consumptive coagulopathy, and possibly by EPCR shedding into circulation.¹⁸ Serum concentrations of soluble EPCR in humans are elevated in sepsis and inflammatory autoimmune disorders consistent with the observations that genetic or acquired defects of the endogenous protein C pathway exacerbate the negative effects of inflammatory disease and sepsis.^18-20

To date, the mechanisms for EPCR-dependent protection against mortality during endoxemia and sepsis by APC are incompletely understood. According to the current paradigm, EPCR plays a central role by localizing APC to caveolin-1 rich microdomains and acting as a cofactor for APC-mediated PAR1 cleavage that results in activation of cytoprotective signaling cascades.^{4, 7} However, the functions of EPCR in normal hemostasis and cytoprotection are likely more complex. For example, APC-mediated down-regulation of monocytes procoagulant activity that is dependent on APC-ligation to Apolipoprotein E Receptor 2 required EPCR but not PAR1, whereas APC protective effects that were dependent on CD11b/CD18 were found to be independent of EPCR.^{21, 22} In vitro, the presence of EPCR helps to confer endothelial barrier integrity and is required for APC-mediated barrier protection that involves transactivation of the sphingosine-1-phosphate receptor 1.^23-25

Since vascular barrier disruption, fluid extravasation and rising interstitial tissue pressures are increasingly considered to be important pathophysiological determinants of organ failure and death,²⁶ we studied the role of EPCR for vascular permeability in vivo using our novel Infrared Fluorescence (IRF) imaging methodology for the determination of Evans Blue (EB) plasma extravasation into organs.²⁷ Application of this methodology to study the contribution of EPCR to vascular leakage, tissue edema and organ injury provide unique insights into the role of EPCR in the pathophysiological induction of vascular leakage during endotoxemia. Low EPCR expression resulted in profound plasma tissue extravasation, lung injury, severe renal hemorrhage and albuminuria. In addition our data point towards an important role for EPCR in the organ-specific modulation of susceptibility to inflammatory disease, demonstrating that EPCR, hence the endogenous cytoprotective PC pathway, provides an important defense mechanism in vivo against vascular leak in inflammation and sepsis.

Results

Determination of lethality in C57Bl/6J mice with LPS from E.coli (serotype 0111:B4)

LPS derivatives from different bacteria as well as different serotype preparations can vary widely in their potency to cause dose-dependent inflammatory reactions and death. Therefore, percent lethality was determined for increasing doses of LPS from E.coli (0111:B4) in C57Bl/6J mice in order to relate degrees of vascular permeability to lethality rates in subsequent experiments. Females (n=186) and males (n=95) were injected with 4-5 dose levels of LPS ranging between 2 and 20 mg/kg. Seven-day lethality for both genders was recorded and expressed in lethality doses (LD) for each dose level of LPS (Figure 1). Since male C57Bl/6J mice were more resistant to LPS than female C57Bl/6J mice, gender-specific dosing for LPS-injection was used to achieve the same percent lethality for males or females in all subsequent experiments (male mice: LD25 (5.8 mg/kg), LD40 (8.9 mg/kg) and LD50 (10.8 mg/kg) and female mice: LD25 (3.7 mg/kg), LD40 (5.2 mg/kg) and LD50 (6.9 mg/kg)). In order to ensure that an inflammatory response to LPS was present even when low doses of LPS were injected, weight loss as a surrogate marker for sickness was determined in C57Bl/6J mice at increasing doses of LPS, 18 hours after i.p. injection. Weight loss was similar at all dose levels of LPS (LD10-LD80) and was present even if mice were not visibly sick. Weight loss was specific to treatment with LPS since it was not observed in control animals injected with similar amounts of denatured protein or rat IgG (Supplemental Figure 1).

Figure 1. Gender specific LPS-induced mortality rates in C57Bl/6J mice.

