Tumor Necrosis Factor-α and Tumor Necrosis Factor Receptors in the Failing Human Heart
Background Tumor necrosis factor-α (TNF-α) is a proinflammatory cytokine that produces negative inotropic effects in the heart. Recently, elevated levels of TNF-α have been reported in patients with advanced congestive heart failure. Although TNF-α is thought to exert its deleterious effects by binding to two cell surface receptors, TNFR1 and TNFR2, the level of expression and regulation of TNF receptors in the heart in cardiac disease states is not known.
Methods and Results We examined mRNA and protein levels for TNFR1, TNFR2, and TNF-α in explanted hearts from organ donors as well as in patients with end-stage dilated cardiomyopathy (DCM) and ischemic heart disease (IHD). Northern blot analysis revealed that mRNA for TNFR1 and TNFR2 was present in nonfailing, DCM, and IHD hearts. TNFR1 and TNFR2 receptor protein levels, as measured by ELISA, were decreased 60% in DCM and IHD patients compared with nonfailing hearts (P<.005). To determine a potential mechanism for the decrease in TNF receptor expression, we measured levels of circulating soluble TNF receptors (sTNFRs) in DCM and IHD patients. This analysis showed that there was a significant one-and-a-half to threefold increase in sTNFRs in DCM (P<.03) and IHD patients (P<.001). Another important finding was that TNF-α mRNA and TNF-α protein were present in the explanted hearts from DCM and IHD patients but not in nonfailing hearts.
Conclusions In summary, the results of this study constitute the initial demonstration that TNF receptor proteins are dynamically regulated in patients with advanced congestive heart failure. Moreover, the observation that failing hearts express elevated levels of TNF-α suggests that overexpression of this cytokine may be one of several different maladaptive mechanisms responsible for the progressive cardiac decompensation that occurs in advanced heart failure.
Tumor necrosis factor-α (TNF-α) is a proinflammatory cytokine that produces left ventricular dysfunction, cardiomyopathy, and pulmonary edema when overexpressed in human subjects.1 2 3 Recent observations4 5 6 7 8 showed that levels of circulating TNF-α are elevated in patients with advanced congestive heart failure. Although the exact clinical significance of this finding is uncertain, given that TNF-α produces negative inotropic effects in cardiac tissue9 10 as well as in cardiac muscle cells,10 11 it has been postulated that the elaboration of this proinflammatory cytokine may contribute to the progressive cardiac decompensation that occurs in advanced congestive heart failure.12
Important to the above hypothesis regarding the pathophysiological role of TNF-α in heart failure is the elucidation of the mechanism(s) by which this proinflammatory cytokine exerts its effects in the heart. Although recent experimental studies from a number of laboratories have begun to explore the basic mechanisms of action of TNF-α at the tissue and cellular levels,9 10 11 13 14 until recently nothing was known regarding the presence or absence of TNF receptors in the adult heart. To this end, we have identified the presence of type 1 (TNFR1) and type 2 (TNFR2) TNF receptors in adult human cardiac myocytes.15 Moreover, we have shown that in isolated contracting cardiac myocytes, the negative inotropic effects are mediated by TNFR1.15 However, our initial studies, which were performed in nonfailing myocardium, may not reflect the level of TNF receptor expression in cardiac disease states such as congestive heart failure.
Given that the levels of TNFR1 and TNFR2 are regulated in a dynamic fashion16 and can be upregulated and/or downregulated17 18 in different pathophysiological conditions, we sought to determine the level of TNF receptor expression in patients with end-stage DCM and end-stage IHD. In addition, recent studies19 demonstrated that under certain forms of stress, the heart is capable of synthesizing biologically active TNF-α. Accordingly, we also sought to determine whether there was evidence for TNF-α biosynthesis in the failing human heart. The results of the current study constitute the initial demonstration that TNF receptor proteins are dynamically regulated in heart disease and are downregulated in patients with advanced congestive heart failure. Moreover, the present study shows that TNF-α mRNA and protein are present in the failing human heart, whereas there was no evidence for TNF-α mRNA or protein biosynthesis in the nonfailing human heart.
Source of Human Myocardium
Explanted hearts from patients with class IV (New York Heart Association) heart failure who underwent cardiac transplantation at Methodist Hospital, Houston, Tex (n=25); the University of Colorado, Denver (n=3); or Temple University Hospital (n=2), Philadelphia, Pa, between June 1992 and June 1994 served as the source for the myocardial samples reported herein. The patient cohort comprised 16 patients with end-stage IHD and 14 patients with idiopathic DCM. The diagnosis of DCM was established on the basis of the finding of a normal coronary angiogram, the absence of valvular or pericardial heart disease by 2D echocardiography, and the absence of a clinical history that would suggest myocarditis. The clinical diagnosis of ischemic cardiomyopathy (IHD) was established on the basis of a history of documented prior myocardial infarction(s) and known coronary heart disease documented by coronary arteriography.
