Inhibition of matrix metalloproteinases prevents peroxynitrite-induced contractile dysfunction in the isolated cardiac myocyte

Abstract

Background and purpose:

The potent oxidant peroxynitrite (ONOO⁻) induces mechanical dysfunction in the intact heart in part through activation of matrix metalloproteinase-2 (MMP-2). This effect may be independent of the proteolytic actions of MMPs on extracellular matrix proteins. The purpose of this study was to examine the effects of ONOO⁻ on contractile function at the level of the single cardiac myocyte and whether this includes the action of MMPs.

Experimental approach:

Freshly isolated ventricular myocytes from adult rats were superfused with Krebs–Henseleit buffer at 21 °C and paced at 0.5 Hz. Contractility was measured using a video edge-detector. ONOO⁻ or decomposed ONOO⁻ (vehicle control) were co-infused over 40 min to evaluate the contraction cease time (CCT). The effects of ONOO⁻ on intracellular [Ca²⁺] were determined in myocytes loaded with calcium green-1 AM. MMP-2 activity was measured by gelatin zymography.

Key results:

ONOO⁻ (30-600 μM) caused a concentration-dependent reduction in CCT. Myocytes subjected to 300 μM ONOO⁻ had a shorter CCT than decomposed ONOO⁻ (14.9+1.5 vs 32.2+3.5 min, n=7–8; P<0.05) and showed increased MMP-2 activity. The MMP inhibitors doxycycline (100 μM) or PD 166793 (2 μM) reduced the decline in CCT induced by 300 μM ONOO⁻. ONOO⁻ caused shorter calcium transient cease time and significant alterations in intracellular [Ca²⁺] homoeostasis which were partially prevented by doxycycline.

Conclusions and implications:

This is the first demonstration that inhibition of MMPs protects the cardiac myocyte from ONOO⁻-induced contractile failure via an action unrelated to proteolysis of extracellular matrix proteins.

Introduction

The overproduction of reactive oxygen species including the potent cellular oxidant peroxynitrite (ONOO⁻) can outstrip cellular antioxidant defences leading to oxidative stress (Cai and Harrison, 2000; Ferdinandy and Schulz, 2003; Ferdinandy, 2006; Pacher et al., 2007). This condition has been implicated in the pathogenesis of many cardiovascular diseases such as hypertension, atherosclerosis and the complications resulting from diabetes and aging, and heart-specific defects in myocarditis, heart failure (Ide et al., 2000; Pacher et al., 2005) and ischaemia–reperfusion injury (Wang and Zweier, 1996; Yasmin et al., 1997; Crestanello et al., 1998).

Overproduction of reactive oxidant species can activate matrix metalloproteinases (MMPs), enzymes best known for their action in remodelling the extracellular matrix (Rajagopalan et al., 1996; Okamoto et al., 2001; Siwik et al., 2001). However, recent studies have unravelled new roles of MMPs, particularly for matrix metalloproteinase-2 (MMP-2), which is ubiquitous to most cell types including cardiac myocytes (Coker et al., 2001; Wang et al., 2002b). These include both physiological and pathological processes such as platelet aggregation, inflammation and neurodegenerative disease (Sawicki et al., 1997; McQuibban et al., 2000; Zhang et al., 2003). In each of these conditions, the biological effect of MMP-2, which is a rapid and early response of the cell to enhanced oxidative stress, occurs by its proteolytic action on novel substrates unrelated to extracellular matrix proteins (Chow et al., 2007; Schulz, 2007). MMPs are implicated in the pathogenesis of many cardiovascular diseases for their actions on both intracellular and extracellular targets in both animal models and in human disease including myocardial infarction (Rohde et al., 1999; Hayashidani et al., 2003; Villarreal et al., 2003), heart failure (Rouet-Benzineb et al., 1999; Spinale et al., 2000; Wilson et al., 2002), ischaemia–reperfusion injury (Fujimura et al., 1999; Cheung et al., 2000; Wang et al., 2002b; Schulze et al., 2003; Sawicki et al., 2005), cytokine-induced cardiac dysfunction (Gao et al., 2003) and sepsis (Yassen et al., 2001; Lalu et al., 2006). In addition, cardiac preconditioning has shown to decrease the deleterious effects caused by MMP-2 in cardiac ischaemia–reperfusion injury, an effect which is lost in hyperlipidemic conditions (Lalu et al., 2002; Giricz et al., 2006).

