The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533
Medicine in focus
Sepsis: current concepts in intracellular signaling Derek Strassheim, Jong Sung Park, Edward Abraham∗ Pulmonary Science and Critical Care, University of Colorado Health Sciences Center, 4200 East 9th Avenue, Box C272, Denver, CO 80262, USA Received 1 November 2001; received in revised form 15 February 2002; accepted 25 March 2002
Abstract Sepsis is the systematic response to infection. In septic patients who develop severe disease, excessive inflammation damages the lungs, liver, kidneys, and cardiovascular system, leading to multiple organ failure and an associated high mortality rate. Sepsis is the leading cause of death in the intensive care unit. The damage to critical organs is primarily due to excessive acute inflammatory response rather than inadequate combat of the infection per se. Impairment of critical organs is closely associated with infiltration of activated neutrophils into those tissues as well as increased activation of endothelial, epithelial, and macrophage populations within the organs to produce a deregulated, overly aggressive inflammatory response. New pharmacological advances hold promise in improving survival from this multi-systemic disorder. Increasing understanding of the signal transduction pathways of inflammatory cells involved in the disease suggests that targeting specific kinases and transcriptional regulatory mechanisms may prove improve outcome from sepsis. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Multi-systemic disorder; Inflammatory response; Septic patients
1. Introduction Sepsis is a disease brought about by the presence of bacteria and bacterial products, such as lipopolysaccharide (LPS) in the bloodstream. The incidence and mortality rate of sepsis remains high, and sepsis is still the most common cause of death in the intensive care unit [1]. The disease differs widely in severity, but for those patients who become critically ill, poor outcome correlates very closely with the severity of the body’s own inflammatory responses to infection [1]. A sufficient inflammatory response is a prerequisite for successfully combating bacterial infection. ∗ Corresponding author. Tel.: +1-303-315-7047; fax: +1-303-315-5632. E-mail address:
[email protected] (E. Abraham).
However, self-inflicted damage to host tissues, such as those of lung, liver, circulatory and renal systems, as a consequence of an overtly aggressive inflammatory response appears to be the key component for a poor patient outcome rather than an inability to fight infection [1]. Macrophages and neutrophils are two innate immune cell types intimately associated with the excessive inflammatory response characterizing severe sepsis. These cells both produce and respond to cytokines, chemokines, and other pro-inflammatory mediators released from endothelial, epithelial, and other cell populations activated by microbial products, such as LPS. Host tissues, as well as infiltrating cells, such as neutrophils, are injured by production of tissue damaging reactive oxygen intermediates, proteases, and general inflammatory environment brought about by these cells (Fig. 1). Stimulated in
1357-2725/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 1 3 5 7 - 2 7 2 5 ( 0 2 ) 0 0 0 4 9 - 3
1528
D. Strassheim et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533
Fig. 1. Migration of neutrophils from vascular networks into major organs where damage to host tissues from hydrolytic enzymes and reactive oxygen species is a major feature of septic shock pathology. (1) Neutrophils in the bloodstream detect presence of chemokines such as IL-8, attach to capillary wall endothelial cells via integrin-mediated processes; (2) binding to endothelial cells, and migration of neutrophil through capillary wall initiates signaling events within neutrophils allowing cytoskeletal shape changes, and priming neutrophils for superoxide formation; (3) once inside host tissues, neutrophils respond to presence of chemokines, cytokines and microbial products by releasing tissue-destroying proteases; (4) and tissue destroying reactive oxygen species; (5) eventually resulting in cell death; (6) of host tissue.
situ macrophages and neutrophils also exacerbate the disease by releasing chemokines and cytokines which set up a positive feedback loop causing even more severe inflammation. If the disease progresses, organ failure develops in severe cases, with a concomitant high mortality rate. This brief review focuses predominantly on the signaling pathways controlling macrophage and neutrophil biology, since limiting the inflammatory response due to these cell types appears to represent a promising avenue for therapeutic intervention in sepsis.
rnediators (Fig. 2, Table 1) [2]. This initiates further cycles of cytokine release from various cell types further promoting the inflammatory response. Both tumor necrosis factor (TNF␣) and interleukin-1 (IL-1) have important roles in enhancing inflammatory responses involved in sepsis. For example, mice in which TNF receptors have been knocked out demonstrate improved survival from sepsis, reviewed in [3]. The signaling aspects, and physiology of mouse receptor knockouts for LPS, TNF␣, IL-1 have been well-described elsewhere [2–4].
