Human myocardial tissue

Experiments were performed on left ventricular myocardium from 18 failing (5 ischemic cardiomyopathy, 13 dilative cardiomyopathy) and 5 nonfailing human hearts obtained during cardiac transplantation. Nonfailing hearts were obtained from organ donors. These hearts were used for pulmonary and aortic valve allograft preparation because there was no appropriate recipient available for cardiac transplantation. Mean age of organ donors (3 men, 2 women) was 34.4 ± 7 years; mean age of organ recipients (16 men, 2 women) was 46.8 ± 3 years. The mean ejection fraction of organ recipients before operation was 24.4 ± 1%. Organ donors and recipients had no signs of bacterial infection or sepsis. Medical therapy of the patients with end-stage heart failure consisted of cardiac glycosides, diuretics, nitrates and angiotensin-converting enzyme (ACE) inhibitors. General anesthesia was performed with flunitrazepam, fentanyl and pancuronium bromide with isoflurane. Cardiac surgery was performed while on cardiopulmonary bypass. Only noninfarcted tissue was used, and scars were carefully trimmed away. Tissue pieces were suspended in ice-cold cardioplegic solution (modified Bretschneider solution containing (in mmol/liter): NaCl 15, KCl 10, MgCl₂4, histidine HCl 180, tryptophan 2, mannitol 30 and potassium dihydrogen oxyglutarate 1) and were delivered immediately from the operation room to the laboratory.

Isolated left ventricular muscle strip preparation and incubation conditions

Left ventricular muscle strips of uniform size with muscle fibers running approximately parallel to the length of the strips were dissected under microscopic control using scissors in ice-cold aerated modified Tyrode solution (composition below). Connective tissue was carefully trimmed away. The muscles were suspended in an organ bath (75 ml) maintained at 37°C and containing a modified Tyrode solution of the following composition (in mmol/liter): NaCl 119.8; KCl 5.4; MgCl₂1.05; CaCl₂1.8; NaHCO₃22.6; NaH₂PO₄0.42; glucose 5.0; ascorbic acid 0.28; EDTA 0.05. The bathing solution was continuously aerated with 95% O₂and 5% CO₂. Tyrode solution was supplemented with either lipopolysaccharide (LPS) from Salmonella typhosa(1 μg/ml), with endotoxin plus N^G-mono-methyl-l-arginine (l-NMMA, 100 μmol/liter), with endotoxin plus hemoglobin (2 μg/ml) and with endotoxin plus 4,5-dihydroxy-1,3-benzene disulfonic acid (Tiron, 10 mmol/liter) or was used without supplementation for control experiments. Under these conditions, left ventricular myocardial preparations were either incubated for 6 or 12 h. Immediately after incubation, some of the ventricular muscle preparations were used for functional experiments, whereas others were frozen by rapid immersion into liquid nitrogen and stored at −80°C.

For control purposes, microscopic analysis of the incubated muscle preparations was performed on hematoxylin-eosin and Luxol fast blue stained cryostat sections. In general, histological examination revealed the microscopic features as expected from the preexisting myocardial disorder (e.g., marked hypertrophy of the cardiomyocytes of patients with dilated cardiomyopathy). There was no morphologic evidence of acute ischemic or toxic myocardial injury or structural alterations in any of the tissue pieces independent from whether myocardial preparations were incubated in unsupplemented Tyrode solution or in Tyrode solution supplemented with endotoxin, endotoxin plus l-NMMA or endotoxin plus Tiron.

Determination of force of contraction

For determination of isometric force of contraction, myocardial preparations were stimulated by two platinum electrodes using field stimulation from a Grass S88 stimulator (Grass, Quincy, Massachusetts; frequency 1 Hz; impulse duration 5 ms; intensity 10% to 20% greater than threshold). Experiments were performed in Tyrode solution at 37°C. Each muscle was stretched to the length at which force of contraction was maximal. Isometric tension was measured with an inductive force transducer (W. Fleck, Mainz, Germany) attached to a Gould recorder (Gould, Cleveland, Ohio). Preparations were allowed to equilibrate for at least 90 min, with the bathing solution being changed once after about 45 min.

