- Email: [email protected]
Sepsis, thrombosis and organ dysfunction
Thrombosis Research 129 (2012) 290–295
Contents lists available at SciVerse ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Mini Review
Sepsis, thrombosis and organ dysfunction Nicola Semeraro ⁎, Concetta T. Ammollo, Fabrizio Semeraro, Mario Colucci Department of Biomedical Sciences and Human Oncology, Section of General, Experimental and Clinical Pathology, University of Bari, Bari, Italy
a r t i c l e
i n f o
Article history: Received 16 September 2011 Received in revised form 16 September 2011 Accepted 14 October 2011 Available online 5 November 2011 Keywords: Endotoxemia Coagulation Fibrinolysis Histones Microvascular thrombosis Multiple organ dysfunction syndrome
a b s t r a c t Sepsis is often associated with haemostatic changes ranging from subclinical activation of blood coagulation (hypercoagulability), which may contribute to localized venous thromboembolism, to acute disseminated intravascular coagulation (DIC), characterized by widespread microvascular thrombosis and subsequent consumption of platelets and coagulation proteins, eventually causing bleeding manifestations. The key event underlying this life-threatening complication is the overwhelming inflammatory host response to the infectious agent leading to the overexpression of inflammatory mediators. The latter, along with the microorganism and its derivatives are now believed to drive the major changes responsible for massive thrombin formation and fibrin deposition, namely 1) the aberrant expression of the TF by different cells (especially monocytes-macrophages), 2) the impairment of physiological anticoagulant pathways, orchestrated mainly by dysfunctional endothelial cells (ECs) and 3) the suppression of fibrinolysis due to overproduction of plasminogen activator inhibitor-1 (PAI-1) by ECs and likely also to thrombin-mediated activation of thrombinactivatable fibrinolysis inhibitor (TAFI). The ensuing microvascular thrombosis and ischemia are thought to contribute to tissue injury and multiple organ dysfunction syndrome (MODS). Recent evidence indicates that extracellular nuclear materials released from activated and especially apoptotic or necrotic cells, e.g. High Mobility Group Box-1 (HMGB-1) and histones, are endowed with cell toxicity, proinflammatory and clot-promoting properties and thus, during sepsis, they may represent late mediators that propagate further inflammation, coagulation, cell death and MODS. These insights into the pathogenesis of DIC and MODS may have implications for the development of new therapeutic agents potentially useful for the management of severe sepsis. © 2011 Elsevier Ltd. All rights reserved.
Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pathogenesis of sepsis-associated thrombus formation . . . . . . . . . . . . Up-regulation . of procoagulant pathways . . . . . . . . . . . . . . . . . . . Impairment . of physiological anticoagulant mechanisms . . . . . . . . . . . . Suppression . of fibrinolysis . . . . . . . . . . . . . . . . . . . . . . . . . Role .of endogenous inflammatory mediators in coagulation/fibrinolysis changes Pathogenesis of multiple organ dysfunction syndrome (MODS) . . . . . . . . Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
290 291 291 292 292 293 293 294 294
Introduction Abbreviations: APC, activated protein C; AT, antithrombin; DIC, disseminated intravascular coagulation; ECs, endothelial cells; EPCR, endothelial protein C receptor; HMGB-1, High Mobility Group Box-1; MODS, multiple organ dysfunction syndrome; PAI-1, plasminogen activator inhibitor-1; PC, protein C; PS, protein S; TAFI, thrombin activatable fibrinolysis inhibitor; TM, thrombomodulin; TF, tissue factor; TFPI, tissue factor pathway inhibitor. ⁎ Corresponding author at: Dipartimento di Scienze Biomediche e Oncologia Umana, Sezione di Patologia Generale, Università degli Studi di Bari - Policlinico, Piazza G. Cesare, 11, 70124 Bari, Italy. Tel.: + 39 080 5478468; fax: + 39 080 5478524. E-mail address: [email protected] (N. Semeraro). 0049-3848/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2011.10.013
Sepsis is almost invariably associated with haemostatic changes ranging from subclinical activation of blood coagulation (hypercoagulability) to systemic clotting activation with massive thrombin and fibrin formation, eventually leading to consumption of platelets and proteins of the haemostatic system (acute disseminated intravascular coagulation, DIC) [1,2]. From a clinical standpoint, septic patients may present with localized thrombotic manifestations, as
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
indicated by the observation that they are at increased risk for venous thromboembolism [3,4]. The most common and dramatic clinical feature, however, is widespread thrombosis in the microcirculation of different organs which may importantly contribute to solitary or multiple organ dysfunction (MODS) [1,2]. In fulminant DIC, the consumption of platelets and coagulation proteins will result in simultaneous bleeding of different severity. DIC is classically associated with Gram negative bacterial infections but it can occur in Gram positive sepsis (with a similar incidence) and in systemic infections with other micro-organisms [1,2]. The pathophysiology of sepsis-associated DIC is extremely complex and still extensively investigated. The key event is the systemic inflammatory response to the infectious agent [5]. Following recognition of unique constituents expressed by the causative microorganism and/or of host-derived factors via specific receptors (pattern recognition receptors, PRRs), particularly the Toll-like receptors (TLRs), immune and other host cells (monocytes-macrophages, platelets and endothelial cells among others) synthesize a number of proteins including proinflammatory cytokines. The latter, together with other mediators generated by the inflammatory cascade, including complement activation products [6], act in concert with the microorganisms and/or their derivatives to trigger the coagulation pathways, DIC and organ dysfunction [1,2,7]. Enzymes generated during the clotting cascade, in turn, interact with specific cellular receptors thus eliciting cell responses that amplify the inflammatory reactions [8]. Inflammation can also result in cell apoptosis or necrosis [5] and recent evidence indicates that products released from dead cells, such as nuclear proteins, are able to propagate further inflammation, coagulation, cell death and organ failure [5,9]. This article briefly summarizes current knowledge on the pathogenesis of DIC and MODS, and the ensuing development of potential therapeutics. Pathogenesis of sepsis-associated thrombus formation In sepsis the causative agent and the associated inflammatory response drive fibrin formation and deposition by several simultaneously acting mechanisms (Fig. 1), namely 1) up-regulation of procoagulant pathways, 2) down-regulation of physiological anticoagulants and 3) suppression of fibrinolysis [1,2,7]. Up-regulation of procoagulant pathways The aberrant in vivo expression of TF plays a pivotal role in sepsisassociated blood clotting activation, as indicated by the following observations: 1) the impairment of the TF pathway by various means prevents coagulation abnormalities (including fibrin deposition in target tissues) and lethality in animal models of sepsis or endotoxemia [2,7,10]; 2) the plasma levels of TF are increased in septic patients
291
and generally associated with raised concentrations of markers of clotting activation [2,7,11]. The cellular source of TF in sepsis, however, still remains an open question. In vitro, endothelial cells (ECs) and mononuclear phagocytes have long been known to synthesize TF in response to a wide variety of stimulating agents or conditions that are of pathophysiological importance in sepsis [2,12]. These cells may also exhibit other clot-promoting properties [2], thus providing a surface onto which the clotting pathways are initiated and propagated, eventually leading to fibrin formation in the cell microenvironment. TF expression has been detected also in human neutrophils in response to inflammatory agents [2,13], in eosinophils upon stimulation and subsequent translocation from their specific granules to the cell surface [14], and in activated platelets after de novo synthesis or release from αgranules [2,15,16]. Other studies, however, suggest that these cells do not synthesize TF but acquire it by binding TF-expressing microparticles (MPs) [2,10,17]. MPs are small membrane vesicles released from activated or apoptotic cells that can adhere to the surface of other cells via specific receptors (for instance PSGL-1 on leukocyte-derived MPs and P-selectin on activated platelets or ECs) making the recipient cell capable of triggering and propagating coagulation [2,12]. Although all mentioned cells might contribute to the aberrant in vivo expression of TF, most available studies point to activated monocytes-macrophages as the main triggers of blood coagulation during sepsis (Fig. 1). In animal models of endotoxemia or sepsis, TF expression is increased in important target organs where fibrin deposition often occurs during DIC, namely lung, kidney, liver, spleen and brain and, at cellular level, it is detected mainly in monocytes present in the microcirculation and in macrophages infiltrating the involved tissues [2,10,12,18]. In the same animals, blood monocytes and macrophages of different origin express strong TF activity [2,10,12,18]. In addition, a selective genetic deficiency of TF expression by hematopoietic cells as well as the deletion of TF gene in myeloid cells was found to reduce LPS-induced coagulation, inflammation, and mortality in mice [10,19]. Increased expression of monocyte-macrophage TF has been also documented in healthy volunteers after administration of low-dose endotoxin [20], in septic or endotoxemic patients, in whom TF was associated with clotting activation, MODS and lethal outcome, and in patients with peritonitis or acute respiratory distress syndrome [2,7,12,18]. Further support for a prominent role of monocytesmacrophages comes from studies on MPs. In endotoxemic mice, levels of MP TF activity were correlated with coagulation activation [21]. In addition, increased numbers of circulating TF-positive MPs of monocyte origin have been detected in patients with meningococcal sepsis and in human low-dose endotoxemia [2,7,22]. Surprisingly, and in contrast with the abundant in vitro evidence, ECs were negative for TF in most animal studies, with very few exceptions [2,10,17,18]. Moreover, the deletion of the TF gene in ECs had no significant effect on clotting activation in endotoxemic mice [10,19], clearly ruling out a major involvement
Fig. 1. Mechanisms contributing to thrombin formation and fibrin deposition in the microcirculation (see text for details). TF, tissue factor; EC, endothelial cell; TM, thrombomodulin; EPCR, endothelial protein C receptor; PS, protein S; TFPI, tissue factor pathway inhibitor; HS, heparan sulphate; PAI-1, plasminogen activator inhibitor-1; TAFI, thrombin activatable fibrinolysis inhibitor.
