β-Adrenergic Blockade in Developing Heart Failure
Effects on Myocardial Inflammatory Cytokines, Nitric Oxide, and Remodeling
Background—Whether β-adrenergic blockade modulates myocardial expression of inflammatory cytokines and nitric oxide (NO) in heart failure is unclear.
Methods and Results—We administered oral metoprolol or no therapy to rats for 12 weeks after large myocardial infarction and subsequently examined left ventricular (LV) remodeling; myocardial tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6 expression; and NO. In untreated rats, echocardiography revealed significant (P<0.001) LV dilatation and systolic dysfunction compared with sham. Papillary muscle studies revealed isoproterenol hyporesponsiveness to be unaltered by NO synthase (NOS) inhibition. Circulating NO metabolites were undetectable. In noninfarcted myocardium, although inducible NOS (iNOS) mRNA was absent, TNF-α, IL-1β, and IL-6 mRNA and protein were markedly elevated compared with sham (P<0.001), with 2-fold higher expression (P<0.025) of IL-6 compared with TNF-α or IL-1β. Metoprolol administration starting 48 hours after infarction (1) attenuated (P<0.02) LV dilatation and systolic dysfunction, (2) preserved isoproterenol responsiveness (P<0.025) via NO-independent mechanisms, and (3) reduced myocardial gene expression and protein production of TNF-α and IL-1β (P<0.025) but not IL-6, which remained high.
Conclusions—During heart failure development, adrenergic activation contributes to increased myocardial expression of TNF-α and IL-1β but not IL-6, and one mechanism underlying the beneficial effects of β-adrenergic blockade may involve attenuation of TNF-α and IL-1β expression independent of iNOS and NO.
Heart failure (HF) is associated with local and systemic elaboration of inflammatory cytokines, including tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6.1 2 3 Although their precise role is unclear, inflammatory cytokines induce myocardial effects similar to the HF phenotype, including myocyte apoptosis,4 myocyte hypertrophy,5 extracellular matrix alterations,6 and contractile depression7 8 9 ascribed to both nitric oxide (NO)-dependent7 and NO-independent mechanisms.8 9 Previous studies have also reported increased myocardial inducible NO synthase (iNOS) expression and activity in HF.10 11 Indeed, because high concentrations of NO attenuate myocyte contraction and catecholamine responses,12 13 one proposed mechanism of myocardial dysfunction in HF is excessive NO production secondary to increased inflammatory cytokines. In support of this concept, studies have shown that NOS blockade improves myocardial β-adrenergic responsiveness in HF.11 14
Recent investigations have shown that in failing myocardium, chronic β-adrenergic blockade improves myocardial function and left ventricular (LV) remodeling, although the cellular mechanisms responsible for these salutary effects have not been fully defined.15 Given that inflammatory cytokines and NO influence β-adrenergic responses in normal myocardium,8 12 we hypothesized that adrenergic nervous system activation and myocardial elaboration of cytokines and NO in HF are interdependent and that the beneficial effects of β-blockade are in part related to modulation of this axis. Thus, we investigated the effects of β-blockade during the development of experimental HF on (1) LV remodeling and function; (2) NO-mediated β-adrenergic hyporesponsiveness, myocardial NOS expression, and systemic elaboration of NO; and (3) myocardial expression of TNF-α, IL-1β, and IL-6.
Infarction Model and Experimental Groups
All studies were performed in compliance with the Guide for the Care and Use of Laboratory Animals (DHHS publication [NIH] 85-23, revised 1996). Wistar-Kyoto rats weighing 200 to 250 g were deeply anesthetized with intramuscular ketamine (52 mg/kg), acepromazine (1.8 mg/kg), and xylazine (10.4 mg/kg) and supported by a rodent ventilator (Harvard Apparatus). Under sterile conditions, a left thoracotomy was performed. A 5.0 prolene ligature was passed over a 3.0 silk lay suture and tied around the proximal left coronary artery. The chest was closed with 4.0 silk suture. Operative mortality was ≈40% owing to acute sudden death. After 48 hours of recovery, the animals were followed up for 12 weeks in 1 of 2 groups: (1) MI-M, receiving 2 g/L metoprolol in their drinking water (average dose 70.7±1.3 mg · kg−1 · d−1) or (2) MI-C, receiving no therapy. Seven animals underwent the full protocol in each group. In addition, 6 animals served as sham-operated controls.