ONOO⁻ is a potent oxidant that activates MMPs by S-glutathiolation of a critical cysteine residue in their propeptide domain resulting in a full-length active MMP (Okamoto et al., 2001). We have previously shown that infusion of ONOO⁻ into the isolated rat heart causes cardiac contractile dysfunction through MMP-2, which is prevented by MMP inhibition (Wang et al., 2002a). However, this previous study did not directly address whether the detrimental effect of MMP-2 occurred via its proteolytic actions on extracellular matrix proteins or on targets specifically associated with the cardiac myocyte (Schulz, 2007).

In this regard, Ishida et al. (1996) showed that the direct administration of ONOO⁻ to isolated neonatal cardiac myocytes caused severe contractile dysfunction, evaluated as the time taken for contractions to cease (contraction cease time, CCT). However, whether MMPs are involved in ONOO⁻-induced contractile failure and alterations of calcium handling in isolated adult cardiac myocytes is unknown. Therefore, we hypothesized that ONOO⁻ produces contractile dysfunction through activation of MMPs independent of their action on extracellular matrix proteins.

Methods

All the experiments were performed according to the recommendations given by the Guide to the Care and Use of Experimental Animals published by the Canadian Council on Animal Care (revised 1993).

Isolation of cardiac myocytes

Calcium-tolerant ventricular myocytes were obtained by enzymatic dissociation as described previously (Bouchard et al., 1993; Light et al., 1998). Briefly, adult male Sprague–Dawley rats (250–350 g) were anaesthetized with an injection of sodium pentobarbital (60 mg kg⁻¹, i.p.). The hearts were isolated and perfused through the aorta in a modified Langendorff perfusion system at constant flow and at 37 °C with oxygenated (95% O₂: 5% CO₂) Krebs–Henseleit solution containing (in mM): NaCl (121), KCl (5), CH₃COONa (1.7), MgCl₂ (0.1), Na₂HPO₄ (0.4), NaHCO₃ (20.8), glucose (5.5), taurine (1) and CaCl₂ (1) for 5 min to clear any blood from the coronary circulation. The solution was then switched to a Ca²⁺-free Krebs–Henseleit solution for 9 min. After this, the solution was replaced for the first digestion process with Krebs–Henseleit solution containing 40 μM Ca²⁺ and 13.3 μg ml⁻¹ Streptomyces collagenase (Yalkut Pharmaceutical, Tokyo, Japan) and 13.3 μg ml⁻¹ Streptomyces protease (Sigma, Oakville, Ontario, Canada) for approximately 10 min. The ventricles were then separated from the atria and great vessels using scissors and chopped into small pieces. For the second digestion process, the chopped tissue was placed into a flask with Krebs–Henseleit solution containing 5.5 mM HEPES, 3% BSA, 100 μM Ca²⁺, and 83.3 μg ml⁻¹ collagenase and 83.3 μg ml⁻¹ protease. The cells were further dissociated by incubation at 37 °C under gentle agitation. A 2 ml aliquot of cell suspension was removed after 10, 20, 30 and 40 min incubation and centrifuged for 25 s at 2000 × g. A volume of 2 ml of the second digestion buffer was replaced in the flask each time after sampling. Each cell suspension was immediately centrifuged and the cells were resuspended in storage buffer (Krebs–Henseleit solution containing 5.5 mM HEPES, 3% BSA and 100 μM Ca²⁺). The aliquots were observed under a light microscope for morphology and aliquots that showed more than 70% rod-shaped cells (typically after 30 min digestion) were used for experiments.

