Both cGMP and peroxynitrite mediate chronic interleukin-6-induced negative inotropy in adult rat ventricular myocytes
- 1Department of Pharmacology and Toxicology 2Department of Pharmaceutical Sciences, University of Arkansas for Medical Sciences, 4301 West Markham Street, Little Rock, AR 72205, USA 3Department of Physiology, Loyola University Medical Center, Stritch School of Medicine, Maywood, IL 60153, USA
- Corresponding author S. J. Liu: Department of Pharmaceutical Sciences, University of Arkansas for Medical Sciences, 4301 West Markham Street MS 522-3, Little Rock, AR 72205, USA. Email: sliu{at}uams.edu
Abstract
We previously showed that chronic exposure to interleukin (IL)-6 decreases contractile and sarcoplasmic reticular (SR) function assessed by postrest potentiation (PRP) via a nitric oxide (NO)-dependent mechanism in adult rat ventricular myocytes (ARVM). Cyclic GMP (cGMP) has been associated with NO-associated negative inotropic effects of IL-6 during acute exposure; however, its role in chronic cardiac effects of IL-6 remains unclear. The present study examined the roles of cGMP and peroxynitrite (ONOO−) in chronic IL-6-induced negative inotropy in ARVM. After ARVM were exposed to IL-6 for 2–24 h, intracellular cGMP contents were time dependently increased; this was mimicked by a NO donor and abolished by 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ), an inhibitor of soluble guanylyl cyclase (sGC), or Rp-8-Br-cGMP, an inhibitor of cGMP-dependent protein kinase G (PKG). Meanwhile, the IL-6-induced decrease in PRP at 2 h was blocked by ODQ or Rp-8-Br-cGMP. By contrast, ODQ or Rp-8-Br-cGMP only attenuated the inhibition of PRP induced by IL-6 after 24 h exposure. Furthermore, IL-6 time dependently increased superoxide anion production and ONOO− formation; the latter was abolished by 5,10,15,20-tetrakis-(4-sulphonatophenyl)-porphyrinato iron (III) (FeTPPS), an ONOO− decomposition catalyst. Interestingly, FeTPPS had no effect on the IL-6-elicited decrease in PRP at 2 h, but attenuated it after 24 h exposure. Moreover, inhibition of sGC/cGMP/PKG, but not ONOO− formation, abolished the IL-6-induced inhibition of kinetics of myocyte contraction during 24 h exposure. We conclude that while the sGC/cGMP/PKG pathway was the primary mechanism for chronic IL-6-induced negative inotropy at 2 h, both sGC/cGMP/PKG and ONOO−, at least in part, mediate the IL-6-induced inhibition of SR function after 24 h exposure.
In vitro studies on the acute cardiac effect of interleukin (IL)-6, a pro-inflammatory cytokine, has shown that IL-6 exerts a negative inotropic effect within 5 min on cardiac papillary muscle (Finkel et al. 1992), embryonic cardiac myocytes (Kinugawa et al. 1994) and adult ventricular myocytes (Sugishita et al. 1999). The acute negative inotropic effect of IL-6 is accompanied by an increase in cell cGMP production (Kinugawa et al. 1994) and blocked by a nonspecific nitric oxide (NO) synthase (NOS) inhibitor (Finkel et al. 1992; Kinugawa et al. 1994; Sugishita et al. 1999). We recently showed that exposure for 2–24 h to IL-6 decreases contractility and sarcoplasmic reticular (SR) function in adult rat ventricular myocytes (ARVM), associated with an enhanced NO production via inducible NO synthase (iNOS) (Yu et al. 2003, 2005). Although the NO/cGMP pathway was suggested to mediate the acute negative inotropic effect of IL-6 (Kinugawa et al. 1994), it is unclear whether this pathway is also responsible or the sole mechanism for the NO-associated decrease in contractile function induced by chronic exposure to IL-6.
The cGMP-dependent mechanism proposes that NO production induced by stimuli, such as proinflammatory cytokines, activates soluble guanylyl cyclase (sGC) and thereby causes an increase in cGMP, which in turn reduces myocardial contractility (Brady et al. 1992; Balligand et al. 1993). For example, the NO-mediated negative inotropic effects of tumour necrosis factor (TNF)-α and IL-1β have been attributed to the induction of cGMP (Harding et al. 1995; Kumar et al. 1999). Studies using NO donors have demonstrated that NO inhibits basal L-type Ca2+-channel current (ICa,L) via a cGMP-dependent (Campbell et al. 1996; Schroder et al. 2003) and cGMP-independent mechanism (Hu et al. 1997). Moreover, application of cGMP has been shown to elicit a negative inotropic effect on cardiac muscles (Shah et al. 1991; Smith et al. 1991). However, neither acute (Sugishita et al. 1999) nor chronic (Yu et al. 2005) exposure to IL-6 has any detectable effect on ICa,L. Thus, it remains unclear whether the cGMP-dependent mechanism is responsible for the IL-6-induced negative inotropy during a long-term exposure.
It is also known that some biological effects of NO can be mediated via a cGMP-independent mechanism, including inhibition of the aerobic energy-producing processes in the heart (Oddis & Finkel, 1995; Tatsumi et al. 2000) and modulation of cardiac ICa,L (Campbell et al. 1996; Hu et al. 1997). A recent study reported that NO induced by a cocktail of cytokines (i.e. IL-1β, interferon (INF)-γ and TNF-α) causes myocardial contractile failure through peroxynitrite (ONOO−) formation (Ferdinandy et al. 2000). This earlier study showed that ONOO− content (measured as 3-nitrotyrosine) was significantly increased in the perfusate of cytokine-treated isolated working rat hearts at the end of 120 min perfusion (Ferdinandy et al. 2000). However, it remains unknown whether ONOO− plays a role in chronic IL-6-induced negative inotropic effects. Thus, the purpose of the present study was to elucidate the possible involvement of sGC/cGMP and ONOO− after 2–24 h of exposure to IL-6 in ARVM. We found that IL-6 increased intracellular cGMP and ONOO− production concomitant with an increase in superoxide free radicals. Both cGMP and ONOO− mediate the IL-6-elicited inhibitory effect on SR function in ARVM in a novel time-dependent manner.