Echocardiography and Serum Collection
Under half the anesthetic dose described above, 2D echocardiography (Acuson 128XP/10) was performed before surgery and 2 and 12 weeks after surgery. LV anteroposterior diameter (D) and short-axis area (A) at the papillary muscle level were measured at end diastole (ED) and end systole (ES). FAC (%) was calculated as [(LVEDA−LVESA)/LVEDA]×100 and was used as an index of LV systolic function. To exclude small infarctions, animals were maintained on protocol only if the 2-week study revealed ≥40% involvement of the LV circumference by infarction. Tail blood was collected after the initial and final studies, with serum stored at −80°C.
Twelve weeks after operation, the rats were deeply anesthetized as described, and median sternotomy was performed. The heart was rapidly excised and placed into a Krebs-Ringer solution containing (in mmol/L) Na+ 152, K+ 3.6, Cl− 135, HCO3− 25, Mg2+ 0.6, H2PO4− 1.3, SO42− 0.6, Ca2+ 2.5, glucose 5.6, and 2,3-butanedione monoxime (BDM) 30, pH 7.4, continuously bubbled with 95% O2/5% CO2 at room temperature, and the noninfarcted LV papillary muscle was harvested as previously described.16 The right and left ventricles were weighed, and noninfarcted tissue was snap-frozen in liquid nitrogen and stored at −80°C. A short-axis section of the LV was stored in formalin for immunohistochemistry.
Papillary Muscle Studies
Papillary muscles were mounted in a water-jacketed tissue bath (Radnoti) superfused with oxygenated Krebs-Ringer solution without BDM at 30°C.16 Field stimulation was performed with parallel platinum electrodes and a Grass SD9 stimulator delivering square-wave pulses (5-ms duration, voltage 30% above threshold) at 0.3 Hz. After 60 minutes of BDM washout, the length corresponding to maximal developed isometric tension (Lmax) was defined, and tension and dT/dt were recorded digitally (PowerLab, ADI Instruments). Superfusion was stopped and isoproterenol concentration-response curves were defined on addition of 10−9 to 10−4 mol/L dl-isoproterenol (Sigma), recording after 3 minutes at each concentration. After this, the NOS inhibitor L-NAME was added at 20 and 100 μmol/L, with measurements repeated after 5 minutes at each concentration. Lmax and muscle diameter were then measured with a calibrated eyepiece. The average muscle diameter was 0.83±0.07 mm. Tension and dT/dtmax were normalized for cross-sectional area.
Serum Nitrite and Nitrate Determinations
Serum nitrite and nitrate concentrations were determined colorimetrically with a commercially available kit (Cayman Chemical) according to the instructions supplied by the manufacturer.
Northern and Western Blotting
Total RNA extraction and Northern blotting, protein extraction and Western blotting, autoradiography, and densitometry were performed as previously described.17 18 For Northern blots, the following cDNA and oligonucleotide probes were used: mouse IL-1β (0.6 kb, BamHI-SmaI), mouse IL-6 (1.0 kb, EcoRI), human TNF-α (1.3 kb, BamHI-HindIII), and mouse iNOS (1.8 kb; Cayman Chemical). Human 28S rRNA (40-base single-stranded oligo; Oncogene Science) was used as an internal control, with results expressed as a ratio of the specific gene to the corresponding 28S rRNA. For Western blots, rabbit anti–rat IL-1β, IL-6, and TNF-α antibodies (Biosource International) were used at a concentration of 3.0 μg/mL (IL-1β, TNF-α) and 5.0 μg/mL (IL-6). Autoradiographic bands were semiquantified by comparison with sham-operated controls.