Detection of iNOS mRNA by in situ hybridization

For histological localization of iNOS by in situ hybridization, 6-μm thin cryosections of the pretreated myocardium were cut and mounted on slides that had been sterilized and pretreated with 3-aminopropyltriethyoxysilan. Slides were used for in situ hybridization immediately or dehydrated by an ethanol series and stored at −80°C. Before hybridization, sections were fixed in 4% paraformaldehyde in phosphate-buffered saline, pH 7.4 (PBS), washed twice with PBS and treated with proteinase K (20 μg/ml in Tris/EDTA buffer) for 7 min, washed in PBS, postfixed in 4% paraformaldehyde in PBS, washed twice in PBS and 0.9% NaCl for 5 min each.

A full-length macrophage inducible NOS cDNA (21), which was kindly donated by Prof. Lowenstein (Baltimore, Maryland), was used for probe preparation. The cDNA was cloned into the Not 1 site of the Strategene vector Bluecript KS+ (Stratagene, Heidelberg, Germany). After linearization with Eco RV, 35S-labeled 500-bp antisense cRNA were generated by in vitro transcription using the T3 RNA polymerase. The 700-bp cRNA probes in sense orientation were transcribed after linearization with Dra I with the T7 RNA polymerase to serve as negative controls in each series of experiments. Incorporated label was purified by gel filtration. Some sections were hybridized with a cRNA probe specific for intercellular adhesion molecule-1 (ICAM-1) as described elsewhere (22).

Sections were prehybridized in hybridization solution (50% formamide, 10% dextransulfate, 10 mmol/liter Denhard solution (Ficoll, polyvinylpyrrolidone, and bovine serum albumin; 1 mg/ml each), 0.9 mol/liter NaCl, 60 mmol/liter Na₂HPO₄, 6 mmol/liter EDTA, 0.5% SDS, and 200 μg/ml tRNA from yeast) for 1 to 2 h in a humid chamber (50% formamide). Denatured cRNA probes (5 min, 65°C) were added to fresh hybridization solution at a concentration of 700,000 dpm/50 μl. Hybridization was performed for 10 to 14 h at 50°C. Sections were washed twice for 10 min in 2× standard saline citrate (SSC) with 20 mmol/liter mercaptoethanol at 50°C, twice for 5 min in 2× SCC with 10 mmol/liter Tris and 1 mmol/liter EDTA (2× SSCTE) at 50°C, for 30 min in 2× SSCTE plus 10 μg/ml RNase A and 10 U RNase T1 at 37°C, once for 5 min 0.2× SSC at 37°C and twice for 15 min to a final stringency of 0.2× SSC at 65°C. The sections were dehydrated in an ethanol series in the presence of 300 mmol/liter ammonium acetate, air-dried for 10 min, and exposed to Kodak X-omat AR X-ray film for 3 days to get an orientation about the signal intensity. Afterwards, sections were coated with undiluted Kodak NTB-2 photoemulsion at 45°C. After 3 to 4 weeks of exposure at 4°C, the emulsion-coated slides were developed in Kodak D-19 (16 g/l), fixed in Kodak Unifix (150 g/l), and counterstained lightly with hematoxylin-eosin.

Determination of iNOS mRNA expression by RT-PCR

The RNA (2 μg) was reverse-transcribed to give cDNA in a final volume of 30 μl containing Tris-HCl 10 mmol/liter (pH 8.3), KCl 50 mmol/liter, MgCl₂5 mmol/liter, random hexamers 1.7 μmol/liter, 0.5 mmol/liter each of dATP, dTTP, dCTP and dGTP, RNase inhibitor 0.5 μmol/liter and Moloney murine leukemia virus reverse transcriptase 3.3 U/μl. The reaction was carried out at 42°C for 1 h, followed by heat inactivation of the enzyme at 75°C for 10 min. The cDNA was stored at −80°C.