The clinical characteristics of five of the seven nonfailing myocardial samples used herein have been reported previously15 (Table⇓); these samples were obtained from organ donors whose hearts initially were considered for cardiac transplantation but subsequently were deemed unsuitable for transplantation either because of blood type or size incompatibility. The hearts of the remaining two female organ donors, aged 31 and 42 years, were also deemed unsuitable for transplantation because of blood type or size incompatibility. The cause of death for these two individuals was a cerebrovascular accident. All donors had normal LV function as documented by 2D echocardiography; moreover, there was no history of primary myocardial disease or evidence of active infection or malignancy at the time of explantation.
All explanted hearts were handled in an identical manner. Immediately after explantation, the LV was rapidly separated from the atrial tissue, and the remaining fat and fibrous tissues were excised; the myocardial samples were then cut into 5×5-cm portions, quickly frozen in liquid nitrogen, and stored at −80°C until the time of analysis.
Demographic, Hemodynamic, and Biochemical Characteristics
Demographic data for the cardiomyopathic patients were obtained from the databases at the Methodist Hospital/Baylor College of Medicine Heart Transplant Center, the University of Colorado, or the Temple University Hospital transplant programs. Hemodynamic data were obtained from a right heart catheterization performed no less than 1 week before explantation with procedures described elsewhere.20
Biochemical data for the cardiomyopathic patients were obtained from serological specimens drawn on the day of transplantation, immediately before explantation of the heart. Plasma samples were available for all 16 of the IHD patients and for 7 of the 14 DCM patients. Frozen (−70°C) aliquots of these plasma samples were used for the determination of circulating levels of TNF-α and sTNFR1 and sTNFR2. Measurements were performed with commercially available kits exactly according to the manufacturer’s suggestions (R&D Systems, Inc). To test for the presence of circulating cytokine inhibitors,21 we performed “spiking experiments” in preliminary control experiments using concentrations of recombinant TNF-α, sTNFR1, and sTNFR2 found in patients with advanced heart failure. These control studies showed that the levels of added recombinant proteins were not quenched by any inhibitory factors in the sera from heart failure patients. To test for the presence of heterophile antibodies,21 we performed serial dilutions (1:2, 1:4, 1:10) on the sera obtained from heart failure patients. These control studies showed that the levels of TNF-α, sTNFR1, and sTNFR2 immunoreactivity declined in a manner parallel to the standard curve.21
Demographic data for the normal subjects were obtained from the review of relevant clinical material. Frozen plasma samples for these patients were not available; therefore pre-explantation measurements of TNF-α, sTNFR1, and sTNFR2 were not performed for these subjects. Accordingly, “normal” levels of TNF-α, TNFR1, and TNFR2 were obtained from age-matched normal volunteers (53.8±8 years old; n=32) who were free of cardiovascular disease or other comorbid conditions.
Myocardial TNF Receptors
Myocardial TNF Receptor Gene Expression
Total RNA was extracted from normal, DCM, and IHD hearts by use of the guanidinium thiocyanate method.22 Total RNA was denatured at 65°C for 10 minutes and size-fractionated on a 1% agarose gel (10 μg/lane) containing 2.2 mol/L formaldehyde. The total RNA samples were then transferred onto a nylon membrane (GeneScreen, Du Pont–New England Nuclear) and hybridized sequentially to random primed cDNA probes.22 The following probes were used for Northern blot analysis: a 1.0-kb EcoR1 fragment of human TNFR1 (a gift from C. Smith, Immunex, Seattle, Wash); a 0.64-kb Not I/Bgl I fragment of human TNFR2 (a gift from C. Smith, Immunex, Seattle, Wash); and a 0.5-kb Xba/HindIII fragment of human GAPDH, which was used as an internal control. The membranes were washed once with standard saline citrate and 0.1% SDS at 55°C for 30 minutes, air dried, and exposed to Kodak X-Omat A film at −70°C.
Myocardial TNF Receptor Proteins
Myocardial TNF receptor proteins were analyzed exactly as described previously.15 Briefly, 0.5- to 1.0-g frozen sections of myocardium were pulverized and suspended in a PBS solution containing protease inhibitors (26.0 mg/dL phenylmethylsulfonyl fluoride, 0.1 mg/dL leupeptin, and 0.2 mg/dL aprotinin) and homogenized for 30 to 60 seconds with a polytron PT-3000 (Brinkmann Instruments, Inc). Myocardial homogenates were centrifuged for 20 minutes at 4°C at 20 000g, and the resultant cell pellet was solubilized according to the method of Stauber et al,23 as described previously.15 The supernatant, which contained the solubilized cell membrane–bound TNF receptor,23 was used for analysis of TNFR1 and TNFR2. The protein content of the membrane fraction was determined by use of a commercially available assay (BCA, Pierce Chemical Co) with bovine serum albumin used as a standard.
Levels of TNFR1 and TNFR2 in the membrane fraction were measured by a “sandwich” enzyme immunoassay (ELISA) with commercially available kits for the detection of human TNFR1 and TNFR2 (Quantikine, R&D Systems, Inc). The antibodies used in these immunoassays have been characterized extensively for specificity by the suppliers, are not influenced adversely by the presence of TNF-α or lymphotoxin-α (formerly called TNF-β), and have a lower limit of detection (0.78 ng/dL) for both receptors. Details of this methodology have been described recently in considerable detail.15 Results were expressed as picograms of TNF receptor per gram of membrane protein and represent the mean value of two separate measurements performed in duplicate.
