Role of nitric oxide and platelet-activating factor in cardiac alterations induced by tumor necrosis factor-α in the guinea-pig papillary muscle

Abstract

Objective: Tumor necrosis factor-α (TNF-α), a proinflammatory cytokine with negative inotropic properties, is implicated in several pathophysiological events. To clarify the mechanism of action of TNF-α on myocardium, we investigated the possible role of platelet-activating factor (PAF) and nitric oxide (NO) as secondary mediators of the depressant effect of this cytokine. Methods: Isometric twitches and intracellular action potentials were recorded from guinea pig papillary muscles. The effects of TNF-α (1–10 ng/ml) were studied in controlled conditions and after treatment with 0.5% Triton X-100, to destroy the endocardial endothelium. N^G-nitro-l-arginine methyl ester (l-NAME), d-NAME (1 mM) and the two different PAF-receptor antagonists WEB 2170 (3 μM) and CV 3988 (5 μM) were used to study the role of NO and PAF in cardiac depression induced by TNF-α. To study the role of NO in cardiac alterations induced by PAF, papillary muscles were pretreated with l-NAME or d-NAME and then challenged with PAF (0.1–1 μM). Nitrite production by papillary muscles challenged with TNF-α alone, TNF-α in the presence of WEB 2170 or CV 3988, or PAF was studied with the Greiss reagent method. PAF production by papillary muscles stimulated by TNF-α was studied by a bioassay method. Results: TNF-α induced an initial, transient positive inotropic effect, then reduced the contractility and the action potential duration in a concentration-dependent manner. Treatment of papillary muscle with Triton X-100 did not modify the response to TNF-α, suggesting that the effect of TNF-α is not mediated by endocardial endothelial cells. Pretreatment with indomethacin reduced the negative effect of TNF-α, while propranolol abolished the initial increase of contractility. The role of PAF and NO as mediators of TNF-α was suggested by: (1) the protective effect of l-NAME, but not of d-NAME, on electrical and mechanical alterations; (2) the stimulatory effect of TNF-α on nitrite production; (3) the inhibitory effect of WEB 2170 and CV 3988, on both the electromechanical alterations and the nitrite production; (4) the synthesis of PAF induced by TNF-α. l-NAME blocked the negative effect of PAF and PAF enhanced nitrite production by papillary muscle. Conclusions: The present results suggest that in cardiac muscle: (1) the release of PAF triggered by TNF-α may account for the stimulation of NO production; (2) both PAF and NO contribute to the development of the electrical and mechanical alterations induced by TNF-α; (3) NO production was down-stream to the synthesis of PAF.

Cardiac action potential, Cardiac contractility, Nitric oxide, Papillary muscle, Platelet-activating factor, Tumor necrosis factor-α

Time for primary review 25 days.

1 Introduction

A number of proinflammatory cytokines have been shown to modify contractile function both in isolated and in situ cardiac preparations [1]. Among these, tumor necrosis factor-α (TNF-α), a multifunctional cytokine [2], was shown to induce rapid (minutes) as well as slow (hours) effects on cardiac muscle. TNF-α was implicated in a number of pathophysiological events in which left ventricular dysfunction occurs, such as septic shock [3], acute viral myocarditis [4], cardiac allograft rejection [5], myocardial infarction [6], and congestive heart failure [7]. Recent studies indicate that human myocardium expresses both mRNA and receptor proteins for TNF-α (TNFR1 and TNFR2). Furthermore, altered levels of these receptors are present in concomitance with elevated plasma levels of TNF-α in patients with advanced heart failure [8].