Gender specific survival curves were constructed by injection of increasing doses of LPS i.p. in female C57Bl/6J mice (n=186; 30-44 mice per group) or male C57Bl/6J mice (n=95; 16-23 mice per group). Lethality increased with increasing doses of LPS and was clearly different between (A) female and (B) male C57Bl/6J mice. Seven-day mortality rates for the various doses of LPS were assessed by Kaplan-Meier log-rank test and plotted against the LPS dose to determine the LD50 for (C) female and (D) male C57Bl/6J mice.

EB dye accumulation in organs of EPCR^low mice

To determine the role of EPCR in the modulation of vascular integrity, EPCR^low and control mice that were matched for age, weight and gender were injected as pairs with i.p saline (baseline; n=12), LPS at LD25 (n=8), LD40 (n=14) or LD50 (n=4). Since LPS doses exceeding LD60-70 caused high mortality (∼70%) during the first 2 days, with animals too sick for vascular leakage experimentation, we limited our vascular permeability studies to LD25 (“low dose LPS”) and LD40-50 (“medium dose LPS”). Organ pairs of EPCR^low and control mice were compared directly and results expressed as relative -fold increase of IRF in the organ of an EPCR^low mouse compared to its matched control (n=119 organ pairs of liver, spleen, lung, kidney, heart and brain).

At baseline (saline injection), EB accumulation in organs of EPCR^low was similar to that of control mice (mean –fold increase 1.1), indicating that low EPCR expression did not result in loss of vascular barrier function during normal physiology or that compensatory mechanisms are sufficient to overcome the deleterious effect of low EPCR expression (Supplemental Figure 2). However, challenging vascular barrier function by administration of LPS resulted in increasing EB accumulation in organs of EPCR^low compared to control mice. At low dose LPS, mean EB accumulation in organs of EPCR^low compared to control mice was increased 1.6-fold (p=0.08) from baseline (saline) and 2.0-fold (p<0.0006) at a medium dose of LPS (Supplemental Figure 2). Methodology analyzing organ pairs of matched EPCR^low and control mice affords certainty that the increased EB accumulation reflects only EPCR-dependent effect and not LPS-induced changes that are similar in EPCR^low and control mice. Furthermore, at a medium dose of LPS 80% (p=0.006) of organ pairs showed significantly increased EB accumulation from baseline in EPCR^low mice compared to control mice (Supplemental Figure 2). To determine in which organ systems vascular barrier function was most affected by low EPCR expression, individual organ systems were compared (Figure 2). Medium dose LPS-induced EB accumulation in EPCR^low mice was significantly increased in the brain (3.4-fold increase; CI 1.4-5.5), followed by kidney, lung (both 2-fold increase; CI 1.2-2.8), and spleen (1.5-fold increase; CI 1.1-1.9). Remarkably, no EPCR-dependent increase of EB accumulation was observed in the liver after medium dose LPS, suggesting that changes in vascular barrier function in the liver induced by LPS are not exacerbated by low EPCR expression (Figure 2).

Figure 2. Organ-specific sensitivity to LPS-induced EB accumulation in EPCR^low mice compared to control mice.

EPCR^low and matched wild-type controls were injected with (A) i.p saline (n=12; 6 pairs), (B) low dose LPS (n=8; 4 pairs) or (C) medium dose LPS (n=18; 9 pairs). Organs were harvested 18 h later, 30′ after i.v. injection of EB. EB accumulation was expressed as the EB fluorescence at 700 nm in relation to organ weight and peak total plasma IRF. Each single data point represents the relative difference of EB dye accumulation in the organ of an EPCR^low mouse compared to its matched control mouse (Y-axis: 1= no difference). Wilcoxon matched-pairs signed rank test was used for analyses of differences. Error bars represent SEM. NS denotes no significant difference between EPCR^low and control mice.