Although the source of TNF-α production in congestive heart failure is not known, previous experimental studies19 24 25 26 showed that under normal conditions, neither TNF-α mRNA nor TNF-α protein is expressed within the heart; however, after certain forms of stress, the heart synthesizes TNF-α mRNA and protein de novo. Accordingly, to determine whether there was TNF-α biosynthesis in the adult human heart, we examined nonfailing and failing myocardial samples for evidence of TNF-α mRNA and protein production.
Myocardial TNF-α Gene Expression
Total RNA from nonfailing, DCM, and IHD hearts was extracted and processed exactly as described above. The total RNA samples were hybridized to a random primed 0.6-kb HindIII/HindIII fragment of human TNF-α (American Tissue Culture Collection) and exposed to Kodak X-Omat A film at −70°C for up to 2 weeks; GAPDH was used as an internal control.
Myocardial TNF-α Protein
To demonstrate the presence or absence of TNF-α protein in nonfailing and failing hearts, two separate studies were performed. First, to determine whether there were measurable levels of TNF-α in nonfailing or failing hearts, we determined intracardiac levels of TNF-α by ELISA, using a modification of the method of Pizarro et al.27 Briefly, frozen sections of nonfailing and failing myocardium were prepared in the same manner as described above for soluble membrane receptor proteins. On the basis of preliminary control experiments that showed that most immunodetectable TNF-α resided in the cytosolic fraction of the whole heart homogenates, we used a commercially available ELISA kit (R&D Systems, Inc) to determine the presence or absence of immunodetectable TNF-α in the cytosolic fractions of the whole heart homogenates. The antibodies employed in this “sandwich” ELISA are not influenced adversely by the levels of TNFR1 and TNFR228 and have a lower limit of detectability of 0.5 ng/dL.
Briefly, a 200-μL sample of equivalent amounts of cytosolic protein was added to assay buffer with the appropriate human TNF-α standards, and the immunoassay was performed according to the manufacturer’s suggestions. Results were analyzed spectrophotometrically at a wavelength of 490 nm with a microtiter plate reader. Final results were expressed as picograms of TNF-α per gram of cytosolic protein. Second, to visualize the anatomic localization of TNF-α within the myocardium, immunohistochemical studies were performed. Frozen tissue was embedded in OCT compound (Miles Inc), sectioned (0.5-μm sections), placed on a probe on plus slides (Allen Fisher & Assoc), and stored at −70°C until immunostaining. Sections were brought to room temperature and fixed with paraformaldehyde (0.04 U) for 20 minutes. Sections were then washed with PBS for 15 minutes. After blocking endogenous peroxidase activity with methanol and hydrogen peroxide (0.003 U), the sections were placed in cold acetone (−20°C) for 5 minutes and washed again for 15 minutes with PBS. To minimize background staining, all sections were first blocked for 30 minutes at room temperature with goat serum. Next, the slides were incubated with a 1:1000 dilution of a primary antibody directed against human TNF-α (Genzyme Corp); control slides were stained with secondary antibody alone (1:1000 dilution). Sections were allowed to incubate in a humidified chamber overnight at 4°C. The slides were then rinsed in PBS five times and incubated for 30 minutes at room temperature with a biotin-conjugated goat anti-rabbit secondary antibody (diluted 1:1000). The slides were stained with an avidin-biotin complex by use of a peroxidase reporter (Vectastain ABC Kit, Vector Labs, Inc). Diaminobenzidine was used as the chromogen to visualize the presence and distribution of TNF-α. Sections were then rinsed in PBS, counterstained in hematoxylin, dehydrated, cleared, and mounted with a synthetic mounting medium. All sections were examined at ×600 magnification.
Values are expressed as mean±SE. Unpaired t tests were used to evaluate mean differences in demographic and biochemical data in the cardiomyopathic patients. One-way ANOVA was used to test for mean differences in the circulating and myocardial levels of TNF-α, TNFR1, and TNFR2. Post-ANOVA comparisons were performed between control and experimental groups by Dunnett’s test or between experimental groups by Newman-Keuls test. A χ2 analysis was performed to test for differences in the proportion of patients with and without elevated levels of intracardiac TNF-α. Regression analysis was used to test for a correlation between circulating and intramyocardial levels of TNF-α. A value of P<.05 indicated that significant differences existed.