Polymerase chain reaction (PCR) was performed using specific primers (Clontech, Heidelberg, Germany) yielding a PCR product of 259-base pair length leading from nucleotide 3589 to nucleotide 3848 of the human iNOS sequency published by Geller et al. (23). Primer sequences were: 5′-CGG TGC TGT ATT TCC TTA CGA GGC GAA GAA GG-3′ and 5′-GGT GCT ACT TGT TAG GAG GTC AAG TAA AGG GC-3′. The PCR was performed in Tris HCl 10 μmol/liter, MgCl₂1.5 μmol/liter, KCl 50 μmol/liter, dNTPs 160 μmol/liter, primers 10 pmol/liter and Taq-polymerase 1.25 IE/50 μl. After an initial cycle of 3 min at 94°C, 2 min at 64°C and 1 min at 72°C, reactions were performed for 35 cycles consisting of a 1-min denaturation step at 94°C, an annealing step of 1 min at 64°C and a synthesis step of 1 min at 72°C. The PCR was finished by a final extension step at 72°C of 7 min. Reactions were performed on a Perkin-Elmer Thermal cycler 480 (Perkin-Elmer, Überlingen, Germany). To control adequate isolation of RNA and conversion to cDNA, amplification of GAPDH was performed using the following primers: 5′-GCT TTT AAC TCT GGT AAA GTG G-3′ and 5′-TCA CGC CAC AGT TTC CCG GAG G-3′. PCR products were visualized by ethidium bromide staining after agarose gel electrophoresis. Identity was also confirmed by automatic sequencing of the PCR product using the following internal primer: 5′ GTG CAG CCC AGC AGC CTG -3′.

Determination of myocardial cGMP content

For determination of myocardial cGMP content, frozen myocardial tissue was pulverized and solubilized in 6% (v/v) trichloroacetic acid containing 10 μmol/liter HCl yielding a final concentration of 100 mg tissue in 1000 μl solution. After centrifugation at 2000 gover 15 min, the supernatant was washed with water-saturated diethylether to remove trichloroacetic acid. Afterwards, the cGMP-containing liquid phase was lyophilized. Determination of the cGMP content in the ventricular myocardium was performed using a commercially available radioimmunoassay (Amersham-Buchler, Braunschweig, Germany). To have a positive control, right atrial trabecula obtained during open heart surgery were acutely exposed to the NO-donator sodium nitroprusside (SNP, 100 μmol/liter). For cGMP content determination, atrial trabecula were essentially handled the same as ventricular preparations.

Materials

Lipopolysaccharides from Salmonella typhosa, N^G-mono-methyl-l-arginine (l-NMMA), 4,5-dihydroxy-1,3-benzene disulfonic acid (Tiron) and hemoglobin were purchased from Sigma-Aldrich (Deisenhofen, Germany). Molecular biology enzymes were all purchased from Boehringer (Mannheim, Germany) if not indicated otherwise. All chemicals used were of analytical grade or the best grade commercially available. All compounds were dissolved in deionized and twice-distilled water.

Statistics

The data shown are means ± SEM; EC₅₀values are given as means ± 95% confidence intervals. For statistical analyses of concentration-response curves, area under the curves (AUCs) were calculated and analyzed by ANOVA. The least significant difference (LSD) test and the Shaffer procedure were used for the multiple pairwise comparisons (i.e., if the F test of the global hypothesis was significant, the one-sided p values of the pairwise comparisons were calculated by the LSD test and compared with Shaffer’s significant levels [α/3, α/3, α/3, α/3, α/2, α]).

Acute effects of endotoxin on force of contraction

The addition of increasing concentrations of endotoxin (0.001 to 3 μg/ml) had no acute effect on basal isometric force of contraction. Also, the concentration-dependent increase in force of contraction in response to increasing concentrations of isoprenaline (0.001 to 3 μmol/liter) was not affected by acute addition of endotoxin (1 μg/ml) into the organ bath.

