The Angiopoietin-Tie2 Pathway in Critical Illness.

Lethal features of sepsis and acute respiratory distress syndrome (ARDS) relate to the health of small blood vessels. For example, alveolar infiltration with proteinaceous fluid is often driven by breach of the microvascular barrier. Spontaneous thrombus formation within inflamed microvessels exacerbates organ ischemia, and in its final stages, erupts into overt disseminated intravascular coagulation. Disruption of an endothelial signaling axis, the Angiopoietin-Tie2 pathway, may mediate the abrupt transition from microvascular integrity to pathologic disruption. This review summarizes preclinical and clinical results that implicate the Tie2 pathway as a promising target to restore microvascular health in sepsis and ARDS.

VWF excess and ADAMTS13 deficiency: a unifying pathomechanism linking inflammation to thrombosis in DIC, malaria, and TTP.

Absent or severely diminished activity of ADAMTS13 (A Disintegrin And Metalloprotease with a ThromboSpondin type 1 motif, member 13) resulting in the intravascular persistence and accumulation of highly thrombogenic ultra large von Willebrand factor (UL-VWF) multimers is the pathophysiological mechanism underlying thrombotic thrombocytopenic purpura. Reduced VWF-cleaving protease levels, however, are not uniquely restricted to primary thrombotic microangiopathy (TMA), e. g. thrombotic thrombocytopenic purpura, but also occur in other life-threatening thrombocytopenic conditions: severely decreased ADAMTS13 activity is seen in severe sepsis, disseminated intravascular coagulation (DIC) and complicated malarial infection. The clinical relevance of these secondary thrombotic microangiopathies is increasingly recognised, but its therapeutic implications have not yet been determined. The presence of a secondary TMA in certain diseases may define patient groups which possibly could benefit from ADAMTS13 replacement or a VWF-targeting therapy. This short-review focuses on the role of UL-VWF multimers in secondary TMA and discusses the potential of investigational therapies as candidates for the treatment of TTP. In conclusion, prospective clinical trials on the effectiveness of protease replacementin vivo seem reasonable. Carefully selected patients with secondary TMA may benefit from therapies primarily intended for the use in patients with TTP.

Synergistic effect of thrombin and CD40 ligand on endothelial matrix metalloproteinase-10 expression and microparticle generation in vitro and in vivo.

OBJECTIVE: Thrombin induces CD40 ligand (CD40L) and matrix metalloproteinases (MMPs) under inflammatory/prothrombotic conditions. Thrombin and CD40L could modulate endothelial MMP-10 expression in vitro and in vivo. METHODS AND RESULTS: Human endothelial cells were stimulated with thrombin (0.1-10 U/mL), CD40L (0.25-1 µg/mL), or their combination (thrombin/CD40L) to assess MMP-10 expression and microparticle generation. Thrombin/CD40L elicited higher MMP-10 mRNA (5-fold; P<0.001) and protein levels (4.5-fold; P<0.001) than either stimulus alone. This effect was mimicked by a protease-activated receptor-1 agonist and antagonized by hirudin, a-protease-activated receptor-1, α-CD40L, and α-CD40 antibodies. The synergistic effect was dependent on p38 mitogen-activated protein kinase and c-Jun N-terminal kinase-1 pathways. Thrombin also upregulated the expression of CD40 in endothelial cell surface increasing its availability, thereby favoring its synergistic effects with CD40L. In mice, thrombin/CD40L further increased the aortic MMP-10 expression. Septic patients with systemic inflammation and enhanced thrombin generation (n=60) exhibited increased MMP-10 and soluble CD40L levels associated with adverse clinical outcome. Endothelial and systemic activation by thrombin/CD40L and lipopolysaccharide also increased microparticles harboring MMP-10 and CD40L. CONCLUSIONS: Thrombin/CD40L elicited a strong synergistic effect on endothelial MMP-10 expression and microparticles containing MMP-10 in vitro and in vivo, which may represent a new link between inflammation/thrombosis with prognostic implications.

Matrix metalloproteinase-10 is upregulated by thrombin in endothelial cells and increased in patients with enhanced thrombin generation.