Demographic, Hemodynamic, and Biochemical Characteristics of Patient Cohort
Three important characteristics of the DCM (n=14) and IHD (n=16) patients are summarized in the Table⇑. First, as shown, there was no significant difference (P>.05) in cardiac index, mean arterial pressure, pulmonary capillary wedge pressure, heart rate, hemoglobin, hematocrit, serum sodium, blood urea nitrogen, or creatinine between DCM and IHD patients. However, age and ejection fraction were each significantly lower for DCM patients compared with IHD patients. Second, although the circulating levels of TNF-α in IHD and DCM patients were not significantly different (P=.99), the circulating levels of TNF-α in the IHD and DCM patients were approximately 10-fold higher than the respective values obtained in control subjects (P<.001 for both). Third, the plasma levels of sTNFR1 and sTNFR2, which represent the “shed” extracellular domains of the membrane-bound TNFR1 and TNFR2,29 30 31 were not significantly different (P<.4 for both) between DCM and IHD patients. However, when the circulating levels of sTNFR1 and sTNFR2 in DCM and IHD patients were compared with circulating levels reported in subjects without heart failure, the levels of sTNFR1 and sTNFR2 were approximately 1.4× to 3× greater (P≤.03).
Mean age for the organ donors without heart failure was 33.9±5.5 years (n=7), which was not significantly different from the age of the DCM patients (P=.32) but which was significantly (P=.007) less than the IHD patients. Although complete hemodynamic data were not available for the normal subjects included in the present study, all subjects were class I (New York Heart Association). 2D echocardiograms obtained before explantation revealed a normal LV ejection fraction for each of the patients examined.
Myocardial TNF Receptors
TNF Receptor Gene Expression
Fig 1⇓ shows a representative Northern blot of total RNA from nonfailing, IHD, and DCM hearts, which used cDNA probes for human TNFR1, TNFR2, and GAPDH. The important finding shown by this figure is that mRNA for both TNFR1 and TNFR2 was detected in nonfailing myocardium as well as in myocardium from IHD and DCM patients. Furthermore, mRNA for TNFR1 and TNFR2 was detected in all of the nonfailing and cardiomyopathic specimens tested. As shown, the intensity of the autoradiographic bands for TNFR1 in nonfailing and failing hearts was greater than that of the respective autoradiographic bands for TNFR2 in nonfailing and failing hearts.
TNF Receptor Protein
Fig 2⇓ summarizes the studies wherein membrane TNF receptor protein levels were quantified by ELISA in nonfailing and failing hearts. The salient finding shown by Fig 2A⇓ is that the level of total TNF receptor protein (TNFR1+TNFR2) was approximately 60% to 70% less in the DCM and IHD hearts compared with total TNF receptor protein levels in nonfailing myocardium. Fig 2⇓ (B and C) shows that when values in failing hearts were compared with values in nonfailing myocardium, there was approximately a 55% to 65% decrease in the expression of TNFR1 and TNFR2 proteins, respectively, in DCM and IHD hearts. ANOVA indicated that there were significant (P≤.03) overall differences in total TNF receptor (Fig 2A⇓), TNFR1 (Fig 2B⇓), and TNFR2 (Fig 2C⇓) levels between nonfailing, DCM, and IHD hearts; ANOVA testing (Dunnett’s test) indicated that the levels of total TNF receptors, TNFR1, and TNFR2 proteins were significantly less (P<.05 for each) in the DCM and IHD hearts compared with the respective values in nonfailing hearts, whereas the mean differences in total TNF receptors, TNFR1, or TNFR2 between DCM and IHD hearts were not different statistically (P>.05 by Newman-Keuls test). Indeed, Fig 2D⇓ shows that the relative distribution of TNF receptor subtypes was similar in nonfailing and failing myocardium; that is, TNFR1 comprised 51.2±4.2%, 44.8±4.4%, and 50.6±1.9%, respectively, of the total TNF receptor protein in nonfailing, DCM, and IHD hearts (P=.37 by ANOVA).
Fig 3⇓ illustrates two salient findings with respect to the presence or absence of TNF-α biosynthesis in nonfailing and failing hearts. First, Northern blot analysis showed that there was no evidence for TNF-α mRNA expression in the nonfailing hearts (n=3), whereas TNF-α mRNA was detectable in myocardium obtained from patients with IHD (n=3) and DCM (n=3). Second, the immunohistochemical studies showed that TNF-α immunostaining was not detected in nonfailing myocardium (Fig 3A⇓), whereas there was obvious TNF-α immunostaining of the cardiac myocytes in the myocardium obtained from IHD (Fig 3C⇓) and DCM (Fig 3E⇓) patients. Fig 3⇓ (B, D, and F) shows that there was no nonspecific myocardial immunostaining with secondary antibody alone in the nonfailing, IHD, or DCM hearts, respectively.