Synthesis of ONOO⁻

Active and decomposed ONOO⁻ were synthesized, and the concentration was verified on the same day of experiments using ultraviolet spectroscopy as described previously (Yasmin et al., 1997). Both active and decomposed ONOO⁻ were diluted with 1 mM NaOH immediately prior to the experiment to achieve the desired working concentration.

Measurement of myocyte contractility

To maintain a constant ONOO⁻ to myocyte ratio, approximately 5000 cardiac myocytes were placed into a perfusion chamber in each experiment. Experiments were performed at 21 °C to allow us to directly compare to the conditions used by Ishida et al. (1996) and to facilitate the accurate temporal measurement of cellular contracture. Freshly isolated rat ventricular myocytes were continuously superfused at 1 ml min⁻¹ with Krebs–Henseleit solution containing 2 mM Ca²⁺ using an infusion pump (Cole-Palmer, Barrington, IL, USA) and electrically paced at 0.5 Hz. Single myocyte contractility was measured using a video edge-detector (Crescent Electronics, Salt Lake City, UT, USA) and data were recorded using pClamp 8.0 software. After 3 min, an infusion of either decomposed or authentic ONOO⁻ was begun (using a microinfusion pump (Baxter, Deerfield, IL, USA) linked via a side arm to the perfusion line delivering Krebs–Henseleit solution), so that the final concentrations of ONOO⁻ in the Krebs–Henseleit solution reached either 30, 100 or 300 μM. The infusion was continued for an experimental observation period of up to a maximum of 40 min or until the time at which the cells ceased to contract (CCT). A second series of cells were perfused in an identical manner except that the MMP inhibitors doxycycline (100 μM; Sigma) or PD 166793 (2 μM; a kind gift from Pfizer, New York, NY, USA), dissolved in Krebs–Henseleit solution, were used to superfuse the cells after the 3 min baseline evaluation period. These MMPs inhibitors at their respective concentrations were chosen based on previous data from isolated rat hearts subjected to either ischaemia–reperfusion (Cheung et al., 2000; Wang et al., 2002b; Schulze et al., 2003; Sawicki et al., 2005), proinflammatory cytokines (Gao et al., 2003) or exogenous ONOO⁻ (Wang et al., 2002a). Parallel experiments were performed using their respective vehicles (water for doxycycline and 0.05% dimethylsulphoxide v/v for PD 166793). Although the previous studies (Cheung et al., 2000; Wang et al., 2002b; Schulze et al., 2003; Sawicki et al., 2005) indicated that doxycycline or PD 166793 does not have any direct effect on myocardial contractile function in intact rat hearts exposed to these drugs for up to 120 min, we also evaluated whether these compounds could have any effect on the contractile function of isolated cells. Infusion of the inhibitors or their vehicles had no effect on contractile function as observed over a 10 min period (data not shown).