Methods
Isolation of ARVM
Single ventricular myocytes were isolated from hearts of adult (3- to 6-month-old) male Sprague-Dawley rats using enzymatic dissociation as previously described (Liu & Schreur, 1995; Yu et al. 2005). Briefly, rats were deeply anaesthetized with ether or isoflurane (Vedco, St Joseph, MO, USA) followed by a thoracotomy. Hearts were rapidly excised and perfused at 37°C via the aorta with a control buffer solution followed by enzymatic and mechanical dissociation. Isolated ARVM were resuspended in culture medium containing antibiotic-free, bicarbonate-buffered culture medium (Gibco, Gaithersurg, MD, USA; pH 7.40 in 5% CO2–95% air at 37°C). After incubation for 3–4 h to allow recovery, ARVM were plated in culture dishes with serum-free l-arginine-containing medium overnight and treated with a fixed concentration (10 ng ml−1) of IL-6 for 2, 4 or 24 h. After the designated exposure duration, quiescent rod-shaped ARVM with clear striations were used for electrophysiological and cell shortening (CS) measurements. All experiments were performed at 37°C. The use of animals was carried out under a protocol approved by the Animal Care and Use Committee at the University of Arkansas for Medical Sciences.
Measurement of intracellular cGMP content
cGMP content in cell lysates was quantified using a cGMP enzyme-immunoassay (EIA) kit (Amersham Pharmacia Biotech, Piscataway, NJ, USA) according to manufacture's instructions. Briefly, ARVM (0.25 × 106 cells) were incubated in culture medium in the presence or absence of 10 ng ml−1 IL-6. In some experiments, 10 μm 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) was added 30 min prior to and during IL-6 treatment. ARVM were then centrifuged at 100 g for 10 min at 4°C; the pellets were resuspended in 250 μl lysis reagent and shaken for 10 min at room temperature. After centrifugation at 1000 g for 3 min at 4°C to remove the debris, part of the cell lysate was used to determine protein concentration using the Bradford method (Bio-Rad, Hercules, CA, USA) with bovine serum album as the standard. The rest of cell lysate was used for the cGMP assay. Intracellular cGMP content was expressed in picomoles per milligram of cell protein.
Measurement of CS
Unloaded CS of ARVM was elicited and recorded as previously described (Liu et al. 2002; Yu et al. 2003, 2005). Briefly, ARVM were superfused with normal Tyrode solution containing (mm): 140 NaCl, 5.4 KCl, 1 CaCl2, 0.8 MgCl2, 10 Hepes/Tris and 5.6 glucose (pH 7.40 at 37°C). CS was then elicited with field-stimulation via bipolar platinum electrodes at a frequency of 0.5 Hz. CS of ARVM was monitored using a selected raster-line segment of a video edge-motion detector system (Crescent Electronics, Sandy, UT, USA). The voltage signal was calibrated to determine actual motion (micrometres). Post-rest potentiation (PRP), which has been used to assess cardiac Ca2+ handling and contractile function (Bers et al. 1998), was measured as the first contraction after 30 or 60 s rest intervals. The relative amplitude of PRP was presented by normalizing its peak amplitude to that of steady-state CS before the rest interval. The measured parameters of contractile function in ARVM included maximum rates of contraction (+dL/dt)max and relaxation (−dL/dt)max of CS in the steady state. They were compared among groups after the amplitude of CS was normalized to 1.
Measurement of ONOO− production
ONOO− production in culture media and cell lysates was measured by detection of 3-nitrotyrosine (3-NT) content using the high-performance liquid chromatography (HPLC)-electrochemical method as previously described (Hensley et al. 1997). The conditioned culture media (50 μl) or cell lysates (50 μl) collected from each sample were incubated with 50 μl of 1 mg ml−1 protease (Sigma, St Louis, MO, USA) at 50°C for 18 h, and were analysed using a CoulArray HPLC instrument (Cambridge, MA, USA). Tyrosine and 3-NT were detected at retention times of approximately 7.1 and 24.5 min, respectively. An analytical column (TOSOHAAS ODS 80-TM C-18 reverse-phase column, Mongtomeryville, PA, USA) was used, and the mobile phase was 50 mm sodium acetate/50 mm citrate/8% (v/v) methanol, pH 3.1. HPLC analysis was performed under isocratic conditions. Tyrosine and 3-NT standards were dissolved together to a final concentration of 1.0 mm of each analyte in 0.01 m HCl and stored in frozen 1.0 ml aliquots at −20°C. The levels of 3-NT were expressed as millimoles of 3-NT in 100 moles of tyrosine.
Determination of superoxide free radical (·O2−) production
The level of ·O2− generation in single ARVM was determined using hydroethidium (HE) staining as previously described by others (Bindokas et al. 1996; Benov et al. 1998). HE, a cell-permeant fluorescent dye, can be oxidized to fluorescent ethidium specifically by ·O2− but not O2, H2O2 or ONOO− (Bindokas et al. 1996). ARVM were incubated with 1 μm HE (Molecular Probes, Inc., Eugene, OR, USA) 15–20 min prior to the end of IL-6 treatment at 37°C. Extracellular dye was then washed away with warm phosphate-buffered saline. Fluorescent images of ARVM were rapidly acquired using identical parameters (i.e. power intensity and exposure duration) with Axiovision 3.1 software (Zeiss) via a colour-chilled CCD camera (C5810; Hamamatsu Photonics, Bridgewater, NJ, USA) connected to a fluorescence microscope (Axioskop; Carl Zeiss, Inc., Thornwood, NY, USA). The excitation and emission for ethidium were 488 and 590 nm, respectively. The fluorescence intensity of free rod-shaped ARVM was analysed using NIH ImageJ software in a blind-examination fashion; fluorescence in the overlapped region of rod-shaped myocytes or in rounded ARVM was excluded.