292
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
of EC TF. Likewise, the role of neutrophils and platelets as a direct source of TF in vivo during sepsis remains controversial [2,10,19,23]. These cells, however, may still participate in the activation of coagulation both by binding TF-positive monocyte MPs [2,10] and by other direct or indirect mechanisms (discussed later). Recently, it has been shown that the selective inhibition of TF expressed by non-hematopoietic cells substantially reduces the clotting activation in endotoxemic mice [10,19] but the precise cellular source of TF is unknown. In endotoxemic and septic animals, TF expression is increased not only in monocytes-macrophages but also in tissue cells, e.g. lung and kidney epithelial cells, and brain astrocytes [2,10,18]. Therefore, also considering that the role of ECs and vascular smooth muscle cells [19] remains uncertain, it is likely that TF up-regulation in parenchymal cells of target organs contributes to clotting coagulation during sepsis. Moreover, the well known increase in vascular permeability and vascular damage occurring during severe inflammation will allow the exposure of normal extravascular (e.g. fibroblast) TF to blood.
protective functions [28]. Indeed, besides exerting anticoagulant and profibrinolytic activity, APC also has important inflammationmodulating effects, including down-regulation of inflammatory cytokines and TF in activated leukocytes, antioxidant properties, anti-apoptotic activity and prevention of loss of endothelial barrier function [29]. During sepsis, plasma antithrombin (AT) is generally decreased because of consumption and low levels are associated with increased mortality [7]. Moreover, inflammatory stimuli can down-regulate the expression of heparan sulphate in cultured ECs [7,24] (Fig. 1) suggesting an additional mechanism for the reduced AT function in sepsis. With respect to TFPI, in animal models of endotoxemia or sepsis, a decreased TFPI expression was found in ECs (Fig. 1) of several organs [24,30]. In addition, anti-TFPI antibodies increased fibrin accumulation in the lungs of septic baboon [30], suggesting that TFPI underexpression, coupled with TF up-regulation, might augment the local procoagulant potential, thus promoting fibrin formation in tissues.
Impairment of physiological anticoagulant mechanisms Suppression of fibrinolysis Under physiological conditions the EC surface expresses various components involved in the anticoagulant pathways, including thrombomodulin (TM), endothelial protein C receptor (EPCR), protein S (PS), tissue factor pathway inhibitor (TFPI) and the heparinlike proteoglycan heparan sulphate. Because the endothelium plays a critical role in orchestrating the host response to sepsis and it is the target of the pathogen and the inflammatory mediators, the behavior of endothelial anticoagulant mechanisms, particularly the protein C (PC) pathway, has been extensively investigated in relation to sepsis. In cultured ECs, inflammatory mediators consistently reduced the expression of TM and EPCR, and the PS secretion [2,7,24,25] (Fig. 1). Although, at variance with these studies, animal experiments on in vivo expression of TM and EPCR by ECs produced rather controversial data [2,7,24,25], a rise in soluble plasma TM and EPCR was consistently observed during experimental endotoxemia, suggesting that endothelial activation-damage by inflammatory mediators does occur in vivo [2,7,24,25]. The importance of the PC anticoagulant pathway for the development of DIC in sepsis is definitely demonstrated by the fact that compromising the PC system resulted in a marked worsening of DIC and in increased morbidity and mortality in different animal models, whereas restoring an adequate APC function (e.g. treatment with APC) prevented the coagulopathy and improved organ failure and survival [7,24]. Interestingly, mice with heterozygous PC deficiency had more severe DIC and a higher mortality than the wild-type controls and mice homozygous for a point mutation of the TM gene that deletes the anticoagulant activity of the protein exhibited 10- to 30-fold greater amounts of fibrin in the microcirculation of several organs than the wild-type mice [2,7]. Studies in human sepsis have in general confirmed the dysfunction of the PC pathway. The plasma levels of soluble TM and EPCR were increased and TM levels were often correlated with disease severity and poor outcome [2,24,26]. Other common findings in septic patients are low levels of PC and PS, due to impaired liver synthesis and possibly to consumption, and low levels of free PS, due to increased C4b-binding protein [7,24]. Acquired severe PC deficiency was associated with early death [27]. Notably, the expression of TM and EPCR on morphologically intact ECs of dermal vessels was reduced in biopsy specimens of purpuric lesions from children with meningococcal sepsis, as compared with control skin-biopsy specimens [26]. Plasma levels of APC remained low in some of these patients even after treatment with PC concentrates, confirming down-regulation of TM in vivo and impaired PC activation. APC plasma levels were found to vary markedly among patients with severe sepsis and were significantly higher in survivors than in non survivors (28-day mortality), suggesting that endogenous APC serves
Numerous studies on cultured ECs challenged with inflammatory stimuli and in animal models of endotoxemia or sepsis have consistently reported the presence of markedly increased levels of plasminogen activator inhibitor-1 (PAI-1) in culture medium and in plasma, respectively. Although a simultaneous increase in tissuetype plasminogen activator (t-PA) often occurs, the net result is a fibrinolytic shut-down because of the very large amounts of PAI-1 [7,24]. In endotoxemic animals a strong elevation of PAI-1 mRNA was found in multiple tissues, including those affected by microthrombi (kidney, adrenals, lung and liver) and PAI-1 expression was detected primarily in ECs at all levels of the vasculature [24]. This suggests that plasma PAI-1 originates from ECs, although a contribution of platelets cannot be excluded [24]. It should be noted that, in some models of endotoxemia or cytokinemia, hypofibrinolysis and fibrin formation in adrenals and/or kidneys were most dependent on a decrease in PAs [1,24]. These findings indicate that suppression of fibrinolysis mediated by PAI-1 increase (Fig. 1) and other tissue- and species-specific alterations, such as decreased PAs in some models, are essential for fibrin deposition in tissue vasculature, at least in experimental sepsis. This concept is supported by studies in mice with targeted disruptions of genes encoding fibrinolytic proteins. Mice with a deficiency of PAs have more extensive fibrin deposition in organs when challenged with endotoxin, whereas PAI-1 knockout mice, in contrast to wildtype controls, have no microvascular thrombosis upon endotoxin challenge [2,24]. In human sepsis, the majority of studies is restricted to circulating fibrinolytic markers. A sustained increase in plasma PAI-1 has been consistently reported by numerous investigators and, in some studies, PAI-1 turned out to be a prognostic marker [2,24]. Plasma t-PA antigen was also elevated [2,24], but the net effect of the changes in t-PA and PAI-1 was definitely antifibrinolytic. The importance of impaired fibrinolysis is supported by the finding that a 4 G/5 G polymorphism in the PAI-1 promoter influencing PAI-1 expression was associated with the clinical outcome of severe sepsis [2,7,24]. More recent evidence indicates that other, thrombin-dependent mechanisms might contribute to hypofibrinolysis during sepsis. Thrombin causes resistance to fibrinolysis by forming more compact and less permeable clots [31], and by activating thrombin-activatable fibrinolysis inhibitor (TAFI), a plasma procarboxypeptidase that, once activated (TAFIa), removes the C-terminal lysines from partially degraded fibrin, thereby reducing plasmin formation [32]. Recent findings suggest that enhanced thrombin generation, the hallmark of sepsis, might influence the fibrin structure. Indeed, ECs stimulated by inflammatory cytokines to express TF cause the production of abnormally
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