Several studies were devoted to investigating the mechanisms by which TNF-α induces cardiac dysfunction, but their results are partially conflicting. Finkel et al. [9]reported that TNF-α induces a reduction of contractile force in isolated hamster papillary muscles, that was blunted by a nitric oxide synthase (NOS) blocker. A similar nitric oxide (NO)-dependent effect of TNF-α was described by Goldhaber et al. [10]on isolated rabbit ventricular myocytes. In contrast, Yokoyama et al. [11]showed that the reduction of feline myocytes shortening induced by TNF-α was not affected by pretreatment with either N^G-nitro-l-arginine or N^G-monomethyl-l-arginine. In all cases, the negative effect was rapid, and took place within a few minutes of treatment, thus excluding the induction of inducible NOS (iNOS). In contrast to these data, however, it has been recently shown that a brief treatment (30 min) with TNF-α had no effect, while a long-lasting incubation (18 h) significantly decreased contractility of isolated guinea pig myocardiocytes, suggesting that, at least in this preparation, the effects of TNF-α are mediated by de novo synthesis of a myocardial iNOS [12]. These latter authors suggested that, in the papillary muscle and in non-purified isolated cardiac cell preparations, the presence of different non-myocardial cells, such as endocardial endothelium or endothelium in microvasculature and intracardiac neurons, which are know to contain constitutive NOS (cNOS), could influence myocardial cells via a paracrine way.

In this context, it has been shown that a number of biological activities of TNF-α are mediated by platelet-activating factor (PAF). TNF-α induces synthesis and release of PAF from macrophages, polymorphonuclear neutrophils and vascular endothelial cells [13]. Similarly to TNF-α, PAF induces a rapid cardiodepressant effect in different animal species, including man [14, 15]. The aim of the present study is to investigate the role of PAF and NO as mediators of the cardiac depressant effect of TNF-α in the isolated guinea pig papillary muscle. The major finding of this study is that the negative inotropic effect of TNF-α depends on a local synthesis of PAF, which in turn promotes the production of NO.

2 Methods

2.1 Isolated papillary muscle

Experiments were performed as detailed previously [15]. The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1985). Adult guinea-pigs were anaesthetised with ether and killed by stunning and cervical dislocation, and the heart was explanted. Papillary muscles obtained from the left ventricle were perfused with Tyrode solution of the following composition, in mM: 154 NaCl, 4 KCl, 2 CaCl₂, 1 MgCl₂, 5.5 d-glucose, 5 (N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid); pH adjusted to 7.38 with NaOH, gassed with 100% O₂ at 37°C. Papillary muscles were electrically driven at a rate of 120 beats/min with a pair of electrodes connected to a 302 T Anapulse Stimulator via a 305-R Stimulus Isolator (W.P. Instruments, New Haven, CT, USA) operating in constant current mode. Isometric twitches and intracellular action potentials were evaluated by a RCA 5734 transducer tube and a floating glass microelectrode, respectively. The maximum rate of rise (+dT/dt) and fall (−dT/dt) of developed tension was obtained by electronic differentiation of the mechanogram; the action potential duration (APD) was measured at 90% repolarization level. During the experiments, the electrical and mechanical activities were continuously recorded into magnetic tape by a 3964 A Hewlett–Packard recorder (Palo Alto, CA, USA), visualised on a Tektronix 2211 digital storage oscilloscope, and reproduced for data analysis by means of a Hewlett–Packard 7470A plotter.

