EVALUATING MYOCARDIAL DEPRESSION IN SEPSIS
2004; Lippincott Williams & Wilkins; Volume: 22; Issue: 1 Linguagem: Inglês
10.1097/01.shk.0000129198.53836.15
ISSN1540-0514
AutoresRichard J. Levy, Clifford S. Deutschman,
Tópico(s)Cardiac Ischemia and Reperfusion
ResumoSepsis is a major cause of mortality in critically ill patients in the United States (1). One hallmark of sepsis is organ system dysfunction. Cardiovascular dysfunction often occurs in patients with sepsis and can present in two ways. After adequate volume resuscitation, patients in hyperdynamic or warm shock are peripherally vasodilated with a high cardiac output. Patients in a hypodynamic state (cold shock) present with increased vascular tone and low cardiac output (2). Adults with sepsis often present in hyperdynamic shock, whereas pediatric patients can present with warm or cold shock (3). Meningococcal sepsis can lead to hypodynamic shock in adults and children (3). Despite an increase in cardiac output during the hyperdynamic phase of sepsis, studies indicate that the myocardium is dysfunctional. Both right and left ventricles can dilate, contractile function may decrease, and ventricular compliance is reduced (2). Work by Parker et al. (4) demonstrated severe depression of ejection fraction in some patients with sepsis despite normal or elevated cardiac index. This dysfunction peaks within a few days of the onset of sepsis and resolves within 7 to 10 days in patients who survive (2). Although myocardial depression in sepsis has been the focus of many investigations, its etiology remains unclear. In an effort to identify potential causes, researchers have studied cardiac function in humans with sepsis and in animal models. The use of clinically relevant animal models is relatively inexpensive, allows for increased sample size and control of confounding variables, and permits variation in study design (5). Unfortunately, different models evoke different hemodynamic responses. For example, the hypodynamic phase occurs between 4 and 7 h after the injection of endotoxin into mice. In contrast, murine hearts enter the hypodynamic phase 20 h after cecal ligation and puncture (CLP) (6, 7). Furthermore, each animal species may respond differently to the same experimental insult. C3H/HeJ mice are hyporesponsive to lipopolysaccharide (LPS), whereas rabbits and Sprague-Dawley rats demonstrate high and low cardiac output derangements in response to varying doses of LPS (5). Small animal models are appealing because they are inexpensive, easy to handle, and can be used in survival studies (5). Mice are particularly useful because investigators can use targeted deletion of specific genes to study pathways of interest. Pigs and nonhuman primates have inflammatory and cardiodynamic responses that are similar to those in humans. However, the latter are expensive and challenging to work with (5, 8). Review of the literature regarding myocardial dysfunction in sepsis is difficult to interpret. Many different models and experimental protocols have been described as representing sepsis and septic insults. This, unfortunately, is misleading and confusing. Some authors describe models that use LPS and endotoxin injection as representative of sepsis. As stated by Wichterman et al. (9), it is important for investigators to recognize that endotoxic shock and sepsis represent different entities. LPS models do not represent models of sepsis, but instead represent models of endotoxicosis. A number of studies have evaluated myocardial dysfunction using models of sepsis, endotoxicosis, and a host of inflammatory insults. Many critics have questioned whether observed cardiac functional abnormalities in experimental sepsis relate to sepsis-associated cardiac dysfunction in humans. Further criticism questions whether these models result in myocardial depression at all. Investigators who evaluate myocardial depression in sepsis face many challenges. Choice of model, identification of crucial time points, and methodology used to assess cardiac function are independent variables that impact on the results of a study. There are limitations with each decision. It is our point of view that many sepsis and sepsis-related models lead to myocardial dysfunction. Evaluation of cardiac dysfunction in experimental sepsis continues to be important independently and in relation to the human scenario. For investigators who endeavor to study sepsis-associated myocardial depression, it is crucial they recognize and navigate the various limitations involved. Thus, the purpose of this review is to highlight the advantages and disadvantages of current approaches used to evaluate myocardial depression in sepsis. Here, we review myocardial contraction, the various methods used to measure and quantify myocardial depression, and we examine the different sepsis and sepsis-related models used to study cardiac function. Finally, we will report the findings of these studies and present the potential etiologies and cellular mechanisms that may lead