Discussion

The present study demonstrates that the NO donor SNP exerts a significant negative inotropic effect on human myocardium. SNP led to a concentration-dependent decrease in basal, isoprenaline- or IBMX-enhanced isometric FOC in isolated atrial trabecula. For the first time, it was demonstrated that this effect holds true also for left ventricular human myocardium. The effects of SNP were similar to those of the stable cGMP analog, 8-bromo-cGMP, which also had a sustained negative inotropic effect on human atrial trabecula. In accordance with previous observations (Brady et al., 1993), this effect developed more slowly than the effect of SNP, which might be because the second messenger penetrates the myocardial tissue more slowly than NO itself. Experiments were performed on isolated atrial trabecula or papillary muscle preparations. In contrast to experiments in vivo or with isolated hearts, changes in coronary blood flow and peripheral vascular resistance by NO or 8-bromo-cGMP were not modulators for the FOC in this experimental setting. Therefore, it can be concluded that NO and cGMP directly depress the cardiac FOC. The observations that the guanylyl cyclase inhibitor methylene blue attenuated the effect of SNP and that exposure of human atrial trabecula to SNP caused an increase in myocardial cGMP content suggest that the negative inotropic effect of SNP and thereby NO is mediatedvia activation of guanylyl cyclase and generation of cGMP.

Our observation that SNP decreases isometric force development in human myocardial preparations is in accordance to previous investigations in other species which also demonstrated a negative inotropic effect of NO and cGMP. Brady et al. (1993) observed in isolated guinea pig cardiac myocytes that length shortening during contraction was attenuated by NO released from the endothelium or by SNP and 8-bromo-cGMP by 20% and that the effect of SNP could be reversed by methylene blue. In contrast to our experimental setting, which does not allow differentiation between the direct effects of NO and 8-bromo-cGMP on cardiac myocytes and those effects which are mediated by other myocardial cell types, the experiments by Brady et al. (1993) demonstrated that both substances directly affect myocyte contractility. Similar effects were observed when the effect of NO and 8-bromo-cGMP on isolated ferret papillary muscle were determined (Smithet al., 1991; Shah et al., 1991). In isolated ejecting hearts of the ferret, the NO-releasing substance bradykinin caused a significant reduction in peak left ventricular pressure (Fort and Lewis, 1993). In humans, it could be demonstrated that intracoronary infusion of SNP suppressed left ventricular pressure development (Paulus et al., 1994), whereas intracoronaryl-NMMA infusion enhanced the positive inotropic response to dobutamine (Hare et al., 1995).

On the other hand, Nawrath et al. (1995) did not observe a negative inotropic effect of various NO donors, including SNP, in isolated atrial muscle strips from various species, including man. Similarly, Weyrich et al. (1994) reported that various NO donors failed to produce any inotropic effect on isolated papillary muscle preparations from cats and rats. Only in the presence of a high concentration of norepinephrine, an unphysiologically high concentration of NO exerted a significant decrease in FOC (Weyrichet al., 1994). The strength of this study was that release of NO into the organ bath was proved by determination of the NO concentration. Therefore, one cannot simply explain the difference between these and our findings with an inappropriate NO donor. This could have been concluded from the findings of Brady et al. (1993), who observed that organic nitrates are possibly not metabolized by mammalian ventricular muscle. Because Nawrath et al.(1995) also performed experiments on human myocardium, one can also not reduce contradictory findings to species differences. Possible explanations might be differences in the specific handling of the myocardium, e.g., tissue preparation in prewarmed Tyrode’s solution might have caused damage to the endothelium, which has been shown to be of distinct relevance for the inotropic effect of NO (Mohan et al., 1996). Also, it has been suggested that effects of NO differ in atrial and ventricular myocardium as has been shown for acetylcholine, which possibly exerts its negative inotropic effect by endothelial release of NO (Endoh and Yashimita, 1981). According to our data, this does not seem to be the case for the effect of NO in humans.

Of special interest is a recent study by Mohan et al. (1996)concerning the inotropic effect of different NO donors and 8-bromo-cGMP on isolated cat papillary muscles. This study suggests that NO and 8-bromo-cGMP exert a positive inotropic effect at low concentrations and a negative inotropic effect at high concentrations. Within the same concentration range and with a very similar experimental setup, a positive inotropic effect of SNP and 8-bromo-cGMP was not observed at either low or high concentrations in human myocardium.