OBJECTIVE: Thrombin is a multifunctional serine protease that promotes vascular proinflammatory responses whose effect on endothelial MMP-10 expression has not previously been evaluated. METHODS AND RESULTS: Thrombin induced endothelial MMP-10 mRNA and protein levels, through a protease-activated receptor-1 (PAR-1)-dependent mechanism, in a dose- and time-dependent manner. This effect was mimicked by a PAR-1 agonist peptide (TRAP-1) and antagonized by an anti-PAR-1 blocking antibody. MMP-10 induction was dependent on extracellular regulated kinase1/2 (ERK1/2) and c-jun N-terminal kinase (JNK) pathways. By serial deletion analysis, site-directed mutagenesis and electrophoretic mobility shift assay an AP-1 site in the proximal region of MMP-10 promoter was found to be critical for thrombin-induced MMP-10 transcriptional activity. Thrombin and TRAP-1 upregulated MMP-10 in murine endothelial cells in culture and in vivo in mouse aorta. This effect of thrombin was not observed in PAR-1-deficient mice. Interestingly, circulating MMP-10 levels (P<0.01) were augmented in patients with endothelial activation associated with high (disseminated intravascular coagulation) and moderate (previous acute myocardial infarction) systemic thrombin generation. CONCLUSIONS: Thrombin induces MMP-10 through a PAR-1-dependent mechanism mediated by ERK1/2, JNK, and AP-1 activation. Endothelial MMP-10 upregulation could be regarded as a new proinflammatory effect of thrombin whose pathological consequences in thrombin-related disorders and plaque stability deserve further investigation.

Linear relationship between ADAMTS13 activity and platelet dynamics even before severe thrombocytopenia.

Von Willebrand factor (VWF) cleaving metalloprotease, ADAMTS13, known for its causative relation to thrombotic thrombocytopenic purpura (TTP), also decreases to variable degree in other clinical conditions associated with thrombocytopenia, indicating a possible contribution of moderate deficiency of ADAMTS13 to platelet dynamics. We measured ADAMTS13 activity along with VWF activity, collagen binding activity (VWF:CB), and thrombin/antithrombin complex (TAT) in plasma drawn from patients with consumptive coagulopathy, in whom the platelet count was closely followed. ADAMTS13 activity was significantly but variably decreased in the patients, and VWF activity and VWF:CB were markedly increased as expected. The platelet count itself was not correlated with ADAMTS13 activity, VWF activity, or VWF:CB. However, the rate of decline of log-scaled platelet count (DeltaLnPLT/day) correlated well with ADAMTS13 activity and VWF:CB. ADADMTS13 activity showed inverse correlation with VWF:CB. Moreover, the correlation between ADAMTS13 and DeltaLnPLT/day was preserved even after VWF:CB was controlled. Multiple regression analysis showed that ADAMTS13 activity was the sole factor explaining DeltaLnPLT/day among ADAMTS13, VWF:CB, TAT, prothrombin time, d-dimer, and fibrinogen. TAT level and d-dimer as indicators of systemic fibrinolytic activity did not correlate with ADAMTS13 activity. In conclusion, we found that the decrease of ADADMTS13 activity in consumptive coagulopathy has stronger relationship to platelet dynamics than has generally been recognized.

Proteolytic cleavage of high mobility group box 1 protein by thrombin-thrombomodulin complexes.

OBJECTIVE: High mobility group box 1 protein (HMGB1) was identified as a mediator of endotoxin lethality. We previously reported that thrombomodulin (TM), an endothelial thrombin-binding protein, bound to HMGB1, thereby protecting mice from lethal endotoxemia. However, the fate of HMGB1 bound to TM remains to be elucidated. METHODS AND RESULTS: TM enhanced thrombin-mediated cleavage of HMGB1. N-terminal amino acid sequence analysis of the HMGB1 degradation product demonstrated that thrombin cleaved HMGB1 at the Arg10-Gly11 bond. Concomitant with the cleavage of the N-terminal domain of HMGB1, proinflammatory activity of HMGB1 was significantly decreased (P<0.01). HMGB1 degradation products were detected in the serum of endotoxemic mice and in the plasma of septic patients with disseminated intravascular coagulation (DIC), indicating that HMGB1 could be degraded under conditions in which proteases were activated in the systemic circulation. CONCLUSIONS: TM not only binds to HMGB1 but also aids the proteolytic cleavage of HMGB1 by thrombin. These findings highlight the novel antiinflammatory role of TM, in which thrombin-TM complexes degrade HMGB1 to a less proinflammatory form.

Degradation of circulating von Willebrand factor and its regulator ADAMTS13 implicates secreted Bacillus anthracis metalloproteases in anthrax consumptive coagulopathy.