Fig 4⇓ shows two important findings with respect to the ELISA analysis of intracardiac TNF-α in nonfailing and failing hearts. First, immunodetectable TNF-α was not present (lower limit of detection <50 pg/g cytosolic protein) in nonfailing hearts (n=7), consistent with previous reports19 25 that showed that neither TNF-α mRNA nor TNF-α protein is elaborated in the heart under nonfailing conditions. In contrast, immunodetectable levels of TNF-α were present in the cytosolic fractions of 55% of the explanted hearts obtained from DCM patients and 40% of the explanted hearts from IHD patients. Although the proportion of patients with elevated TNF-α levels was significantly greater in the DCM (P<.02) and IHD (P<.04) groups compared with nonfailing subjects, the proportion of patients with elevated TNF-α levels was not significantly different (P=.86) between DCM and IHD patients. Moreover, there was no significant difference (P=.07) between the level of intracardiac TNF-α in the DCM (218.5±64 pg/g cytosolic protein) and IHD patients (137.7±41 pg/g cytosolic protein). To determine whether there was a concordance between TNF-α gene and protein expression in the failing hearts, we compared the presence or absence of TNF-α mRNA and immunodetectable TNF-α (ie, >50 pg/g cytosolic protein) in the failing hearts. This analysis showed that whenever TNF-α mRNA was evident by Northern analysis, immunodetectable TNF-α protein was present in the cytosolic fractions of the failing ventricles (n=6 hearts). However, we were unable to detect TNF-α mRNA by Northern blot analysis in 12 hearts in which there was immunodetectable TNF-α protein in the cytosolic extracts of the failing hearts.
To determine whether there was a significant relationship between the intracardiac level of TNF-α at the time of explantation and the plasma level of TNF-α that was drawn several hours before cardiac explantation, we performed a linear regression analysis. As shown in Fig 5⇓, there was no significant correlation between the levels of TNF-α in the heart and in the plasma (slope P=.65; r=.10 [P=.35]). Thus, the presence of an elevated level of TNF-α in the plasma level did not necessarily predict which patients would have elevated TNF-α in their myocardium. However, as shown in Fig 5⇓, the converse of this statement did not hold true. That is, except for one IHD patient, all of the patients with immunodetectable levels of intracardiac TNF-α had plasma TNF-α levels that were beyond the values reported for nonfailing subjects.
To determine whether the observed decrease in TNF receptor proteins in failing hearts may have been causally related to the presence of elevated levels of TNF-α in the myocardium (homologous desensitization), we compared the amount of total TNF receptor proteins (TNFR1+TNFR2) in the DCM and IHD hearts with and without immunodetectable intracardiac TNF-α. This analysis showed that there was no significant difference (P=.35) in the levels of total TNF receptor proteins in the cardiomyopathic hearts with (102.9±8.7 ng/g protein) or without (80.6±9.8 ng/g protein) immunodetectable TNF-α.
One of the major conclusions to be drawn from this study is that the level of expression of TNF receptor proteins is decreased in patients with advanced heart failure. Two lines of evidence support this statement. First, levels of membrane-bound TNF receptor proteins were approximately 60% to 70% less in hearts from DCM and IHD patients compared with levels in nonfailing myocardium (Fig 2A⇑). Moreover, the decrease in TNF receptor proteins in DCM patients did not appear to be the result of age-related differences in this group, since there was no significant difference (P=.32) in age between DCM patients and nonfailing organ donors. Second, the decrease in total TNF receptor protein in the failing hearts was not the result of a selective diminution in a particular TNF receptor subtype but rather the result of a general decrease in both receptor subtypes, TNFR1 and TNFR2 (Fig 2B⇑ and 2C⇑). Taken together, the above findings suggest that TNF receptors are dynamically regulated in heart disease and are downregulated in patients with heart failure.
Quite apart from the novelty of the above findings with respect to TNF receptor proteins, the results of the present study are important for a second reason. Although the source of TNF-α production in advanced heart failure is not known, recent studies from this and other laboratories suggest that under certain forms of stress, the heart is capable of synthesizing TNF-α mRNA and protein.19 25 26 Thus, the finding that TNF-α mRNA and protein were present in the hearts of DCM and IHD patients (Figs 3⇑ and 4⇑) suggests but does not prove that the heart itself may be a source of TNF-α production in advanced congestive heart failure. This intriguing point of view is also supported by the observation that except for one patient, all of the patients with immunodetectable levels of intracardiac TNF-α had elevated circulating levels of TNF-α in their sera (Fig 5⇑). Moreover, as noted, the notion that the failing heart may produce TNF-α is entirely consistent with experimental studies19 24 25 26 that showed that under certain forms of stress, the heart synthesizes TNF-α de novo. The above arguments notwithstanding, our data suggest that the myocardium may not be the sole source for TNF-α production in congestive heart failure. Approximately 50% of the patients with elevated levels of TNF-α in their sera did not have immunodetectable TNF-α in their hearts. Thus, there may be multiple sites of TNF-α production in advanced heart failure. In this regard, increased levels of serum neopterin (a marker for monocyte activation) have been reported in patients with advanced heart failure.32 Whether the elevated serum neopterin levels reflect mononuclear cell activation secondary to increased levels of circulating TNF-α or whether instead there is primary activation of mononuclear cells with the secondary elaboration of TNF-α remains to be determined.