Evaluation of MMP activity

The cell storage buffer was removed from an aliquot of freshly isolated cardiac myocytes and replaced with Krebs–Henseleit buffer containing 100 μM Ca²⁺ (oxygenated with 95% O₂: 5% CO₂) to reduce hypercontracture by three separate wash and centrifugation steps (2000 × g for 25 s). Cardiac myocytes (∼200 000 cells) in 3 ml buffer were then exposed to the bolus addition of either 300 μM ONOO⁻ or decomposed ONOO⁻ followed by 5 min incubation at 21 °C. The addition of ONOO⁻ or decomposed ONOO⁻ had no significant effect on buffer pH. Cells were separated from the media by centrifugation (2000 × g for 40 s) and the latter was concentrated 20 times using Amicon Ultra-4 centricon tubes (Millipore, Bedford, MA, USA). Cell homogenates were prepared in 350 μl of homogenization buffer containing 50 mM Tris-HCl (pH 7.4), 3.1 mM sucrose, 1 mM dithiothreitol (Fisher Scientific, Ottawa, Ontario, Canada), 1:1000 protease cocktail inhibitor (P-8340; Sigma) and 0.1% Triton X-100 and homogenized on ice using a sonicator (Misonix, Farmingdale, NY, USA, 3 × 10 s with 60 s between each cycle). Gelatin zymography was performed as described by Cheung et al. (2000). Samples from cell homogenates and media were applied to 8% polyacrylamide gels co-polymerized with 2 mg ml⁻¹ gelatin. After electrophoresis, gels were rinsed three times for 20 min each in 2.5% Triton X-100 to remove sodium dodecyl sulphate. Then the gels were washed twice in incubation buffer (50 mM Tris-HCl, 5 mM CaCl₂, 150 mM NaCl and 0.05% NaN₃) for 20 min each at room temperature and then incubated in a fresh aliquot of the same at 37 °C. The gels were stained in 2% Coomassie Brilliant Blue G/25% methanol/10% acetic acid for 2 h and then destained for 1 h in 2% methanol/4% acetic acid. Gels were scanned using a GS-800-calibrated densitometer (Bio-Rad, Mississauga, Ontario, Canada) and band densities were measured using SigmaGel software (Jandel, San Rafael, CA, USA).

Evaluation of intracellular [Ca²⁺]

Cardiac myocytes placed in storage buffer were loaded for 30 min at room temperature and then for 30 min at 37 °C with the Ca²⁺-sensitive fluorescent probe calcium green-1 AM (4 μM, dissolved in a mixture of dimethyl sulphoxide and pluronic acid, 1:1 v/v; Molecular Probes, Eugene, OR, USA). After loading, cells were washed and centrifuged twice (2000 × g for 25 s) with storage buffer and placed on coverslips for observation at × 200 with an inverted microscope (CK40; Olympus, Center Valley, PA, USA) while being superfused with the same Krebs–Henseleit buffer used for contractility measurements, and paced at 0.5 Hz at 21 °C. An infusion of either decomposed or 300 μM ONOO⁻ was started after the baseline recording over a period of 20 min via a side arm using a microinfusion pump. Some cardiac myocytes were exposed to doxycycline (100 μM) dissolved in Krebs–Henseleit buffer 5 min before starting the infusion of 300 μM ONOO⁻ and maintained during the 20 min exposure period. A Photon Technology International (Model 814; Lawrenceville, NJ, USA) photomultiplier detection system and Clampex software (version 8.1) was used for data acquisition and analysis. Calcium Green-1 AM was excited at 480 nm and the emitted light intensity at 520 nm was digitized and stored. The parameters evaluated were the percentage of change in the calcium transient amplitude vs baseline, the time to cease calcium transients and the time for maximal increase in diastolic calcium.

Data analysis and statistical procedures

Data are expressed as means±s.e.mean. Student's t-test (unpaired) or one-way ANOVA followed by post hoc analysis using Tukey's multiple comparison test was used for statistical analysis, as appropriate. P<0.05 was the criterion for significance. The n refers to the number of individual cardiac myocytes tested. In all experiments, cells were tested from ⩾4 separate isolations of myocytes.

Results

Concentration-dependent contractile dysfunction by ONOO⁻

To directly evaluate the effects of ONOO⁻ on contractile dysfunction in the isolated cardiac myocyte, we used the CCT as a measure of severely impaired contractility. This measure was previously used in neonatal cardiac myocytes exposed to ONOO⁻ (Ishida et al., 1996), although the effects of ONOO⁻ on adult rat ventricular myocytes have not been documented. The administration of decomposed ONOO⁻ did not affect the CCT, compared to control conditions (32.5±3.5 vs 35.1±3.0 min; P>0.05, n=8 and n=9, respectively) (Figure 1a). However, in the presence of ONOO⁻, there was a concentration-dependent decrease in CCT (Figure 1b). As pH is a potential variable in these experiments due to the alkaline ONOO⁻ buffer, we monitored the effects of up to 300 μM ONOO⁻ or its vehicle, decomposed ONOO⁻, on superfusate pH and observed no change in pH.