2. Pathogenesis
2.2. MAPK pathways activated by LPS, cytokines, and chemokines
2.1. Initial signaling responses in sepsis Recent studies show that among patients with sepsis, the numbers infected with gram-negative or grampositive bacteria are approximately equal. Both LPS from gram-negative bacteria and gram-positive bacterial products, including peptidoglycans and lipotechoic acid, interact with toll like receptors (TLR) on cell surfaces, activating kinases that lead to enhanced transcription of cytokines and other pro-inflammatory
LPS, cytokines, and chemokines activate various mitogen activated protein kinase (MAPK) signaling networks, thereby leading to transcription of cytokine genes and have been described elsewhere [4]. Among the most important of the kinases are pathways involving p38, ERK 1/2 (p42/p44), and other MAPKs. Indeed the importance of such pathways has been demonstrated by the removal of one such macrophage MAPK component, Tpl2, which was found to protect against LPS-induced sepsis [5].
D. Strassheim et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533
1529
Fig. 2. Signaling events controlling neutrophil and macrophage biology. Numerous, complex and highly interconnected signaling pathways control neutrophil and macrophage physiology. These pathways control chemotaxis of neutrophils from the peripheral bloodstream to sites of inflammation; release of secretory granules containing anti-microbial proteases; production of anti-microbial reactive oxygen species (ROS); release of chemokines, pro- and anti-inflammatory cytokines from both neutrophils and macrophages; and delay of the constitutive apoptotic pathway in neutrophils. LPS activates many transcription factors via mitogen activated protein kinase pathways, thereby stimulating release of TNF␣, a primary initiator of the inflammatory response. Chemokines activate G protein coupled receptors in macrophages and neutrophils, resulting in activation of effectors such as phosphoinositide 3-kinase (PI3K␥), phospholipase C2 (PLC2) and PLC3, and phospholipase A2 (PLA2 ). PI3K␥, PLC2, and PLC3 are essential for production of reactive oxygen species, superoxide (O2 − ) in response to chemokines. PI3K␥ is also involved in chemotaxis towards chemokines. PLA2 modulates cell signaling in autocrine and paracrine manners by controlling the production of prostaglandins, leukotrienes, and thromboxanes. Pro-inflammatory cytokines such as IL-1, and anti-inflammatory cytokines, such as IL-10, modulate gene expression and further cytokine production. The duration and extent of neutrophil activation correlates with the degree of host tissue damage and severity of septic shock.
2.3. Transcription factors NF-B plays a central important role both in inflammation and sepsis [2–4]. NF-B is critical for the transcription of many molecules involved in inflammatory responses, including cytokines, chemokines, adhesion molecules, nitric oxide synthase, and complement components. Although the classic transcriptionally active form of NF-B is a hetero-dimer including p65 (RelA) and p50 subunits, other compositional forms of NF-B exist, and utilize RelB, c-Rel, or p52 as hetero- or homo-dimers.
Tolerance or desensitization of cells, develops as a negative feedback response to the continued exposure to LPS and manifests as diminished production of TNF␣ or other cytokines under the positive control of NF-B. p50−/− mice fail to develop tolerance to LPS [6]. Macrophages from p50−/− mice, unlike their wild type counterparts, do not show diminished synthesis of TNF␣ after extensive prior exposure to LPS. The mechanism for the lack of LPS tolerance in this setting appears to reflect the inhibitory role that p50/p50 homodimers normally play in suppressing NFB-dependent transcription in vivo. By preventing
1530
D. Strassheim et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533
Table 1 Summary of protein function Protein
Function
Toll like receptors
Receptor family, at least 10 members, TLR4 present on most cell types binds LPS, resulting in activation of NFB and pro-inflammatory gene transcription. Interleukin-1 is a major pro-inflammatory cytokine, produced by macrophages, activates NFB. Tumor necrosis factor ␣ is a major pro-inflammatory cytokine, released from macrophages, controlling cellular survival/apopotosis. Generally small peptide hormones, which bind to G-protein coupled receptors present on many cell types, have profound effects of leukocyte biology, stimulating chemotaxis, proteolytic enzyme release and superoxide production. The mitogen activated protein kinase family composed of 12 members grouped into five subfamilies. These kinases are activated by many signaling pathways, regulating gene transcription, and many other functions. A MKKKs mitogen activated protein kinase kinase kinases, activate the kinases which activate MAPKs. 