Chemicals
Recombinant rat IL-6 was purchased from Pepro Tech, Inc. (Rocky Hill, NJ, USA); cGMP EIA system was purchased from Amersham Pharmacia Biotech. Rp-8-Br-cGMP, 5,10,15,20-tetrakis-(4-sulphonatophenyl)-porphyrinato iron (III) (FeTPPS) and (±)-S-nitroso-N-acetyl-penicillamine (SNAP) were purchased from Calbiochem (San Diego, CA). All other chemicals wee from Sigma. Stock solutions of ODQ, SNAP and NAP were prepared in dimethylsulfoxide (DMSO). The final concentration of DMSO in extracellular solutions was less than 0.1%.
Statistical analysis
In all experiments, data in response to IL-6 were compared with the time-control, and expressed as a ratio or percentage of each time-control value before combining for statistical analysis. Values are presented as means ± s.e.m. (standard errors of means). Data of intracellular cGMP levels were averaged from duplicate measurements taken from 3–6 hearts. Statistical significance (P < 0.05) was evaluated by one-way analysis of variance (ANOVA) with Duncan's or Student–Newman–Keuls multiple comparisons test.
Results
Intracellular cGMP levels in response to chronic exposure to IL-6
We previously showed that IL-6 increases NO production in ARVM after 2–24 h of exposure (Yu et al. 2003, 2005). In the present study, we first examined intracellular cGMP contents in response to stimulation with 10 ng ml−1 IL-6 for 2, 4 and 24 h. The basal level of intracellular cGMP in ARVM was 0.44 ± 0.06 pmol (mg protein)−1, which remained stable during 24 h in culture in the absence of IL-6 (Fig. 1). This value is in agreement with that reported by others for rat left ventricle (Schulz et al. 1992). Figure 1 also shows that after incubation for 2, 4 or 24 h, IL-6 increased cellular cGMP levels in a time-dependent manner, i.e. cGMP content was increased to 0.77 ± 0.06, 1.20 ± 0.06 and 4.40 ± 0.52 pmol (mg protein)−1 (n = 3–6, P < 0.05) after 2, 4 and 24 h, respectively. Incubation of ARVM with 100 μm SNAP, a NO donor serving as a positive control, also increased cGMP content to 1.92 ± 0.45, 3.11 ± 0.08 and 7.30 ± 1.16 pmol (mg protein)−1 (n = 3–5, P < 0.05) at 2, 4 and 24 h, respectively. By contrast, the cGMP content was not altered in the presence of 0.1% (v/v) DMSO (vehicle for SNAP and NAP) (Fig. 2) or 100 μm NAP, a nonfunctional parent compound of SNAP (Shimojo et al. 1999) (data not shown). These data suggest that IL-6 increased cGMP production in accord with the increased NO production reported previously.
Effect of inhibition of sGC on IL-6-induced cGMP production
To further examine the role of sGC in the IL-6-induced cGMP production, experiments shown in Fig. 1 were repeated in the presence of 10 μm ODQ, a specific inhibitor of sGC. Figure 2 shows that ODQ markedly diminished the time-dependent increase in cGMP levels in ARVM treated with IL-6 for 2 (Fig. 2A), 4 (Fig. 2B) and 24 h (Fig. 2C). Note that 10 μm ODQ alone had no significant effect on intracellular cGMP contents at each time point. These results support the suggestion that the sGC/cGMP pathway is activated by IL-6 during chronic exposure.
Role of sGC/cGMP in IL-6-induced decrease in PRP
We previously showed that IL-6 suppresses SR function (examined by PRP and caffeine-induced contraction), which is primarily mediated by iNOS activation (Yu et al. 2005). Thus, we further examined whether the increased cGMP, which is known to activate PKG, was responsible for the IL-6-induced suppression of PRP. ARVM were treated with 10 μm ODQ or 100 μm Rp-8-Br-cGMP, a specific protein kinase G (PKG) inhibitor, in the absence or presence of IL-6. Figure 3 shows that IL-6 suppressed PRP of CS following a 60 s rest interval (PRP-60) after 2 h (Fig. 3A and B) and 24 h (Fig. 3C and D) exposure, respectively, as reported previously (Yu et al. 2005). The combined data show that IL-6 caused a 46.4% reduction in PRP60 after 2 h exposure to IL-6 (Fig. 3B). In the presence of ODQ, this inhibitory effect of IL-6 on PRP60 was completely abolished (Fig. 3A and B). Similarly, Fig. 3D shows that IL-6 resulted in a 34.8% decrease in PRP60 after 24 h exposure. Under these conditions, ODQ reduced the IL-6-induced decrease in PRP60 from 34.8 to 21.7%, i.e. a 37.8% attenuation in the inhibitory effect of IL-6 (Fig. 3D). Similar results of ODQ attenuating the IL-6 actions were seen in PRP after a 30 s rest interval (data not shown). Note that 0.1% (v/v) DMSO or ODQ alone had no significant effect on PRP. In other experiments, 5 μm ODQ caused a 30% inhibition of the IL-6-induced decrease in PRP60 (n = 6–9). These results suggest that the sGC/cGMP pathway contributes differentially to the inhibitory effect of IL-6 on SR function during 2–24 h of exposure.