Pathology data from the anthrax animal models show evidence of significant increases in vascular permeability coincident with hemostatic imbalances manifested by thrombocytopenia, transient leucopenia, and aggressive disseminated intravascular coagulation. In this study we hypothesized that anthrax infection modulates the activity of von Willebrand factor (VWF) and its endogenous regulator ADAMTS13, which play important roles in hemostasis and thrombosis, including interaction of endothelial cells with platelets. We previously demonstrated that purified anthrax neutral metalloproteases Npr599 and InhA are capable of cleaving a variety of host structural and regulatory proteins. Incubation of human plasma with these proteases at 37 degrees C in the presence of urea as a mild denaturant results in proteolysis of VWF. Also in these conditions, InhA directly cleaves plasma ADAMTS13 protein. Npr599 and InhA digest synthetic VWF substrate FRETS-VWF73. Amino acid sequencing of VWF fragments produced by InhA suggests that one of the cleavage sites of VWF is located at domain A2, the target domain of ADAMTS13. Proteolysis of VWF by InhA impairs its collagen binding activity (VWF:CBA) and ristocetin-induced platelet aggregation activity. In plasma from anthrax spore-challenged DBA/2 mice, VWF antigen levels increase up to 2-fold at day 3 post-infection with toxigenic Sterne 34F(2) strain, whereas VWF:CBA levels drop in a time-dependent manner, suggesting dysfunction of VWF instead of its quantitative deficiency. This conclusion is further supported by significant reduction in the amount of VWF circulating in blood in the ultra-large forms. In addition, Western blot analysis shows proteolytic depletion of ADAMTS13 from plasma of spore-challenged mice despite its increased expression in the liver. Our results suggest a new mechanism of anthrax coagulopathy affecting the levels and functional activities of both VWF and its natural regulator ADAMTS13. This mechanism may contribute to hemorrhage and thrombosis typical in anthrax.

Elevated plasma levels of fibrin degradation products by granulocyte-derived elastase in patients with disseminated intravascular coagulation.

Plasma levels of granulocyte-derived elastase (GE-XDP), D-dimer, and soluble fibrin (SF) were examined in 177 patients with disseminated intravascular coagulation (DIC) of various etiologies. Plasma levels of GE-XDP and D-dimer, but not SF, were significantly high in patients with sepsis and solid cancer. The ratio of GE-XDP/ D-dimer was significantly high in patients with trauma, burn, and sepsis, suggesting that fibrinolysis due to GE-XDP may be dominant in DIC. Plasma levels of GE-XDP and D-dimer, but not SF, were significantly high in patients with overt DIC and correlated with DIC score. Plasma levels of GE-XDP, but not SF, correlated significantly with D-dimer. Plasma levels of D-dimer, but not SF, correlated significantly with plasmin plasmin inhibitor complex (PPIC). Plasma levels of GE-XDP and D-dimer, but not SF, were significantly high in nonsurvivors. Plasma levels of GE-XDP, but not SF, correlated significantly with sepsis-related organ failure assessment (SOFA) score. These results suggest that GE-XDP is a potentially useful marker for the diagnosis of overt-DIC and as a predictor of organ failure-related outcome.

Temporal changes in factors associated with neutrophil elastase and coagulation in intensive care patients with a biphasic waveform and disseminated intravascular coagulation.

Summary. The biphasic waveform is an early marker of disseminated intravascular coagulation (DIC). Neutrophil elastase (NE) cleaves coagulation factors; thus, elevated elastase levels or its dysregulation by alpha-1-protease inhibitor (Alpha1PI) may be linked to DIC. Time courses over a period were determined for factors associated with NE and coagulation in 14 Intensive Care Unit patients with a biphasic waveform who developed DIC. The data were analyzed using a random coefficient linear regression model to predict the variables' mean values on day 0 and their mean rates of change over the period in which the biphasic waveform appeared. The biphasic waveform was normal on day 0, maximized on day 1, and approached normal again by day 4. Alpha1PI/NE complex levels were 2.5-fold greater than normal for the entire period. The A1PI activity, antigen, and specific activity levels were normal on day 0 and increased thereafter by 21.0, 10.5, and 8.9% of normal per day, respectively. Factor II, V, VII, IX, and X activity levels were, respectively, 57, 46, 46, 77, and 46% of normal on day 0, whereas factor VIII and fibrinogen levels were normal. All coagulation factor levels trended upward with time but not significantly. The prothrombin time, but not the activated partial thromboplastin time, was prolonged, and the platelet counts and hematocrits were below normal on day 0 and remained so thereafter. We conclude that events associated with neutrophil activation, elastase release, and perturbations of coagulation precede both the appearance of the biphasic waveform and the diagnosis of DIC in these patients.