Paraffin-embedded sections 5 μm thick were stained for ecNOS and nitrotyrosine with an immunoenzymatic staining kit (DAKO PAP, System 40) as previously described.18 The following primary antibodies were used: rabbit anti–human ecNOS (2.0 μg/mL; Santa Cruz Biotechnology) and rabbit anti-nitrotyrosine as an indirect measure of peroxynitrite (1.0 μg/mL; Upstate Biotechnology). Immunoreactivity was evaluated by light microscopy and graded on a semiquantitative scale.
Comparisons of data between experimental groups were made with the unpaired t test with Bonferroni correction for multiple comparisons. Given 3 experimental groups, a value of P<0.025 was considered significant. Comparisons of dT/dtmax before and after L-NAME within groups were made with the paired t test. A value of P<0.05 was considered significant. Group data are expressed as mean±SEM.
Effects of β-Adrenergic Blockade on LV Remodeling and Systolic Function
Figure 1⇓ shows short-axis echocardiographic views from 1 animal at baseline and 12 weeks after myocardial infarction (MI). After MI, there was significant LV dilatation and systolic dysfunction, indicated by the reduced fractional area change (FAC). The Table⇓ shows LV remodeling data at 12 weeks. Compared with sham, both MI groups had significant LV dilatation (LV end-diastolic diameter, P<0.001; LV end-diastolic area, P<0.002), LV hypertrophy (LV/body weight ratio, P<0.001 for MI-control [MI-C] and P<0.02 for MI-metoprolol [MI-M] without differences in body weight), and reduced systolic function (FAC, P<0.001). However, compared with MI-C, the MI-M animals had significantly less LV dilatation (LV end-diastolic diameter, P<0.01) and hypertrophy (P<0.005) and improved systolic function (P<0.02), indicating that metoprolol attenuated detrimental LV remodeling. Although net heart rate reduction at 12 weeks was lower in the MI-M group (27±7 versus 16±9 bpm), this was not statistically significant.
Effects of β-Adrenergic Blockade on Myocardial Catecholamine Sensitivity, NO, and NOS
Figure 2⇓ displays isoproterenol concentration-response curves from papillary muscle studies. Contractility was assessed by use of dT/dtmax normalized to each respective baseline before isoproterenol. Absolute values of dT/dtmax at baseline (mN/s · mm2) were not statistically different between the 3 groups. As seen in Figure 2A⇓, sham-operated animals’ muscles displayed a progressive increase in contractile response with increasing isoproterenol, whereas MI-C muscles displayed a flat isoproterenol response and minimal increase in dT/dtmax (P<0.025 versus sham at micromolar concentrations and higher). Conversely, the MI-M group displayed preservation of the catecholamine response (P<0.025 versus MI-C) and increased sensitivity to nanomolar concentrations compared with sham (P<0.025).
To determine the influence of NO on catecholamine sensitivity, the contractile response to 10−3 mol/L isoproterenol was assessed before and after NG-nitro-l-arginine methyl ester (L-NAME). As seen in Figure 2B⇑, 20 μmol/L L-NAME had no significant effect on contractile response in any group. Similar results were seen with 100 μmol/L L-NAME (data not shown). Furthermore, serum nitrites and nitrates (either at baseline or 12 weeks after surgery) were not detectable in any group, and Northern blotting revealed no detectable myocardial iNOS mRNA. Finally, as seen in Figure 3⇓, immunohistochemistry for endothelial constitutive NOS (ecNOS) and nitrotyrosine in noninfarcted regions revealed only mild to moderate staining in vascular tissue and endothelial cells and none in myocytes, a constant pattern across all 3 experimental groups. No significant leukocytic cell infiltration was appreciated. Together, these data confirmed lack of upregulation of the myocardial NO axis and that NOS inhibition had no immediate impact on myocardial function.