Moreover, data of kinetics of steady-state CS showed that IL-6 reduced maximum rates of contraction and relaxation after 2 h exposure (Fig. 3E), as reported previously (Yu et al. 2005). Under these conditions, ODQ completely abolished the IL-6-induced reductions in maximum rates of contraction and relaxation (Fig. 3E). When ARVM were exposed for 24 h to IL-6, Fig. 3F shows that maximum rates of contraction and relaxation of steady-state CS were suppressed more in comparison with those observed after 2 h exposure (Fig. 3E). However, in contrast to its partial inhibition on PRP, ODQ also completely abolished the IL-6 effect on the kinetic of CS after 24 h exposure (Fig. 3F).
The effect of inhibition of PKG by Rp-8-Br-cGMP on the chronic IL-6 action was also examined. Figure 4A shows that the IL-6-induced 39% decrease in PRP60 after 2 h exposure was also completely abolished in the presence of 100 μm Rp-8-Br-cGMP. By contrast, after 24 h exposure, Fig. 4B shows that Rp-8-Br-cGMP reduced the IL-6-induced 42.2% decrease in PRP60 to 20.6%, an approximate 51% inhibition of the IL-6 actions. In other experiments, 2 μm KT5823, another PKG inhibitor, caused a 49% attenuation in the IL-6-induced suppression of PRP60 after 24 h exposure (n = 5–6). Thus, these data support previous suggestion that activation of sGC/cGMP/PKG mediates the IL-6-elicited suppression of SR function at 2 h but contributes partially to the inhibitory effect of IL-6 at 24 h. Moreover, data of kinetics of steady-state CS also showed that the inhibitory effect of IL-6 on maximum rates of contraction and relaxation of CS after 2 h (Fig. 4C) and 24 h (Fig. 4D) exposure was diminished in the presence of Rp-8-Br-cGMP, an effect similar to that induced by ODQ.
ONOO− formation during chronic IL-6 stimulation
ONOO−, a highly damaging reactive agent, has been shown to be involved in cytokine-induced reductions in myocardial contractile function (Ferdinandy et al. 1999, 2000). We then examined whether chronic exposure to IL-6 induces the formation of ONOO− in ARVM. The cellular level of 3-NT has been considered as an important marker for the formation of ONOO− in vivo (Beckman, 1996). The basal levels of 3-NT were 61 ± 18 and 3 ± 3 mmol (100 mol tyrosine)−1 in cell lysates and culture medium, respectively. Both of the basal 3-NT levels remained constant during 24 h in culture in the absence of IL-6 (Fig. 5A and B). Exposure to IL-6 caused an increase in 3-NT in cell lysates to 7-, 12- and 59-fold of time-control levels (n = 5, P < 0.05) at 2, 4 and 24 h, respectively (Fig. 5A). Meanwhile, 3-NT levels in culture media were also significantly increased to 264 ± 112, 188 ± 107 and 676 ± 116 mmol (100 mol tyrosine)−1 (n = 4–5, P < 0.05) after 2, 4 and 24 h exposure to IL-6, respectively (Fig. 5B). Furthermore, these increases in IL-6-induced rise in 3-NT contents were completely blocked by 50 μm FeTPPS, a ONOO− decomposition catalyst (Ferdinandy et al. 2000), after exposure to IL-6 for 2 (Fig. 6A), 4 (Fig. 6B) and 24 h (Fig. 6C). Note that FeTPPS (50 μm) alone had no significant effect on 3-NT levels compared with time-controls during 2–24 h of exposure (Fig. 6A–C). These results suggest that there was a time-dependent increase in ONOO− formation during 24 h stimulation with IL-6.
·O2− production during chronic IL-6 stimulation
It is well known that ONOO− is formed by the reaction of NO with ·O2−. We have shown that IL-6 increases NO production during 24 h incubation (Yu et al. 2005). To further confirm the IL-6-induced increase in ONOO− formation, we examined whether the ·O2− level of ARVM was increased during IL-6 stimulation. ARVM were incubated in the absence and presence of 10 ng ml−1 IL-6 for 24 h and then loaded for 20 min with a ·O2−-sensitive dye. Figure 7A and B shows representative phase-contrast and respective fluorescent images of time-control and IL-6-treated ARVM at low and high magnifications, respectively. In the time-control, rod-shaped ARVM displayed a slightly different degree of fluorescent intensities reflecting different levels of ·O2− production, whereas rounded cells had much higher ·O2− levels (upper panels of Fig. 7A and B). This is consistent with the notion that injured or damaged myocytes undergo various degrees of oxidative stress. These figures also show that ·O2− levels in rod-shaped IL-6-treated myocytes appeared to be greater than those in time-controls (bottom versus top panels in Fig. 7A, and middle versus top panels in Fig. 7B).
Figure 7C and D shows histograms of ·O2− levels in rod-shaped ARVM of time-control (open bars) and IL-6-treated (filled bars) groups after 2 h and 24 h incubation, respectively. Each histogram was curve-fit by the Gaussian distribution (dashed lines, control; continuous lines, IL-6-treated). Figure 7C shows that the ·O2− levels in the time-control and IL-6-treated ARVM after 2 h incubation were not significantly different; for example, the average intensity was 14.1 ± 0.4 and 14.6 ± 0.5 fluorescence units in control (n = 246) and IL-6-treated (n = 191) ARVM, respectively. By contrast, Fig. 7D shows that 24 h exposure to IL-6 resulted in a significant increase in ·O2− production (i.e. 27.9 ± 0.6 fluorescence units, n = 181 in IL-6-treated ARVM versus 18.7 ± 0.4, n = 187 in time-controls). Figure 7E shows combined data from four hearts in which IL-6 increased the ·O2− generation in ARVM by 68% after 24 h of incubation (P < 0.001), while there was only a 13% increase in ·O2− levels after 2 h exposure. Thus, these data demonstrated that the dramatic increase in ONOO− formation observed after 24 h exposure to IL-6 (Fig. 5) was supported by apparent increases in both NO and ·O2− production.