to cardiac dysfunction in sepsis. MYOCARDIAL CONTRACTION Before discussing the various models and methods used to evaluate myocardial dysfunction in sepsis, it is important to review the physiology and biochemistry of cardiomyocyte contraction. The basic functions of the cardiomyocyte are contraction and relaxation. The contractile unit of each cell, the sarcomere, is composed of thick and thin filaments (Fig. 1). Thick filaments are composed of myosin. Thin filaments are composed of monomers of α-actin intertwined into two long filaments and anchored at the Z-discs.Fig. 1.: The contractile unit. The thick filaments are composed of myosin, whereas the thin filaments are composed of actin. Thin filaments are anchored at the z-discs. The sarcomere is the contractile unit of the cell.Upon depolarization, L-type calcium channels generate a slow inward calcium current (10). The small increase in intracellular calcium leads to activation of the ryanodine receptor, leading to release of a large amount of calcium from the sarcoplasmic reticulum (11). The myosin head binds to actin, hydrolyzing ATP, and undergoes a conformational change. The sarcomeres shorten as thin and thick filaments slide past each other and myosin binds to actin. Tropomyosin, located within the groove between the two actin filaments, and the troponin complex regulate myosin binding (10) (Fig. 2). The troponin complex is made up of troponin I, troponin C, and troponin T (10). Troponin I inhibits actin-myosin binding. Troponin C binds calcium and releases the inhibition caused by troponin I (Fig. 2). Troponin T binds the troponin complex to tropomyosin.Fig. 2.: Tropomyosin and the troponin complex regulate actin-myosin binding. Tropomyosin is represented by the black line within the groove of the actin filaments. ○, Troponin C; ○, Troponin I; and the black oval represents Troponin T.Resequestration of calcium from the cytosol by the sarcoplasmic reticulum ATPase terminates contraction and initiates relaxation. Phospholamban regulates this ATPase and inhibits it when dephosphorylated (10). In addition, the sodium-calcium exchanger can remove a small amount of intracellular calcium (10). Autonomic receptors also regulate calcium influx into cardiomyocytes (11). The stimulated β-adrenergic receptor in association with a G-protein activates adenylate cyclase, leading to conversion of ATP to cAMP. This in turn phosphorylates the L-type calcium channel, leading to increased release of calcium from the sarcoplasmic reticulum (11). Abnormal cardiomyocyte function, termed myocardial depression, is defined by contractile dysfunction, impaired relaxation, or both. These intrinsic cardiomyocyte abnormalities lead to decreased systolic and/or diastolic cardiac performance. MEASUREMENTS OF CARDIAC FUNCTION The heart, as a whole, is a pump that circulates blood from a venous reservoir to the arterial system (12, 13). Pump function depends on preload, afterload, heart rate, and contractility. Preload, the end diastolic volume, leads to elongation of sarcomere and fiber length (14). Afterload, the forces opposing ventricular ejection, leads to wall stress based on chamber size and wall thickness (14). Myocardial contractility, determined by the calcium-contractile protein interaction, cannot be assessed in vivo independent of preload and afterload (14). Thus, measurement of cardiac performance becomes the evaluation of pump function of the entire heart (14). Furthermore, examination of cardiac performance in vivo disregards the internal aspects of cardiac function. These include hormonal and autonomic influences (12). Myocardial contractility, however, can be assessed in isolated myocardium or in isolated cells within controlled environments (14). Myocardial force-length relationships can be measured directly with known levels of preload and afterload (12). Ventricular pressure-volume relationships are the surrogate measure of contractility in the intact heart (12). Preload can be altered by occlusion of the inferior vena cava and afterload can be altered with the use of vasoconstrictors and dilators or by partial aortic occlusion (12). When preload is varied to generate a series of pressure-volume loops, the end-systolic points lie in an almost linear curve (13). The slope of this line, called the end-systolic slope, is used to determine the contractile state (12, 15) (Fig. 3). This slope is very sensitive to changes in contractility (13). An increase in slope represents an increase in contractility.Fig. 3.: End-systolic pressure-volume relationship. Serial pressure-volume loops are depicted. The end-systolic slope is represented by the line connecting the end-systolic points (modified from Ref. 15).Fractional shortening and mean velocity of shortening are two other measures used to assess intrinsic myocardial contractility. Both are independent of afterload, but fractional shortening is preload dependent (12). When both measures are evaluated together, contractility can be assessed accurately (12). Velocity