Figure 1.

Concentration-dependent effect of a continuous infusion of ONOO⁻ for 40 min on contraction cease time (CCT) in isolated cardiac myocytes. There was no effect of infusions of decomposed ONOO⁻ (dec. ONOO⁻) on the CCT. ^*P<0.05 vs control, n=4–9 cells from a total of 6 independent preparations. ONOO⁻, peroxynitrite.

MMP inhibitors do not alter baseline contractile properties of isolated cardiac myocytes

Cardiac myocytes were treated with two different MMP inhibitors (doxycycline or PD 116793) as well as their respective vehicle controls (water for doxycycline and 0.05% dimethylsulphoxide for PD 166793; n=5 per group from four independent isolations). During the 10-min perfusion period with MMP inhibitors, the cells contracted regularly and none of them ceased to contract. Moreover, there was no evidence of early depolarizations or afterdepolarizations, as shown by the lack of alterations of the contractile waveform of the traces or absence of significant differences in cell shortening or the time to 50% relaxation (data not shown).

MMP inhibitors prevent contractile dysfunction caused by ONOO⁻

Cardiac myocytes were exposed to a continuous infusion of 300 μM ONOO⁻ in the presence or absence of the MMP inhibitors (doxycycline or PD 166793) or their respective vehicles. Administration of 300 μM ONOO⁻ caused a rapid onset in contractile dysfunction in myocytes, which was revealed as irregular contractions with early afterdepolarizations and hypercontracture as the infusion of ONOO⁻ progressed (Figure 2a). Myocytes challenged with 300 μM ONOO⁻ in the presence of doxycycline or PD 166793 exhibited a significantly delayed onset of contractile dysfunction in comparison to ONOO⁻ alone (Figures 2b and c). These data are summarized in Figure 2d.

Figure 2.

Effects of MMP inhibitors on contraction cease time (CCT) during 300 μM ONOO⁻ exposure. (a–c) Representative traces of cell shortening in cardiac myocytes subjected to ONOO⁻ alone (a) or with 100 μM doxycycline (b) or 2 μM PD 166793. (c) In each panel, the expanded time scale of the cell-shortening traces are shown at (i) baseline recording and (ii) 10 min after infusion of 300 μM ONOO⁻. (d) Summary analysis in terms of the CCT. n=5–6 cells per group from a total of five independent preparations. ^*P<0.05 vs ONOO⁻. MMP, matrix metalloproteinase; ONOO⁻, peroxynitrite.

Impairment in Ca²⁺ homoeostasis caused by ONOO⁻ is attenuated by MMPs inhibitors

Alteration in calcium homoeostasis is a common mechanism for dysfunctional excitation coupling in cardiac myocytes. Therefore, to evaluate whether the protective effect of MMP inhibition was related to possible effects on Ca²⁺ homoeostasis, we measured calcium transients in cardiac myocytes subjected to 300 μM ONOO⁻. The generation, amplitude and shape of Ca²⁺ transients in cardiac myocytes exposed to decomposed ONOO⁻ were unaltered over the duration of the experimental protocol (Figure 3a). In contrast, 300 μM ONOO⁻ caused a progressive increase in diastolic [Ca²⁺] as well as a time-dependent reduced peak systolic transient amplitude (Figure 3b). Doxycycline treatment delayed the onset of changes in Ca²⁺ homoeostasis but did not prevent their occurrence, as alterations of Ca²⁺ transient shape in the form of premature Ca²⁺ release were observed during the experiment (Figure 3c), although at a later time. However, in the presence of doxycycline, the peak Ca²⁺ transient amplitude was better maintained in comparison to ONOO⁻ alone.