Effect of endotoxin treatment on force of contraction in failing myocardium

Six and 12 hours’ incubation of myocardial preparations in Tyrode solution supplemented with LPS led to a significant attenuation of the concentration-dependent increase in force of contraction in response to isoprenaline (0.001 to 3 μmol/liter) compared to preparations incubated in unsupplemented Tyrode solution. Six hours’ endotoxin treatment of failing myocardium decreased the maximum inotropic effect of isoprenaline by 54% (p = 0.0087; Fig. 1, upper panel); 12-h treatment decreased it by 69% (p = 0.0023; Fig. 1, lower panel). There was no change in the potency of isoprenaline due to endotoxin treatment as indicated by EC₅₀values (12-h incubation, control vs. endotoxin: 0.044 (0.030 to 0.064) μmol/liter and 0.065 (0.039 to 0.108) μmol/liter).

1. Download : Download high-res image (179KB)
2. Download : Download full-size image

Figure 1. Effect of treatment of isolated human left ventricular myocardial preparations from terminally failing hearts with lipopolysaccharides (Endo, 1 μg/ml) on the concentration-dependent inotropic response to isoprenaline. Upper panelshows the effect of 6 h and lower panelthe effect of 12 h of incubation. Also, the effects of co-incubation of isolated myocardial preparations with either l-NMMA (100 μmol/liter) or hemoglobin (Hb, 2 μg/ml) are demonstrated. Ordinate gives force of contraction in mN, abscissa concentration of isoprenaline in μmol/liter. Data are given as mean ± SEM.

Effect of l-NMMA and hemoglobin on force of contraction

To investigate whether the decreased inotropic response to isoprenaline and the decrease in basal contractility following endotoxin treatment might be mediated by activation of NOS, the effect of co-incubation of the ventricular preparations with endotoxin and the NOS-inhibitor l-NMMA and of co-incubation with endotoxin and the NO-scavenger hemoglobin was determined. l-NMMA significantly attenuated the effects of endotoxin on myocardial force generation. Similarly, hemoglobin attenuated the effect of endotoxin treatment, although this effect did not reach statistical significance (Fig. 1).

Effect of endotoxin treatment on force of contraction in nonfailing myocardium

The relative concentration-dependent effect of isoprenaline on isometric force of contraction was more pronounced in nonfailing than in failing myocardium, the maximum increase in force of contraction in response to isoprenaline being 523 ± 58% in nonfailing (n = 10) and 339 ± 37% (n = 17) in failing myocardium (p = 0.009) (Fig. 2, A). Twelve hours of endotoxin treatment of nonfailing left ventricular myocardium decreased the maximum inotropic response to isoprenaline by 66% (p < 0.001 endotoxin vs. control). The effect was similar to the endotoxin effect in failing myocardium (Fig. 2, B). Co-incubation of nonfailing myocardium with l-NMMA significantly attenuated the endotoxin effect. Again, there was no effect of endotoxin treatment on the potency of isoprenaline (EC₅₀values: control 0.027 (0.016 to 0.046) μmol/liter vs. endotoxin 0.046 (0.028 to 0.074) μmol/liter).

1. Download : Download high-res image (113KB)
2. Download : Download full-size image

Figure 2. Comparison of the effect of lipopolysaccharide treatment for 12 h on the concentration-dependent inotropic response of failing and nonfailing left ventricular myocardial preparations to isoprenaline. (A)This shows the inotropic response to isoprenaline after incubation of myocardial preparations in normal tyrode (control); (B)shows the inotropic response to isoprenaline after incubation in tyrode supplemented with lipopolysaccharides (Endo, 1 μg/ml). Data are given as mean ± SEM.

Detection of iNOS mRNA in incubated left ventricular myocardial preparations

Also, to investigate whether the decrease in myocardial force development could be explained by an enhanced expression of iNOS in response to endotoxin treatment, presence of iNOS mRNA in control myocardium and endotoxin-treated myocardium was determined by RT-PCR and by in situ hybridization.

The RT-PCR with specific primers for iNOS led to the amplification of a cDNA segment corresponding in its electrophoretic migration characteristics to the expected iNOS cDNA segment of 259-bp length (Fig. 3). Identity of the PCR product with the expected cDNA fragment was confirmed by sequence analysis (not shown). In failing myocardium, iNOS mRNA was detectable in all samples (6/6) independent from whether samples were treated with endotoxin or not. Similarly, in the majority of samples from nonfailing hearts (4/6), iNOS mRNA was already detectable in those samples that had not been exposed to endotoxin. In samples from two nonfailing hearts in which iNOS mRNA was not already present in tyrode-incubated controls, a iNOS PCR product was obtained after treatment with endotoxin.