Increased intravascular volume contraction in EPCR^low mice after LPS

In order to investigate if EB dye accumulation in organs of EPCR^low mice was due to direct plasma extravasation into tissues as opposed to blood pooling or dehydration, the degree of intravascular plasma contraction in relation to weight loss was determined. Compared to control mice (n=9) where hematocrits before and after LPS (medium dose) were not different (mean hematocrit 44.5% vs. 45.6%), hematocrits in EPCR^low mice (n=9) rose significantly from a mean of 47.3% to 57.7% (p=0.0005) (Figure 3). No significant loss of intravascular plasma volume was present after injection of low dose LPS (data not shown). Remarkably, in 4 of the 9 EPCR^low mice hematocrits exceeded 60%. In parallel, average total plasma volumes decreased significantly in EPCR^low mice from 34.3 mL/kg before LPS to 27.5 mL/kg after LPS (p=0.004). In contrast, plasma volumes remained similar in control mice (36.1 mL/kg vs. 35.3 mL/kg). Percent weight loss after LPS was similar for EPCR^low and control mice (6.5 % vs.7.5%; p=0.6) (Figure 3). Taken together, the significantly decreased intravascular plasma volumes in EPCR^low mice after LPS at the same degree of weight loss compared to wild-type mice is diagnostic of plasma extravasation into tissues. Neither dehydration nor blood pooling could have been important mechanisms of EB tissue accumulation since dehydration would have been associated with increased weight loss, and blood pooling would have not affected the hematocrit. Thus, low EPCR expression results in a significant increase in hematocrits and contraction of intravascular plasma volumes after LPS challenge.

Figure 3. EPCR^low mice demonstrated significant intravascular volume contraction in response to LPS.

EPCR^low and control mice (each group n=9) were injected with a medium dose LPS. (A) Hematocrit and (C) weight were determined before and 18 h after LPS and served to calculate (B) plasma volumes. Student t-test was used for analyses of differences (* denotes p<0.01). Error bars represent SEM.

EPCR^low mice suffered from severe renal injury and albuminuria

To define the pathology associated with EPCR-dependent plasma extravasation in response to LPS, histology was performed on kidneys, lung, and liver. Kidneys of EPCR^low and matched control mice receiving LPS (medium dose) or saline (baseline) were harvested and processed for histology immediately after IRF imaging. Compared to control mice, kidneys of EPCR^low mice were found to have significantly increased EB dye accumulation after LPS (Figure 2) and demonstrated pronounced renal injury, which was absent in wild-type mice (Figure 4). No differences in organ histology were present at baseline (data not shown). After LPS, kidneys of EPCR^low mice were visually swollen and had a significantly increased kidney-to-body weight ratio compared to control mice (Figure 4). Histological analysis revealed profound renal hemorrhage at the cortico-medullary junction in kidneys of EPCR^low mice that was not obvious in kidneys of control mice receiving LPS (Figure 4). In addition, EPCR^low mice had significantly increased albuminuria (2.6-fold (p=0.03)) compared to controls after LPS, whereas no difference was observed at baseline (Figure 5). Albuminuria, expressed as IRF/mL urine, was assessed at baseline and after LPS (medium dose) by determination of EB accumulation in harvested bladders in relation to bladder weight (= urine volume) and peak total plasma IRF. Since albuminuria under physiological conditions is negligible, the appearance of albumin in urine after LPS is therefore pathognomonic of glomerular permeability.

Figure 4. EPCR^low mice suffered from severe renal injury following LPS.

Shown is a representative experiment where EPCR^low mice and matched control mice received medium dose LPS (n=18; 9 pairs). Kidneys were harvested 18 h later, 30′ after i.v. EB (25 mg/kg). (A) Odyssey scans at 700 nm of kidneys cut in half (top) and EB quantitation in relation to organ wet weight and peak total plasma IRF (bottom). (B) Kidneys of EPCR^low mice appeared swollen and their kidney/body weight ratios were higher compared to wild-type mice. Histology of kidneys and stains with (C) Carstairs and (D) H/E revealed severe hemorrhage in the kidneys of EPCR^low mice while kidneys of control mice appeared unaffected. On both stains, red cells stain orange red (arrow indicates islands of red cell extravasation). Student t-test was used for analyses of differences (* denotes p≤0.05). Error bars represent SEM.

Figure 5. EPCR^low mice developed significant albuminuria after LPS-challenge.