Effects of β-Adrenergic Blockade on Myocardial Inflammatory Cytokine Expression
Figure 4⇓ shows autoradiograms and corresponding densitometry of total myocardial RNA (noninfarcted region, 12 weeks after ligation or sham operation) by Northern blotting for TNF-α, IL-1β, and IL-6. In the sham group, TNF-α and IL-1β mRNAs were undetectable and IL-6 mRNA was detected only at low levels. In the MI-C group, mRNA expression of each of these cytokines markedly increased, with 2-fold higher expression of IL-6 than either TNF-α or IL-1β (P<0.025). Compared with MI-C, the MI-M group displayed significant, 2.9-fold reductions in both TNF-α and IL-1β mRNA expression (P<0.025 versus MI-C) but no change in the augmented expression of IL-6 mRNA, which remained markedly elevated (5- to 6-fold higher) compared with either TNF-α or IL-1β. Figure 5⇓ shows autoradiograms and corresponding densitometry of Western blots 12 weeks after infarction or sham operation. Protein production of TNF-α, IL-1β, and IL-6 mirrored mRNA expression and was modulated in a similar fashion. Both MI groups displayed significantly increased protein levels of each compared with sham, with selective reduction of myocardial TNF-α and IL-1β in the MI-M group (≈4-fold, P<0.025 versus MI-C) but not of IL-6, which remained elevated. Thus, postinfarct LV dysfunction was associated with increased mRNA and protein of all 3 inflammatory cytokines, whereas treatment with metoprolol selectively decreased TNF-α and IL-1β.
This study demonstrates several novel findings in an established stage of LV dysfunction after MI in the rat. First, β-adrenergic blockade with metoprolol selectively decreases gene expression and protein production of TNF-α and IL-1β by noninfarcted myocardium but has no effect on IL-6. Second, the marked increase in myocardial proinflammatory cytokine expression is not associated with increased expression of myocardial iNOS or ecNOS, myocardial nitrotyrosine, or circulating NO metabolites. Third, increased myocardial NO production does not contribute to reduced myocardial catecholamine responsiveness, and β-blockade preserves catecholamine responsiveness via NO-independent mechanisms. These results suggest that activation of the adrenergic nervous system during HF development contributes to increased myocardial expression of TNF-α and IL-1β but not IL-6 and that one mechanism by which β-adrenergic blockade confers benefit in failing myocardium may be related to attenuation of myocardial TNF-α and IL-1β expression independent of iNOS and activity of NO.
Inflammatory Cytokines and NO in Postinfarct LV Dysfunction
Several lines of evidence advance the concept that inflammatory cytokines play a central pathophysiological role in HF. Elevated circulating and myocardial levels of TNF-α, IL-1β, and IL-6 have been reported in patients,1 2 3 with plasma levels correlating with severity of disease. Inflammatory cytokines induce biological effects similar to the phenotypic changes of HF, including contractile depression,7 8 9 myocyte growth and induction of a fetal gene program,5 myocyte apoptosis,4 and extracellular matrix alterations.6 Mechanisms of cytokine-induced contractile depression proposed include altered β-receptor coupling to adenylyl cyclase,8 increased NO and peroxynitrite formation,7 19 and alterations in intracellular Ca2+ handling.9
In our study, postinfarction LV dysfunction was associated with robust gene expression and protein production of TNF-α, IL-1β, and IL-6 in noninfarcted myocardium (Figures 4⇑ and 5⇑), the site of ongoing myocardial remodeling. The time period of study corresponded to a late stage after infarction when cellular inflammatory infiltration had abated (Figure 3⇑). Given their known toxic myocardial effects, these findings implicate TNF-α, IL-1β, and IL-6 as important mediators of adverse cardiac remodeling and decompensation in HF. These data extend the work of Irwin et al,20 who showed that TNF-α mRNA and protein are persistently expressed by myocytes in noninfarcted rat myocardium from 1 day to 5 weeks after infarction, and Ono et al,21 who demonstrated that myocardial gene expression of TNF-α, IL-1β, and IL-6 is elevated up to 20 weeks after coronary ligation.