1. Download : Download high-res image (234KB)
2. Download : Download full-size image

Figure 3. Result of iNOS mRNA detection by RT-PCR in left ventricular myocardial preparations from two nonfailing hearts incubated for 12 h either in normal tyrode solution (control) or in tyrode plus lipopolysaccharides (1 μg/ml). Amplification products of iNOS-PCR were loaded on the left side, those of GAPDH-PCR on the right side of the gel. Notice that only in two of six nonfailing hearts shown in this figure, iNOS mRNA was not already detectable in control samples, and that iNOS mRNA was detectable in all samples from failing hearts (n = 6).

In situ hybridization revealed specific signals for iNOS mRNA in cardiac myocytes as well as in non-myocyte cells (Fig. 4). The signal obtained from cardiac myocytes was much denser than the signal obtained from other cell types, indicating that iNOS is predominantly expressed in cardiac myocytes. In contrast, the hybridization signal for intercellular adhesion molecule-1 (ICAM-1) used as a control was mainly expressed in endothelial cells along intramyocardial blood vessels and was not reliably detectable in cardiac myocytes (not shown). Thus, the predominant expression of iNOS mRNA within cardiac myocytes is not due to a selective loss of endothelial cells during the incubation procedure.

1. Download : Download high-res image (218KB)
2. Download : Download full-size image

Figure 4. Result of iNOS mRNA detection in human myocardial preparations by in situ hybridization. Hybridized slides is shown from endotoxin-treated myocardium. Notice that the specific iNOS hybridization signal can mainly be depicted from cardiac myocytes.

Determination of myocardial cGMP content

Myocardial cGMP content was determined in left ventricular myocardial preparations that were incubated in parallel to muscle preparations used for functional experiments. Neither after 6 h nor after 12 h of incubation was an increase in myocardial cGMP level detectable (Fig. 5).

1. Download : Download high-res image (69KB)
2. Download : Download full-size image

Figure 5. Bar graph demonstrating the effect of endotoxin treatment (Endo, 1 μg/ml, 6 and 12 h) of human ventricular myocardial preparations in the absence or presence of l-NMMA (100 μmol/liter) on myocardial cGMP content. Acute treatment of atrial trabecula with sodium nitroprusside (SNP, 100 μmol/liter) was performed as a control. Six hours of incubation: n = 6 per condition; 12 h of incubation: n = 10 per condition, atrial preparations: n = 10 per condition. Ordinate: cGMP concentration in fmol/mg wet weight. Data are given as mean ± SEM.

Effect of co-incubation of myocardial preparations with endotoxin and Tiron

To investigate whether oxygen-derived free radicals (O₂⁻) and formation of peroxynitrite (ONOO⁻) contribute to endotoxin-induced contractile dysfunction, the effect of the superoxide (O₂⁻) scavenger Tiron was investigated. Also, in this independent series of experiments, endotoxin treatment of human myocardial preparations for 6 h led to a pronounced decrease in the concentration-dependent positive inotropic response to isoprenaline by a maxim of 50% (p = 0.035, ttest) (Fig. 6). The addition of Tiron to the incubation medium abolished the effect of endotoxin on myocardial contractility (Fig. 6). Incubation with Tiron alone had no effect on the inotropic response to isoprenaline (Fig. 6).

1. Download : Download high-res image (196KB)
2. Download : Download full-size image

Figure 6. Effect of co-incubation of human left ventricular myocardial preparations for 6 h either with lipopolysaccharides (1 μg/ml, n = 5), with lipopolysaccharides and the radical-scavenger Tiron (10 mmol/liter, n = 5), with Tiron (10 mmol/liter, n = 5) alone or in normal Tyrode (control, n = 5). Upper panelshows the concentration-dependent positive inotropic effect of isoprenaline. Ordinate gives force of contraction in mN, abscissa gives isoprenaline concentrations in μmol/liter. Bar graph in the lower panelsummarizes the maximum isoprenaline effects on isometric force of contraction in the different groups.