2.2 Experimental protocol

Papillary muscles were equilibrated in Tyrode solution for at least 40 min before each challenge. All solutions containing TNF-α (1, 5 and 10 ng/ml; Sigma, St. Louis, MO, USA) or the other drugs were prepared immediately before the experiments and were not recirculated. To study the role of endocardial endothelium in myocardial alterations induced by TNF-α, papillary muscles were immersed in Triton X-100 (0.5% in Tyrode solution) for 1–2 s, then challenged with TNF-α (10 ng/ml). Previous studies have shown that this technique completely destroys the endothelial layer within mammalian papillary muscle [9, 16]. In selected experiments, papillary muscles were pretreated with propranolol (0.2 μM; Sigma) or indomethacin (1 μM; Sigma) before challenge with TNF-α (10 ng/ml) in order to block β-adrenergic response or cyclooxygenase, respectively. N^G-nitro-l-arginine methyl ester (l-NAME, 1 mM; Sigma) was applied for 2 h before challenge to block the synthesis of NO [17]. The biologically inactive enantiomer of l-NAME, d-NAME (1 mM; Sigma), was used as control. WEB 2170 (3 μM; Boehringer Ingelheim, Germany) and CV 3988 (5 μM; Takeda Chemical Industries, Kyoto, Japan), two chemically unrelated PAF-receptor antagonists [18, 19], were used to block PAF receptors; WEB 2170 and CV 3988 were administered to papillary muscles starting 15 min before, and during all the period of treatment with TNF-α (5 and 10 ng/ml). PAF (Bachem Feinchemikalien, Bubendorf, Switzerland) was first dissolved in physiological solution containing bovine serum albumin, then the appropriate aliquots of the stock solution were added to the Tyrode solution to reach the concentrations of 0.1 and 1 μM. Treatment with TNF-α or PAF lasted 30 min, then the perfusion was switched to control, drug-free Tyrode solution, to study the reversibility of the effect.

2.3 Nitrite assay

To measure nitrite production by isolated papillary muscles, small aliquots (150 μl) of perfusate were mixed with a equal volume of 1% sulfanilamide–1% N-(1-naphthyl)ethylenediamine dihydrochloride in 2% phosphoric-free acid (Greiss acid) at room temperature for 10 min [20]. Nitrite concentrations were calculated by comparison with optical density of standard solution of sodium nitrite prepared in Tyrode solution. All values were background corrected for nitrite values obtained in non-conditioned Tyrode solution. Optical density at 550 nm was measured using a Microplate Reader Model 450 (Bio-Rad, Hercules, CA, USA). To assess the effects of TNF-α and the role of PAF on the synthesis of NO, we measured nitrite production from papillary muscles challenged with 10 ng/ml TNF-α, without or after pretreatment with WEB 2170 (3 μM) or CV 3988 (5 μM); further experiments were performed in papillary muscles challenged with 1 μM PAF.

2.4 PAF assay

PAF extracted and purified from papillary muscles was quantified by bioassay on washed rabbit platelets [13]. PAF bioactivity, tested after extraction and purification by thin layer chromatography (TLC) and HPLC, was characterised by comparison with synthetic PAF according to the following criteria: induction of platelet aggregation by a pathway independent of both ADP and arachidonic acid/thromboxane A₂; specificity of platelet aggregation as inferred from the inhibitory effect of PAF-receptor antagonist WEB 2170 (5 μM); TLC and HPLC chromatographic behaviour and physicochemical characteristics such as inactivation by strong bases and 5 min heating in boiling water. The methods used have been described in detail [21].

2.5 Statistical analysis

Data are expressed as the mean±S.E. The experimental groups were compared using one-way analysis of variance. If a significant F-value resulted from the analysis of variance (P<0.05), the Newman–Keuls’ multiple range test was applied to determine where differences were located among the groups.

3 Results

3.1 Effect of TNF-α on myocardial performance

As shown in Figs. 1 and 2, perfusion of papillary muscles with Tyrode solution containing TNF-α (1, 5, 10 ng/ml) induced concentration-dependent effects on the electrical and mechanical activities. While 1 ng/ml TNF-α caused no significant alteration, the higher concentrations of TNF-α induced a biphasic effect, characterised by a transient increase of contractile force (T_max), which reached 115.2±5.7% of the control value (Fig. 1), of the maximum rate of rise (+dT/dt) and fall (−dT/dt) of developed tension, followed by a slow decrease of contractility (Figs. 1 and 2). The initial positive inotropic effect of TNF-α lasted a few (3–5) minutes, thereafter T_max progressively decreased, until it reached a stable level 25 min after the challenge. This effect was already evident at 5 ng/ml, and became highly significant at 10 ng/ml TNF-α. In contrast, the time to peak tension was not modified by TNF-α (not shown). The observed reduction of contractility was accompanied by a concentration-dependent reduction of the APD (Figs. 1 and 2). The resting membrane potential (E_r), the overshoot (Os) and the maximum rate of depolarisation (dV/dt max) of the action potential, however, were not affected by TNF-α (Table 1). An almost complete recovery of mechanical and electrical properties of papillary muscles was observed 20–30 min after the perfusion was switched to the control TNF-α-free Tyrode solution (T_max=92.8±5.1; +dT/dt=96.5±4.2; −dT/dt=96.3±2.2; APD=100.3±1.3% of the control, prechallenge value).