Albuminuria was assessed in EPCR^low and matched control mice at baseline (i.p. saline (n=12; 6 pairs) and after injection of medium dose of LPS (n=14; 7 pairs). EB in harvested bladders was quantified 18 hours after LPS-injection and 30′ after injection of EB by IRF imaging in relation to bladder weight (= urine volume; RFI/mL urine) and peak total plasma IRF and expressed as RFI/mL urine. (A) Relative change in albuminuria between pairs of EPCR^low and control mice at baseline (saline injection) and after LPS. Each single data point represents the relative difference of EB IRF in the urine of an EPCR^low mouse compared to its matched control mouse (Y-axis: 1= no difference). (B) IRF imaging and photographs depicting the bladders of EPCR^low and control mice after LPS. Wilcoxon matched-pairs signed rank test for non-parametric data was used for analyses of differences (* denotes p<0.01). Error bars represent SEM.

In addition to kidneys, also lungs showed increased EPCR-dependent EB accumulation in response to LPS (Figure 2). Histological analysis was consistent with more pronounced lung edema and injury to lungs of EPCR^low mice compared to controls. Lungs of EPCR^low mice showed evidence of septal thickening with accumulation of polymorphonuclear leukocytes, karyorhectic debris and, increased alveolar fibrillar materials that were not obvious in lungs of control mice receiving LPS (Figure 6).

Figure 6. EPCR^low mice suffered from increased lung permeability and injury following LPS.

Shown are two representative experiments where EPCR^low and matched control mice received medium dose LPS (n=8; 4 pairs). Lungs were harvested 18 h later, 30′ after i.v. EB (25 mg/kg). (A) Photographs demonstrated increased EB color in lungs of EPCR^low compared to wild-type mice. (B) EB was quantified by IRF spectroscopy for each pair of mice in relation to organ wet weight and peak total plasma IRF. (C) Odyssey scans at 700 nm of the lungs. (D) Histology (H/E) of lungs revealed increased septal thickening with accumulation of polymorphonuclear leukocytes and karyorhectic debris (black arrow), and increased alveolar fibrillar material (blue arrow) consistent with more pronounced lung edema in EPCR^low mice.

Remarkably, in contrast to kidney and lung, no EPCR-dependent differences in relative EB accumulation were observed in the liver after LPS. Recently we found that in C57Bl/6J (control) mice areas of LPS-induced EB dye extravasation in the liver corresponded to focal areas of liver necrosis.²⁷ Histological analysis of livers of EPCR^low mice injected with LPS (medium dose) showed similar findings of focal areas of liver necrosis corresponding to areas of EB dye extravasation (Supplemental Figure 3). Moreover, at baseline (saline injection) livers of EPCR^low and control mice appeared similar (data not shown). Thus, histological findings of relatively similar liver pathology in EPCR^low and control mice after LPS are consistent with the observed lack of any EPCR-dependent difference in EB accumulation in the liver. Notwithstanding these similarities, histology of the liver did reveal a major difference between EPCR^low and control mice after LPS. Whereas the necrotic areas of the liver were infiltrated with polymorphic mononuclear cells in control mice, infiltration of corresponding necrotic areas with polymorphic mononuclear cells in livers of EPCR^low mice was notably less (Supplemental Figure 3).