Figure 3.

Effects of ONOO⁻ and doxycycline on calcium homoeostasis. Representative traces of cells exposed to either decomposed ONOO⁻ (a) or 300 μM ONOO⁻ alone (b) or with addition of 100 μM doxycycline. (c) Expanded time scale traces of calcium transients are shown at (i) baseline recording and (ii) 10 min after infusion of ONOO⁻. ONOO⁻, peroxynitrite.

Cardiac myocytes exposed to 300 μM ONOO⁻ displayed a significant increase in Ca²⁺ transient amplitude at 5 min in comparison to decomposed ONOO⁻. Doxycycline treatment normalized this variable to values equal to those obtained with exposure to decomposed ONOO⁻ (Figure 4a). These data are consistent with early alteration in Ca²⁺ homoeostasis induced by ONOO⁻. Furthermore, cardiac myocytes challenged with ONOO⁻ had a significant reduction in the time to cease calcium transients that was attenuated by doxycycline treatment (Figure 4b). Represented in terms of the time to reach maximum diastolic [Ca²⁺] during the experimental protocol, cardiac myocytes challenged with 300 μM ONOO⁻ showed a reduced time to maximum increase in diastolic [Ca²⁺] level compared with decomposed ONOO⁻. Doxycycline, however, did not alter this variable (Figure 4c). Furthermore, we performed an analysis of the decay rates by fitting to a monoexponential function and found no differences in the values of half-life (data not shown).

Figure 4.

Quantitative analysis of the effects of ONOO⁻ and doxycycline (doxy) on calcium homoeostasis. (a) Percent change in calcium transient amplitude at 5 min ONOO⁻ infusion. (b) Time to cease calcium transients. (c) Time to reach the maximum increase in diastolic calcium level. ^*P<0.05 vs decomposed ONOO⁻ (dec. ONOO⁻), †P<0.05 vs ONOO⁻, n=4 cells per group from a total of four independent preparations. ONOO⁻, peroxynitrite.

ONOO⁻ activates myocyte MMP-2

To evaluate whether ONOO⁻ was able to activate MMPs when given as a bolus to an aliquot of cardiac myocytes, gelatin zymography was performed on the concentrated incubation media from cardiac myocytes to examine changes in MMP-2 or MMP-9 activities. After 5 min exposure to 300 μM ONOO⁻, there was a significant increase in 62 kDa MMP-2 activity in the media, compared with the effects of exposure to decomposed ONOO⁻ and no evidence of 72 kDa activity (Figure 5). We did not observe any evidence of MMP-9 activity (data not shown).

Figure 5.

Gelatin zymography of media from cardiac myocytes exposed to either decomposed ONOO⁻ (dec. ONOO⁻) or 300 μM ONOO⁻. Upper panel shows a representative zymogram. Lower panel shows summary data of the densitometric analysis of 62 kDa MMP-2 activity. Note that the HT1080 cell supernatant used as a control is human MMP-2 which consists of both 72 and 64 kDa forms, the latter which is slightly higher in molecular weight than the 62 kDa rat MMP-2. ^*P<0.05 vs dec. ONOO⁻, n=3 experiments. ONOO⁻, peroxynitrite.

Discussion and conclusions

This study is the first demonstration that inhibition of MMPs reduces ONOO⁻-induced contractile dysfunction in isolated cardiac myocytes, independent of the canonical actions of MMPs on extracellular matrix proteins. Previously, we had demonstrated that MMP-2 is activated in isolated rat hearts challenged with ONOO⁻ and that a MMP inhibitor prevented this dysfunction (Wang et al., 2002a). However, this study did not determine whether the effect of the activation of MMP-2 was independent of proteolysis of extracellular matrix proteins.