12 members in four subfamilies. Src is a non-receptor cytoplasmic tyrosine kinase; src family includes hck, fgr, are widely expressed regulating many signaling pathways. A protein tyrosine kinase of the Axl/UFO family apparently negatively regulating the transcription of TNF␣ gene in macrophages. Transcription factor particularly important in inflammation, transcription of many pro-inflammatory gene products requires NFB, believed to play important role in sepsis. Janus kinase-signal transducers and activators of transcription are tyrosine phosphorylated transcription factors activated by JAKs activated by various cytokine receptors. Four JAKs, and seven STATs are known. Phospholipase C catalyses the hydrolysis of phospahidylinositol 4,5 bisphosphate to the second messengers diacylglycerol and inositol 1,4,5 trisphosphate which activate protein kinase C and release of internal calcium stores respectively. Activated by chemokines in neutrophils and macrophages. Catalyze conversion of phosphatidyl inositol 4,5 bisphosphate to membrane delimited second messenger phosphatidyl inositol-3,4,5-trisphosphate. Several forms are known regulated by distinct mechanisms, intimately involved in neutrophil activation. Protein kinase B is ubiquitously expressed kinase involved in many signaling pathways, appears to be very important in neutrophil activation. PKB binds and is activated by the phospholipid second messenger PIP3 the product of PI3K activity. A family of some 20 members, transduce the activity of seven transmembrane G protein coupled receptors. G protein coupled receptors, a superfamily of probably more than 1000, regulating most cellular functions. Intimately involved in activation neutrophils and macrophages by chemokines.
IL-1 TNF␣ Chemokines
MAPK Tpl2 Src Mer NFB JAK-STAT PLC
PI3K
PKB
G proteins GPCRs
transcriptionally active hetero-dimers of NF-B, such as p50/p65, from interacting with B binding sites in the promoters of TNF␣ and other pro-inflammatory genes under the regulatory control of NF-B, increased levels of p50/p50 appears to result in a cellular state where there is reduced cellular activation in response to endotoxin. STAT3 is a member of the signal transducers and activators of transcription (STAT) transcription factor family, which functions together with the janus kinases (JAK), in JAK-STAT signaling networks commonly activated by cytokines. Activation of STAT3 appears to be essential for the anti-inflammatory actions of IL-10 [7]. The serum concentrations of TNF␣, and the cytokines IL-10, IL-6, IFN␥, and IL-1 were all substantially elevated after LPS administration
to STAT3−/− mice [7]. In these STAT3−/− mice, STAT3 expression was specifically knocked out in macrophages and neutrophils by use of the cre-lox P system. The mutant mice were highly susceptible to LPS-induced sepsis, presumably because the anti-inflammatory actions of IL-10 on macrophages and neutrophils were greatly suppressed. Additional STAT family members show importance in sepsis. In mice the cecal ligation and puncture procedure (CLP) induces severe septic peritonitis. This model of severe sepsis contrasts with that caused by injection of LPS, in that the animal must combat live bacteria carried by fecal matter into the body cavity. STAT4−/− mice and STAT6−/− mice subject to CLP show resistance to sepsis lethality, with 50% dead on day four versus 50% dead on day 2 in wild type
D. Strassheim et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533
mice [8]. STAT4 and STAT6 are activated by the cytokines IL-12 and IL-13 respectively. In STAT6−/− mice bacterial levels were much lower than in wild type mice, after cecal ligation and puncture (CLP). STAT6−/− mice had higher peritoneal II-12, TNF␣, macrophage derived chemokine (MDC), and C10 after CLP, mediators known to enhance bacterial clearance. In STAT4−/− mice, renal injury induced by infection was reduced, with lower renal levels of the pro-inflammatory chemokines MIP-2 and KC without any alterations in expression of the anti-inflammatory cytokines IL-10 or IL-13. Thus, in contrast to the beneficial role that STAT3 appears to occupy in sepsis, STAT4 and STAT6 may have deleterious downstream effects in that STAT4−/− and STAT6−/− mice appear to benefit from diminished activity of STAT4 and STAT6, by way of increased bacterial clearance, and diminished tissue damage, respectively. 2.4. Protein tyrosine kinases: Src family, integrins and Mer Src was the first described member of the src family of non-receptor protein tyrosine kinases. Neutrophil adhesion to endothelial cells, the first step allowing migration of neutrophils from capillaries into host tissues, is a critical component of tissue inflammation, and leads to the activation of neutrophils to increase release of ROS and proteolytic enzymes. The neutrophil expressed Src family tyrosine kinases Hck and Fgr are critical for integrin-mediated cell adhesion [9]. Both 2-intracellular adhesion molecule 1 (ICAM-1) and 1-vascular cell adhesion molecule 1 (VCAM-1) have important roles in mediating adhesion between neutrophils and endothelial cells [9]. Hck−/− Fgr−/− double knockout mice showed reduced neutrophil accumulation in liver, and suffered only mild liver damage in contrast to control mice after LPS administration [10]. The receptor tyrosine kinase, Mer, is expressed in macrophages, and is a member of the Axl/Mer/Eyk/ Tyro3/Rek family of protein tyrosine kinases. Mer kinase deficient mice (Merkd ) showed vastly increased susceptibility to LPS-induced cell death, 88% versus 7% for wild type mice [11]. Administration of anti-TNF␣ Ab reduced the death rate, suggesting that Mer kinase activity inhibits the process of TNF␣ release from macrophages. LPS-induced greater