of shortening can be estimated in the intact heart using the first derivative of developed pressure, dp/dt, during isovolumic ventricular contraction (13). The major assumption is that there is no change in chamber size or shape during isovolumic contraction (13). Many investigators have used dp/dt/Pmax because it is independent of afterload (13). Evaluation of myocardial relaxation can also be challenging. During isovolumic relaxation, ventricular volume and wall thickness are assumed to be constant (12). Thus, change in pressure should reflect change in force. Pressure decreases exponentially during diastole (12). The first derivative of the peak rate of pressure decrease, dp/dtmin, and the time constant (tau) are used to determine adequacy of relaxation and diastolic function (12). To evaluate myocardial depression using different models of sepsis, investigators must be able to measure various cardiovascular parameters. These may involve direct assessments or evaluation of parameters that are surrogates for cardiac performance. This can be achieved using a variety of techniques. Invasive monitoring During sepsis in critically ill patients, invasive hemodynamic monitoring via pulmonary artery and arterial catheterization often is used (16). Using these tools, clinicians monitor cardiac output, stroke volume, and mixed venous oxygen saturation directly. In addition, invasive determination of indirect parameters via measurement of intracardiac or intravascular pressures that are believe to correlate with cardiac performance are assessed. To allow correlation with the clinical condition, a majority of animal studies have used invasive monitoring to evaluate cardiac function during sepsis. Pulmonary arterial catheters have been used in various baboon and sheep models to evaluate cardiac output, mixed venous oxygen saturation, or pulmonary artery pressures (17–20). The aorta has also been instrumented in a baboon model of sepsis (19). In canine endotoxemia, invasive cannulae have been placed to measure cardiac output and intracardiac pressures using ultrasonic crystal dimension analyses (21). Small animals have also been instrumented to evaluate cardiac function during sepsis. Yang et al. (7) have cannulated the right jugular vein and carotid artery of mice and rats to determine cardiac output using indocyanine green dilution. Such cannulae also allow measurement of pressure changes during ventricular systole and diastole to estimate cardiac performance (22). Tao et al. (15) placed a conductance catheter directly into the left ventricle of mice and determined end systolic slope and changes in maximum and minimum dp/dt. Echocardiography Invasive monitoring can be technically difficult and requires exposure to additional anesthesia. Thus, noninvasive techniques for measuring cardiac function are appealing. Echocardiography is an inexpensive and portable technique (23) that uses sound waves to image structures in vivo (24). Using two-dimensional and M-mode echocardiography, cardiac output and shortening fraction can be estimated (24). Shortening fraction is an index of contractility and is calculated by the equation: where EDD is end diastolic dimension and ESD is end systolic dimension (25). Stroke volume is estimated by the following equation (26): Cardiac output can be calculated by multiplying stroke volume by heart rate. A variety of studies using 12- to 15-mHz transducers have successfully estimated cardiac function in different mouse models of sepsis (6, 25, 26). However, subjective interpretation of acquired images is a major limitation of echocardiographic determination of cardiac performance (23). Magnetic resonance imaging (MRI) MRI creates spatial images from radio wave frequencies generated by different tissues within a magnetic field (27). Because of its dimensional accuracy and high resolution, MRI can precisely and noninvasively measure cardiac output, stroke volume, and ejection fraction (27). MRI has been used in rat and mouse models to assess cardiac function during healthy and pathologic states (23, 28). We have recently reported decreased cardiac output in mice at 48 h after CLP using electrocardiogram-gated MRI (29). In vitro, whole heart evaluation The isolated, perfused heart is a reproducible and inexpensive preparation. Originally described by Langendorff, the isolated heart preparation is denervated and removed from the influence of circulating neurohormonal factors (30). Although viewed as a disadvantage by many authors, isolation of the heart allows assessment of the effect of specific agents on cardiac function without the confounding involvement of other organ systems (30). Most commonly, cardiac function in rats has been investigated using this technique (30). In the Langendorff heart preparation, retrograde oxygenated perfusate is delivered via the aorta to the coronary arteries (30). An intraventricular balloon is inserted through the left atrium into the left ventricle and inflated (30). Once inflated in position, left