dense fibrin networks that resist fibrinolysis [33]. Moreover, activated platelets, commonly found in sepsis, not only alter the fibrin structure and reduce susceptibility to lysis via the direct interaction between fibrin and αIIbβ3 integrin [33] but also release, together with the causative micro-organism, inorganic polyphosphates, which modify the fibrin architecture and reduce the binding of t-PA and plasminogen to fibrin, thus increasing fibrin resistance to fibrinolysis [34]. With respect to TAFI, evidence is accumulating that it might be involved in sepsis-associated hypofibrinolysis (Fig. 1). In vitro, LPSstimulated human monocytes inhibit fibrinolysis through a TFmediated enhancement of TAFI activation. Moreover, clots generated by TF-positive monocytes are resistant to the profibrinolytic activity of heparins [35]. Because TF-expressing monocytes-macrophages drive clotting activation in sepsis, these findings provide an additional mechanism whereby these cells may favor fibrin accumulation in the microcirculation. In animal models of endotoxemia or sepsis, TAFI levels were reduced, likely because of activation and consumption. In addition, blocking TAFIa with synthetic inhibitors or inhibiting thrombin-TM-dependent TAFI activation enhanced the rate of fibrin degradation thus reducing fibrin deposition in target tissues [2]. In human studies, TAFI levels were consistently decreased in septic patients and in healthy volunteers with low-grade endotoxemia [2]. More important, in patients with severe meningococcal infection, [2,36] the levels of TAFI activation markers were increased in patients with DIC as compared with those without, were significantly higher in non-survivors versus survivors and strongly correlated with severity scores of the disease. It thus appears that TAFI activation does occur in severe sepsis and that the measurement of TAFI activation markers may be clinically useful. The relevance of TAFI is further supported by the fact that a single nucleotide polymorphism in the TAFI gene that causes the substitution Thr325Ile and produces increased TAFIa stability and activity was associated with a poor outcome in meningococcal sepsis [2]. Role of endogenous inflammatory mediators in coagulation/fibrinolysis changes While in vitro the pathogen, its derivatives and the main inflammatory mediators are able to influence to variable extent each of the mechanisms described above, their role during experimental and human sepsis or endotoxemia is more difficult to establish. With respect
293
to the up-regulation of procoagulant pathways, various inflammatory mediators are likely involved, besides the micro-organism itself and its products (e.g. LPS). Although the main cytokines, i.e. TNF, IL-1 and IL6, can activate blood coagulation in humans and primates most probably via the TF pathway, neutralization studies with specific antibodies would suggest a major role of endogenous IL-6 and, to a lesser extent, of IL-1 [37]. With respect to the anticoagulant pathways and fibrinolysis, the few available studies suggest that, in different animal models of sepsis or endotoxemia, TNF and IL-1 are involved in the TM and PC downregulation, and in PAI-1-mediated suppression of fibrinolysis [37]. Recently, in a baboon model of severe sepsis, a potent inhibitor of C3 convertase was shown to attenuate the inflammatory response, DIC, and the dysfunction of several organs [38]. The simultaneous decrease in TF and PAI-1 expression in multiple tissues, and the attenuation of TFPI and TM down-regulation in ECs clearly point to a key role of complement-derived mediators in sepsis-associated pathophysiological changes. Altogether, these findings indicate that, as predicted by in vitro studies, inflammatory stimuli are indeed able to elicit the main mechanisms responsible for in vivo thrombin generation and fibrin deposition during sepsis.
Pathogenesis of multiple organ dysfunction syndrome (MODS) MODS is the hallmark of severe sepsis and septic shock and represents the main cause of the high mortality in these conditions. Several closely interlinked mechanisms have been proposed to explain this dramatic event, all of which result from the global hyper-inflammatory response to the triggering pathogen [39]. An important role of DIC is supported by several lines of evidence [2,7]: a) numerous histological studies in septic patients and in animals with sepsis or endotoxemia have documented the presence of thrombi in small and mid-size vessels of multiple organs and the relation of these thrombi with organ ischemia and dysfunction; b) in animal experiments, amelioration of DIC by various interventions improves organ failure and, in some cases, mortality; c) DIC has been shown to be an independent predictor of organ dysfunction and mortality in septic patients. Other widely recognized mechanisms contributing to MODS are the release of reactive oxygen and nitrogen species, and proteolytic enzymes by neutrophils recruited at tissue level and the presence in the interstitial space of high concentrations of selected cytokines that might be directly toxic
Fig. 2. Extracellular release of histones and mechanisms of thrombus formation. Histones are nuclear proteins released from activated neutrophils and dying cells and can promote thrombus formation by inducing endothelial injury, platelet activation with consequent expression of procoagulant properties, and impairment of TM-mediated protein C activation. NETs, neutrophil extracellular traps; PolyP, polyphosphates; PC, protein C; APC, activated PC; TM, thrombomodulin.
294
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
to vulnerable parenchyma and play a role particularly in sepsis with severe leukopenia [39]. Increasing evidence indicates the existence of new players in sepsis-associated organ dysfunction-failure, namely extracellular nuclear proteins that originate mainly from dying cells and therefore can be considered as late mediators of MODS. High Mobility Group Box-1 (HMGB-1) is released actively by stimulated cells of innate immunity and passively by necrotic cells into the extracellular environment where it becomes a lethal mediator of systemic inflammation [40]. HMGB-1 increases in the circulation in different animal models of sepsis and in septic patients. In animal experiments, HMGB-1 proved to be lethal and to promote the development of microvascular thrombosis by stimulating monocyte TF expression and by reducing the activity of thrombin-TM complex thereby inhibiting protein C activation [41]. Notably, anti-HMGB-1 antibodies conferred significant protection against lethal endotoxemia, sepsis and LPS-induced lung injury [40]. More recently, it has been shown that extracellular histones too (particularly histones H3 and H4) are major mediators of injury during sepsis [9]. Besides being toxic for human ECs and other cells in vitro, histones mimicked the manifestations of sepsis, including microvascular thrombosis, organ failure and death when injected in mice. More important, antibodies against histone H4 protected mice in different models of endotoxemia and sepsis. Recombinant APC was shown to degrade histones thus lowering their toxicity towards ECs in vitro and preventing lethality in histone-treated animals. Of note, histones and their APC-induced degraded forms were detected in plasma from septic animals and patients. Possible sources of histones during sepsis are activated inflammatory cells and dying cells (Fig. 2). Neutrophils can be induced by different stimuli, including activated platelets, to release the so-called neutrophil extracellular traps (NETs), complex structures containing histones, DNA and granule proteins [2,41]. Besides killing micro-organisms, NETs induce tissue damage and, at least in vitro, thrombus formation [42]. Another, likely more important source of histones is massive cell apoptosisnecrosis overwhelming the clearance capacity of mononuclear phagocytes, thereby permitting histones to enter the circulation. Histones have recently been shown to promote thrombin formation by different mechanisms (Fig. 2) among which platelet activation resulting in the release of polyphosphates and the expression of other platelet procoagulant properties [43], and the impairment of TM-mediated protein C activation [44]. The release of nuclear proteins during the late stages of sepsis can thus amplify inflammation, coagulation, cell death and MODS. In conclusion, considerable progress has been made in our knowledge on the mechanisms underlying sepsis-associated DIC and MODS that help to develop new therapeutic agents for the management of severe sepsis. However, the use of TF inhibitors, which would be the most logical treatment considering the pivotal role of TF in clotting activation during sepsis, is still debated. A phase III trial with recombinant TFPI did not show an overall survival benefit in septic patients [7]. Likewise, treatment with antithrombin concentrates failed to significantly reduce the mortality of septic patients in a large-scale clinical trial [1,7]. So far, the only anticoagulant drug showing promising results is recombinant human APC, which has been evaluated in several clinical trials and is indicated for the reduction of mortality in adult patients with severe sepsis at high risk of death [1]. The beneficial effects of this drug have been attributed not only to the restoration of the protein C anticoagulant pathway but also to its wellknown anti-inflammatory action [25,29]. The finding that APC degrades histones [9] adds a new mechanism whereby the drug exerts its beneficial effects especially in severe sepsis. Because histones seem to be critical mediators of organ dysfunction and death in septic patients, an attractive approach to treat MODS and prevent death could be the development of effective histone antagonists which might prove therapeutic without the bleeding complications that can result from APC therapy.
Conflict of interest The authors state that they have no conflict of interest.