1531
activation of NF-B in macrophages from Merkd mice compared to nontransgenic controls, implicating that Mer may directly antagonize LPS activation of NF-B. 2.5. Phospholipid signaling; phosphoinositide 3-kinase, phospholipase C β2/3 and phospholipase A2 Signals that up-regulate inflammation, such as LPS, tend to delay neutrophil apoptosis, providing a potential mechanism to further promote inflammation. Conversely, apoptosis of neutrophils is believed to be a dynamic process that limits tissue injury at sites of inflammation. Phosphoinositide 3-kinase (PI3K) has been implicated in both neutrophil chemotaxis [12] and antiapoptotic processes [13]. PI3K converts phosphatidyl inositol 4,5-bisphosphate to phosphatidylinositol 3,4,5-trisphosphate (PIP3 ) PIP3 is a bona fide second messenger, activating protein kinase B (PKB/Akt) enzymes, thereby regulating numerous downstream signaling pathways. Gi -activating chemokine receptors are generally believed to activate only the PI3K␥ isoenzyme since only this PI3K␥ interacts with the ␥ subunits of activated Gi proteins. In concordance with these ideas, neutrophils from PI3K␥−/− mice show greatly diminished in vitro chemotaxis in response to the Gi -activating stimuli fMLP and MIP-1␣ [14]. Additionally, these neutrophils do not generate superoxide in response to fMLP, implicating PI3K␥ as an important potential drug target. However, despite previous data showing linkage between PI3K␥ and G protein coupled receptors, recent data from our laboratory demonstrate that LPS signaling also results in activation of PI3K␥ in neutrophils [15]. Additionally, PI3K␥ plays an important role in activation of NF-B and transcription of pro-inflammatory cytokines in LPS stimulated neutrophils [15]. Neutrophil chemokine receptors are known to activate predominantly Gi type G proteins, which activate phospholipase C (PLC), with the resultant activation of PKC and calcium-dependent processes. Of the four G protein activated mammalian PLC1-4, neutrophils express only PLC2 and PLC3. Interestingly, PLC2/3−/− , double knockout mice failed to produce any superoxide in response to, fMLP, thereby implicating PLC2/3 along with PI3K␥ as important potential drug targets [13].
1532
D. Strassheim et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533
Phospholipase A2 activated by cell surface receptors, such as those for chemokines and cytokines, hydrolyses phosphatidyl choline to lysophosphatidyl choline, and arachidonic acid. Arachidonic acid is the intermediate used for synthesis of a whole range of prostaglandins, leukotrienes, and thromboxanes, which are fatty acid based auto- and paracrine cell agonists. These agents acutely regulate the inflammatory response, typically via activation of G protein coupled receptors. PLA2 −/− mice were found to have significantly reduced pulmonary edema, neutrophil sequestration, and deterioration of gas exchange caused by LPS-induced sepsis [16]. A recently initiated clinical trial is examining the utility of inhibition of sPLA2 in the treatment of patients with severe sepsis or septic shock. 2.6. Signaling initiated by reactive oxygen species ROS affect many signaling events, protein kinase networks, and downstream transcription factors. Activated neutrophils, such as those infiltrating the lungs and other organs of septic animals produce large amounts of ROS [17]. Under these conditions, neutrophil derived ROS may initiate signaling events in other cell populations, in addition to their direct destructive effects on host tissues. Recent evidence demonstrates that activated PKB/ Akt protects lung tissue from oxidant-induced lung injury and delays death of mice [18]. It is believed that hyperoxic pulmonary damage is caused by production of ROS [17]. Expression of constitutively active PKB in the lungs via adenovirus gene transfer protected lungs from hyperoxic pulmonary damage, induced in mice by exposure to 100% oxygen [18]. It should also be noted that the ROS, H2 O2 , potently activates PKB in a rapid and transient manner in vascular smooth muscle cells [19]. PKB is well known as an anti-apoptotic pro-survival agent, and it is therefore logical to expect it may play a pivotal role in protecting host tissue from damage induced by sepsis.