Inotropic effects of endotoxin in human nonfailing and failing myocardium

Treatment of isolated human left ventricular myocardium for 6 or 12 h with endotoxin led to a decrease in basal and isoprenaline-enhanced force of contraction. The effect was attenuated by the NO-synthase inhibitor l-NMMA and abolished by the superoxide scavenger Tiron, indicating that negative inotropism of endotoxin is mediated by enhanced release of NO and possible interaction with oxygen-derived radicals.

Suffredini et al. (24)were the first to demonstrate that endotoxin-treatment led to a depression of left ventricular contractility in humans. The limitation of their examination was that administration of endotoxin in vivo did not allow them to discriminate between direct effects of endotoxin in the myocardium and indirect effects such as alterations in peripheral vascular resistance or toxic effects of released catecholamines or cytokines. In contrast, the present study provides evidence that prolonged endotoxin treatment is sufficient to directly induce contractile dysfunction in human myocardium.

The negative effect of endotoxin treatment on basal force of contraction and the inotropic response to isoprenaline was present in both nonfailing and failing myocardium. Thus, endotoxin also leads to a decrease in force of contraction in myocardial preparations in which force of contraction and especially the inotropic response to isoprenaline is already impaired (25)and in which NOS activity is already enhanced 26, 27, 28as has been reported previously for failing human myocardium. Therefore, enhanced NOS expression and activity in heart failure does not seem to be a yes or no effect, but a gradual and modifiable effect that progressively leads to an impairment of cardiac function.

Molecular evidence for iNOS mRNA expression in response to endotoxin treatment

Findings in animal models of endotoxin- or cytokine-induced cardiomyopathy indicate that induction of iNOS is one key mechanism to produce cardiodepressant effects 7, 9, 10, 11, 14. In accordance, iNOS mRNA could be detected within human myocardium by RT-PCR and in situ hybridization in the present study. However, no significant difference was seen between iNOS mRNA signals from endotoxin-treated and from control myocardium as determined either by in situ hybridization or by RT-PCR. In failing myocardium in which enhancement of iNOS expression and activity has already been demonstrated 26, 27, 28, one might have expected this finding. However, also in the nonfailing myocardium, iNOS mRNA was already detectable in most of the samples treated in Tyrode without endotoxin supplementation. One explanation might be that, according to a previous report (27), iNOS gene expression can be observed in more than 50% of nonfailing donor hearts, which might be due to medical treatment before, or handling of the hearts following, explanation. Alternatively, one has to consider that endotoxin contamination of incubation basins and media could also have contributed to iNOS induction in control preparations.

The next question was in which myocardial cells does iNOS expression occur and whether NO effects on cardiac myocytes are rather paracrine or autocrine. Results of in situ hybridization for iNOS mRNA indicate that mRNA expression occurs in all cell types including fibroblasts and endothelial cells, but especially within cardiac myocytes from which the strongest signal was obtained. The observation that iNOS mRNA is primarily present in cardiac myocytes in human myocardium is in accordance with recent findings by Haywood et al. (27), who also detected iNOS within cardiac myocytes by immunohistochemical staining of human myocardium. Thus, NO can be generated within cardiac myocytes themselves, and NO effects on myocyte contractility are rather autocrine than paracrine.

Functional evidence for enhanced NOS activity

Because RT-PCR and in situ hybridization did not allow us to quantify differences between iNOS and mRNA levels, the question whether endotoxin treatment of human myocardium leads to an enhanced iNOS activity had to be addressed by functional experiments. In this context, the effects of the NOS-inhibitor l-NMMA and of the NO scavenger hemoglobin were examined. The addition of both the NOS inhibitor and of the NO-scavenger to the incubation medium attenuated the endotoxin effect in failing and nonfailing myocardium. This observation from functional experiments indicates that endotoxin effects are mediated by enhancement of NOS activity and formation of NO. Because contractile dysfunction in response to endotoxin did not develop acutely, but only after prolonged incubation, one has to assume that endotoxin leads to an increase especially in iNOS gene expression, which might have been overlooked by the molecular methods chosen for this study.