ventricular systolic and diastolic pressures can be measured (30). The working heart preparation involves an initial Langendorff preparation, however, an oxygenated perfusate is pumped antegrade into the left atrium (30). The advantage of this preparation is that it is possible to study cardiac function while varying the preload and afterload (30). A major limitation is that coronary perfusion must match the metabolic needs of the heart (30). After LPS injection, rat and mouse hearts have been isolated and perfused to evaluate myocardial dysfunction (25, 31, 32). Also, mouse Langendorff preparations have been used to observe the effect of sepsis-related circulating factors on cardiac dysfunction (33). In vitro, isolated cell system evaluation Another method to evaluate myocardial dysfunction examines the contractile behavior of isolated cardiomyocytes. Left atrial myocyte preparations from endotoxic guinea pigs demonstrated depressed contractile tension and reduced +dP/dtmax by 2 h postendotoxin (34). Contractility returned toward baseline after 24 h (34). Similar findings were seen in guinea pig left ventricular papillary muscles isolated 16 h postendotoxin (34). In other work, rabbit cardiomyocytes were incubated with LPS and depressed cell shortening was observed (35). In other studies, skinned myocardial fibers isolated from endotoxemic rats generated decreased tension and were unresponsive to inotropes (36, 37). MODELS Lipopolysaccharide (LPS) LPS is a glycolipid molecule located within the cell wall of gram-negative bacteria. Endotoxin, although sometimes used interchangeably with LPS, contains a combination of cell wall proteins, lipids, and polysaccharides, including LPS. Because LPS has been isolated from serum of patients with sepsis, it is believed to contribute to or be causative of the sepsis syndrome by some authors (5). In fact, when injected into humans at low dosage, LPS leads to increased cardiac output and peripheral vasodilation as seen in sepsis (5). Injection of LPS and endotoxin into various animals leads to hemodynamic compromise and endotoxin shock (9). LPS can be administered intravenously or into the intraperitoneal space. Intravenously, it can be given as a single bolus or by continuous intravenous infusion. Small doses of LPS usually lead to a hyperdynamic response, whereas larger doses can be highly lethal and cause a hypodynamic state. Aggressive volume resuscitation after larger doses of LPS can prevent the development of the hypodynamic state. Intraperitoneal injection of 5 or 25 mg/kg of LPS into mice leads to significantly decreased shortening fraction determined by echography and significantly depressed dp/dtmax 6 h after injection (38, 25). Diastolic compliance determined by –dp/dtmax is also significantly decreased. (25). At 6 h after an intraperitoneal injection of 4 mg/kg of LPS into Sprague-Dawley rats, cardiac work and efficiency are decreased in a working heart Langendorff preparation (39). Chronic low-dose infusion of 5 μg/kg/h of LPS intravenously in canines initially produces an increase in cardiac output 1 h after initiation of infusion with a return to baseline by 4 h and maximal decrease in function at 24 h (21). The cardiac depression in this model is characterized by ventricular dilatation and decreased ventricular compliance as assessed by invasive pressure measurement and ultrasonic crystal dimension analysis (21). After a 5 mg/kg intravenous injection of LPS in rabbits, decreased cardiac output and an increased in vascular tone are observed (5). However, after a 1 mg/kg intravenous LPS bolus in rabbits, a hyperdynamic alteration is seen (5). Furthermore, after 10 mg/kg of intravenous endotoxin, rats demonstrate reduced dp/dtmax and –dp/dtmax by 4 h postinjection using a Langendorff preparation (40). Thus, LPS and endotoxin injection models lead to fulminant cardiac dysfunction rapidly after administration in a variety of animal species. An advantage of this model is the stability of LPS compared with infused bacteria (41). LPS is easily stored and accurate doses can be administered. However, a major disadvantage is that LPS models lack a focus of infection and thus do not represent clinical sepsis. Bacterial infusion When designing a model to mimic human sepsis syndrome, many researchers critique the practice of delivering a pathogen bolus as a single load (5). Instead, a continuous infusion of bacteria is believed to be more representative of the human disease. After a 2-h intravenous infusion of Escherichia coli, baboons demonstrate a hyperdynamic cardiovascular response as determined by pulmonary artery catheterization and thermodilution determined cardiac output (17). Intravenous infusion of E. coli into dogs leads to an immediate hyperdynamic response with increased cardiac output and vasodilation as determined by thermodilution and invasive pressure measurement (18). This response occurs without changes in ventricular contractility or compliance as assessed by end-systolic slope and –dp/dtmax (18). After a 4-h intravenous E. coli infusion in pigs, there is a decrease in myocardial contraction measured in isolated right ventricular trabeculae (42). In a sheep bacteremia model, decreased cardiac output with vasodilation are evident within 30 min of infusing E. coli (43). Another hyperdynamic model uses infused Pseudomonas aeruginosa in sheep and a newborn model infused group B Streptococci into newborn pigs (44, 45). These models are fulminant and cause variable hemodynamic responses. There are several limitations with this type of model. First, large doses of bacteria are required to overcome host defenses (41). Second, intravenous infusion of bacteria does not represent sepsis because most patients with sepsis have an infectious source from which pathogens originate (41). Finally, serum cytokine responses increase transiently and may not represent the clinical response seen in patients (41). Peritoneal inoculation In an effort to simulate the clinical manifestations of human gram-negative peritonitis, researchers have implanted bacteria-laden fibrin clots into the peritoneal cavities of research animals. Advantages of this model include administration of a known quantity of bacteria and delayed morbidity and mortality due to fibrin impedance of early systemic absorption (5). Implantation of an E. coli-laden fibrin clot containing 1.9-6.7 × 1011 CFU/kg into baboon peritoneum causes a hyperdynamic cardiovascular disturbance (19). Natanson and colleagues (46) examined the cardiovascular effect of implanting E. coli-infected clots into the peritoneum of conscious canines. They reported that systolic and diastolic performance were severely reduced leading to ventricular dilation, rightward shift of the diastolic volume-pressure relationship, and maintenance of stroke volume in survivors (47). Furthermore, they noted that nonsurvivors had smaller ventricle size, less compliant ventricles, and reduced stroke volume (47). Compared with LPS intravenous injection of 5.6 mg/kg in Sprague-Dawley rats where cardiac output decreases dramatically and SVR increases, peritonitis with E. coli-laden fibrin clot leads to an increase in cardiac output with peripheral vasodilation (48). Mathiak and colleagues (48) state that peritoneal inoculation better mimics human sepsis and is less fulminant than LPS. In this model, a focus of infection leads to bacteremia and a cytokine response similar to patients with sepsis. This is a clear advantage over LPS and bacterial infusion models. Furthermore, it is simplistic and specific quantities of bacteria can be used. CLP The CLP model of sepsis was developed by Wichterman, Baue, and Chaudry in the 1980s (9). Historically, a number of species, including mice, rats, and sheep have been studied using this model (7, 49, 50). Because CLP leads to polymicrobial peritonitis caused by leakage of fecal material through ischemic bowel, it is difficult to control for the amount of bacterial contamination. However, standardizing the size of the needle used for puncture and the length of cecum ligated leads to predictable mortality (51). An obvious advantage of this model is that it mimics the human scenario of a perforated intra-abdominal viscus. After CLP, survival curves of 6- to 16-week-old mice correlate with those of 10- to 17-year-old humans with sepsis (52). Aged mice have worse survival than younger mice after CLP (52). Furthermore, antibiotic administration in younger mice leads to improved survival but is not beneficial in older mice (52). An early hyperdynamic phase with increased cardiac output and stroke volume occurs within 2 to 10 h in mice and rats (7, 53). In these studies, the authors assessed hemodynamics using radioactive microspheres and indocyanine dye (7, 53). At 24 h post-CLP, there is progression to a hypodynamic phase with decreased cardiac output and stroke volume and vasoconstriction (7). Eighteen hours after single, 20-gauge puncture CLP in female mice, end-systolic slope and dp/dtmax and –dp/dtmax are decreased (22). Twelve hours after sepsis is induced in mice via CLP (followed by fluid resuscitation and antibiotic administration), cardiac output increases and the vasculature dilates. The hypodynamic response is delayed beyond 48 h (26). Thus, CLP, as a model of sepsis, replicates the human state. Using fluid resuscitation and antibiotics in this model alters the cardiovascular response. The CLP model has many advantages. Like peritoneal inoculation, it is a simple procedure and leads to cytokine responses similar to septic patients. Unlike peritoneal inoculation, there is no need to quantify the amount of bacteria. Furthermore, peritoneal contamination after CLP leads to polymicrobial peritonitis. Because CLP resembles a common human cause of sepsis, its use as a model of sepsis is appealing. Cell culture Work with animal models can be technically challenging and is subject to interanimal variability. Thus, many investigators have chosen to study tissue and cell cultures (54, 55). Cell culture provides