References [1] Levi M, Schultz M, van der Poll T. Disseminated intravascular coagulation in infectious disease. Semin Thromb Hemost 2010;36:367–77. [2] Semeraro N, Ammollo CT, Semeraro F, Colucci M. Sepsis-associated disseminated intravascular coagulation and thromboembolic disease. Mediterr J Hematol Infect Dis 2010;2(3):e2010024. [3] Smeeth L, Cook C, Thomas S, Hall AJ, Hubbard R, Vallance P. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075–9. [4] Alikhan R, Spyropoulos AC. Epidemiology of venous thromboembolism in cardiorespiratory and infectious disease. Am J Med 2008;121:935–42. [5] Cinel I, Opal SM. Molecular biology of inflammation and sepsis: a primer. Crit Care Med 2009;37:291–304. [6] Markiewski MM, DeAngelis RA, Lambris JD. Complexity of complement activation in sepsis. J Cell Mol Med 2008;12:2245–54. [7] Levi M. The coagulant response in sepsis. Clin Chest Med 2008;29:627–42. [8] Monroe DM, Key NS. The tissue factor-factor VIIa complex: procoagulant activity, regulation, and multitasking. J Thromb Haemost 2007;5:1097–105. [9] Xu J, Zhang X, Pelayo R, Monestier M, Ammollo CT, Semeraro F, et al. Extracellular histones are major mediators of death in sepsis. Nat Med 2009;15:1318–21. [10] Pawlinski R, Mackman N. Cellular sources of tissue factor in endotoxemia and sepsis. Thromb Res 2010;125:S70–3. [11] Gando S, Nanzaki S, Sasaki S, Kemmotsu O. Significant correlation between tissue factor and thrombin markers in trauma and septic patients with disseminated intravascular coagulation. Thromb Haemost 1998;79:1111–5. [12] Osterud B, Bjorklid E. Sources of tissue factor. Semin Thromb Hemost 2006;32: 11–23. [13] Maugeri N, Brambilla M, Camera M, Carbone A, Tremoli E, Donati MB, et al. Human polymorphonuclear leukocytes produce and express functional tissue factor upon stimulation. J Thromb Haemost 2006;4:1323–30. [14] Moosbauer C, Morgenstern E, Cuvelier SL, Manukyan D, Bidzhekov K, Albrecht S, et al. Eosinophils are a major intravascular location for tissue factor storage and exposure. Blood 2007;109:995–1002. [15] Camera M, Frigerio M, Toschi V, Brambilla M, Rossi F, Cottell D, et al. Platelet activation induces cell-surface immunoreactive tissue factor expression, which is modulated differently by antiplatelet drugs. Arterioscler Thromb Vasc Biol 2003;23:1690–6. [16] Panes O, Matus V, Saez CG, Quiroga T, Pereira J, Mezzano D. Human platelets synthesize and express functional tissue factor. Blood 2007;109:5242–50. [17] Østerud B. Tissue factor expression in blood cells. Thromb Res 2010;125:S31–4. [18] Semeraro N, Colucci M. Tissue factor in health and disease. Thromb Haemost 1997;78:759–64. [19] Pawlinski R, Wang JG, Owens III AP, Williams J, Antoniak S, Tencati M, et al. Hematopoietic and nonhematopoietic cell tissue factor activates the coagulation cascade in endotoxemic mice. Blood 2010;116:806–14. [20] Franco RF, de Jonge E, Dekkers PE, Timmerman JJ, Spek CA, van Deventer SJ, et al. The in vivo kinetics of tissue factor messenger RNA expression during human endotoxemia: relationship with activation of coagulation. Blood 2000;96: 554–9. [21] Wang JG, Manly D, Kirchhofer D, Pawlinski R, Mackman N. Levels of microparticle tissue factor activity correlate with coagulation activation in endotoxemic mice. J Thromb Haemost 2009;7:1092–8. [22] Aras O, Shet A, Bach RR, Hysjulien JL, Slungaard A, Hebbel RP, et al. Induction of microparticle- and cell-associated intravascular tissue factor in human endotoxemia. Blood 2004;103:4545–53. [23] Rondina MT, Schwertz H, Harris ES, Kraemer BF, Campbell RA, Mackman N, et al. The septic milieu triggers expression of spliced tissue factor mRNA in human platelets. J Thromb Haemost 2011;9:748–58. [24] Semeraro N, Colucci M. Endothelial cell perturbation and disseminated intravascular coagulation. In: ten Cate H, Levi M, editors. Molecular mechanisms of disseminated intravascular coagulation. Georgetown: Landes Bioscience; 2003. p. 156–90. [25] Esmon CT. Inflammation and thrombosis. J Thromb Haemost 2003;1:1343–8. [26] Faust SN, Levin M, Harrison OB, Goldin RD, Lockhart MS, Kondaveeti S, et al. Dysfunction of endothelial protein C activation in severe meningococcal sepsis. N Engl J Med 2001;345:408–16. [27] Macias WL, Nelson DR. Severe protein C deficiency predicts early death in severe sepsis. Crit Care Med 2004;32:S223–8. [28] Liaw PC, Esmon CT, Kahnamoui K, Schmidt S, Kahnamoui S, Ferrell G, et al. Patients with severe sepsis vary markedly in their ability to generate activated protein C. Blood 2004;104:3958–64. [29] Mosnier LO, Zlokovic BV, Griffin JH. The cytoprotective protein C pathway. Blood 2007;109:3161–72. [30] Tang H, Ivanciu L, Popescu N, Peer G, Hack E, Lupu C, et al. Sepsis-induced coagulation in the baboon lung is associated with decreased tissue factor pathway inhibitor. Am J Pathol 2007;171:1066–77. [31] Wolberg AS. Thrombin generation and fibrin clot structure. Blood Rev 2007;21: 131–42.
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295 [32] Mosnier LO, Bouma BN. Regulation of fibrinolysis by thrombin activatable fibrinolysis inhibitor, an unstable carboxypeptidase B that unites the pathways of coagulation and fibrinolysis. Arterioscler Thromb Vasc Biol 2006;26:2445–53. [33] Campbell RA, Overmyer KA, Selzman CH, Sheridan BC, Wolberg AS. Contributions of extravascular and intravascular cells to fibrin network formation, structure, and stability. Blood 2009;114:4886–96. [34] Mutch NJ, Engel R, Uitte de Willige S, Philippou H, Ariëns RA. Polyphosphate modifies the fibrin network and down-regulates fibrinolysis by attenuating binding of tPA and plasminogen to fibrin. Blood 2010;115:3980–8. [35] Semeraro F, Ammollo CT, Semeraro N, Colucci M. Tissue factor-expressing monocytes inhibit fibrinolysis through a TAFI-mediated mechanism, and make clots resistant to heparins. Haematologica 2009;94:819–26. [36] Emonts M, de Bruijne EL, Guimarães AH, Declerck PJ, Leebeek FW, de Maat MP, et al. Thrombin-activatable fibrinolysis inhibitor is associated with severity and outcome of severe meningococcal infection in children. J Thromb Haemost 2008;6:268–76. [37] van der Poll T, de Jonge E, Levi M. Regulatory role of cytokines in disseminated intravascular coagulation. Semin Thromb Hemost 2001;27:639–51. [38] Silasi-Mansat R, Zhu H, Popescu NI, Peer G, Sfyroera G, Magotti P, et al. Complement inhibition decreases the procoagulant response and confers organ protection in a baboon model of E. coli sepsis. Blood 2010;116:1002–10.
295
[39] Wang H, Ma S. The cytokine storm and factors determining the sequence and severity of organ dysfunction in multiple organ dysfunction syndrome. Am J Emerg Med 2008;26:711–5. [40] Sunden-Cullberg J, Norrby-Teglund A, Treutiger CJ. The role of high mobility group box-1 protein in severe sepsis. Curr Opin Infect Dis 2006;19:231–6. [41] Papayannopoulos V, Metzler KD, Hakkim A, Zychlinsky A. Neutrophil elastase and myeloperoxidase regulate the formation of neutrophil extracellular traps. J Cell Biol 2010;191:677–91. [42] Fuchs TA, Brill A, Duerschmied D, Schatzberg D, Monestier M, Myers Jr DD, et al. Extracellular DNA traps promote thrombosis. Proc Natl Acad Sci U S A 2010;107: 15880–5. [43] Semeraro F, Ammollo CT, Morrissey JH, Dale GL, Friese P, Esmon NL, et al. Extracellular histones promote thrombin generation through platelet-dependent mechanisms: involvement of platelet TLR2 and TLR4. Blood 2011;118:1952–61. [44] Ammollo CT, Semeraro F, Xu J, Esmon NL, Esmon CT. Extracellular histones increase plasma thrombin generation by impairing thrombomodulin-dependent protein C activation. J Thromb Haemost 2011;9:1795–803.