Elastase mediated fibrinolysis in acute promyelocytic leukemia.

The bleeding syndrome of acute promyelocytic leukemia (APL) is complex and consists of disseminated intravascular coagulation (DIC) and hyperfibrinolysis. Elastase, derived from malignant promyelocytes, is believed to mediate the fibrinogeno- and fibrinolysis by aspecific proteolysis. In this study we measured the role of elastase in fifteen patients with APL by using an assay for elastase degraded fibrin(ogen) and the results were compared with those obtained in patients with sepsis induced DIC. High levels of elastase were observed in sepsis and APL. The levels of fibrinogen and fibrin degradation products were significantly higher in APL patients compared to patients with sepsis induced DIC. Nevertheless, the level of elastase degraded fibrin(ogen) was higher in the sepsis group (635.3 ng/ml, compared to 144.3 ng/ml in APL; p <0.0001). So, the enormous increase in fibrin and fibrinogen degradation products in APL cannot be explained by elastase activity. This study suggests a minor role for elastase mediated proteolysis in the hemorrhagic diathesis in APL patients.

Neutrophil elastase may play a key role in developing symptomatic disseminated intravascular coagulation and multiple organ failure in patients with head injury.

OBJECTIVE: To study the mechanism associated with the development of symptomatic disseminated intravascular coagulation (DIC) after head injury. METHODS: Plasma parameters were analyzed in patients with symptomatic (group A, n = 10) and asymptomatic DIC (group B, n = 15) induced by head injury, and in patients in whom DIC was caused by sepsis (group C, n = 10). RESULTS: Levels of fibrinogen, alpha2PI-plasmin complex and platelets in group A (58.1 mg/dL, 22.4 microg/mL, 16.0 x 10(4)/ mm3) and group B (98.3, 22.1, 16.6 x 10(4)) were comparable, but differed significantly from those in group C (297.4, 2.4, 6.3 x 10(4)). Significant differences were observed between groups A and B in both neutrophil-elastase (1,528 vs. 293 microg/ml) and D-dimer (42.1 vs. 17.6 microg/mL). CONCLUSION: Neutrophil elastase may be implicated in the development of symptomatic DIC after head injury, whose characteristics include "enhanced fibrinolysis with minimal platelet loss."

Soluble thrombomodulin increases in patients with disseminated intravascular coagulation and in those with multiple organ dysfunction syndrome after trauma: role of neutrophil elastase.

OBJECTIVE: Our goal was to investigate the role of soluble thrombomodulin (TM) and neutrophil elastase in patients with trauma. DESIGN: This study is a prospective case-control study. MATERIALS AND METHODS: Forty-seven trauma victims, 14 with disseminated intravascular coagulation (DIC), 5 with multiple organ dysfunction syndrome (MODS), and 28 control patients without DIC or MODS were the participants. Soluble TM and neutrophil elastase (elastase-alpha1-proteinase inhibitor complex) were measured on the day of the injury, and on the first, third, and fifth days after admission. The results of these measurements and demographic data were compared among the groups, and correlations between the soluble TM and the neutrophil elastase were examined. The DIC patients were classified into subgroups of survivors (n = 5) and nonsurvivors (n = 9), and the changes of the soluble TM between the subgroups were then studied. MEASUREMENTS AND MAIN RESULTS: A high incidence of DIC patients encountered MODS complications (12 of 14, 86%). The DIC patients had higher Injury Severity Scores (ISSs) than the other patients. The levels of soluble TM and neutrophil elastase significantly increased on the day of admission in the patients with DIC and also in those with MODS (p < 0.05 vs. control patients) and continued to show markedly high values until the fifth day of admission in the patients with DIC. In the DIC patients, the levels of soluble TM were higher in the nonsurvivors than in the survivors (p < 0.05 on the third and the fifth days of admission). In all patients, there was weak but statistically significant correlation between peak levels of soluble TM and ISS (r2 = 0.125, p < 0.025). Comparison of the levels of soluble TM and neutrophil elastase in the patients with DIC or MODS demonstrated an excellent correlation (r2 = 0.718 and r2 = 0.714, respectively). CONCLUSIONS: Soluble TM as a novel endothelial cell injury marker increases in patients with DIC and also in those with MODS after trauma. Neutrophil elastase may be involved in the pathogenesis of the injury. Soluble TM is a marker of the severity of injury and is a good predictor of MODS.