A key point of this study is that despite elevated myocardial cytokine expression, there was no increased activity of the NO axis, with no detectable myocardial iNOS, no increase in myocardial ecNOS, no detectable circulating NO metabolites, and no increase in myocardial nitrotyrosine in the infarcted animals compared with sham (Figure 3⇑). In addition, papillary muscle studies revealed no modulation of the isoproterenol contractile response with NOS blockade (Figure 2B⇑). These results indicate that, at least in this established stage of LV dysfunction after infarction, the biological effects of inflammatory cytokines leading to myocardial dysfunction, reduced β-adrenergic responsiveness, and myocardial remodeling occur via NO-independent mechanisms. This is perhaps surprising given previous studies reporting increased myocardial NOS activity11 and iNOS expression10 in HF and improved myocardial β-adrenergic responsiveness with NOS blockade.11 14 However, other investigators have reported the opposite, ie, reduced cardiac NO production during HF development22 and no impact of NOS inhibition on contractile function in failing myocardium.23 Thus, the precise role of NO as a mediator of contractile dysfunction and cardiac remodeling in HF is controversial.
The reasons for these discrepancies are not fully clear but may be related to differences in the temporal stage of the HF phenotype when studied. Indeed, in a rabbit MI model, Akiyama et al24 demonstrated that cardiac iNOS and plasma/cardiac nitrite and nitrate increase early (within 3 days) after infarction and then decline rapidly over the course of 2 weeks. The source of cardiac iNOS was exclusively the infarcted or at-risk region with no expression in the contralateral normal zone. This pattern of cardiac iNOS release mirrors the time course of inflammatory cytokine gene expression in infarcted rat myocardium described by Ono et al,21 which peaked at 1 week, whereas cytokine levels in noninfarcted myocardium remained elevated up to 20 weeks after infarction. Thus, it would appear that after infarction, there is an initial increase in NO secondary to increased cardiac iNOS and inflammatory cytokines from the infarcted myocardium, possibly related to active inflammation, which then decreases during later stages of LV dysfunction in which inflammation has subsided. Similarly, reports of increased iNOS in myocardial tissue from patients with severe HF10 raise the possibility that myocardial iNOS may be reexpressed as cardiac decompensation progresses to more advanced stages, perhaps in relation to more pronounced elevations of cytokines at this time point.2
Effect of β-Adrenergic Blockade on Myocardial Inflammatory Cytokine Expression, Function, and Remodeling After MI
Although β-adrenergic blockade in HF has been shown to improve myocardial function and remodeling in clinical and experimental studies,15 the precise mechanisms underlying these beneficial effects are not well defined. Our study provides the first demonstration of the impact of β-blockade on myocardial inflammatory cytokine expression in developing HF together with concurrent effects on chamber remodeling. As seen in the Table⇑, postinfarction LV remodeling was attenuated by metoprolol with less LV dilatation and hypertrophy and improved LV systolic function. As seen in Figures 4⇑ and 5⇑, these salutary effects were accompanied by selective reductions in myocardial gene expression and protein production of TNF-α and IL-1β but not IL-6. Selective reduction of TNF-α and IL-1β, rather than generalized reduction of all 3 cytokines, argues against the possibility that this simply represents an epiphenomenon associated with improved LV remodeling and more favorable wall stress. Instead, these data raise the interesting possibility that prolonged β-adrenergic activation during HF development is a stimulus for myocardial TNF-α and IL-1β expression but not for IL-6, which may be regulated by other factors. Although this novel hypothesis has not been directly investigated previously, there is recent evidence for modulation of myocyte cytokine responses by catecholamines,25 and this new concept warrants further exploration.
These data also suggest distinct roles and control mechanisms for IL-6, as opposed to TNF-α and IL-1β, in the process of LV remodeling. IL-6 expression was consistently higher in MI-C animals than either TNF-α or IL-1β, which increased by comparable degrees. Also, unlike those of TNF-α and IL-1β, IL-6 levels remained persistently elevated in the metoprolol-treated group despite significant reductions in detrimental LV remodeling. Indeed, previous studies have demonstrated that IL-6 has unique functional aspects compared with other inflammatory cytokines. Using IL-6–deficient transgenic mice, Xing et al26 demonstrated that unlike TNF-α or IL-1β, which are clearly proinflammatory in nature, IL-6 confers a variety of anti-inflammatory effects and controls the extent of tissue inflammatory response. Studies in postischemic reperfused myocardium from this laboratory17 18 have shown that whereas the time courses of gene expression of TNF-α and IL-1β parallel each other, peaking early (1 hour) after reperfusion, IL-6 expression remains persistently elevated for 6 hours, an effect attributable to differences in transcriptional regulation of IL-6. Analogous to these studies, our data indicate that there are also significant differences in IL-6 behavior compared with that of TNF-α and IL-1β during the development of LV dysfunction after MI. TNF-α and IL-1β appear to be influenced by adrenergic activation to a greater extent and have greater association with adverse ventricular remodeling than IL-6. The mechanisms underlying these differences remain to be determined.