Involvement of ONOO− in the IL-6-induced decrease in PRP during chronic exposure
We then examined the role of ONOO− in IL-6-induced decrease in the contractile function of ARVM. Myocytes were treated with 50 μm FeTPPS prior to and during exposure to IL-6 to prevent ONOO− formation. Figure 8A and B shows that in the presence of FeTPPS, IL-6 suppressed PRP60 by 55.6% after 2 h exposure, compared with 41.2% inhibition in the absence of FeTPPS. By contrast, after 24 h exposure, FeTPPS reduced the IL-6-induced 34.4% decrease in PRP60 (Fig. 8D) to 21.9%, an approximate 36% attenuation in the IL-6 action. Similar results were also observed in the IL-6-induced suppression of PRP30 (data not shown). Again, FeTPPS alone had a tendency to increase but not significantly PRP of CS during 24 h incubation when compared with time-controls (Fig. 8C and D). These data suggest that the amount of ONOO− formation at 2 h is not sufficient to be involved in the decreased SR function induce by IL-6. However, the increased ONOO− after 24 h exposure contributes partially but significantly to the inhibitory effect of IL-6 on SR function.
Moreover, the effects of FeTPPS on the kinetics of steady-state CS after 2 and 24 h exposure to IL-6 are shown in Fig. 8E and F, respectively. Interestingly, these results show no significant effect of FeTPPS on the IL-6-induced reduction in maximum rates of contraction and relaxation after 2 or 24 h exposure.
Discussion
We have previously shown that chronic exposure to IL-6 for 2–24 h results in a negative inotropy mediated primarily by NO/iNOS activation (Yu et al. 2005) via a JAK2/STAT3 signalling (Yu et al. 2003). We also showed that this chronic IL-6-induced negative inotropy results primarily from suppression of SR function but not L-type Ca2+ channels (Yu et al. 2005). The present study further examined whether NO-activated sGC/cGMP/PKG pathway is the primary or sole mechanism underlying the IL-6-induced negative inotropy ((+dL/dt)max, (−dL/dt)max and PRP) during chronic exposure. We demonstrated that during 24 h exposure: (1) IL-6 increases intracellular cGMP contents time dependently; (2) inhibition of sGC/cGMP/PKG abolishes the IL-6-induced decrease in PRP at 2 h but only attenuates it at 24 h; (3) IL-6 increases ONOO− formation time dependently, accompanied by an increase in ·O2− at 24 h; (4) inhibition of ONOO− formation has no effect on IL-6-induced decrease in PRP at 2 h, but partially blocks it at 24 h; and (5) inhibition of cGC/cGMP/PKG but not ONOO− formation abolishes IL-6-induced decreases in (+dL/dt)max, (−dL/dt)max of myocyte contraction.
The cardiac effects of NO have been extensively studied but remain somewhat controversial (for review, see Kelly & Smith, 1997; Massion et al. 2003). For example, NO induced by proinflammatory cytokines (e.g. IL-1β and TNF-α) depresses cardiac contractile function (Kelly & Smith, 1997); however, studies using transgenic mice reveal that the inotropic effect of NO on the basal state is bimodal (Massion et al. 2003). Nevertheless, mechanisms reported for cytokine-associated NO-induced negative inotropic effects include activation of sGC/cGMP (Balligand et al. 1993), direct inhibition of electron transport of mitochondria (Wang et al. 1996; Tatsumi et al. 2000), production of oxidants such as ONOO− (Ferdinandy et al. 2000), and S-nitrosylation of cellular proteins (Xu et al. 1998).
cGMP-dependent mechanism for chronic IL-6-induced negative inotropic effects
Acute negative inotropic effects of IL-6 (i.e. occurring in <5 min) on cardiac papillary muscle (Finkel et al. 1992), chick embryonic cardiomyocytes (Kinugawa et al. 1994) and adult ventricular myocytes (Sugishita et al. 1999) have been attributable to a NO-dependent mechanism. However, cellular NO levels were not measured during acute exposure to IL-6 in these studies. Using NOS inhibitors, these studies suggested that NO/cGMP mediates the acute inotropic effect of IL-6 (Finkel et al. 1992; Kinugawa et al. 1994). However, the increased cGMP contents, which were detected at 5 min and returned to the basal level at 1 h, appeared to occur after the observed decrease in contractility and [Ca2+]i in chick embryonic cardiomyocytes (Kinugawa et al. 1994). The same study also showed a second increase in cGMP contents detected at 24 h incubation with IL-6; however, the role of cGMP in IL-6-induced decrease in contractile function during the chronic exposure is not clearly defined.
IL-6/NO/sGC/cGMP/PKG signalling.
We have shown that upon IL-6 stimulation no increase in NO production is detectable at the first hour until 2 h, in accord with the time-dependent expression of iNOS (Yu et al. 2003, 2005). After 2 h exposure, the increased NO is primarily responsible for chronic IL-6-induced negative inotropy because IL-6 effects are completely blocked by NG-monomethyl-l-arginine (l-NMMA), a NOS inhibitor, or AMT, a selective iNOS inhibitor (Yu et al. 2005). Under the same conditions of exposure to IL-6, we found a time-dependent increase in cGMP, which can be mimicked by a NO donor SNAP but not NAP. The IL-6-induced activation of sGC/cGMP/PKG is closely correlated with its negative inotropic actions because both effects of IL-6 are abolished by either sGC or PKG inhibitors (Figs 2–4). The involvement of the cGMP-dependent mechanism in IL-6-induced decrease in PRP is in agreement with our previous findings (Yu et al. 2005) demonstrating that inhibition of NO production by l-NMMA or AMT blocks the suppression of SR function induced by 2 h exposure to IL-6. Taken together, these results support our hypothesis that IL-6 increases NO production via iNOS mediated by activation of JAK2/STAT3 signalling, which subsequently activates the sGC/cGMP/PKG pathway, and thereby causes a negative inotropy (Fig. 9).
cGMP-dependent effects of IL-6 on SR verse on basal contraction.