Fig. 2

Mechanical and electrical effects of a 30 min treatment with TNF-α (1, 5 and 10 ng/ml; n=8 for each concentration) in the isolated papillary muscle. To study the role of endothelial cells, four papillary muscles were treated with 0.5% Triton X-100 for 1 s, then challenged with 10 ng/ml TNF-α. Data are expressed as the mean±S.E.% of the control, prechallenge value. When the role of endocardial endothelium was studied, we assumed as control (100%) the electrical and mechanical parameters of papillary muscles treated with 0.5% Triton X-100 and allowed to stabilize (15–20 min). Statistical significance: * p<0.05; ** p<0.01. No significant difference was found between control papillary muscles and papillary muscles treated with Triton X-100 after the challenge with TNF-α. Baseline control values were: peak tension (T_max)=245±2.7 mg; maximum rate of rise of developed tension (+dT/dt)=4.1±0.2 g/s; maximum rate of fall of developed tension (−dT/dt)=3.3±0.1 g/s; action potential duration, measured at 90% of repolarization, (APD)=238±9.1 ms.

Fig. 1

Effects of TNF-α (10 ng/ml) on contractile force (T_max) and action potential duration (APD) in the isolated guinea pig papillary muscle. The early (5 min), transient increase, was followed by a slow decrease in contractility and APD. Data represent the mean±S.E.% of the control, prechallenge value (n=8). Statistical significance: * p<0.05; ** p<0.01. Baseline control values were: peak tension (T_max)=230±3.6 mg; action potential duration (APD)=227±11.3 ms.

Table 1

Electrical effects of TNF-α (1, 5 and 10 ng/ml; n=8 for each concentration of TNF-α) in the isolated papillary muscle

	Er	Os	dV/dt
Control	84.4±0.6	32.1±4.3	134.8±2.3
TNF-α 1 ng/ml	87.4±1.5	32.1±1.0	123.7±4.2
TNF-α 5 ng/ml	86.9±0.8	32.8±0.6	145.2±3.6
TNF-α 10 ng/ml	85.5±0.9	32.7±0.6	137.4±2.9

E_r=resting membrane potential (mV); Os=overshoot (mV); dV/dt_max=maximum rate of depolarisation (V/s). All values were expressed as the mean±S.E of at least six observations for each experiment.

3.2 Role of endocardial endothelium

Treatment of papillary muscles with Triton X-100 to remove the endocardial endothelium [16]reduced cardiac contractility to about 40%. After treatment with Triton X-100, the negative inotropic effect and the related alterations of contraction and action potential induced by 10 ng/ml TNF-α were not significantly different from those induced in untreated papillary muscles (Fig. 2).

3.3 Role of β-adrenoreceptors and cyclooxygenase

Pretreatment with propranolol (0.2 μM; n=5) completely prevented the early positive inotropism, but had no effect on the reduction of contractile force and APD induced by 10 ng/ml TNF-α (T_max=59.5±3.2; +dT/dt=52.5±3.2; −dT/dt=55.0±4.9; APD=83.0±2.1% of the control, prechallenge value). Indomethacin (1 μM; n=5) did not influence the early positive inotropic effect (T_max=110.5±1.7), but partially prevented the negative effect on inotropism and APD (T_max=85.5±2.1; +dT/dt=84.5±1.2; −dT/dt=87.3±3.5; APD=90.1±2.3% of the control, prechallenge value) induced by 10 ng/ml TNF-α.