Together with reduction of myocardial expression of TNF-α and IL-1β and improved LV remodeling and systolic function, metoprolol administration after large infarction also restored the myocardial isoproterenol response (Figure 2⇑), confirming the results of clinical studies.27 28 This effect was unrelated to NO, given the lack of functional impact of NOS inhibition. Although we did not specifically examine alternative mechanisms (aside from NO) for changes in β-adrenergic responsiveness in this study, this finding may be related to reversal of β-adrenergic signal transduction abnormalities, as suggested by others.15 27 28 Specifically, metoprolol treatment in HF has been shown to restore27 28 and recouple28 downregulated and uncoupled β-adrenergic receptors. Given that inflammatory cytokines can produce uncoupling of the β-receptor from adenylyl cyclase,8 improvement in myocardial β-adrenergic responsiveness with metoprolol may also be related to reductions in expression of TNF-α and IL-1β.
In summary, this study demonstrates that LV dysfunction after MI in the rat is associated with marked increases in myocardial gene expression and protein production of TNF-α, IL-1β, and IL-6 in the noninfarcted zone without increased expression of myocardial iNOS or ecNOS. Although there is marked associated β-adrenergic hyporesponsiveness, this is not related to ongoing increased myocardial NO production. β-Adrenergic blockade with metoprolol in this setting improves LV remodeling and systolic function, restores isoproterenol sensitivity via NO-independent mechanisms, and selectively decreases myocardial gene expression and protein production of TNF-α and IL-1β but not IL-6. These results suggest that activation of the adrenergic nervous system during HF development contributes to increased myocardial expression of TNF-α and IL-1β but not IL-6 and that one mechanism underlying the salutary effects of β-blockade in HF may relate to attenuation of myocardial expression of TNF-α and IL-1β, independent of iNOS expression and NO.
This work was supported by an Established Investigator Grant (Dr Prabhu) and Grant-in-Aid (Dr Murray) from the American Heart Association, the Research Service of the Department of Veterans Affairs (Dr Prabhu, Dr Freeman), a grant from the South Texas Health Research Center (Dr Prabhu), and a grant from the San Antonio Area Foundation (Dr Prabhu). The authors gratefully acknowledge the excellent technical assistance of Danny Escobedo, Teri Frosto, and Mindy Kaplan.
- Received August 5, 1999.
- Revision received November 12, 1999.
- Accepted November 29, 1999.
- Copyright © 2000 by American Heart Association
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.
Wagner DR, Kubota T, Sanders VJ, McTiernan CF, Feldman AM. Differential regulation of cardiac expression of IL-6 and TNF-α by A2- and A3-adenosine receptors. Am J Physiol. 1999;276:H2141–H2147.
Krown KA, Page MT, Nguyen C, Zechner D, Gutierrez V, Comstock KL, Glembotski CC, Quintana PJE, Sabbadini RA. Tumor necrosis factor alpha-induced apoptosis in cardiac myocytes: involvement of the sphingolipid signaling cascade in cardiac cell death. J Clin Invest. 1996;98:2854–2865.
Thaik CM, Calderone A, Takahashi N, Colucci WS. Interleukin-1β modulates the growth and phenotype of neonatal rat cardiac myocytes. J Clin Invest. 1995;96:1093–1099.
Yue P, Massie BM, Simpson PC, Long CS. Cytokine expression increases in nonmyocytes from rats with postinfarction heart failure. Am J Physiol. 1998;275:H250–H258.
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.