Expression of TNF Receptors
Recent studies from this laboratory15 identified the presence and functional significance of TNFR1 and TNFR2 in adult cardiac myocytes. Although regulation of TNF receptors has been characterized extensively in other cell types, regulation of TNF receptors has not been characterized previously in the healthy or diseased heart. With respect to TNF receptor regulation in other cell types, several different mechanisms have been proposed, including proteolytic cleavage of TNF receptors from the cell membrane (“receptor shedding”)33 34 35 36 ; transcriptional regulation of TNFR1 and TNFR2, which, although poorly understood, appears to involve at least the protein kinase A and C pathways16 37 ; posttranslational modification of the TNFR1 and TNFR2 proteins16 ; or internalization of TNF receptors induced by homologous38 or heterologous ligands.39 40 Nonetheless, it should be recognized that the biological consequence of TNF receptor downregulation is uncertain, since neither TNF receptor number or affinity appears to modulate the susceptibility of TNF-α–sensitive cell types to growth-inhibitory or cytotoxic effects of this proinflammatory cytokine.41 42 43 44 45 46
Although the present study has not clearly identified the mechanism for TNF receptor downregulation, several potential explanations warrant further discussion. First, there was no obvious, clear-cut evidence for transcriptional downregulation of myocardial TNF receptor proteins in the failing hearts. Although the Northern blot analysis used herein provides only a semiquantitative assessment of mRNA levels, the results show that TNFR1 and TNFR2 mRNA was present in all of the failing myocardial samples tested. Nonetheless, on the basis of the available data, we cannot rigorously exclude the possibility that transcriptional downregulation of TNFR1 and TNFR2 may have occurred. A second possible mechanism for the decreased levels of TNF receptor protein in the failing hearts is that elaboration of TNF-α within the myocardium might have led to ligand binding and internalization of the TNF-α ligand–TNF receptor complex (homologous desensitization). However, as noted, levels of TNF receptor protein were not significantly different in patients with and without immunodetectable TNF-α in their hearts. Thus, although we cannot entirely discount the possibility that homologous desensitization of the TNF receptors occurred, it appears unlikely. Perhaps the most likely explanation for the decreased levels of intracardiac TNF receptor proteins is that there was ongoing receptor shedding in the patients with advanced heart failure.33 34 35 36 This view is supported by the finding that levels of circulating sTNFR1 and sTNFR2 were significantly increased in patients with advanced failure (Table⇑). It should be recognized, however, that the levels of circulating sTNFR1 and sTNFR2 reflect a generalized shedding of TNF receptors from a variety of different cell types, not just those receptors shed by cells that reside within the myocardium. Finally, it is entirely possible that some combination of all three of the above mechanisms may have been responsible for the decreased expression of myocardial TNF receptors. Nonetheless, although the above studies with TNF receptors showed that these receptors are downregulated in heart failure, the absolute degree of TNF receptor downregulation in heart failure patients (relative to nonfailing control subjects) may be difficult to address precisely in a clinical study of this nature because of difficulties inherent in obtaining normal myocardium.
Although the elaboration of TNF-α in congestive heart failure was originally proposed as a potentially important mechanism for the cachexia that frequently occurs in this syndrome,4 there is an increasing awareness that TNF-α may play a much broader pathophysiological role in congestive heart failure than was originally posited.12 47 Many of the clinical hallmarks of congestive heart failure, including LV dysfunction, cardiomyopathy, and pulmonary edema, can be explained by the known biological effects of TNF-α in humans.1 2 3 Thus, although the exact clinical significance of elevated levels of TNF-α in advanced heart failure is still uncertain, what is clear is that elevated levels of TNF-α can produce a number of the classic features of congestive heart failure.
The results of the present study provide several potential new insights into the pathophysiological role for TNF-α in heart failure. First, the observation that myocardial TNF receptor proteins are downregulated in patients with elevated TNF-α levels suggests that the heart itself is a target organ for TNF-α in advanced heart failure. Although the decrease in TNF receptor expression may be viewed teleologically as a potentially beneficial adaptive response of the myocardium, it remains to be determined whether the magnitude of the decrease in TNF receptor expression in advanced heart failure is sufficient to desensitize the heart to the negative inotropic effects of TNF-α. A second potentially important finding of the present study is that the level of soluble “shed” TNF receptors was increased significantly in patients with advanced congestive heart failure. Given that there was greater than a 500-fold molar excess of soluble binding proteins to circulating TNF-α in patients with heart failure and that experimental studies from this laboratory48 have shown that a 100-fold molar excess of soluble binding proteins is sufficient to completely abrogate the negative inotropic effects of TNF-α in vitro, the shedding of soluble TNF receptors may act to buffer the heart against the untoward effects of TNF-α in advanced heart failure. However, it is not known whether circulating soluble TNF receptors in the periphery are capable of neutralizing the effects of TNF-α synthesized centrally within the myocardium. Furthermore, it is possible that soluble TNF receptors may stabilize TNF-α in the circulation and may act therefore as a reservoir that slowly releases biologically active TNF-α into the circulation. Thus, the clinical significance of elevated circulating levels of TNF receptors in heart failure patients remains uncertain at present. The third and perhaps most intriguing aspect of the present study is the observation that there was evidence for TNF-α mRNA and/or protein biosynthesis in 50% of the failing hearts, whereas there was no evidence for TNF-α biosynthesis in nonfailing hearts. Although this finding does not establish myocardial TNF-α biosynthesis as the proximate cause of heart failure, the observation that failing hearts express elevated levels of a cytokine with negative inotropic properties raises the interesting possibility that overexpression of myocardial TNF-α may be one of several potentially important maladaptive mechanisms that contribute to the progressive cardiac decompensation that occurs in advanced heart failure.
