Induction of Cyclooxygenase-2 and Activation of Nuclear Factor-κB in Myocardium of Patients With Congestive Heart Failure
Background—Chronic heart failure is associated with induction of secondary inflammatory mediators, including prostanoids. The latter exert diverse functional and morphological effects on cardiac myocytes. Induction of cyclooxygenase (COX), the enzyme responsible for generating prostanoids, requires activation of nuclear factor-κB (NF-κB). The aim of the present study was to determine the expression of COX-2 and activation of NF-κB in the failing human heart.
Methods and Results—Myocardial tissue from 27 patients with end-stage heart failure (various etiologies: ischemic heart disease, n=16; idiopathic dilated cardiomyopathy, n=10; and valvular heart disease, n=1), 2 septic patients, and 8 normal control subjects was immunostained with antisera to COX-2 and NF-κB. Western blotting was performed and showed high anti–COX-2 antibody specificity and the presence of COX-2 protein in the sample tissues. In situ hybridization and immunohistochemistry showed little or no expression of COX-2 and NF-κB in the control hearts. In contrast, there was abundant expression of COX-2 mRNA and protein in myocytes and inflammatory cells in areas of fibrotic scar compared with regions of normal morphology in all cases of heart failure, except the cases with sepsis, which showed an abundance of COX-2 throughout the myocardium. Sites of NF-κB activation were associated with those of COX-2 expression.
Conclusions—We demonstrate induction of COX-2 and activation of NF-κB in the myocardium of failing human hearts. Induction of both molecules appears to be associated with the presence of inflammation and scar formation.
Prostanoids (prostaglandins and thromboxane A2) are the metabolic products of the membrane phospholipid arachidonic acid via the COX pathway. Presently, 2 forms of COX enzyme are recognized: the constitutive enzyme COX-1, which is normally expressed in most tissues; and the newly discovered COX-2, which is induced in many cell types in response to various stimuli, including cytokines.1 A number of cytokines have been shown to be expressed in the failing human heart. These cytokines exert their actions directly on the myocardium or modulate the expression and release of other mediators, such as prostanoids. Many inflammatory mediators use NF-κB as one of their mechanisms of induction and perpetuation.2 3 Recent studies have shown that COX-2 is induced by inflammatory cytokines such as tumor necrosis factor-α , which is produced in heart failure,4 via the transcription factor NF-κB.2 In addition to cytokines, hypoxia, an important feature of ischemic myocardium, induces COX-2 (via NF-κB) in cultured endothelial cells independently of other stimuli.5 Although there have been several studies that demonstrated induction of COX-2 and its association with NF-κB activation in various cells,2 3 and despite the numerous studies describing the production of prostanoids in myocytes,6 7 whether or not COX-2 is expressed in the human myocardium remains to be elucidated. Interest in COX-2 expression in the failing heart is heightened by the diverse inotropic and morphological effects of prostaglandins on the heart.6 8 We therefore sought to determine expression of COX-2 and activation of NF-κB in the failing human heart by immunohistochemistry, Western blotting, and in situ hybridization.
Tissues from failing human hearts were collected at the time of transplantation from patients with IHD (n=16 males; mean age, 51.6±1.8 years; EF, 18.8±3.0%; LVEDP, 27.7±2.1 mm Hg; LVEDD, 71.0±4.0 mm), DCM (n=10; 8 females; mean age, 44.5±4.0 years; EF, 14.4±1.8%; LVEDP, 27.6±2.5 mm Hg; LVEDD, 72.5±5.4 mm), and valvular heart disease (one 63-year-old man; EF, 20%; LVEDD, 75 mm); tissues were also collected at autopsy from 2 patients with sepsis (mean age, 56.0±8.0 years). All patients had heart failure (NYHA class 3 or 4) and had received treatment with antiarrhythmic drugs, diuretics, digoxin, nitroglycerin, ACE inhibitors, β-receptor blockers, or Ca2+ channel blockers. Duration of the disease ranged from 3 months to 9 years. Pretransplant complications included diabetes, hyperlipidemia, hypertension, and renal disease. Normal control hearts were collected at surgery (unused donor hearts, n=5) or autopsy (n=3). Slices of the left ventricle, from both infarcted and noninfarcted regions, were cut from the whole heart and placed in either 4% paraformaldehyde or 10% formalin. The study conformed to the ethics committee requirements of the Montreal General Hospital.
Immunohistochemistry and Western Blotting
Using a previously described method,9 we immunostained the tissues with polyclonal antisera to human COX-2 and NF-κB (Boehringer Mannheim). Each complete field of cardiac myocytes was graded in a blinded fashion from 0 to 4 (where 0=no staining, 1=focal staining, 2=diffuse weak staining, 3=diffuse moderate staining, and 4=diffuse strong staining).9 In addition to immunohistochemistry, adjacent left ventricular tissues were homogenized and analyzed by Western blot with the Novex Xcell II system (Novex) to confirm the specificity of the COX-2 antiserum.