EB dye accumulation in organs of mice treated with anti-mouse EPCR antibody

We next determined whether EPCR-dependent sensitivity to LPS-induced vascular leak could also be revealed in C57Bl/6J wild-type mice using a blocking antibody against mouse EPCR (rcr-16)^{28, 29}. Wild-type mice that were matched for age, weight and gender received rcr-16 or vehicle control (saline) 2 hours before LPS at LD50 (n= 8 pairs; 16 mice). Organ pairs of antibody-treated and control mice were compared directly and results expressed as relative -fold increase of IRF in the organ of an antibody-treated mouse compared to its matched control. As EPCR^low mice, wild-type mice treated with blocking EPCR antibodies demonstrated increased EB accumulation in organs compared to control mice. The pattern of organ specific EPCR-dependent sensitivity to LPS-induced vascular leak was attenuated, but generally similar to that observed in EPCR^low mice. (Supplemental Figure 4). Notably, significant albuminuria and EB accumulation were found in kidneys of mice treated with blocking EPCR antibody compared to saline. A 2.3-fold (CI 1.4-3.5) increase of EB accumulation in kidneys of mice treated with EPCR blocking antibody compared to saline was present, and albuminuria was 3-fold (CI 1.6-4.4) increased, respectively. These results were comparable to the renal findings in EPCR^low mice. Thus, blocking EPCR function in wild-type mice or using EPCR^low mice identified the kidney as an organ with high sensitivity to EPCR-dependent protection against LPS-induced vascular leak. This corroborates that the observed discriminative effects between EPCR^low and control mice in the kidney involved EPCR-dependent mechanisms and were not the result of a developmental defect caused by low EPCR expression or an unknown genotype difference between EPCR^low and wild-type mice.

Discussion

This study emphasizes the important protective role EPCR plays in vivo in vascular barrier stabilization during inflammation.^{23, 24} While EPCR was demonstrated to be required for endothelial barrier protection in vitro, this function of EPCR has not been previously established in vivo. The role of EPCR for vascular leakage in vivo was determined using our recently developed IRF technology for quantification of EB accumulation in organs.²⁷ Pair wise analysis of EPCR^low and control mice that were matched for gender, age and weight allowed for the specific detection of EPCR-dependent differences in vascular barrier function in an LPS-induced mouse model of endotoxemia. Proof of plasma leak as the predominant cause of EB tissue accumulation was provided by demonstration of profound intravascular volume loss at the same extent of weight loss in EPCR^low mice compared to wild-type mice, thus excluding dehydration for the volume contraction, or blood pooling for accumulation of EB dye in central organs.

Kidney and lung were the solid organs that showed the highest discriminative EB accumulation between EPCR^low and control mice in response to LPS, suggesting that EPCR plays an important role in maintaining vascular integrity during inflammation especially in these organs. The increased susceptibility to LPS-induced vascular deterioration due to low EPCR expression (<10%) was supported by histological findings of vascular permeability, such as renal hemorrhage and pulmonary alveolar edema. In contrast, at the LPS dose that induced EPCR-dependent vascular deterioration in kidneys and lungs, focal necroses were the major pathological findings in the liver. No EPCR-dependent differences in susceptibility to focal necrosis were found as livers of both EPCR^low and control mice were affected to a similar extent. These findings coincided with a similar degree of EB dye accumulation in the livers of these mice, supporting the absence of a clear role of EPCR for modulating vascular integrity in the liver in endotoxemia. In summary, these data suggest that the relative importance of EPCR-dependent protection from vascular leak is organ specific.

The organ system most severely affected by low EPCR expression was the kidney as evident from both the EPCR^low mice and wild-type mice treated with blocking EPCR antibodies. In addition to significant EB dye accumulation in the kidneys of EPCR^low mice compared to control mice, EPCR^low mice suffered from severe kidney injury in response to LPS that was not obvious in control mice. Findings consisted of kidney swelling, albuminuria and profound renal hemorrhage at the cortico-medullary junction indicative of the most extreme expression of vascular breach. These EPCR-mediated effects on renal integrity in endoxemia are consistent with previously reported roles of the protein C pathway for renal protection in vitro, diabetic nephropathy injury models in vivo, and patients with chronic renal failure on hemodialysis.^30-34

In the brain, significantly increased EB dye accumulation was detected in EPCR^low compared to control mice after LPS. These findings provide further evidence that EPCR may be important to prevent blood-brain-barrier disruption during inflammation, extending findings where mice overexpressing EPCR were protected against brain edema in a model of venous sinus thrombosis.³⁵ EPCR is present on vascular endothelial cells in the central nervous system and facilitates the transport of APC across the blood-brain-barrier to provide cytoprotective activities on neurons and microglia cells.^{36, 37} Although encephalopathy is a well recognized and alarming complication of sepsis associated with poor outcomes, the pathophysiological processes underlying inflammatory blood-brain-barrier breakdown remain poorly understood.³⁸ Our results indicate that low EPCR expression enhanced susceptibility to LPS-induced blood-brain-barrier breakdown, providing new potential leads to elucidate the mechanisms leading to sepsis-associated encephalopathy. Furthermore, these new insights suggest the blood-brain-barrier (and the brain) as a previously underappreciated target for the endogenous protein C pathway in sepsis, consistent with the well documented neuroprotective effects of pharmacological APC applications in ischemic stroke.³⁹