Note Added in Proof
Similar to the findings in the present study with respect to soluble TNF receptors, Ferrari and colleagues (Circulation. 1995;92:1479-1486) have shown that patients with advanced heart failure have elevated levels of sTNFR1 and sTNFR2 in their peripheral circulation. Moreover, these authors showed that elevated levels of sTNFR2 were a marker of poor prognosis for patients with advanced heart failure.
Selected Abbreviations and Acronyms
|IHD||=||ischemic heart disease|
|LV||=||left ventricle (left ventricular)|
|TNF||=||tumor necrosis factor|
|TNFR1||=||tumor necrosis factor receptor type 1|
|TNFR2||=||tumor necrosis factor receptor type 2|
This research was supported by research funds from the Department of Veterans Affairs, the National Institutes of Health (P50 HL-O6H), and an educational grant from Merck and Company. The authors gratefully acknowledge the secretarial assistance of Jana Grana and Adrienne Chee as well as the technical assistance of Dorellyn Lee-Jackson. We would also like to acknowledge the kind gift of two myocardial specimens (DCM) from Dr Pia Pollock and to thank Dr Andrew I. Schafer for his past and present guidance and support. Finally, we would like to thank the reviewers for their thoughtful suggestions for improving this manuscript.
Reprint requests to Douglas L. Mann, MD, Cardiology Section, VA Medical Center, 2002 Holcombe Blvd, Houston, TX 77030.
Guest editor for this article was Victor Dzau, MD, Stanford (Calif) University School of Medicine.
- Received May 22, 1995.
- Revision received August 16, 1995.
- Accepted September 25, 1995.
- Copyright © 1996 by American Heart Association
Hegewisch S, Weh HJ, Hossfeld DK. TNF-induced cardiomyopathy. Lancet. 1990;2:294-295.
Levine B, Kalman J, Mayer L, Fillit HM, Packer M. Elevated circulating levels of tumor necrosis factor in severe chronic heart failure. N Engl J Med. 1990;223:236-241.
McMurray J, Abdullah I, Dargie HJ, Shapiro D. Increased concentrations of tumour necrosis factor in ‘cachectic’ patients with severe chronic heart failure. Br Heart J. 1991;66:356-358.
Matsumori A, Yamada T, Suzuki H, Matoba Y, Sasayama S. Increased circulating cytokines in patients with myocarditis and cardiomyopathy. Br Heart J. 1994;72:561-566.
Dutka DP, Elborn JS, Delamere F, Shale DJ, Morris GK. Tumour necrosis factor-alpha in severe congestive cardiac failure. Br Heart J. 1993;70:141-143.
Katz SD, Rao R, Berman JW, Schwarz M, Demopoulos L, Bijou R, LeJemtel TH. Pathophysiological correlates of increased serum tumor necrosis factor in patients with congestive heart failure: relation to nitric oxide–dependent vasodilation in the forearm circulation. Circulation. 1994;90:12-16.
Finkel MS, Oddis CV, Jacob TD, Watkins SC, Hattler BG, Simmons RL. Negative inotropic effects of cytokines on the heart mediated by nitric oxide. Science. 1992;257:387-389.
Yokoyama T, Vaca L, Rossen RD, Durante W, Hazarika P, Mann DL. Cellular basis for the negative inotropic effects of tumor necrosis factor-alpha in the adult mammalian heart. J Clin Invest. 1993;92:2303-2312.
Gulick TS, Chung MK, Pieper SJ, Lange LG, Schreiner GF. Interleukin 1 and tumor necrosis factor inhibit cardiac myocyte β-adrenergic responsiveness. Proc Natl Acad Sci U S A. 1989;86:6753-6757.
Chung MK, Gulick TS, Rotondo RE, Schreiner GF, Lange LG. Mechanism of action of cytokine inhibition of β-adrenergic agonist stimulation of cyclic AMP in rat cardiac myocytes: impairment of signal transduction. Circ Res. 1990;67:753-763.
Torre-Amione G, Kapadia S, Lee J, Bies RD, Lebovitz R, Mann DL. Expression and functional significance of tumor necrosis factor receptors in human myocardium. Circulation. 1995;92:1487-1493.
Kapadia S, Lee JR, Torre-Amione G, Birdsall HH, Ma TS, Mann DL. Tumor necrosis factor gene and protein expression in adult feline myocardium after endotoxin administration. J Clin Invest. 1995;96:1042-1052.
Young JB, Leon CA, Weilbacher DA. Endomyocardial biopsy in critically ill patients: the procedure and diagnostic and prognostic potential. In: Majid P, Kirby RR, Taylor RW, eds. Problems in Critical Care: Advances in Interventional Cardiology. Philadelphia, Pa: JB Lippincott Co; 1988:433-443.