Contents lists available at SciVerse ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Mini Review
Sepsis, thrombosis and organ dysfunction Nicola Semeraro ⁎, Concetta T. Ammollo, Fabrizio Semeraro, Mario Colucci Department of Biomedical Sciences and Human Oncology, Section of General, Experimental and Clinical Pathology, University of Bari, Bari, Italy
a r t i c l e
i n f o
Article history: Received 16 September 2011 Received in revised form 16 September 2011 Accepted 14 October 2011 Available online 5 November 2011 Keywords: Endotoxemia Coagulation Fibrinolysis Histones Microvascular thrombosis Multiple organ dysfunction syndrome
a b s t r a c t Sepsis is often associated with haemostatic changes ranging from subclinical activation of blood coagulation (hypercoagulability), which may contribute to localized venous thromboembolism, to acute disseminated intravascular coagulation (DIC), characterized by widespread microvascular thrombosis and subsequent consumption of platelets and coagulation proteins, eventually causing bleeding manifestations. The key event underlying this life-threatening complication is the overwhelming inflammatory host response to the infectious agent leading to the overexpression of inflammatory mediators. The latter, along with the microorganism and its derivatives are now believed to drive the major changes responsible for massive thrombin formation and fibrin deposition, namely 1) the aberrant expression of the TF by different cells (especially monocytes-macrophages), 2) the impairment of physiological anticoagulant pathways, orchestrated mainly by dysfunctional endothelial cells (ECs) and 3) the suppression of fibrinolysis due to overproduction of plasminogen activator inhibitor-1 (PAI-1) by ECs and likely also to thrombin-mediated activation of thrombinactivatable fibrinolysis inhibitor (TAFI). The ensuing microvascular thrombosis and ischemia are thought to contribute to tissue injury and multiple organ dysfunction syndrome (MODS). Recent evidence indicates that extracellular nuclear materials released from activated and especially apoptotic or necrotic cells, e.g. High Mobility Group Box-1 (HMGB-1) and histones, are endowed with cell toxicity, proinflammatory and clot-promoting properties and thus, during sepsis, they may represent late mediators that propagate further inflammation, coagulation, cell death and MODS. These insights into the pathogenesis of DIC and MODS may have implications for the development of new therapeutic agents potentially useful for the management of severe sepsis. © 2011 Elsevier Ltd. All rights reserved.
Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pathogenesis of sepsis-associated thrombus formation . . . . . . . . . . . . Up-regulation . of procoagulant pathways . . . . . . . . . . . . . . . . . . . Impairment . of physiological anticoagulant mechanisms . . . . . . . . . . . . Suppression . of fibrinolysis . . . . . . . . . . . . . . . . . . . . . . . . . Role .of endogenous inflammatory mediators in coagulation/fibrinolysis changes Pathogenesis of multiple organ dysfunction syndrome (MODS) . . . . . . . . Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
290 291 291 292 292 293 293 294 294
Introduction Abbreviations: APC, activated protein C; AT, antithrombin; DIC, disseminated intravascular coagulation; ECs, endothelial cells; EPCR, endothelial protein C receptor; HMGB-1, High Mobility Group Box-1; MODS, multiple organ dysfunction syndrome; PAI-1, plasminogen activator inhibitor-1; PC, protein C; PS, protein S; TAFI, thrombin activatable fibrinolysis inhibitor; TM, thrombomodulin; TF, tissue factor; TFPI, tissue factor pathway inhibitor. ⁎ Corresponding author at: Dipartimento di Scienze Biomediche e Oncologia Umana, Sezione di Patologia Generale, Università degli Studi di Bari - Policlinico, Piazza G. Cesare, 11, 70124 Bari, Italy. Tel.: + 39 080 5478468; fax: + 39 080 5478524. E-mail address: [email protected] (N. Semeraro). 0049-3848/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2011.10.013
Sepsis is almost invariably associated with haemostatic changes ranging from subclinical activation of blood coagulation (hypercoagulability) to systemic clotting activation with massive thrombin and fibrin formation, eventually leading to consumption of platelets and proteins of the haemostatic system (acute disseminated intravascular coagulation, DIC) [1,2]. From a clinical standpoint, septic patients may present with localized thrombotic manifestations, as
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
indicated by the observation that they are at increased risk for venous thromboembolism [3,4]. The most common and dramatic clinical feature, however, is widespread thrombosis in the microcirculation of different organs which may importantly contribute to solitary or multiple organ dysfunction (MODS) [1,2]. In fulminant DIC, the consumption of platelets and coagulation proteins will result in simultaneous bleeding of different severity. DIC is classically associated with Gram negative bacterial infections but it can occur in Gram positive sepsis (with a similar incidence) and in systemic infections with other micro-organisms [1,2]. The pathophysiology of sepsis-associated DIC is extremely complex and still extensively investigated. The key event is the systemic inflammatory response to the infectious agent [5]. Following recognition of unique constituents expressed by the causative microorganism and/or of host-derived factors via specific receptors (pattern recognition receptors, PRRs), particularly the Toll-like receptors (TLRs), immune and other host cells (monocytes-macrophages, platelets and endothelial cells among others) synthesize a number of proteins including proinflammatory cytokines. The latter, together with other mediators generated by the inflammatory cascade, including complement activation products [6], act in concert with the microorganisms and/or their derivatives to trigger the coagulation pathways, DIC and organ dysfunction [1,2,7]. Enzymes generated during the clotting cascade, in turn, interact with specific cellular receptors thus eliciting cell responses that amplify the inflammatory reactions [8]. Inflammation can also result in cell apoptosis or necrosis [5] and recent evidence indicates that products released from dead cells, such as nuclear proteins, are able to propagate further inflammation, coagulation, cell death and organ failure [5,9]. This article briefly summarizes current knowledge on the pathogenesis of DIC and MODS, and the ensuing development of potential therapeutics. Pathogenesis of sepsis-associated thrombus formation In sepsis the causative agent and the associated inflammatory response drive fibrin formation and deposition by several simultaneously acting mechanisms (Fig. 1), namely 1) up-regulation of procoagulant pathways, 2) down-regulation of physiological anticoagulants and 3) suppression of fibrinolysis [1,2,7]. Up-regulation of procoagulant pathways The aberrant in vivo expression of TF plays a pivotal role in sepsisassociated blood clotting activation, as indicated by the following observations: 1) the impairment of the TF pathway by various means prevents coagulation abnormalities (including fibrin deposition in target tissues) and lethality in animal models of sepsis or endotoxemia [2,7,10]; 2) the plasma levels of TF are increased in septic patients
291
and generally associated with raised concentrations of markers of clotting activation [2,7,11]. The cellular source of TF in sepsis, however, still remains an open question. In vitro, endothelial cells (ECs) and mononuclear phagocytes have long been known to synthesize TF in response to a wide variety of stimulating agents or conditions that are of pathophysiological importance in sepsis [2,12]. These cells may also exhibit other clot-promoting properties [2], thus providing a surface onto which the clotting pathways are initiated and propagated, eventually leading to fibrin formation in the cell microenvironment. TF expression has been detected also in human neutrophils in response to inflammatory agents [2,13], in eosinophils upon stimulation and subsequent translocation from their specific granules to the cell surface [14], and in activated platelets after de novo synthesis or release from αgranules [2,15,16]. Other studies, however, suggest that these cells do not synthesize TF but acquire it by binding TF-expressing microparticles (MPs) [2,10,17]. MPs are small membrane vesicles released from activated or apoptotic cells that can adhere to the surface of other cells via specific receptors (for instance PSGL-1 on leukocyte-derived MPs and P-selectin on activated platelets or ECs) making the recipient cell capable of triggering and propagating coagulation [2,12]. Although all mentioned cells might contribute to the aberrant in vivo expression of TF, most available studies point to activated monocytes-macrophages as the main triggers of blood coagulation during sepsis (Fig. 1). In animal models of endotoxemia or sepsis, TF expression is increased in important target organs where fibrin deposition often occurs during DIC, namely lung, kidney, liver, spleen and brain and, at cellular level, it is detected mainly in monocytes present in the microcirculation and in macrophages infiltrating the involved tissues [2,10,12,18]. In the same animals, blood monocytes and macrophages of different origin express strong TF activity [2,10,12,18]. In addition, a selective genetic deficiency of TF expression by hematopoietic cells as well as the deletion of TF gene in myeloid cells was found to reduce LPS-induced coagulation, inflammation, and mortality in mice [10,19]. Increased expression of monocyte-macrophage TF has been also documented in healthy volunteers after administration of low-dose endotoxin [20], in septic or endotoxemic patients, in whom TF was associated with clotting activation, MODS and lethal outcome, and in patients with peritonitis or acute respiratory distress syndrome [2,7,12,18]. Further support for a prominent role of monocytesmacrophages comes from studies on MPs. In endotoxemic mice, levels of MP TF activity were correlated with coagulation activation [21]. In addition, increased numbers of circulating TF-positive MPs of monocyte origin have been detected in patients with meningococcal sepsis and in human low-dose endotoxemia [2,7,22]. Surprisingly, and in contrast with the abundant in vitro evidence, ECs were negative for TF in most animal studies, with very few exceptions [2,10,17,18]. Moreover, the deletion of the TF gene in ECs had no significant effect on clotting activation in endotoxemic mice [10,19], clearly ruling out a major involvement
Fig. 1. Mechanisms contributing to thrombin formation and fibrin deposition in the microcirculation (see text for details). TF, tissue factor; EC, endothelial cell; TM, thrombomodulin; EPCR, endothelial protein C receptor; PS, protein S; TFPI, tissue factor pathway inhibitor; HS, heparan sulphate; PAI-1, plasminogen activator inhibitor-1; TAFI, thrombin activatable fibrinolysis inhibitor.