During myocardial ischaemia–reperfusion injury, ONOO⁻ is rapidly generated during the first few minutes of reperfusion, which precedes the activation and release of MMP-2 from the heart (Yasmin et al., 1997; Cheung et al., 2000). In the current study, we observed an increase in 62 kDa MMP-2 activity in the conditioned cell medium following exposure to ONOO⁻. The release of MMP-2 from heart muscle is understood to be a consequence of its activation by oxidative stress (Cheung et al., 2000; Wang et al., 2002a, 2002b) and may be a means for the cell to protect itself from further proteolytic stress. MMP-2 activation as a result of ischaemia–reperfusion injury in the intact heart results in the cleavage of troponin I and myosin light chain 1 (Wang et al., 2002b; Sawicki et al., 2005). Whether the same event occurs during ONOO⁻ exposure to the isolated myocytes is unknown. Our data suggest that the effect of ONOO⁻ was in part mediated by increased MMP activity as the contractile dysfunction was diminished by two structurally independent MMPs inhibitors, doxycycline or PD 166793, and was independent of an action of MMPs to cleave extracellular matrix proteins in the isolated myocytes studied here. We found that 100 μM doxycycline was superior to 3 μM PD 166793 in preventing the contractile dysfunction caused by ONOO⁻. This could be, to some extent, due to proposed ROS scavenging properties of some tetracyclines at high concentrations (Whiteman and Halliwell, 1997), in comparison to the lower concentration used here at which it has marked MMP inhibitor activity in heart tissue (Cheung et al., 2000). To the best of our knowledge, ROS scavenging properties have not been reported for PD 166793.

We focused our study on the effects of ONOO⁻ and activation of MMPs on the contractile function of isolated cardiac myocytes to rule out effects of MMPs on extracellular matrix proteins. Our experimental model and the functional parameters measured have some resemblance to the situation in myocardial ischaemia–reperfusion injury, which is characterized by an increased [Ca²⁺]_i implicated in contractile dysfunction. As reported in this study, alterations in calcium handling and contractile function were observed in isolated cardiac myocytes. In addition, in isolated perfused rat hearts subjected to ischaemia–reperfusion injury (endogenous increase of ONOO⁻), we found enhanced MMP-2 activity and degradation of troponin I and myosin light chain-1 (Yasmin et al., 1997; Cheung et al., 2000; Wang et al., 2002b; Sawicki et al., 2005) as also observed in isolated guinea pig hearts exposed to ONOO⁻ (Rork et al., 2006). Therefore, our results from single cardiomyocytes suggest a possible intracellular action of MMPs and provide a plausible mechanism for the deleterious effects of ONOO⁻-mediated MMPs activation during ischaemia–reperfusion injury at the whole heart level.