3. Therapy Any therapeutic strategies for sepsis treatment should consider limiting the extent and/or duration of the activation of neutrophils, macrophages, and other
activated cell populations, but without removing their essential anti-microbial activity completely. Relevant targets in neutrophils might include reducing chemotaxis, secretion of anti-microbial proteases and other hydrolytic enzymes, production of ROS, release of chemokines and pro-inflammatory cytokines1 or increasing the release of anti-inflammatory cytokines. Additionally, the recently reported [20,21] positive results, showing that administration of recombinant activated protein C reduced mortality after severe sepsis and septic shock, suggest that modulation of coagulation is an appropriate downstream target for anti-sepsis therapies. Similar results have been obtained with tissue factor pathway inhibitor (TFPI), an endogenous inhibitor of tissue factor associated coagulation cascade [22]. Therapies targeting specific cytokines, such as IL-1 and TNF␣, have failed to demonstrate significant benefit for patients [23]. In particular antibodies or soluble receptors capable of sequestering circulating inflammatory cytokines, have proven ineffective even though initial promise was shown in animal models [23]. As discussed in this review, there are several kinase pathways that may be appropriate targets for therapy in sepsis. Among these are the src family tyrosine kinases hck and fgr, PI3K, PLC2, and PLC3. Other kinases that participate in LPS-induced signaling and whose activation results in modulation of inflammatory responses include p38 and ERK (p42/p44). Transcription factors that participate in regulation of expression of inflammatory mediators may also be appropriate targets for antisepsis therapies. Much interest has centered on NF-B, but because of its multiple effects on immunoregulatory pathways, toxicities with therapies directed against this transcription factor may be significant. STAT3, apparently essential for the action of the anti-inflammmatory cytokine IL-10, might be a target whose activity should be increased. By contrast, strategies to decrease STAT4 and STAT6 activity may prove beneficial.
References [1] D.C. Angus, R.S. Wax, Epidemiology of sepsis: an update, Crit. Care. Med. 29 (Suppl. 7) (2001) S109–116. [2] M. Guha, N. Mackman, LPS induction of gene expression in human monocytes, Cell. Signall. 13 (2001) 85–94.
D. Strassheim et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1527–1533 [3] J.C. Gutierrez-Ramos, H. Bluethmann, Molecules and mechanisms operating in septic shock: Lessons from knockout mice, Immunol. Today 18 (1997) 329–334. [4] J.M. Kyriakis, J. Avruch, Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation, Physiol. Rev. 81 (2001) 808–860. [5] C.D. Dumitru, J.D. Ceci, C. Tsatsanis, D. Kontoyiannis, K. Stamatakis, J.H. Lin, C. Patriotis, N.A. Jenkins, N.G. Copeland, G. Kollias, P.N. Tsichlis, TNF-␣ induction by LPS is regulated posttranscriptionally via a Tpl2/ERK-dependent pathway, Cell 103 (2000) 1071–1083. [6] J. Bohuslav, W. Kravchenko, G.C.N. Parry, J.H. Erlich, S.N. Gerondakis, N. Mackman, R.J. Ulevitch, Regulation of an essential innate immune response by the p50 subunit of NFB, J. Clin. Invest. 