292
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
of EC TF. Likewise, the role of neutrophils and platelets as a direct source of TF in vivo during sepsis remains controversial [2,10,19,23]. These cells, however, may still participate in the activation of coagulation both by binding TF-positive monocyte MPs [2,10] and by other direct or indirect mechanisms (discussed later). Recently, it has been shown that the selective inhibition of TF expressed by non-hematopoietic cells substantially reduces the clotting activation in endotoxemic mice [10,19] but the precise cellular source of TF is unknown. In endotoxemic and septic animals, TF expression is increased not only in monocytes-macrophages but also in tissue cells, e.g. lung and kidney epithelial cells, and brain astrocytes [2,10,18]. Therefore, also considering that the role of ECs and vascular smooth muscle cells [19] remains uncertain, it is likely that TF up-regulation in parenchymal cells of target organs contributes to clotting coagulation during sepsis. Moreover, the well known increase in vascular permeability and vascular damage occurring during severe inflammation will allow the exposure of normal extravascular (e.g. fibroblast) TF to blood.
protective functions [28]. Indeed, besides exerting anticoagulant and profibrinolytic activity, APC also has important inflammationmodulating effects, including down-regulation of inflammatory cytokines and TF in activated leukocytes, antioxidant properties, anti-apoptotic activity and prevention of loss of endothelial barrier function [29]. During sepsis, plasma antithrombin (AT) is generally decreased because of consumption and low levels are associated with increased mortality [7]. Moreover, inflammatory stimuli can down-regulate the expression of heparan sulphate in cultured ECs [7,24] (Fig. 1) suggesting an additional mechanism for the reduced AT function in sepsis. With respect to TFPI, in animal models of endotoxemia or sepsis, a decreased TFPI expression was found in ECs (Fig. 1) of several organs [24,30]. In addition, anti-TFPI antibodies increased fibrin accumulation in the lungs of septic baboon [30], suggesting that TFPI underexpression, coupled with TF up-regulation, might augment the local procoagulant potential, thus promoting fibrin formation in tissues.
Impairment of physiological anticoagulant mechanisms Suppression of fibrinolysis Under physiological conditions the EC surface expresses various components involved in the anticoagulant pathways, including thrombomodulin (TM), endothelial protein C receptor (EPCR), protein S (PS), tissue factor pathway inhibitor (TFPI) and the heparinlike proteoglycan heparan sulphate. Because the endothelium plays a critical role in orchestrating the host response to sepsis and it is the target of the pathogen and the inflammatory mediators, the behavior of endothelial anticoagulant mechanisms, particularly the protein C (PC) pathway, has been extensively investigated in relation to sepsis. In cultured ECs, inflammatory mediators consistently reduced the expression of TM and EPCR, and the PS secretion [2,7,24,25] (Fig. 1). Although, at variance with these studies, animal experiments on in vivo expression of TM and EPCR by ECs produced rather controversial data [2,7,24,25], a rise in soluble plasma TM and EPCR was consistently observed during experimental endotoxemia, suggesting that endothelial activation-damage by inflammatory mediators does occur in vivo [2,7,24,25]. The importance of the PC anticoagulant pathway for the development of DIC in sepsis is definitely demonstrated by the fact that compromising the PC system resulted in a marked worsening of DIC and in increased morbidity and mortality in different animal models, whereas restoring an adequate APC function (e.g. treatment with APC) prevented the coagulopathy and improved organ failure and survival [7,24]. Interestingly, mice with heterozygous PC deficiency had more severe DIC and a higher mortality than the wild-type controls and mice homozygous for a point mutation of the TM gene that deletes the anticoagulant activity of the protein exhibited 10- to 30-fold greater amounts of fibrin in the microcirculation of several organs than the wild-type mice [2,7]. Studies in human sepsis have in general confirmed the dysfunction of the PC pathway. The plasma levels of soluble TM and EPCR were increased and TM levels were often correlated with disease severity and poor outcome [2,24,26]. Other common findings in septic patients are low levels of PC and PS, due to impaired liver synthesis and possibly to consumption, and low levels of free PS, due to increased C4b-binding protein [7,24]. Acquired severe PC deficiency was associated with early death [27]. Notably, the expression of TM and EPCR on morphologically intact ECs of dermal vessels was reduced in biopsy specimens of purpuric lesions from children with meningococcal sepsis, as compared with control skin-biopsy specimens [26]. Plasma levels of APC remained low in some of these patients even after treatment with PC concentrates, confirming down-regulation of TM in vivo and impaired PC activation. APC plasma levels were found to vary markedly among patients with severe sepsis and were significantly higher in survivors than in non survivors (28-day mortality), suggesting that endogenous APC serves
Numerous studies on cultured ECs challenged with inflammatory stimuli and in animal models of endotoxemia or sepsis have consistently reported the presence of markedly increased levels of plasminogen activator inhibitor-1 (PAI-1) in culture medium and in plasma, respectively. Although a simultaneous increase in tissuetype plasminogen activator (t-PA) often occurs, the net result is a fibrinolytic shut-down because of the very large amounts of PAI-1 [7,24]. In endotoxemic animals a strong elevation of PAI-1 mRNA was found in multiple tissues, including those affected by microthrombi (kidney, adrenals, lung and liver) and PAI-1 expression was detected primarily in ECs at all levels of the vasculature [24]. This suggests that plasma PAI-1 originates from ECs, although a contribution of platelets cannot be excluded [24]. It should be noted that, in some models of endotoxemia or cytokinemia, hypofibrinolysis and fibrin formation in adrenals and/or kidneys were most dependent on a decrease in PAs [1,24]. These findings indicate that suppression of fibrinolysis mediated by PAI-1 increase (Fig. 1) and other tissue- and species-specific alterations, such as decreased PAs in some models, are essential for fibrin deposition in tissue vasculature, at least in experimental sepsis. This concept is supported by studies in mice with targeted disruptions of genes encoding fibrinolytic proteins. Mice with a deficiency of PAs have more extensive fibrin deposition in organs when challenged with endotoxin, whereas PAI-1 knockout mice, in contrast to wildtype controls, have no microvascular thrombosis upon endotoxin challenge [2,24]. In human sepsis, the majority of studies is restricted to circulating fibrinolytic markers. A sustained increase in plasma PAI-1 has been consistently reported by numerous investigators and, in some studies, PAI-1 turned out to be a prognostic marker [2,24]. Plasma t-PA antigen was also elevated [2,24], but the net effect of the changes in t-PA and PAI-1 was definitely antifibrinolytic. The importance of impaired fibrinolysis is supported by the finding that a 4 G/5 G polymorphism in the PAI-1 promoter influencing PAI-1 expression was associated with the clinical outcome of severe sepsis [2,7,24]. More recent evidence indicates that other, thrombin-dependent mechanisms might contribute to hypofibrinolysis during sepsis. Thrombin causes resistance to fibrinolysis by forming more compact and less permeable clots [31], and by activating thrombin-activatable fibrinolysis inhibitor (TAFI), a plasma procarboxypeptidase that, once activated (TAFIa), removes the C-terminal lysines from partially degraded fibrin, thereby reducing plasmin formation [32]. Recent findings suggest that enhanced thrombin generation, the hallmark of sepsis, might influence the fibrin structure. Indeed, ECs stimulated by inflammatory cytokines to express TF cause the production of abnormally
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