Plasma levels of granulocyte elastase-alpha 1-proteinase inhibitor complex in patients with disseminated intravascular coagulation: pathophysiologic implications.

To investigate the role of neutrophil activation in the pathophysiology and sequelae of disseminated intravascular coagulation (DIC), we measured plasma levels of granulocyte elastase-alpha 1-proteinase inhibitor complex (GEPIC) in 41 patients with DIC and 27 patients with similar underlying conditions but without DIC. Mean GEPIC levels were significantly higher in patients with DIC (421.0 +/- 45.6 ng/ml) than in patients without DIC (246.1 +/- 41.9 ng/ml, P < 0.01). Significant differences were also noted in DIC patients with or without infection (474.7 +/- 61.2 ng/ml vs. 302.4 +/- 48.9 ng/ml, P < 0.04), with or without organ dysfunction (546.6 +/- 72.7 ng/ml vs. 305.6 +/- 42 ng/ml, P < 0.01), and with or without respiratory failure (640.0 +/- 91.2 ng/ml vs. 328.1 +/- 55.1 ng/ml, P < 0.01). No significant difference was found in mean GEPIC levels in DIC patients with or without renal failure, heart failure, hepatic failure, or gastrointestinal bleeding. The frequency of respiratory failure correlated with rising plasma levels of GEPIC. Mortality was higher in patients with GEPIC levels > 500 ng/ml (53.8%) than in patients with GEPIC levels < 500 ng/ml (28.6%). This correlation was particularly strong in patients with DIC, infection, and respiratory failure. Based on these data, we suggest that neutrophil activation, triggered by the coagulation cascade and perhaps augmented by endotoxin or cytokine release with infection, significantly contributes to respiratory failure and mortality in patients with DIC.

Factors relating to coagulation, fibrinolysis and hepatic damage after liver resection.

A survey of the blood of twenty-two patients who had undergone hepatic resection was performed. Serum levels of alpha-2 plasmin inhibitor-plasmin complex initially decreased from 1.58 +/- 0.31 microgram/ml on the preoperative day (PREOP), to 0.92 +/- 0.14 mu/ml on the first postoperative day (POD 1), and then increased to 3.13 +/- 0.92 micrograms/ml on the seventh postoperative day (POD 7) (mean +/- SE)). Thrombin-anti-thrombin III complex (14.2 +/- 4.3 ng/ml on PREOP and 26.0 +/- 4.1 ng/ml on POD 7 (mean +/- SE)) and D-dimer (335 +/- 96 ng/ml on PREOP and 1859 +/- 258 ng/ml on POD 7 (mean +/- SE)) increased in the early postoperative stage. The level of 6-keto-prostaglandin F1 alpha increased after the operations (from 13.2 +/- 1.8 pg/ml on PREOP to 37.8 +/- 12.8 pg/ml on POD 7 (mean +/- SE)). The level of thromboxane B-2 decreased at first, and then gradually increased and returned to its preoperative level on POD 7 (144.7 +/- 43.8 pg/ml on PREOP, 57.6 +/- 27.5 pg/ml on POD 1 and 152.5 +/- 58.4 pg/ml on POD 7 (mean +/- SE)). Superoxide dismutase activity increased at first, and then gradually decreased, postoperatively (2.8 +/- 0.5 NU/ml on PREOP, 4.8 +/- 0.8 NU/ml on POD 1 and 2.6 +/- 0.3 NU/ml on POD 7 (mean +/- SE)). That is, biodefensive reactions which protect patients against the shift to disseminated intravascular coagulation (DIC) were inferred with by the increase in antiplatelet aggregation, despite the activation of coagulation and fibrinolytic mechanisms after hepatic resection.

Polymorphonuclear leucocyte elastase in Plasmodium falciparum malaria.