102 (1998) 1645–1652. [7] K. Takeda, B.E. Clausen, T. Tsujimura, N. Terada, I. Forster, Enhanced Th1 activity and development of chronic enterocolitis devoid of STAT3 in macrophages and neutrophils, Immunity 10 (1999) 39–49. [8] A. Matsukawa, M.H. Kaplan, C.M. Hogaboam, N.W. Lukacs, S.L. Kunel, Pivitol role of signal transducer and activator of transcription (STAT4) and STAT6 in the innate immune response during sepsis, J. Exp. Med. 193 (2001) 679–688. [9] N.A. Essani, M.L. Bajt, A. Farhood, S.L. Vonderfecht, H. Jaeschke, Transcriptional activation of vascular cell adhesion molecule-1 gene in vivo and its role in the pathophysiology of neutrophil-induced liver injury in murine endotoxin shock, J. Immunol. 158 (1997) 5941–5948. [10] C.A. Lowell, G. Berton, Resistance to endotoxic shock and reduced neutrophil migration in mice deficient for the Src family kinases Hck and Fgr, Proc. Natl. Acad. Sci. U.S.A. 95 (1998) 7580–7584. [11] T.D. Camenisch, B.H. Koller, H.S. Earp, G.K. Matsushima, A novel receptor tyrosine kinase, Mer, inhibits TNF␣ production and lipopolysaccharide-induced endotoxic shock, J. Immunol. 162 (1999) 3498–3503. [12] P. Ricket, O.D. Weiner, F. Wang, H.R. Bourne, G. Servent, Leukocytes navigate by compass: roles of PI3K␥ and its lipid products, Trends Cell Biol. 10 (2000) 466–474. [13] A.M. Condliffe, K.A. Cadwallader, T.R. Walker, R.C. Rintoul, A.S. Corburn, E.R. Chilvers, Phosphoinositide 3-kinase: Pulmonary cells, Respir. Res. 1 (2000) 24–29.
1533
[14] Z. Li, H. Jiang, W. Xie, Z. Zhang, A.V. Smrcka, D. Wu, Roles of PLC-2 and PLC-3 and PI3K␥ in chemoattractantmediated signal transduction, Science 287 (2000) 1046–1049. [15] H.-K. Yum, J. Arcaroli, J. Kupfer, R. Shenkar, J.M. Penninger, T. Sasaki, K.-Y. Yang, E. Abraham, Involvement of PI3-K in neutrophil activation and the development of acute lung injury, J Immunol. 167 (2001) 6601–6608. [16] T. Nagase, N. Uozumi, S. Ishli, K. Kume, T. Izumi, Y. Ouchi, T. Shimizu, Acute lung injury by sepsis and acid aspiration: a key role for cytosolic phospholipase A2, Nat. Immunol. 1 (2000) 42–46. [17] S.R. Baldwin, R.H. Simon, C.M. Grum, L.H. Ketai, L.A. Boxer, L.J. Devall, Oxidant activity in expired breath of patients with adult respiratory distress syndrome, Lancet 1 (1986) 11–14. [18] Y. Lu, L. Parkyn, L.E. Otterbein, Y. Kureishi, K. Walsh, A. Ray, P. Ray, Activated Akt protects the lung from oxidantinduced injury and delays death of mice, J. Exp. Med. 193 (2001) 545–549. [19] M. Ushio-Fukai, R.W. Alexander, M. Akers, Q.Q. Yin, Y. Fujio, K. Walsh, K.K. Griendling, Reactive oxygen spedes mediate the activation of Akt/protein kinase B by angiotensin II in vascular smooth muscle cells, J. Biol. Chem. 274 (1999) 22699–22704. [20] G.R. Bernard, J.-L. Vincent, R.-F. Laterre, S.P. La Rosa, J.-F. Dhainaiut, A. Lopez-Rodriguez, J.S. Steingrub, G.E. Garber, J.D. Helterbrand, E.W. Ely, C.J. Fischer, Efficacy and safety of recombinant human activated protein C for severe sepsis, New Engl. J. Med. 344 (2001) 699–709. [21] E. Abraham, K. Reihart, P. Svoboda, A. Seibert, D. Olthoff, A. Dal Nogare, R. Postier, G. Hempelmann, T. Butler, E. Martin, C. Zwingelstein, S. Percell, V. Shu, A. Leighton, A.A. Creasey, Assessment of the safety of recombinant tissue factor pathway inhibitor in patients with severe sepsis: a multicenter, randomized, placebo-controlled, single-blind, dose escalation study, Crit. Care Med. 29 (2001) 2081–2089. [22] E. Abraham, Tissue factor inhibition and clinical trial results of tissue factor pathway inhibitor in sepsis, Crit. Care Med. 28 (Suppl. 9) (2000) S31–33. [23] P. Arndt, E. Abraham, Immunological therapy of sepsis: Experimental therapies, Intensive Care Med. 27 (2001) S104–S115.