Since the vascular barrier integrity in endotoxemia is EPCR-dependent, we speculate that the molecular pathway utilizing EPCR for barrier protection is the endogenous protein C pathway. We base this assumption on our understanding of the well described EPCR-dependent activation of protein C and EPCR-dependent cytoprotective activities of pharmacologic APC on cells and tissues.^{6, 12, 40} Additional compelling support for the protective role of the endogenous protein C pathway is also provided by the results from previous animal studies in sepsis or endotoxemia where modulation of important players of the endogenous protein C system, such as protein C expression, EPCR expression, or antibody blockade of (A)PC or EPCR, all similarly modulate survival.^{13, 15, 16}. For instance, pharmacological APC administration reduced LPS-induced mortality in wild-type mice but not in EPCR^low mice.^{10, 11} In addition, rescued EPCR-deficient mice had a prothrombotic phenotype and expressed lower levels of APC, whereas mice overexpressing EPCR demonstrated higher levels of circulating APC after thrombin infusion^{16, 41}. However, additional and/or alternative mechanisms remain possible, especially since EPCR is expressed on vascular endothelial and on blood and immune cells. While our results demonstrating profound organ-specific vascular leakage are consistent with low expression of EPCR on vascular endothelial cells, additional functions of EPCR mediated by blood and immune cells may contribute to host defense and anti-inflammatory effects. Our observation of diminished polymorphic mononuclear infiltrate in livers of EPCR^low but not control mice after LPS underlines this notion and expands the hypothesis that EPCR may be involved in cell trafficking during inflammation.^42-44

In summary, our results demonstrate that EPCR is required to maintain vascular integrity and prevent tissue edema in vivo during endotoxemia. Responses to inflammation seem to be organ-specific with kidneys, lungs, and brains strongly affected by EPCR-deficiency. Tissue edema and vascular leak in sepsis and inflammation are increasingly recognized as a deleterious entity that predisposes to organ failure and mortality.²⁶ Understanding extravasation, tissue edema, and vascular barrier stabilization on a molecular level has just begun, and may open new avenues for life saving barrier-protective agents in sepsis, such as strategies aimed at improving or retaining EPCR expression on vascular endothelial and blood cells during inflammation.

Significance.

The study demonstrates that the endothelial protein C receptor (EPCR), important for protein C activation and activated protein C (APC) cytoprotective effects on cells, is required to maintain organ-specific vascular integrity during endotoxemia. Using highly sensitive novel infrared fluorescence imaging techniques and measurements of intravascular volume depletion EPCR-dependent susceptibility to LPS-induced vascular leakage was determined in various organs. The kidneys, lungs, and brains of EPCR^low mice or wild-type mice treated with blocking anti-EPCR-antibodies were shown to be particular vulnerable to endotoxemia-induced vascular leakage compared to wild-type control mice. This study supports the important protective role for the endogenous cytoprotective PC pathway against vascular leakage during inflammation and demonstrates that EPCR-dependent vascular protection is organ specific. Since vascular leakage is increasingly recognized as deleterious within the pathophysiology of sepsis, strategies aimed at improving or retaining EPCR expression on vascular endothelial and blood cells during inflammation may prove to be beneficial.

Non-standard Abbreviations and Acronyms

APC

activated protein C

EB

Evans blue

EPCR

endothelial protein C receptor

EPCR^low

EPCR^δ/δ mice expressing <10% EPCR

i.p.

intraperitoneal

LD

lethality dose

IRF

Infrared Fluorescence

LPS

lipopolysaccharide

PAR1

protease activated receptor 1

PC

protein C

RFI

relative fluorescence intensities