Moreover, the present study showed that the cGMP-dependent pathway appears to be the primary mechanism for IL-6-induced reduction in kinetics of basal CS over 2–24 h (Fig. 3E and F, and Fig. 4C and D), consistent with findings reported by others (Kinugawa et al. 1994). By contrast, interestingly we found that the cGMP-dependent mechanism is responsible for IL-6-induced decrease in SR function entirely only at 2 h but partially after longer exposure (Fig. 3B versus D, and Fig. 4A versus B). In addition, after 24 h exposure, the attenuating effect due to inhibition of PKG by Rp-8-Br-cGMP or KT5823 seems greater than that resulting from inhibition of sGC by ODQ (i.e. 50 versus 38%). One possible explanation is that ODQ does not completely inhibit cGMP production after 24 h exposure to IL-6 because cGMP contents are found to be ∼twofold of control levels (Fig. 2C). Nevertheless, these partial inhibitions on the IL-6 action suggest that cGMP-independent mechanism(s) are involved in the long-term (i.e. >2 h) IL-6-induced inhibition of the SR.
The involvement of the cGMP-dependent mechanism in the IL-6-induced decreases in (+dL/dt)max of basal contraction is attributable to its reduction in Ca2+ influx via L-type Ca2+ channels, SR Ca2+ release from ryanodine-receptor channels, and/or Ca2+ sensitivity of myofilaments. Our previous findings have shown that IL-6-induced decreases in Ca2+o responsiveness and the (+dL/dt)max of CS do not result from inhibition of Ca2+ influx via L-type Ca2+ channels (Yu et al. 2005). A piece of indirect evidence in this earlier study suggested lack of changes in the Ca2+ sensitivity of myofilaments after 2 h exposure to IL-6. However, this possibility cannot be excluded for the negative inotropic effect of IL-6 observed after longer exposure (e.g. at 24 h). In addition, studies on acute effects of NO donors showed that activation of the NO/cGMP/PKG pathway reduces myofilament responsiveness to Ca2+ with no effect on Ca2+ transients (Vila-Petroff et al. 1999; Layland et al. 2002); an effect is mediated by PKG phosphorylation of troponin I (Layland et al. 2002). Although the Ca2+ transient was not measured in the present study, a decrease in Ca2+ transients has been reported in chick embryonic cardiomyocytes exposed to IL-6 for 24 h (Kinugawa et al. 1994). Thus, the NO/cGMP/PKG-dependent mechanism is likely to reduce SR Ca2+ release and myofilament Ca2+ sensitivity. A study using ryanodine receptor Ca2+ release channels purified from rabbit skeletal muscle showed that a serine residue of these channels can be phosphorylated by PKG and protein kinase A (PKA) (Suko et al. 1993); however, how this phosphorylation correlates with the functional change in channel activities remains undefined. The discrepancy of effects of inhibition of the c-GMP-dependent pathway on kinetics of basal contraction and on the SR is not entirely clear. One plausible explanation is that IL-6-induced decrease in SR function involves both cGMP-dependent and -independent mechanisms as described above, and the cGMP-dependent portion is sufficiently responsible for Ca2+ cycling during myocyte contraction on a beat-to-beat basis. It is also conceivable that changes in the relative contribution of the cGMP-dependent mechanism and in contractile protein expression and phosphorylation status could occur after long-term (i.e. 24 h) exposure to IL-6. Thus, cellular mechanisms underlying NO/cGMP/PKG-associated alterations in contractile and SR function in the presence of IL-6 require further investigations.
cGMP-independent mechanism for chronic IL-6-induced negative inotropic effects
Involvement of ONOO− in chronic IL-6 cardiac effects.
The variety of NO redox forms has been suggested to mediate NO-associated, cGMP-independent cytotoxicity (Stamler et al. 1992). Formation of ONOO− with ·O2− is one of principal biological reactions of NO besides the activation of sGC/cGMP (Beckman & Koppenol, 1996). For example, ONOO− formation has been shown to cause myocardial contractile failure following an increase in NO induced at the end of 120 min perfusion with a cytokine cocktail (IL-1β, INF-γ and TNF-α) (Ferdinandy et al. 2000). ONOO− can elicit nitration of tyrosine residues in proteins, and yield nitrotyrosine (NT) that has been a convenient marker for detecting ONOO− production (Beckman & Koppenol, 1996). Although other reactive nitrogen oxidants could also produce NT, our data showing complete inhibition of 3-NT contents by FeTPPS (Fig. 6) support the notion that ONOO− is the most likely source for the 3-NT formation inside the cell (Beckman & Koppenol, 1996).
We also found that during 24 h IL-6 stimulation, time-dependent increases in ONOO− formation in cell lysates and culture media are in accord with the increase in NO production as reported previously (Yu et al. 2003, 2005). ONOO− formation requires both NO and ·O2− and is determined by competition of NO and superoxide dismutase (SOD) for ·O2− (Beckman & Koppenol, 1996; Ferdinandy & Schulz, 2003). However, we detected a significant increase in cytosolic ·O2− production only at 24 h but not at 2 h, suggesting that ·O2− production is not necessarily induced by the increased NO. One plausible interpretation for the discrepancy of changes in NO/ONOO− and ·O2− levels at 2 h exposure to IL-6 is that NO/ONOO− measurements are determined from whole cell populations, whereas ·O2− levels are assessed only from rod-shaped ARVM. The lack of effects of FeTPPS on IL-6-induced suppression of kinetics of CS and the SR at 2 h is consistent with no detectable increase in cellular ·O2− levels because functional measurements are also obtained only from rod-shaped ARVM. This supports our suggestion that the cGMP-dependent pathway is the sole mechanism for the early phase of chronic IL-6-induced cardiac effects. Moreover, the increased NO/ONOO− in culture media, which is probably due to releases from injured myocytes, at 2 h of exposure to IL-6 appears to be insufficient to exert a significant inotropic effect.