Sixty-one patients with falciparum malaria were studied prospectively to determine the plasma concentrations of the lysosomal proteinase, polymorphonuclear leucocyte elastase (PMN-elastase) and their relationship to disease severity. The patients were divided into 3 groups; severe (parasitaemia > 5%) or vital organ dysfunction (n = 23), moderate (parasitaemia 1%-5% without complications) (n = 15), and mild (parasitaemia < 1%) (n = 23). The mean plasma PMN-elastase level in 10 healthy Thai volunteers was 49.5 (SD = 21.6) ng/ml (range 33-65 ng/ml). Plasma PMN-elastase concentrations on admission were elevated (> 2 x SD above normal) in all patients with severe malaria and were above 100 ng/ml in 86.6% and 65% of the moderately severe and mild patients respectively. PMN-elastase levels during the first 3 hospital days were significantly higher in severe malaria compared with the other 2 groups (P = < 0.001-0.013). The levels decreased as the patients became afebrile and aparasitaemic. Admission plasma concentrations of PMN-elastase correlated directly with bilirubin (rs = 0.50, P < 0.001), serum glutamic oxalacetic transaminase (rs = 0.54, P0.001), parasite count (rs = 0.62, P < 0.001), blood urea nitrogen (rs = 0.54, P < 0.001) and inversely with antithrombin III activity (rs = 0.54, P < 0.001) and the platelet count (rs = 0.58, P < 0.001). Polymorphonuclear leucocyte activation may contribute to the pathogenesis of severe malaria.

Change in the acid protease activity in plasma of patients with disseminated intravascular coagulation.

Lysosomal protease was determined in the serum of patients with disseminated intravascular coagulation (DIC) to clarify whether the platelet count is an appropriate diagnostic index which allows the early initiation of treatment. The platelet count and the serum level of cathepsin D, a lysosomal protease, were determined in 60 patients diagnosed to have DIC. The cathepsin D activity could not be detected in the sera of healthy individuals with a platelet count of 100,000 or above, but was detectable in the serum of DIC patients with a platelet count of 100,000, and this activity increased as the platelet count decreased to 70,000 and 50,000, and was about 5 times higher at a platelet count of 30,000 than at 70,000. In DIC patients, the decrease in the platelet count was correlated with the increase in the serum lysosomal protease activity. The appearance of cathepsin D activity in the serum of DIC patients is considered to reflect the release of lysosomal enzyme activities from damaged organs, and the treatment for DIC must be initiated before the platelet count decrease below 100,000, and cathepsin D activity then appears in the serum. At a platelet count of 30,000 or less, DIC becomes established, and no therapeutic effects can be expected because of the associated multiple organ failure.

Plasma active renin, inactive renin and kallikrein in patients with disseminated intravascular coagulation.

To investigate the role plasma kallikrein plays in the in vivo activation of inactive renin, we measured plasma active renin, inactive renin, kallikrein and prekallikrein levels in 10 patients with disseminated intravascular coagulation (DIC), with 16 normal persons as controls. The plasma active renin concentration was expressed by the angiotensin I generation rate after the addition of sheep renin substrate. Plasma inactive renin was activated by trypsin. The plasma total kallikrein level was measured by an assay of kallikrein activity on synthetic substrate S-2302 after the addition of a prekallikrein activator. Plasma kallikrein was assayed by its activity on S-2302 without addition of the activator. The prekallikrein level was obtained by subtracting the kallikrein activity from the total kallikrein activity. A significant decrease in the plasma prekallikrein concentration was observed in DIC patients, as compared to that of controls (p less than 0.01). There was no significant difference in plasma levels of kallikrein, inactive renin, and the proportion of active renin between DIC patients and normal controls, but the active renin level was higher in DIC patients. There was no significant correlation between the level of plasma kallikrein and the proportion of active renin in either normal controls or DIC patients. These results are compatible with, but do not prove, the theory that plasma kallikrein plays a role in the in vivo activation of inactive renin.

Activation and complexation of protein C and cleavage and decrease of protein S in plasma of patients with intravascular coagulation.

Activated protein C (APC) is inhibited by two major plasma inhibitors (PCIs). To find evidence for in vivo complexation of APC, immunoblotting studies were performed on plasmas of 85 patients with suspected disseminated intravascular coagulation (DIC). Samples from 62 of these patients contained 5% to 35% of protein C antigen in APC:inhibitor complexes, indicating that protein C activation and inhibition had occurred. In 24 normal plasmas, no detectable APC:PCI complexes were observed (less than 5%). Patients with higher levels of complexes had more abnormal coagulation test data for DIC. The major band of APC complexes detected by anti-protein C antibodies did not react with antibodies to the heparin-dependent protein C inhibitor (PCI-1) previously described. Rather, APC was complexed with another recently described plasma protein C inhibitor, PCI-2. Immunoblotting studies for protein S, the cofactor for APC, revealed that the majority of the DIC patient plasmas contained a higher than normal proportion of protein S in cleaved form, suggesting that protein S may have been proteolytically inactivated. Protein S total antigen levels were also found to be low in DIC patients, excluding those with malignancy. These studies support the hypothesis that the protein C pathway is activated during DIC.