Cannon JG, Nerad JL, Poutsiaka DD, Dinarello CA. Measuring circulating cytokines. J Appl Physiol. 1993;75:1897-1902.
Maniatis T, Fritsch EF, Sambrook J. Molecular Cloning: A Laboratory Approach. Cold Spring Harbor, NY: Cold Spring Harbor Press; 1989.
Stauber GB, Aiyer RA, Aggarwal BB. Human tumor necrosis factor-α receptor: purification by immunoaffinity chromatography and initial characterization. J Biol Chem. 1988;263:19098-19104.
Giroir BP, Johnson JH, Brown T, Allen GL, Beutler B. The tissue distribution of tumor necrosis factor biosynthesis during endotoxemia. J Clin Invest. 1992;90:693-698.
Giroir BP, Horton JW, White JD, McIntyre KL, Lin CQ. Inhibition of tumor necrosis factor prevents myocardial dysfunction during burn shock. Am J Physiol. 1994;267:H118-H124.
Seckinger PL, Isaaz S, Dayer JM. A human inhibitor of tumor necrosis factor α. J Exp Med. 1988;167:1511-1516.
Peetre C, Thysell H, Grubb A, Olsson I. A tumor necrosis factor binding protein is present in human biological fluids. Exp Hematol. 1988;41:414-419.
Engelmann H, Aderka M, Rubenstein M, Rotman D, Wallach D. A tumor necrosis factor-binding protein purified to homogeneity from human urine protects cells from tumor necrosis factor toxicity. J Biol Chem. 1989;264:11974-11980.
Pennica D, Lam VT, Mize NK, Weber RF, Lewis M, Fendly BM, Lipari MT, Goeddel DV. Biochemical properties of the 75-kDa tumor necrosis factor receptor. J Biol Chem. 1992;267:21172-21178.
Aggarwal BB, Totpal K, Pocsik E, Graff K, Liao W, Higuchi M. Role of protein kinases and phosphatases in receptor biosynthesis, shedding and mediation of cellular responses to tumor necrosis factor. In: Fiers WS, Burman WA, eds. Tumor Necrosis Factor: Molecular and Cellular Biology and Clinical Relevance. Basel, Switzerland: Karger; 1993:65-80.
Crowe PD, Vanarsdale TL, Goodwin RG, Ware CF. Specific induction of 80-kDa tumor necrosis factor receptor shedding in t-lymphocytes involves the cytoplasmic domain and phosphorylation. J Immunol. 1993;151:6882-6890.
Scheurich P, Kobrich G, Pfizenmaier K. Antagonistic control of tumor necrosis factor receptors by protein kinases A and C: enhancement of TNF receptor synthesis by protein kinase A and transmodulation of receptors by protein kinase C. J Exp Med. 1989;170:947-958.
Johnson SE, Baglioni C. Tumor necrosis factor receptors and cytocidal activity are down-regulated by activators of protein kinase C. J Biol Chem. 1988;263:5686-5692.
Ding AH, Sanchez E, Srimal S, Nathan CF. Macrophages rapidly internalize their tumor necrosis factor receptors in response to bacterial lipopolysaccharide. J Biol Chem. 1989;264:3924-3929.
Aggarwal BB, Eessalu TE. Effect of phorbol esters on down-regulation and redistribution of cell surface receptors for tumor necrosis factor-α. J Biol Chem. 1987;262:16450-16455.
Kull F, Jacobs S, Cuatrecasas P. Cellular receptor for 125I-labeled tumor necrosis factor: specific binding, affinity labeling, and relationship to sensitivity. Proc Natl Acad Sci U S A. 1985;82:5756-5760.
Michishita M, Yoshida Y, Uchino H, Kazuhiro N. Induction of tumor necrosis factor-α and its receptors during differentiation in myeloid leukemic cells along the monocyte pathway. J Biol Chem. 1990;265:8751-8759.
Scheurich P, Ucer U, Kronke M, Pfizenmaier K. Quantification and characterization of high-affinity membrane receptors for tumor necrosis factor on human leukemic cell lines. Int J Cancer. 1994;38:127-133.
Peetre C, Gulberg U, Nilsson E, Olsson I. Effects of recombinant tumor necrosis factor on proliferation and differentiation of leukemic and normal hemopoietic cells in vitro: relationship to cell surface receptor. J Clin Invest. 1986;78:1694-1700.
Munker R, DiPersio J, Koeffler HP. Tumor necrosis factor: receptors on hematopoietic cells. Blood. 1987;70:1730-1734.
Rubin BY, Anderson SL, Sullivan SA, Williamson BD, Carswell EA, Old LJ. Nonhematopoietic cells selected for resistance to tumor necrosis factor produce tumor necrosis factor. J Exp Med. 1986;164:1350-1355.
Kapadia S, Torre-Amione G, Yokoyama T, Mann DL. Soluble tumor necrosis factor binding proteins modulate the negative inotropic effects of TNF-α in vitro. Am J Physiol. 1995;37:H517-H525.