for purity of cell type, the ability to evaluate short- and long-term incubation strategies, and the opportunity to investigate the effect of exposure to specific agents (55). A variety of different cell lines have been used in sepsis research. To simulate the inflammatory response, cell cultures are incubated with a host of inflammatory mediators. These inciting agents include cytokines such as TNF-α, IL-1β, IFN-γ, and IL-6, as well as nitric oxide (NO) donors and LPS (56–62). When cultured rat cardiomyocytes are incubated with ultrafiltrate from patients with sepsis, a decrease in stimulated contraction frequency is seen (63). Furthermore, reduced amplitude and speed of contraction are observed when rat cardiac myocytes are exposed to the supernatant of serum from children with meningococcal sepsis (64). Studying cultured cardiomyocytes exposed to various mediators of sepsis is useful. An obvious limitation is that these experiments are performed ex vivo. ETIOLOGY A number of mediators and pathways have been shown to be associated with myocardial depression in sepsis, but a unifying cause has yet to be found. Although not intended to be all inclusive, the potential etiologies reviewed in this section are the most commonly cited and reported contributors to myocardial dysfunction in sepsis and sepsis-related models. An early theory of cardiac depression in sepsis involved decreased oxygen delivery to the heart. This was disproved with two human studies evaluating coronary hemodynamics. These demonstrated that global cardiac perfusion in sepsis is normal or increased (65, 66). Coronary blood flow also increased during sepsis after intraperitoneal fecal inoculum in rats (67). High-energy phosphates Another hypothesis contends that sepsis leads to cytopathic hypoxia. This means that an uncoupling of oxidative phosphorylation disrupts high-energy phosphate production (68). The literature is confusing regarding ATP availability, however. Many studies have demonstrated preserved ATP levels in dysfunctional septic myocardium (69–71). On the other hand, decreased ATP levels have been reported in cardiomyocytes after CLP, administration of endotoxin, or various manipulations in cell culture (49, 60, 72, 73). Although important, measurement of ATP levels does not necessarily provide information about ATP production and the integrity of oxidative phosphorylation. In fact, when cardiomyocytes are exposed to hypoxic conditions, down-regulated contractile activity and oxygen consumption maintain ATP levels (74). This process is called myocardial hibernation (74). Because of this phenomenon, investigators have begun to examine the kinetic activities of electron transport chain complexes in septic myocardium. For instance, a 50% reduction in complex I and III activity is seen in rabbit heart after subcutaneous LPS injection (75). Complexes I through IV demonstrate decreased activity by 18 h in rat myocardium after CLP (49). We have shown that myocardial cytochrome oxidase (complex IV) is competitively inhibited in the early stages of CLP in mice (76). This inhibition becomes irreversibly inhibited during the hypodynamic phase (76). Myocardial depressant substance (MDS) Another potential etiology of cardiac dysfunction in sepsis is circulating MDS or factor. In support of this theory, ultrafiltrate from patients with sepsis decreases contraction frequency in rat cardiomyocytes (63). A heat stable, proteinaceous, 10- to 25-kD MDS has been identified in patients with meningococcal sepsis (64). A number of different proteins, including cytokines such as TNF-α (17 kD), have been identified as potential circulating MDS (2). TNF-α infusion decreases canine left ventricular contractility by 23% after 1 h and by 52% after 5 h (77). Another protein, lysozyme, has recently been identified as an MDS in a canine E. coli bacteremia model (78). Others have suggested that an MDS originates from the gut during critical illness (79). Cellular mechanisms Although the etiology of cardiac depression during sepsis is probably multifactorial, identifying individual contributors is important. Thus, a host of pathways and disruptions in cellular homeostasis have been examined in septic myocardium. Cytokines LPS as well as TNF-α and IL-1β directly depress cardiac contractility (38, 80). TNF-α and IL-1β produced by cardiomyocytes also may lead to cardiac dysfunction (38). LPS first binds to CD-14, a glycoprotein expressed on the surface of cardiac myocytes. CD-14 is closely related to the Toll-like receptor-4 (TLR4) and the glycoprotein, MD-2 (38) (Fig. 4). The intracellular pathway linked to CD-14/TLR4 is shared with the type I IL-1 receptor pathway (80). The adapter molecule, MyD88 recruits IL-1 receptor associated kinase-1 (IRAK 1) and IRAK 2 (81). IRAK 1 and IRAK 2 interact with TNF receptor associated factor 6, bridging them to transforming growth factor-β-activated kinase and nuclear factor (NF) κB-inducing kinase (NIK) (80, 81). NIK activates the Inhibitory κB (IκB) kinas
Referência(s)