dense fibrin networks that resist fibrinolysis [33]. Moreover, activated platelets, commonly found in sepsis, not only alter the fibrin structure and reduce susceptibility to lysis via the direct interaction between fibrin and αIIbβ3 integrin [33] but also release, together with the causative micro-organism, inorganic polyphosphates, which modify the fibrin architecture and reduce the binding of t-PA and plasminogen to fibrin, thus increasing fibrin resistance to fibrinolysis [34]. With respect to TAFI, evidence is accumulating that it might be involved in sepsis-associated hypofibrinolysis (Fig. 1). In vitro, LPSstimulated human monocytes inhibit fibrinolysis through a TFmediated enhancement of TAFI activation. Moreover, clots generated by TF-positive monocytes are resistant to the profibrinolytic activity of heparins [35]. Because TF-expressing monocytes-macrophages drive clotting activation in sepsis, these findings provide an additional mechanism whereby these cells may favor fibrin accumulation in the microcirculation. In animal models of endotoxemia or sepsis, TAFI levels were reduced, likely because of activation and consumption. In addition, blocking TAFIa with synthetic inhibitors or inhibiting thrombin-TM-dependent TAFI activation enhanced the rate of fibrin degradation thus reducing fibrin deposition in target tissues [2]. In human studies, TAFI levels were consistently decreased in septic patients and in healthy volunteers with low-grade endotoxemia [2]. More important, in patients with severe meningococcal infection, [2,36] the levels of TAFI activation markers were increased in patients with DIC as compared with those without, were significantly higher in non-survivors versus survivors and strongly correlated with severity scores of the disease. It thus appears that TAFI activation does occur in severe sepsis and that the measurement of TAFI activation markers may be clinically useful. The relevance of TAFI is further supported by the fact that a single nucleotide polymorphism in the TAFI gene that causes the substitution Thr325Ile and produces increased TAFIa stability and activity was associated with a poor outcome in meningococcal sepsis [2]. Role of endogenous inflammatory mediators in coagulation/fibrinolysis changes While in vitro the pathogen, its derivatives and the main inflammatory mediators are able to influence to variable extent each of the mechanisms described above, their role during experimental and human sepsis or endotoxemia is more difficult to establish. With respect
293
to the up-regulation of procoagulant pathways, various inflammatory mediators are likely involved, besides the micro-organism itself and its products (e.g. LPS). Although the main cytokines, i.e. TNF, IL-1 and IL6, can activate blood coagulation in humans and primates most probably via the TF pathway, neutralization studies with specific antibodies would suggest a major role of endogenous IL-6 and, to a lesser extent, of IL-1 [37]. With respect to the anticoagulant pathways and fibrinolysis, the few available studies suggest that, in different animal models of sepsis or endotoxemia, TNF and IL-1 are involved in the TM and PC downregulation, and in PAI-1-mediated suppression of fibrinolysis [37]. Recently, in a baboon model of severe sepsis, a potent inhibitor of C3 convertase was shown to attenuate the inflammatory response, DIC, and the dysfunction of several organs [38]. The simultaneous decrease in TF and PAI-1 expression in multiple tissues, and the attenuation of TFPI and TM down-regulation in ECs clearly point to a key role of complement-derived mediators in sepsis-associated pathophysiological changes. Altogether, these findings indicate that, as predicted by in vitro studies, inflammatory stimuli are indeed able to elicit the main mechanisms responsible for in vivo thrombin generation and fibrin deposition during sepsis.
Pathogenesis of multiple organ dysfunction syndrome (MODS) MODS is the hallmark of severe sepsis and septic shock and represents the main cause of the high mortality in these conditions. Several closely interlinked mechanisms have been proposed to explain this dramatic event, all of which result from the global hyper-inflammatory response to the triggering pathogen [39]. An important role of DIC is supported by several lines of evidence [2,7]: a) numerous histological studies in septic patients and in animals with sepsis or endotoxemia have documented the presence of thrombi in small and mid-size vessels of multiple organs and the relation of these thrombi with organ ischemia and dysfunction; b) in animal experiments, amelioration of DIC by various interventions improves organ failure and, in some cases, mortality; c) DIC has been shown to be an independent predictor of organ dysfunction and mortality in septic patients. Other widely recognized mechanisms contributing to MODS are the release of reactive oxygen and nitrogen species, and proteolytic enzymes by neutrophils recruited at tissue level and the presence in the interstitial space of high concentrations of selected cytokines that might be directly toxic
Fig. 2. Extracellular release of histones and mechanisms of thrombus formation. Histones are nuclear proteins released from activated neutrophils and dying cells and can promote thrombus formation by inducing endothelial injury, platelet activation with consequent expression of procoagulant properties, and impairment of TM-mediated protein C activation. NETs, neutrophil extracellular traps; PolyP, polyphosphates; PC, protein C; APC, activated PC; TM, thrombomodulin.
294
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295
to vulnerable parenchyma and play a role particularly in sepsis with severe leukopenia [39]. Increasing evidence indicates the existence of new players in sepsis-associated organ dysfunction-failure, namely extracellular nuclear proteins that originate mainly from dying cells and therefore can be considered as late mediators of MODS. High Mobility Group Box-1 (HMGB-1) is released actively by stimulated cells of innate immunity and passively by necrotic cells into the extracellular environment where it becomes a lethal mediator of systemic inflammation [40]. HMGB-1 increases in the circulation in different animal models of sepsis and in septic patients. In animal experiments, HMGB-1 proved to be lethal and to promote the development of microvascular thrombosis by stimulating monocyte TF expression and by reducing the activity of thrombin-TM complex thereby inhibiting protein C activation [41]. Notably, anti-HMGB-1 antibodies conferred significant protection against lethal endotoxemia, sepsis and LPS-induced lung injury [40]. More recently, it has been shown that extracellular histones too (particularly histones H3 and H4) are major mediators of injury during sepsis [9]. Besides being toxic for human ECs and other cells in vitro, histones mimicked the manifestations of sepsis, including microvascular thrombosis, organ failure and death when injected in mice. More important, antibodies against histone H4 protected mice in different models of endotoxemia and sepsis. Recombinant APC was shown to degrade histones thus lowering their toxicity towards ECs in vitro and preventing lethality in histone-treated animals. Of note, histones and their APC-induced degraded forms were detected in plasma from septic animals and patients. Possible sources of histones during sepsis are activated inflammatory cells and dying cells (Fig. 2). Neutrophils can be induced by different stimuli, including activated platelets, to release the so-called neutrophil extracellular traps (NETs), complex structures containing histones, DNA and granule proteins [2,41]. Besides killing micro-organisms, NETs induce tissue damage and, at least in vitro, thrombus formation [42]. Another, likely more important source of histones is massive cell apoptosisnecrosis overwhelming the clearance capacity of mononuclear phagocytes, thereby permitting histones to enter the circulation. Histones have recently been shown to promote thrombin formation by different mechanisms (Fig. 2) among which platelet activation resulting in the release of polyphosphates and the expression of other platelet procoagulant properties [43], and the impairment of TM-mediated protein C activation [44]. The release of nuclear proteins during the late stages of sepsis can thus amplify inflammation, coagulation, cell death and MODS. In conclusion, considerable progress has been made in our knowledge on the mechanisms underlying sepsis-associated DIC and MODS that help to develop new therapeutic agents for the management of severe sepsis. However, the use of TF inhibitors, which would be the most logical treatment considering the pivotal role of TF in clotting activation during sepsis, is still debated. A phase III trial with recombinant TFPI did not show an overall survival benefit in septic patients [7]. Likewise, treatment with antithrombin concentrates failed to significantly reduce the mortality of septic patients in a large-scale clinical trial [1,7]. So far, the only anticoagulant drug showing promising results is recombinant human APC, which has been evaluated in several clinical trials and is indicated for the reduction of mortality in adult patients with severe sepsis at high risk of death [1]. The beneficial effects of this drug have been attributed not only to the restoration of the protein C anticoagulant pathway but also to its wellknown anti-inflammatory action [25,29]. The finding that APC degrades histones [9] adds a new mechanism whereby the drug exerts its beneficial effects especially in severe sepsis. Because histones seem to be critical mediators of organ dysfunction and death in septic patients, an attractive approach to treat MODS and prevent death could be the development of effective histone antagonists which might prove therapeutic without the bleeding complications that can result from APC therapy.
Conflict of interest The authors state that they have no conflict of interest.