Chung MK, Gulick TS, Rotondo RE, Schreiner GF, Lange LG. Mechanism 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.
Yokoyama T, Vaca L, Rossen RD, Durante W, Hazarika P, Mann DL. Cellular basis for the negative inotropic effects of tumor necrosis factor-α in the adult mammalian heart. J Clin Invest. 1993;92:2303–2312.
Haywood GA, Tsao PS, von der Leyen HE, Mann MJ, Keeling PJ, Trinidade PT, Lewis NP, Byrne CD, Rickenbacher PR, Bishopric NH, Cooke JP, McKenna WJ, Fowler MB. Expression of inducible nitric oxide synthase in human heart failure. Circulation. 1996;93:1087–1094.
Yamamoto S, Tsutui H, Tagawa H, Saito K, Takahashi M, Tada H, Yamamoto M, Katoh M, Egashira K, Takeshita A. Role of myocyte nitric oxide in β-adrenergic hyporesponsiveness in heart failure. Circulation. 1997;95:1111–1114.
Balligand J-L, Kelly RA, Marsden PA, Smith TW, Michel T. Control of cardiac muscle cell function by an endogenous nitric oxide signaling system. Proc Natl Acad Sci U S A. 1993;90:347–351.
Brady AJB, Warren JB, Poole-Wilson PA, Williams TJ, Harding SE. Nitric oxide attenuates cardiac myocyte contraction. Am J Physiol. 1993;265:H176–H182.
Hare JM, Loh E, Creager MA, Colucci WS. Nitric oxide inhibits the positive inotropic response to β-adrenergic stimulation in humans with left ventricular dysfunction. Circulation. 1995;92:2198–2203.
Chandrasekar B, Mitchell DA, Colston JT, Freeman GL. Regulation of CCAAT/enhancer binding protein, interleukin-6, interleukin-6 receptor, and gp130 expression during myocardial ischemia/reperfusion. Circulation. 1999;99:427–433.
Cheng XS, Shimokawa H, Momii H, Oyama J, Fukuyama N, Egashira K, Nakazawa H, Takeshita A. Role of superoxide anion in the pathogenesis of cytokine-induced myocardial dysfunction in dogs in vivo. Cardiovasc Res. 1999;42:651–659.
Irwin MW, Mak S, Mann DL, Qu R, Penninger JM, Yan A, Dawood F, Wen WH, Shou Z, Liu P. Tissue expression and immunolocalization of tumor necrosis factor-α in postinfarction dysfunctional myocardium. Circulation. 1999;99:1492–1498.
Ono K, Matsumori A, Shioi T, Furukawa Y, Sasayama S. Cytokine gene expression after myocardial infarction in rat hearts: possible implication in left ventricular remodeling. Circulation. 1998;98:149–156.
Recchia FA, McConnell PI, Bernstein RD, Vogel TR, Xu X, Hintze TH. Reduced nitric oxide production and altered myocardial metabolism during the decompensation of pacing-induced heart failure in the conscious dog. Circ Res. 1998;83:969–979.
Harding SE, Davies CH, Money-Kyrle AM, Poole-Wilson PA. An inhibitor of nitric oxide synthase does not increase contraction or β-adrenoreceptor sensitivity of ventricular myocytes from failing human heart. Cardiovasc Res. 1998;40:523–529.
Kan H, Xie Z, Finkel MS. Norepinephrine-stimulated MAP kinase activity enhances cytokine-induced NO production by rat cardiac myocytes. Am J Physiol. 1999;276:H47–H52.
Heilbrunn SM, Shah P, Bristow MR, Valantine HA, Ginsburg R, Fowler MB. Increased β-receptor density and improved hemodynamic response to catecholamine stimulation during long-term metoprolol therapy in heart failure from dilated cardiomyopathy. Circulation. 1989;79:483–490.
Hall JA, Kaumann AJ, Bown MJ. Selective β1-adrenoreceptor blockade enhances positive inotropic responses to endogenous catecholamines mediated through β2-adrenoreceptors in human atrial myocardium. Circ Res. 1990;66:1610–1623.