3.4 Role of NO in cardiac alterations induced by TNF-α

In order to study the role of NO as mediator of the effects of TNF-α, papillary muscles were pretreated with the NOS inhibitor l-NAME (1 mM) or with its biologically inactive enantiomer d-NAME (1 mM) for 2 h, and then challenged with TNF-α (5 and 10 ng/ml). The incubation of papillary muscles with l-NAME had no effect per se. However, when these papillary muscles were challenged with TNF-α, l-NAME completely blocked the negative effects of TNF-α on T_max, +dT/dt, −dT/dt and APD. In contrast, pretreatment with d-NAME did not alter the electrical and mechanical responses of papillary muscle to TNF-α (Fig. 3).

Fig. 3

Role of NO on cardiac alterations induced by TNF-α. The effects caused by a 30 min treatment with 5 and 10 ng/ml TNF-α in control papillary muscles (n=8 for each concentration) are compared with those induced in papillary muscles pretreated with l-NAME (1 mM; n=4) or d-NAME (1 mM; n=4). Data are expressed as the mean±S.E.% of the values recorded in control conditions or after treatment with l-NAME or d-NAME, respectively. To perform statistical analysis, the effects of TNF-α on untreated papillary muscles were compared with those induced after pretreatment with l-NAME or d-NAME (* p<0.05; ** p<0.01). Baseline values for these latter groups of papillary muscles were comparable to those reported in the legend for Fig. 2.

3.5 Role of PAF in cardiac alterations induced by TNF-α

Several studies indicate that TNF-α may stimulate the synthesis and release of PAF by various cell types [13, 22]. Therefore, we tested the possibility that PAF mediates some of the cardiac effects of TNF-α. For this purpose, papillary muscles were incubated with WEB 2170 (3 μM) or CV 3988 (5 μM), two chemically unrelated PAF-receptor antagonists, 15 min before the treatment with TNF-α (5 and 10 ng/ml). These concentrations of WEB 2170 and CV 3988, tested on the isolated papillary muscle, did not induce any significant mechanical and electrical changes (data not shown). In the presence of WEB 2170 or CV 3988, TNF-α induced a positive inotropic effect, which was similar to that observed in control papillary muscles (135.0±14.1% of the control value), but was persistent. In contrast, the reduction of contractility as well as the shortening of the APD induced by TNF-α were completely blocked (Fig. 4).

Fig. 4

Role of PAF on cardiac alterations induced by TNF-α. The effects caused by a 30 min treatment with 5 and 10 ng/ml TNF-α in control papillary muscles (n=8 for each concentration) are compared with those induced in papillary muscles pretreated with WEB 2170 (3 μM; n=8) or CV 3988 (5 μM; n=8). Data are expressed as the mean±S.E.% of the values recorded in control conditions or after treatment with WEB 2170 or CV 3988, respectively. To perform statistical analysis, the effects of TNF-α on untreated papillary muscles were compared with those induced after pretreatment with WEB 2170 or CV 3988 (* p<0.05; ** p<0.01). Baseline values for these latter groups of papillary muscles were comparable to those reported in the legend for Fig. 2.

3.6 Relationship between PAF and NO

These results indicate that both the generation of NO and production of PAF contribute to the development of the electrical and mechanical alterations induced by TNF-α. To evaluate whether a relationship exists between PAF and NO production, we compared the effects of PAF administration to control papillary muscles with those caused by PAF in papillary muscles pretreated with l-NAME (1 mM) or d-NAME (1 mM). In control experiments, PAF (0.1 and 1 μM) induced a biphasic effect, with a transient, initial increase of contractile force (T_max=122.4±1.8% of the control value at 3 min), followed by a concentration-dependent reduction of T_max, +dT/dt and −dT/dt, as well as of APD (Fig. 5), whereas E_r, Os and dV/dt_max of the action potential parameters were not significantly affected (Table 2). The time-course of these effects was similar to that induced by TNF-α.