In Situ Hybridization
A nonradioactive technique using the DIG system (Boehringer Mannheim) was employed. Briefly, cryostat sections were permeabilized as described elsewhere10 then hybridized with the labeled probe at 42°C for 16 hours. This was followed by several washes in SSC (×4 to 0.1) and RNase A solutions. Sections were incubated with alkaline phosphatase conjugated sheep anti-digoxigenin Fab fragments. Signals were visualized with 4-nitro-blue-tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate in 10% polyvinyl alcohol–treated equalization buffer.
There was a gradient for COX-2 expression in the myocardium of patients with heart failure secondary to IHD, with the strongest signal in the subendocardium (infarcted zone=3.1±0.2, noninfarcted zone=1.8±0.4), the weakest in the subepicardium (infarcted zone=2.6±0.2, noninfarcted zone=1.9±0.6), and a midlevel intensity in the midmyocardium (infarcted zone=2.9±0.1, noninfarcted zone=1.9±0.6) (Figure 1⇓). The most apparent expression of COX-2 was seen in myocytes and inflammatory cells. Expression of COX-2 in endothelial cells of the endocardium and intramural coronary arteries ranged from weak focal to strong diffuse (Figure 1B⇓). There was a very apparent immunoreactivity for NF-κB in the myocytes of patients with IHD in both infarcted and noninfarcted regions (Figure 2B⇓ and 2F⇓). In general, cells with apparent COX-2 expression showed nuclear staining for NF-κB. Patients with DCM showed unequal signal intensity for COX-2, ranging from none to a moderate diffuse signal (subendocardium=2.4±0.3, midmyocardium=2.1±0.4, and subepicardium=2.2±0.3). Weak to moderate expression of COX-2 in the myocytes and inflammatory cells in DCM was associated with the presence of extensive myocardial fibrosis (Figure 1C⇓). No expression of the enzyme was seen in other cell types. Immunoreactivity for NF-κB was also seen in these cases; however, it was far less evident than in IHD cases (Figure 2C⇓). Patients with heart failure secondary to sepsis showed moderate to strong expression of COX-2 in the myocytes, endocardium, and inflammatory cells (subendocardium=3.5±0.5, midmyocardium=2.9±0.5, and subepicardium=1.9±0.6) (Figure 1E⇓ and 1G⇓). Similarly, there was strong nuclear immunostaining for NF-κB in the myocytes and inflammatory cells in septic hearts (Figure 2A⇓). The patient with valvular heart disease showed little expression of COX-2 in the myocytes and endocardium and no signal in the coronary arteries. There was only sparse focal immunoreactivity for NF-κB (Figure 2D⇓ and 2E⇓). Normal control hearts showed little or no expression of either factor (Figure 1F⇓). Negative control experiments did not show any nonspecific signal (Figure 1D⇓ and 1H⇓). Western blot analysis revealed that the COX-2 antiserum showed high specificity for the activated mouse macrophage lysate and also detected COX-2 protein in failing hearts of patients with IHD and DCM. There was no significant correlation between the expression of these molecules and age, sex, duration of illness, complications, type of treatment, or cardiac function.
In the present study, we demonstrate for the first time induction of COX-2 and activation of NF-κB in the failing human heart. Our investigations showed abundant expression of COX-2 and activation of NF-κB in myocytes and inflammatory cells in the infarcted myocardium compared with the noninfarcted myocardium of patients with heart failure secondary to IHD. Abundant expression of the enzyme was seen in septic hearts. In contrast, expression of the enzyme in DCM was only seen in areas of myocardial fibrosis. Activation of NF-κB was seen in cells associated with induction of COX-2. Both molecules were rarely seen in normal control hearts. These findings demonstrate increased expression of COX-2 and activation of NF-κB in the failing myocardium and suggest that induction of COX-2 and increased formation of prostanoids may contribute to the pathophysiology of heart failure.
Little is known about the expression and regulation of COX-2 in the myocardium. Previous studies have shown increased release of thromboxane A2 in myocardial ischemia and production of arachidonic metabolites in the ischemic myocardium and in cardiac myocytes under physiological and stimulated conditions.7 8 10 In addition, Liu et al11 have recently shown the presence of both COX-1 and COX-2 in the normal and lipopolysaccharide-stimulated rat heart. The latter study supports our findings of induced COX-2 expression in the myocytes and inflammatory cells of patients with heart failure secondary to sepsis. Beside lipopolysaccharide, there are several other mediators known to induce COX-2 in other cells that may contribute to the induction of the enzyme in myocytes of failing hearts. For example, hypoxia and tumor necrosis factor-α, important features of IHD, have been shown to induce prostanoid formation in several cell types, including myocytes.1 5 In part, the pathway of COX-2 induction and the subsequent formation of prostanoids involve activation of NF-κB.2 5 Interestingly, there are 2 NF-κB binding sites in the promoter region of COX-2.12 In the present study, we demonstrate induction of COX-2 and translocation of NF-κB from the cytoplasm to the nuclei in cardiomyocytes and inflammatory cells of the failing human heart. Our data suggest that ischemia and inflammatory mediators are responsible for activation of NF-κB and induction of COX-2 in this disease condition.