One previously reported effect of NO donors and of 8-bromo-cGMP on myocardial contractile performance is a shortening of the contractile twitch with a hastening of myocardial relaxation. In the present study, SNP and 8-bromo-cGMP caused a slight shortening of time to peak tension with an earlier onset of relaxation. Maximum contraction velocities were not altered compared with controls. This observation is in accordance with the in vivo data in humans of Paulus et al. (1994), who observed a decrease in left ventricular electromechanical systole time because of an earlier onset of left ventricular isovolumic relaxation during intracoronary SNP infusion. In this study, SNP even caused a decrease in the absolute value for the velocity of pressure decrease. In the guinea pig isolated heart experiments of Grocott-Mason et al. (1994), left ventricular pressure decline was not only premature but also accelerated, at least in early diastole after SNP infusion. Fort and Lewis (1993) also described a significant reduction in the velocity of pressure decline during bradykinin infusion of isolated ejecting ferret hearts. However, this effect was independent of the absence or presence of a NOS inhibitor and thus possibly unrelated to NO. Taken together, there seem to be slight differences concerning changes in the time-dependent variables of the contractile cycle caused by NO donors, which might depend on the species chosen for the experiments and the experimental setting (in vivo, isolated ejecting heart, isolated atrial/papillary muscle experiments). However, the concurring result of all these studies is that NO, besides a decrease in peak force or pressure development, leads to an abbreviation of the contractile cycle.

Data concerning the effect of cGMP on the time course of the contractile cycle are very similar. This is important if one wants to strengthen the hypothesis that the effects of NO on cardiac contractility are mediated by cGMP. Shah et al. (1991)reported that in isolated ferret papillary muscle, 8-bromo-cGMP within a concentration range similar to that in our experiments reduced time to peak tension to a similar extent as peak developed tension. This led to an earlier onset of mechanical relaxation and a reduction of overall twitch duration. Similarly, a reduction in myocyte twitch amplitude and time to peak shortening by roughly 20% without changes in shortening velocity was observed in isolated adult rat ventricular myocytes after exposure to 8-bromo-cGMP (Shah et al., 1994). In this study, the effects of cGMP on time-dependent variables resembled very much those of SNP, which suggested an identical negative inotropic mechanism. On the other hand, in accordance with earlier observations in ferret cardiac muscle (Shah et al., 1991), 8-bromo-cGMP led to a significant decrease in half-maximal relaxation time in human atrial trabecula in this study. This effect was not observed after exposure of atrial trabecula to SNP. One might speculate whether the effect of cGMP on relaxation time is concentration-dependent and whether the observed discrepancy between 8-bromo-cGMP and SNP is caused by the fact that the effect of 100 μmol/l SNP can not be compared directly with the effect of 100 μmol/l 8-bromo-cGMP. These two substances do not depend on each other in a stoichiometric way; therefore, concentrations of both substances are difficult to compare. This might also explain why the negative inotropic effect of 8-bromo-cGMP was much more pronounced than the effect of an equimolar concentration of SNP. On the other hand, differences between cGMP and NO donors concerning their effects on the time course of the contractile twitch might also occur, because NO has other effects in addition to stimulation of guanylyl cyclase.

The results of myocardial cGMP quantification support the theory that NO effects on cardiac contractile performance are mediatedvia guanylyl cyclase activation and generation of cGMP. Not only was there a remarkable increase in myocardial cGMP concentrations after exposure to SNP, which proves that NO leads to an activation of the guanylyl cyclase in human myocardium, but also the effects of SNP on atrial and ventricular contractility and on myocardial cGMP concentrations could both be attenuated by the guanylyl cyclase inhibitor methylene blue. This finding is in agreement with observations in other studies (Brady et al., 1993; Mohanet al., 1996). Interestingly, the effect of SNP on both FOC and myocardial cGMP concentrations could not be completely antagonized by methylene blue. This might indicate that mechanisms other than activation of the guanylyl cyclase-cGMP pathway are involved as suggested previously (Stamler et al., 1993). These mechanisms include inhibition of enzyme activity by S-nitrosylation (as has been shown for cathepsin B, aldolase, γ-glutamylcysteinyl synthetase, alcohol and aldehyde dehydrogenases and glyceraldehyde-3-phosphate dehydrogenase) and by the binding of NO to Fe(II) or Fe(III) in heme groups (as shown for cytochrome P-450 or for NOS itself) (Stamler, 1994). Indeed, NO has led to increased superoxide anion (O₂ ⁻) and superperoxide (H₂O₂) production in rat heart mitochondria (Poderoso et al., 1996). Also, experiments with rat heart mitochondria preparations have demonstrated that NO interferes with mitochondrial electron transport either directly or indirectlyvia formation of peroxynitrite (Cassina and Radi, 1996). That another substance released from SNP is responsible for the remaining negative inotropic effect in the presence of methylene blue seems to be rather unlikely, because potassium cyanide, which is released from SNP together with NO and leads to an inhibition of oxidative phosphorylation of mitochondria, showed no lasting effect on atrial trabecula contractility.

There has been an ongoing discussion whether cGMP and cGMP-elevating substances affect basal contractility or only contractility in the presence of cAMP-elevating substances. Latter findings supported the theory that cGMP influenced cardiac contractility by opposing the effect of cAMP (Goldberg et al., 1975). In this study, the effect of SNP on atrial trabecula contractility was slightly, but not significantly more pronounced in the presence of isoprenaline and IBMX. Interestingly, in the ventricular myocardium, this difference between the negative inotropic effect of SNP in the absence or presence of cAMP-increasing substances was significant (P < .05) for papillary muscle preparations from nonfailing, but not from failing hearts. This observation might be from the fact that the generation of cAMP in response to isoprenaline is reduced because of the well-characterized desensitization of the adenylyl cyclase (Bristowet al., 1982; Böhm et al., 1990). Thus, the negative inotropic effect of NO, at least in ventricular myocardium, seems to be stronger in, but not dependent on the presence of elevated cAMP levels.