Fig. 5

Role of NO on cardiac alterations induced by PAF. The effects caused by a 30 min treatment with 0.1 and 1 μM PAF in control papillary muscles (n=8 for each concentration) are compared with those induced in papillary muscles pretreated with l-NAME (1 mM; n=4) or d-NAME (1 mM; n=4). Data are expressed as the mean±S.E.% of the values recorded in control conditions or after treatment with l-NAME or d-NAME, respectively. To perform statistical analysis, the effects of PAF in untreated papillary muscles were compared with those induced after pretreatment with l-NAME or d-NAME (* p<0.05; ** p<0.01). Baseline values for these papillary muscles were comparable to those reported in the legend for Fig. 2.

Table 2

Electrical effects of PAF (0.1 and 1 μM; n=8 for each concentration of PAF) in the isolated papillary muscle

	Er	Os	dV/dt
Control	76.1±1.0	27.8±0.8	92.3±2.8
PAF 0.1 μM	75.6±1.6	28.2±1.4	98.7±6.6
PAF 1 μM	78.2±1.9	28.6±0.7	96.2±6.2

E_r=resting membrane potential (mV); Os=overshoot (mV); dV/dt_max=maximum rate of depolarisation (V/s). All values were expressed as the mean±S.E. of at least six observations for each experiment.

When PAF was administered to papillary muscles pretreated for 2 h with l-NAME (1 mM), the transient positive inotropic effect persisted, whereas the reduction of contractility and the shortening of the action potential were abrogated. d-NAME completely failed to protect papillary muscles against the negative effect of PAF (Fig. 5).

3.7 Nitrite production induced by TNF-α and PAF

TNF-α (10 ng/ml) enhanced the basal nitrite production; the maximal increase (51.5±14.0% over the control value) was recorded 3–5 min after the challenge with TNF-α. This effect was completely abrogated by pretreatment of papillary muscles with the PAF-receptor antagonists WEB 2170 or CV 3988 (Fig. 6).

Fig. 6

Nitrite production from papillary muscles induced by TNF-α (10 ng/ml), PAF (1 μM) and TNF-α in the presence of the PAF-receptor antagonists WEB 2170 (3 μM) or CV 3988 (5 μM) (n=4 for each group; ** p<0.01). Data are mean±S.E.% of the values recorded in control conditions or after treatment with WEB 2170 or CV 3988, respectively. Baseline nitrite concentration in the perfusate was 541±13 nM.

A significant amount (20.4±11.6 ng/g of tissue) of PAF was extracted from papillary muscles treated with 10 ng/ml TNF-α, but not from untreated controls. These results suggest that a local production of PAF triggered by TNF-α may account, at least in part, for the stimulation of NO production. Indeed, similarly to TNF-α, PAF (1 μM) was found to stimulate NO production from papillary muscle (Fig. 6).

4 Discussion

Our study demonstrates that, in the isolated guinea-pig papillary muscle, the reduction of contractility and of APD induced by TNF-α are mediated both by PAF and NO. Moreover, the production of NO induced by TNF-α is consequent to the production of PAF. PAF was indeed shown to induce biphasic, concentration-dependent effects on myocardial performance [15, 23]. Previous studies have shown that NO reduces myocardial contractility [24, 25]and ONOO⁻ impairs the myocardial muscle respiration [26]. A reduction of transmembranal calcium current was invoked to explain the shortening of the APD and the reduced contractility induced by NO [27]and PAF [23, 28]. The following results support the role of PAF as mediators of the negative inotropic effect of TNF-α: (1) the blocking effect of WEB 2170 and CV 3988, two chemically different PAF-receptor antagonists; (2) the synthesis of PAF within the cardiac muscle triggered by TNF-α. The present study does not provide conclusive results on the cell origin of PAF synthesised by isolated guinea-pig papillary muscle. However, isolated cardiac cells [29]as well as macrophages and endothelial cells [13, 22]have been shown to be able to synthesize PAF and are therefore candidates for the production of this mediator in our experimental conditions. The evidence that the biological effects of TNF-α on the cardiac muscle persisted despite the destruction of endothelium tended to exclude this source.