After 24 h exposure to IL-6, the increased ·O2−, NO and ONOO− levels are accompanied by a decrease in contractile and SR function. The mechanism underlying this ·O2− generation remains unclear and is probably NO independent because of little effect of l-NMMA on ·O2− levels at 24 h exposure to IL 6 (M.-Y. G. Liu, unpublished data). Under these conditions, decomposition of ONOO− only partially inhibits the IL-6-induced decrease in PRP, suggestive of the involvement of ONOO− in suppression of SR function. This is consistent with findings of studies in vascular smooth muscle, which showed that exogenous ONOO− produces a concentration-dependent reduction in Ca2+ uptake by the Ca2+ pump in SR-enriched membrane fractions (Grover et al. 2003b) and in membrane vesicles prepared from cells overexpressing vascular SERCA2b (Grover et al. 2003a). Studies in the cardiac system reported that ONOO− desensitizes myofilament responsiveness to Ca2+, an effect partially mediated by the cGMP-dependent mechanism (Brunner & Wolkart, 2003). By contrast, studies using ONOO− donors suggested that ONOO− increases ICa,L (Malan et al. 2003) and myocardial contractility in isolated rat hearts (Paolocci et al. 2000) independently of cGMP. The positive inotropy of isolated hearts induced by ONOO− donors could result from an increase in coronary blood flow due to the vascular effect of NO/ONOO−. Since differential effects of ONOO− on kinetics of contraction and the SR are different from those induced by the cGMP/PKG mechanism, our results favour the possibility that ONOO− actions are cGMP independent in ARVM. This leads to our suggestion that both ONOO− and sGC/cGMP/PKG contribute to inhibitory effects of IL-6 on the SR in ARVM after 24 h exposure (Fig. 9).
Other possible mechanisms.
Studies using SR membrane vesicles prepared from skeletal muscle showed that exogenous NO inhibits Ca2+ release from ryanodine-receptor release channels (Mészáros et al. 1996) and Ca2+ uptake by Ca2+ ATPase (SERCA1) (Ishii et al. 1998). These studies suggested a direct effect of NO on SR function because neither effect is not altered by either Rp-8-Br-cGMP or cGMP. Moreover, studies on acute effects of NO donors on channel recordings showed that different NO donors cause increases in cardiac ryanodine-receptor channel activities purified from canine hearts (Xu et al. 1998) and SR Ca2+ leak in skeletal muscle fibres (Pouvreau et al. 2004). These studies suggested that NO regulates SR Ca2+ release channels by S-nitrosylation and/or oxidation of target proteins in a concentration-dependent and direct manner. Although it is unknown whether SERCA2 is S-nitrosylated or oxidized during chronic exposure to IL-6, our data are in agreement with NO-induced suppression of SR function probably because the amount of NO production induced by IL-6 is moderate. In addition, other possible mechanisms that could be involved in chronic IL-6-induced negative inotropy include downregulation of gene and protein expression of SERCA2, as observed in neonatal cardiomyocytes after IL-6 exposure for 30 and 48 h, respectively (Villegas et al. 2000), and NO-associated inhibition of mitochondrial respiratory enzyme activities (Poderoso et al. 1996) and energy production as observed during chronic IL-1β exposure (Tatsumi et al. 2000).
In summary, the present study is the first report to show that while the sGC/cGMP/PKG pathway is the sole mechanism for the early phase of IL-6-induced negative inotropy, at least dual mechanisms (sGC/cGMP/PKG and ONOO−) mediate the IL-6-elicited cardiac actions during longer exposure in ARVM. The chronic IL-6-induced decreases in SR and contractile function are closely associated with generation of reactive nitrogen and oxygen species. Modulations of target proteins associated with excitation–contraction coupling and mitochondrial function by phosphorylation, S-nitrosylation and/or oxidation occur in a unique concentration- and time-dependent manner. During chronic exposure to IL-6 which activates multiple signalling pathways, NO-associated alterations in cellular function could be protective (e.g. anti-apoptotic) or detrimental (e.g. pro-apoptotic), probably dependent on distinct time- and concentration-dependent sensitivities of each target and the balance of its signalling networks. Comprehending different time-dependent signalling mechanisms for cardiac effects of IL-6 could provide insights into developing specific therapeutic strategies for treatment of IL-6-associated heart diseases.
Time-dependent increase in intracellular cGMP contents in adult rat ventricular myocytes (ARVM) during interleukin (IL)-6 stimulation Intracellular cGMP levels in time-controls (□) and in myocytes treated with 10 ng ml−1 IL-6 (▪) were measured at 2, 4 and 24 h of incubation. During these periods, exposure of ARVM to 100 μm (±)-S-nitroso-N-acetylpenicillamine (SNAP; ▿), an NO donor, also increased cGMP levels. Data represent means ± s.e.m. of 3–6 experiments. *P < 0.05, compared with time-controls (ANOVA).
Effect of inhibition of soluble guanylyl cyclase (sGC) by ODQ on IL-6-induced increase in intracellular cGMP contents Myocytes were pretreated with 10 μm 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ), a specific sGC inhibitor, 30 min prior to and during incubation with 10 ng ml−1 IL-6 for 2 (A), 4 (B) and 24 h (C). DMSO (solvent for ODQ) or ODQ alone had no effect on cGMP levels. Data represent means ± s.e.m. of 3–6 experiments. *P < 0.05, compared with the time-control.