References [1] Levi M, Schultz M, van der Poll T. Disseminated intravascular coagulation in infectious disease. Semin Thromb Hemost 2010;36:367–77. [2] Semeraro N, Ammollo CT, Semeraro F, Colucci M. Sepsis-associated disseminated intravascular coagulation and thromboembolic disease. Mediterr J Hematol Infect Dis 2010;2(3):e2010024. [3] Smeeth L, Cook C, Thomas S, Hall AJ, Hubbard R, Vallance P. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075–9. [4] Alikhan R, Spyropoulos AC. Epidemiology of venous thromboembolism in cardiorespiratory and infectious disease. Am J Med 2008;121:935–42. [5] Cinel I, Opal SM. Molecular biology of inflammation and sepsis: a primer. Crit Care Med 2009;37:291–304. [6] Markiewski MM, DeAngelis RA, Lambris JD. Complexity of complement activation in sepsis. J Cell Mol Med 2008;12:2245–54. [7] Levi M. The coagulant response in sepsis. Clin Chest Med 2008;29:627–42. [8] Monroe DM, Key NS. The tissue factor-factor VIIa complex: procoagulant activity, regulation, and multitasking. J Thromb Haemost 2007;5:1097–105. [9] Xu J, Zhang X, Pelayo R, Monestier M, Ammollo CT, Semeraro F, et al. Extracellular histones are major mediators of death in sepsis. Nat Med 2009;15:1318–21. [10] Pawlinski R, Mackman N. Cellular sources of tissue factor in endotoxemia and sepsis. Thromb Res 2010;125:S70–3. [11] Gando S, Nanzaki S, Sasaki S, Kemmotsu O. Significant correlation between tissue factor and thrombin markers in trauma and septic patients with disseminated intravascular coagulation. Thromb Haemost 1998;79:1111–5. [12] Osterud B, Bjorklid E. Sources of tissue factor. Semin Thromb Hemost 2006;32: 11–23. [13] Maugeri N, Brambilla M, Camera M, Carbone A, Tremoli E, Donati MB, et al. Human polymorphonuclear leukocytes produce and express functional tissue factor upon stimulation. J Thromb Haemost 2006;4:1323–30. [14] Moosbauer C, Morgenstern E, Cuvelier SL, Manukyan D, Bidzhekov K, Albrecht S, et al. Eosinophils are a major intravascular location for tissue factor storage and exposure. Blood 2007;109:995–1002. [15] Camera M, Frigerio M, Toschi V, Brambilla M, Rossi F, Cottell D, et al. Platelet activation induces cell-surface immunoreactive tissue factor expression, which is modulated differently by antiplatelet drugs. Arterioscler Thromb Vasc Biol 2003;23:1690–6. [16] Panes O, Matus V, Saez CG, Quiroga T, Pereira J, Mezzano D. Human platelets synthesize and express functional tissue factor. Blood 2007;109:5242–50. [17] Østerud B. Tissue factor expression in blood cells. Thromb Res 2010;125:S31–4. [18] Semeraro N, Colucci M. Tissue factor in health and disease. Thromb Haemost 1997;78:759–64. [19] Pawlinski R, Wang JG, Owens III AP, Williams J, Antoniak S, Tencati M, et al. Hematopoietic and nonhematopoietic cell tissue factor activates the coagulation cascade in endotoxemic mice. Blood 2010;116:806–14. [20] Franco RF, de Jonge E, Dekkers PE, Timmerman JJ, Spek CA, van Deventer SJ, et al. The in vivo kinetics of tissue factor messenger RNA expression during human endotoxemia: relationship with activation of coagulation. Blood 2000;96: 554–9. [21] Wang JG, Manly D, Kirchhofer D, Pawlinski R, Mackman N. Levels of microparticle tissue factor activity correlate with coagulation activation in endotoxemic mice. J Thromb Haemost 2009;7:1092–8. [22] Aras O, Shet A, Bach RR, Hysjulien JL, Slungaard A, Hebbel RP, et al. Induction of microparticle- and cell-associated intravascular tissue factor in human endotoxemia. Blood 2004;103:4545–53. [23] Rondina MT, Schwertz H, Harris ES, Kraemer BF, Campbell RA, Mackman N, et al. The septic milieu triggers expression of spliced tissue factor mRNA in human platelets. J Thromb Haemost 2011;9:748–58. [24] Semeraro N, Colucci M. Endothelial cell perturbation and disseminated intravascular coagulation. In: ten Cate H, Levi M, editors. Molecular mechanisms of disseminated intravascular coagulation. Georgetown: Landes Bioscience; 2003. p. 156–90. [25] Esmon CT. Inflammation and thrombosis. J Thromb Haemost 2003;1:1343–8. [26] Faust SN, Levin M, Harrison OB, Goldin RD, Lockhart MS, Kondaveeti S, et al. Dysfunction of endothelial protein C activation in severe meningococcal sepsis. N Engl J Med 2001;345:408–16. [27] Macias WL, Nelson DR. Severe protein C deficiency predicts early death in severe sepsis. Crit Care Med 2004;32:S223–8. [28] Liaw PC, Esmon CT, Kahnamoui K, Schmidt S, Kahnamoui S, Ferrell G, et al. Patients with severe sepsis vary markedly in their ability to generate activated protein C. Blood 2004;104:3958–64. [29] Mosnier LO, Zlokovic BV, Griffin JH. The cytoprotective protein C pathway. Blood 2007;109:3161–72. [30] Tang H, Ivanciu L, Popescu N, Peer G, Hack E, Lupu C, et al. Sepsis-induced coagulation in the baboon lung is associated with decreased tissue factor pathway inhibitor. Am J Pathol 2007;171:1066–77. [31] Wolberg AS. Thrombin generation and fibrin clot structure. Blood Rev 2007;21: 131–42.
N. Semeraro et al. / Thrombosis Research 129 (2012) 290–295 [32] Mosnier LO, Bouma BN. Regulation of fibrinolysis by thrombin activatable fibrinolysis inhibitor, an unstable carboxypeptidase B that unites the pathways of coagulation and fibrinolysis. Arterioscler Thromb Vasc Biol 2006;26:2445–53. [33] Campbell RA, Overmyer KA, Selzman CH, Sheridan BC, Wolberg AS. Contributions of extravascular and intravascular cells to fibrin network formation, structure, and stability. Blood 2009;114:4886–96. [34] Mutch NJ, Engel R, Uitte de Willige S, Philippou H, Ariëns RA. Polyphosphate modifies the fibrin network and down-regulates fibrinolysis by attenuating binding of tPA and plasminogen to fibrin. Blood 2010;115:3980–8. [35] Semeraro F, Ammollo CT, Semeraro N, Colucci M. Tissue factor-expressing monocytes inhibit fibrinolysis through a TAFI-mediated mechanism, and make clots resistant to heparins. Haematologica 2009;94:819–26. [36] Emonts M, de Bruijne EL, Guimarães AH, Declerck PJ, Leebeek FW, de Maat MP, et al. Thrombin-activatable fibrinolysis inhibitor is associated with severity and outcome of severe meningococcal infection in children. J Thromb Haemost 2008;6:268–76. [37] van der Poll T, de Jonge E, Levi M. Regulatory role of cytokines in disseminated intravascular coagulation. Semin Thromb Hemost 2001;27:639–51. [38] Silasi-Mansat R, Zhu H, Popescu NI, Peer G, Sfyroera G, Magotti P, et al. Complement inhibition decreases the procoagulant response and confers organ protection in a baboon model of E. coli sepsis. Blood 2010;116:1002–10.
295
[39] Wang H, Ma S. The cytokine storm and factors determining the sequence and severity of organ dysfunction in multiple organ dysfunction syndrome. Am J Emerg Med 2008;26:711–5. [40] Sunden-Cullberg J, Norrby-Teglund A, Treutiger CJ. The role of high mobility group box-1 protein in severe sepsis. Curr Opin Infect Dis 2006;19:231–6. [41] Papayannopoulos V, Metzler KD, Hakkim A, Zychlinsky A. Neutrophil elastase and myeloperoxidase regulate the formation of neutrophil extracellular traps. J Cell Biol 2010;191:677–91. [42] Fuchs TA, Brill A, Duerschmied D, Schatzberg D, Monestier M, Myers Jr DD, et al. Extracellular DNA traps promote thrombosis. Proc Natl Acad Sci U S A 2010;107: 15880–5. [43] Semeraro F, Ammollo CT, Morrissey JH, Dale GL, Friese P, Esmon NL, et al. Extracellular histones promote thrombin generation through platelet-dependent mechanisms: involvement of platelet TLR2 and TLR4. Blood 2011;118:1952–61. [44] Ammollo CT, Semeraro F, Xu J, Esmon NL, Esmon CT. Extracellular histones increase plasma thrombin generation by impairing thrombomodulin-dependent protein C activation. J Thromb Haemost 2011;9:1795–803.