Our observation that pretreatment with indomethacin attenuated the effects of TNF-α on cardiac contractility is in agreement with in vivo and in vitro studies, showing that pretreatment with a cyclooxygenase inhibitor abolishes many of the cardiovascular changes induced by sublethal doses of TNF-α [30], and that indomethacin reduces the effects of TNF-α on skeletal muscle contractility [31]. Furthermore, previous studies in our laboratory demonstrated that pretreatment with indomethacin significantly lessens the effects of PAF on the isolated papillary muscle [15]. The observation that pretreatment of papillary muscles with propranolol prevented the early positive inotropism induced by TNF-α is in accordance with previous studies indicating that the positive inotropic effect induced by PAF in guinea-pig [23]and in human [15]papillary muscle, as well as in isolated rat atria [32]is blocked by this drug. This finding suggests that the transient increase of contractility induced by TNF-α is consequent to a PAF-induced release of endogenous catecholamines within cardiac tissue.

PAF was found to act as a secondary mediator of TNF-α in several other experimental conditions [22], including its angiogenic effects [21]. On the other hand, several lines of evidence indicate that the negative inotropic effects of TNF-α depends on the generation of NO [1]. However, the available results are conflicting; in fact, certain studies indicate the involvement of cNOS [9, 10], whereas other studies suggest the involvement of iNOS [12]. The results of the present study indicate that TNF-α induces an early production of NO, as detected as nitrate generation. This evidence suggests an involvement of cNOS, rather than iNOS, in TNF-α triggered NO production. Previous studies have shown that cardiac myocytes express cNOS and that cNOS mediates the NO-dependent parasympathetic signaling [33]. However, in pathological conditions, such as septic shock, a persistent production of TNF-α and PAF occurs; therefore, it is possible that in these conditions iNOS contributes to the NO generation. Indeed, in experimentally induced endotoxin septic shock in rats, an activation of iNOS was observed [34]. Moreover, we confirm that NO generation is critical in mediating the negative inotropic effect of TNF-α, as l-NAME, but not d-NAME, prevented both the mechanical and the electrical alterations triggered by TNF-α. The new finding of the present study is that the production of NO triggered by TNF-α is dependent on the synthesis of PAF. Indeed, PAF-receptor antagonists inhibit generation of nitrite triggered by TNF-α and PAF directly stimulates the production of nitrite by the papillary muscle, mimicking the electrical and mechanical alterations induced by TNF-α. Thus, TNF-α seems to induce NO generation through the synthesis of PAF rather than directly. These results are in agreement with the finding that also the angiogenic effect of TNF-α depends on the production of PAF and on PAF-induced NO generation [35]. Previous studies have shown that PAF contributes to the induction of NOS by bacterial lipopolysaccharides (LPS). Indeed, it has been shown that PAF-receptor antagonists inhibit the induction of calcium-independent NOS in the lungs of rats treated with LPS, but do not interfere with the in vitro activity of the enzyme [34]. These experiments suggest that the synthesis of PAF induced by LPS stimulates a subsequent production of NO. Moreover, it has been shown that the vasoactive and hypotensive effects of PAF are dependent on NO generation [34, 36]. Several studies suggest that TNF-α and PAF have a critical role in endotoxin/septic shock [34, 37]. The myocardial infarction is another pathological condition in which the role of TNF-α and PAF has been studied [6, 38–40]. In these clinical settings, the maldistributive shock due to peripheral vasodilatation coexists with a central depression of the cardiac function. The results of the present study further support a link between TNF-α and PAF in the development of these pathological conditions. Moreover, our findings indicate that NO acts as a final mediator for a PAF-dependent stimulus, such as TNF-α.

Acknowledgements

This study was supported by grants of the Ministero dell’Università e della Ricerca Scientifica (MURST) to GA and GC, AIRC, Consiglio Nazionale delle Ricerche (CNR), Targeted Project on Biotechnology and Istituto Superiore di Sanità AIDS project n.30A.0.09 to GC.