The precise mechanism by which NO and cGMP could affect cardiac contractility has not been completely understood so far (for review, see Lohmann et al., 1991). There is evidence from experiments with myocytes of different species that cGMP decreases I_Ca (Nawrath, 1977; Levi et al.; 1989;Méry et al., 1991). In contrast to the functional data obtained in this study, in these electrophysiological studies, cGMP inhibited only cAMP-stimulated, but not basal I_Ca (Lohmannet al., 1977). This decrease in I_Ca might be caused by a stimulation of cAMP hydrolysis by cGMP (Endoh, 1979;Flitney and Singh, 1981; Walter, 1984). Evidence for the latter hypothesis comes from the observation that IBMX largely reverses the inhibitory effect of cGMP on cAMP-stimulated I_Ca (Hartzell and Fischmeister, 1986). Again, this is in contrast to our finding that in the presence of IBMX the effect of SNP on atrial contractility was even more pronounced. Another mechanism leading to a decrease in I_Ca independent from cAMP has to be taken into account,e.g., modulation of L-type Ca⁺⁺ -channel activity by cGMP-dependent protein kinase (Levi et al., 1989; Thakkar et al., 1988; Méry et al., 1991). In contrast to these considerations, Kirstein et al. (1995) demonstrated that the NO donor SIN-1 within a concentration range of 1 pmol/l to 10 nmol/l had a stimulatory effect on I_Ca in isolated human atrial myocytes, which was found to be caused by cGMP-induced inhibition of cGMP-inhibited phosphodiesterase. I_Ca was suppressed only at higher concentrations. In the present study, the lowest SNP concentration studied was 10 nmol/l. A negative inotropic effect started at a concentration of 100 nmol/l and higher. Thus, it might be possible that the stimulatory effect of NO/cGMP on I_Ca might have been missed in our experiments. A third mechanism explaining the negative inotropic effect of cGMP has been demonstrated only recently by Shahet al. (1994). They observed that 8-bromo-cGMP reduces the myofilament response to Ca⁺⁺ in intact adult ventricular myocytes. Most likely, this effect is mediated viacGMP-dependent protein kinase. This latter mechanism could explain why, despite an increase in I_Ca and low NO concentrations, there is no increase in FOC.

The physiological relevance of the negative inotropic effect of NO has repeatedly been a matter of controversy. Obviously, even at high SNP concentrations (100 μmol/l), the reduction in contractile force was significantly smaller than the decrease in contractility caused by other negative inotropic substances such as carbachol or R-PIA. However, an acute reduction in peak force development by roughly 10 to 20% under basal conditions and in the presence of cAMP-elevating positive inotropic substances is not negligible. Also, the results of the present study indicate that a significant negative inotropic effect occurs already at lower concentrations of SNP, which indicates that physiological concentrations of NO also decrease cardiac contractility. So far, in vivo data on NO concentrations within the coronary system or within the heart under different pathological conditions are missing. However, there are data from healthy volunteers (Gilmer et al., 1996) and from women with preeclampsia (Lyell et al., 1995) on serum nitrate or nitrite concentrations, respectively, which both reflect serum NO levels. Both studies indicate that under physiological as well as pathological conditions, NO serum concentrations are within a concentration range similar to the negative inotropic SNP concentrations used in this study. Recent findings of an induction of NOS and an increase in NOS activity in congestive heart failure (de Belder et al., 1993; Haywood et al., 1996), septic cardiomyopathy (Thoeneset al., 1996) or human cardiac allografts (Lewis et al., 1996) and the observation that there is a relation between iNOS expression or activity and contractile dysfunction (Lewis et al., 1996), also provide strong evidence that the observedin vitro effects also play a role in modulation of cardiac contractility in vivo and are of important clinical relevance in a variety of cardiovascular disorders. In accordance, inhibition of NO release with use of the NO synthase inhibitorl-NMMA has already been tested as a new therapeutic approach in human septic shock (Petrous et al., 1994). Unfortunately, despite beneficial effects, l-NMMA did not lead to an increase but to a decline in cardiac output in this study, which might be because of an impairment of myocardial perfusion caused by decreased NO release.

In summary, the results in this study show that NO viastimulation of guanylyl cyclase and generation of cGMP exerts a negative inotropic effect on human atrial and ventricular myocardium. This inotropic effect of NO might be of functional relevance for the control of cardiac contractility by the endothelium. The negative inotropic potency of NO might be especially important in pathological conditions, when NO synthase activity is increased and high concentrations of NO are generated within the myocardium. Control of NO release, e.g., by inhibitors of NO synthase, might be an important therapeutic target in cardiovascular disorders such as septic cardiomyopathy, congestive heart failure after myocarditis or cardiac allograft rejection.