The current study raises the question of the pathophysiological significance of COX-2 induction and NF-κB activation in chronic heart failure. It is well known that prostaglandin formation is increased in patients with chronic heart failure and that inhibition of the COX metabolic pathway results in adverse effects on systemic vascular resistance, cardiac output, renal blood flow, glomerular filtration, and calf vascular resistance.13 14 Whether these effects can be attributed to inhibition of cardiac COX-2 induction remains to be elucidated. On the other hand, activation of NF-κB may be associated with induction of other mediators known to be present in heart failure, including NO. Recent reports have demonstrated cytotoxic and negative inotropic effects of NO on cardiac myocytes.15 Accordingly, any factor that modulates NO expression may potentially have deleterious effects on the heart.
In conclusion, we have demonstrated that induction of COX-2 in failing human hearts is associated with the presence of myocardial scarring (inflammation and fibrosis) as well as sepsis. We have also demonstrated activation of NF-κB in sites of induced COX-2 expression. We suggest a possible role for NF-κB in the ischemia-mediated induction of COX-2 in myocytes, inflammatory cells, and endothelial cells of failing human myocardium. Increased production of prostanoids by ischemic cells may have important pathological consequences in IHD and sepsis. The use of specific COX-2 inhibitors in animal models of heart failure will determine the exact role of the enzyme in this fatal disease.
Selected Abbreviations and Acronyms
|IHD||=||ischemic heart disease|
|LVEDD||=||left ventricular end-diastolic diameter|
|LVEDP||=||left ventricular end-diastolic pressure|
This study was supported by the Heart and Stroke Foundations of Canada and Quebec. We thank Dr B. Kennedy for his help during the study. We also acknowledge the support of the Endothelium Network of the Fonds de la recherche en santé du Québec (FRSQ).
- Received January 6, 1998.
- Revision received May 10, 1998.
- Accepted May 14, 1998.
- Copyright © 1998 by American Heart Association
Smith WL, Garavito RM, DeWitt DL. Prostaglandin endoperoxide H synthases (cyclo-oxygenase)-1 and -2. J Biol Chem. 1996;271:33157–33160.
Yamamoto K, Arakawa T, Ueda N, Yamamoto S. Transcriptional roles of nuclear factor kB and nuclear factor-interleukin-6 in the tumor necrosis factor α-dependent induction of cyclo-oxygenase-2 in MC3T3–E1 cells. J Biol Chem. 1995;270:31315–31320.
Torre-Amione G, Kapadia S, Lee J, Durand J-B, Bies RD, Young JB, Mann DL. Tumor necrosis factor-α and tumor necrosis factor receptors in the failing human heart. Circulation. 1996;93:704–711.
Schmedtje JF Jr, Ji YS, Liu WL, DuBois RN, Runge MS. Hypoxia induces cyclo-oxygenase-2 via the NF-kB p65 transcription factor in human vascular endothelial cells. J Biol Chem. 1997;272:601–608.
Linssen MCJG, Engels W, Lemmens PJMR, Heijnen VVTH, Bilsen MV, Reneman RS, van der Vusse GJ. Production of arachidonic acid metabolites in adult rat cardiac myocytes, endothelial cells, and fibroblast-like cells. Am J Physiol. 1993;264(Heart Circ Physiol 33):H973–H982.
Lai J, Jin HK, Yang RH, Winer JLW, Yen R, King KL, Zeigler F, Ko A, Cheng J, Bunting S, Paoni NF. Prostaglandin F2-alpha induces cardiac myocyte hypertrophy in vitro and cardiac growth in vivo. Am J Physiol. 1996;40:H2197–H2208.
Curtis MJ, Pugsley MK, Walker MJ. Endogenous chemical mediators of ventricular arrhythmias in ischemic heart disease. Cardiovasc Res. 1993;27:703–719.
Appleby SB, Ristimaki A, Neilson K, Narko K, Hla T. Structure of the human cyclo-oxygenase-2 gene. Biochem J. 1994;302:723–727.
Dzau VJ, Packer M, Lilly LS, Swartz SL, Hollenberg NK, Williams GH. Prostaglandins in severe congestive heart failure. N Engl J Med. 1982;10:347–352.
Townend JN, Doran J, Lote CJ, Davies MK. Peripheral hemodynamic effects of inhibition of prostaglandin synthesis in congestive heart failure and interaction with captopril. Br Heart J. 1995;73:434–441.
Brady AJB, Poole-Wilson PA, Harding SE, Warren JB. Nitric oxide production within cardiac myocytes reduces their contractility in endotoxemia. Am J Physiol. 1992;263(Heart Circ Physiol 32):H1963–H1966.