Conversely, NOS inhibition did not lead to a complete blockade of the endotoxin effects on human myocardium. This distinguishes the recent findings from observations in isolated myocytes 9, 10, 14or isolated perfused hearts (29)in which effects of endotoxin or cytokine on contractility were abolished by NOS inhibition. There may be species differences or even other mechanisms not yet known that contribute to endotoxin-induced contractile dysfunction, which are independent from release of NO. However, the most likely explanation is that penetration of l-NMMA into the multicellular myocardial preparations used for the present experiments was less effective than penetration in isolated myocytes used in previous experiments.

Myocardial cGMP levels in response to endotoxin

Also, in contrast to observations in other species in which endotoxin- or cytokine-induced contractile dysfunction was accompanied by an increase in myocardial or myocyte cGMP levels 10, 13, no increase occurred in myocardial cGMP levels in response ro endotoxin treatment in human myocardium. One explanation for this discrepency with previous reports is that the peak increase in NO release and myocardial cGMP levels was missed because it occurs earlier as after 6 or 12 h of endotoxin exposure. In this context, it has been demonstrated that once established, cytokine-induced contractile dysfunction persists even if NOS is inhibited later on (29). One also has to consider that there are unrelated effects of endotoxin that lower cGMP levels or that expression of iNOS mRNA in most of the examined samples led to an elevation of baseline cGMP levels. However, the most likely explanation is that enhanced iNOS activity and NO formation impair myocardial contractility by a mechanism that is independent from activation of guanylyl cyclase. This view is supported by the recent observation that the NO-induced reduction in the contractility of isoprenaline-stimulated rat hearts could not be modified by a guanylyl cyclase inhibitor (30).

Involvement of O₂⁻and ONOO⁻in endotoxin-induced contractile dysfunction

One alternative mechanism by which NO leads to a decrease in force of contraction is the formation of peroxynitrite (ONOO⁻) (31), which is formed by the reaction of NO with superoxide anions (O₂⁻) (32)and has been demonstrated to be released from activated macrophages as a result of the simultaneous generation of NO and O₂⁻(33). There is evidence that similar to NO (34)ONOO⁻inhibits rat heart mitochondrial respiration (35). In contrast to the effects of NO, effects of ONOO⁻on cardiac contractility and mitochondrial respiration show no rapid reversibility; thus, ONOO⁻might be the more pathological radical 36, 37. In response to endotoxin treatment, ONOO⁻formation in association with enhanced iNOS activity has been demonstrated in rat aorta in vivo (38).

The hypothesis that endotoxin-induced contractile dysfunction in human myocardium could be mediated by the formation of ONOO⁻is supported by the finding that the effect of endotoxin was abolished in the presence of Tiron, a chelate complex forming and antioxidant substance that scavenges O₂⁻and ONOO⁻36, 39. Obviously, one has to consider that a radical scavenger also unselectively scavenges NO. However, there is evidence that Tiron does not interfere with the effects of NO or hydroxyl radicals (OH⁻) 36, 39. Because the effects of endotoxin could be antagonized by both NOS inhibition and scavenging of O₂⁻and ONOO⁻, one might assume that endotoxin leads to an increased release of NO and by reaction of NO with O₂⁻to the formation of ONOO⁻. Formation of ONOO⁻would also explain the missing increase in myocardial cGMP levels in response to endotoxin treatment and NO release because reaction of NO with O₂⁻might have channeled NO away from its target guanylyl cyclase, as has been demonstrated in human artery and vein preparations in vitro (40).

Conclusions

The results of this study demonstrate that endotoxin exposure of human myocardium impairs cardiac contractility independently from systemic endotoxin effects. This impairment of myocardial performance appears to be mediated by an enhanced release of NO and a consecutive enhanced formation of ONOO⁻. The use of specific scavengers of O₂⁻and ONOO⁻might lead to the development of new therapeutic strategies for cardiac contractile dysfunction in septic shock and inflammatory heart diseases.