Effect of inhibition of sGC on IL-6-induced decrease in postrest potentiation (PRP) A and C, representative traces of PRP of cell shortening (CS) following a 60 s rest interval (PRP60) in a DMSO time-control (a) and ARVM pretreated with 10 ng ml−1 IL-6 (b), 10 μm ODQ (c), and 10 ng ml−1 IL-6 in the presence of 10 μm ODQ (d) after 2 and 24 h incubation, respectively. The relative amplitude of PRP60 normalized to the amplitude of prerest steady-state CS in each group after 2 (B) and 24 h (D) exposure. E and F, effects of ODQ on IL-6-induced decrease in kinetics of CS at 2 and 24 h, respectively. Data represent means ± s.e.m. Numbers in the parentheses represent the number of experiments. *P < 0.01, compared with the time-control group. †P < 0.05, compared with ODQ or IL-6 alone.
Effect of Rp-8-Br-cGMP on IL-6-induced decrease in PRP60 Myocytes were pretreated with 100 μm Rp-8-Br-cGMP, a selective protein kinase G (PKG) inhibitor, 30 min prior to and during exposure to 10 ng ml−1 IL-6 for 2 (A) and 24 h (B). C and D, effects of Rp-8-Br-cGMP on IL-6-induced decrease in kinetics of CS at 2 and 24 h, respectively. Data represent means ± s.e.m. *P < 0.05 and **P < 0.01, compared with time-controls; †P < 0.05, compared with Rp-8-Br-cGMP alone.
Time-dependent increase in 3-nitrotyrosine (3-NT) levels in cell lysate and culture medium in response to IL-6 stimulation Concentrations of 3-NT in cell lysates (A) and culture medium (B) were measured in the time-control (□) and IL-6-treated ARVM (▪) at 2, 4 and 24 h. Data represent means ± s.e.m. of 5 experiments. *P < 0.05, compared with the time-controls (ANOVA).
Effect of FeTPPS on IL-6-induced increase in cellular 3-NT levels Myocytes were pretreated with 50 μm 5,10,15,20-tetrakis-(4-sulphonatophenyl)-porphyrinato iron (III) (FeTPPS; a peroxynitrite decomposition catalyst) for 30 min in the absence and presence of 10 ng ml−1 IL-6 for 2 (A), 4 (B) and 24 h (C). FeTPPS completely blocked the increase in 3-NT levels in cell lysates treated with IL-6, whereas alone it had no effect on cellular 3-NT levels. Data represent means ± s.e.m. of 3–4 experiments. *P < 0.05, compared with the time-control group.
Effect of IL-6 on superoxide anion (·O2−) production in ARVM Representative images (phase-contrast, left panels; and fluorescent, right panels) of control and IL-6-treated (10 ng ml−1 for 24 h) ARVM stained with hydroethidium (HE) were acquired at low (A) and high (B) magnifications. Parameters of data acquisition (e.g. exposure duration and source power intensity) were kept identical in each experiment. ARVM that were not stained with HE served as negative controls (N.C., bottom panels in B). Histograms of fluorescent intensity of ARVM in time-control and IL-6-treated groups isolated from the same heart after exposure for 2 (C) and 24 h (D). E, combined data of ·O2− levels in ARVM after IL-6 treatment for 2 (light grey) and 24 h (dark grey) were collected from 4 hearts. Data represent means ± s.e.m. *,†P < 0.01 when compared with time-control and ARVM treated for 2 h, respectively.
Effect of FeTPPS on IL-6-induced decrease in PRP Representative traces of PRP60 of CS in a time-control (a), and ARVM pretreated with 10 ng ml−1 IL-6 (b), 50 μm FeTPPS (c), and 10 ng ml−1 IL-6 in the presence of 50 μm FeTPPS (d) after 2 (A) and 24 h (C) of exposure. The trace break in each group represents a 50 s break of the 60 s rest interval. B and D, combined data from ARVM treated with FeTPPS 30 min before and during incubation with or without IL-6 for 2 h and 24 h, respectively. FeTPPS had no effect at 2 h but partially inhibited IL-6-induced decrease in PRP60 at 24 h. E and F, effects of FeTPPS on IL-6-induced decrease in kinetics of CS at 2 h and 24 h, respectively. Data represent means ± s.e.m. *P < 0.05 and **P < 0.01, compared with time-controls; †P < 0.05, ‡P < 0.01 compared with FeTPPS alone. §P < 0.05, compared with IL-6 alone.
A summary diagram of possible cellular mechanisms mediating chronic inotropic effects of IL-6 in ARVM in a time-dependent manner We have previously shown that IL-6 induces an iNOS/NO-dependent negative inotropic effect on ARVM during 2–24 h of exposure, which is mediated through activation of JAK2/STAT3 (Yu et al. 2003, 2005). Results in the present study fill the gap of signalling transduction between NO and its effectors. During an early phase of IL-6 stimulation, only sGC/cGMP/PKG is probably responsible for its negative inotropy (at 2 h, dashed arrows). When AVRM are exposed to IL-6 for longer periods of time (e.g. at 24 h, continuous arrows), dual mechanisms mediate the IL-6-induced decrease in contractile and SR function.
Acknowledgements
We thank Drs Syed F. Ali, Syed Z. Imam and Kenneth Muldrew for their assistance with 3-NT measurements, and Ms Kerry A. Roberto for her expert technical assistance. This work was supported in part by grants from National Heart, Lung, and Blood Institute of National Institutes of Health (R01HL62226), American Heart Association/Heartland Affiliate (AHA/HA) and the UAMS Graduate Student Research Fund.
Footnotes
-
- Accepted May 5, 2005.
- Received March 24, 2005.
- Revision received April 28, 2005.
- The Physiological society 2005