ONOO⁻ can activate MMPs by different mechanisms. One of these is the S-glutathiolation of the propeptide domain which has been observed for MMP-1, -8 and -9 (Okamoto et al., 2001). Moreover, it was suggested that activation of MMPs by ONOO⁻ could be mediated by nitration of tyrosine residues in the hinge region and further unfolding of the pro-MMP (Rajagopalan et al., 1996) or by S-nitrosylation of the cysteine residue in the propeptide domain, as observed with MMP-9 on its activation by either ischaemia–reperfusion or following administration of nitric oxide donors (Gu et al., 2002). The effect of ONOO⁻ on MMP activity is biphasic as higher concentrations of ONOO⁻ have clearly been shown to inactivate MMPs (Owens et al., 1997; Okamoto et al., 2001). However, ONOO⁻ was shown to inhibit tissue inhibitors of MMPs (TIMP-1 and TIMP-2), which could increase net gelatinolytic activity in the cells (Frears et al., 1996; Chakraborti et al., 2004). Owing to technical limitations of the whole heart experiments, we could not determine the intracellular compartment(s) in which activation of MMPs occurred on exposure to ONOO⁻. However, ONOO⁻ may cause cellular injury through several mechanisms including protein modifications and the inactivation of proteins involved in Ca²⁺ handling, such as the nitration and inactivation of sarcoplasmic reticulum Ca²⁺ ATPase in porcine cardiac sarcoplasmic reticulum microsomes, with either 0.3 mM or 1 mM ONOO⁻ (Lokuta et al., 2005). Furthermore, ONOO⁻ inhibited the Na⁺/Ca²⁺ exchanger in microsomes prepared from bovine smooth muscle cells, an event accompanied by the activation of MMP-2 and inhibition of TIMP-2 (Chakraborti et al., 2004). Moreover, reactive oxygen species have deleterious effects on Ca²⁺ homoeostasis in cardiac myocytes through their action on a number of proteins including phospholamban and calcium channels (Guerra et al., 1996; Sulakhe et al., 1997; Eigel et al., 2004; Sharikabad et al., 2004). In this regard, we observed alterations in both diastolic and systolic calcium homoeostasis in the presence of ONOO⁻ and this effect was partially reversed by MMPs inhibitors. However, our study cannot determine which structure(s) related to calcium handling are susceptible to degradation by MMPs. Therefore, these results are consistent with the notion that ONOO⁻ alters intracellular calcium handling and contractility through mechanisms that involve activation of MMPs.

In addition, ONOO⁻ is known to alter proteins involved in cardiac contractility including myofibrillar, cytoskeletal and mitochondrial respiratory chain enzymes. These changes include post-translational modifications of several amino acids, such as cysteine oxidation of creatine kinase, tyrosine nitration of cytoskeletal α-actinin and desmin along with modifications in cysteine and tyrosine residues and Fe–S bridges of mitochondrial proteins (see for review Pacher et al., 2007). These alterations are likely to contribute to contractile dysfunction that cannot be prevented by MMPs inhibition, as our study reports. Recently, we identified α-actinin as a further target of MMP-2 degradation in isolated hearts infused with ONOO⁻ (Sung et al., 2007) and it is tempting to speculate that ONOO⁻-induced post-translational modifications may enhance the targeting of susceptible proteins to proteolysis by MMP-2.

Ishida et al. (1996) showed that infusion of 200 μM ONOO⁻ caused contractile dysfunction in isolated neonatal cardiac myocytes and this effect was accompanied by impaired Ca²⁺ homoeostasis. In our model using freshly isolated adult cardiac myocytes, we also observed abnormalities in Ca²⁺ homoeostasis due to ONOO⁻. The MMP inhibitor doxycycline delayed the onset but did not prevent the impairment in Ca²⁺ homoeostasis caused by ONOO⁻. However, we observed in our experiments that the time to cease calcium transient was delayed by doxycycline treatment and also that there was a significant increase in Ca²⁺ transient amplitude at 5 min, suggesting that the contractile dysfunction observed in this model is related to systolic Ca²⁺ abnormalities and also to a loss of sensitivity to Ca²⁺ in the contractile machinery. The loss of sensitivity of the cardiac myofibrils for Ca²⁺ upon ONOO⁻ challenge has been shown in isolated perfused rat hearts (Brunner and Wolkart, 2003).

In conclusion, we have demonstrated that inhibition of MMPs partially reduces contractile dysfunction in isolated cardiac myocytes resulting from their direct exposure to ONOO⁻. The effect of MMP activation in these cells is independent of a proteolytic action on extracellular matrix proteins. This provides further insight into the pathological mechanisms related to myocardial injury following oxidative stress. Detoxifying ONOO⁻ or inhibiting the activation of MMPs or their enzymatic activity may be an effective therapeutic strategy to lessen the impact of oxidative stress on cardiomyocyte function in pathological conditions such as ischaemia–reperfusion injury.

Abbreviations

CCT

contraction cease time

MMP-2

matrix metalloproteinase-2

MMPs

matrix metalloproteinases

Conflict of interest

The authors state no conflict of interest.