Cardioprotective effects of K_ATP channel activation during hypoxia in goldfish Carassius auratus

Many studies suggest that the activation of cardiac muscle ATP-sensitive potassium (K_ATP) channels can play a beneficial role during hypoxia and ischemia in mammals (Gross and Peart,2003; Grover and Garlid,2000; O'Rourke,2000). This conclusion is in accord with the original hypothesis of Noma (1983), who was the first to characterize sarcolemmal K_ATP (sarcK_ATP)channels. These are activated when intracellular oxygen levels and ATP/ADP ratios are low, although the precise function of sarcK_ATP channels in cardiac muscle is unknown (Flagg et al.,2004). Since the subsequent discovery that K_ATPchannels are also located in the mitochondrial inner membrane(mitoK_ATP; Inoue et al.,1991), there has been considerable debate as to the relative roles played by sarcK_ATPvs mitoK_ATP channels in cardioprotection and ischemic preconditioning(Gross and Peart, 2003; Toyoda et al., 2000). K_ATP channel activation under depleted oxygen conditions is often linked to the prior production of nitric oxide (NO), which acts through a cGMP-dependent mechanism (Dawn and Bolli,2002; Wang et al.,2001).

The observation that hypoxia-induced K_ATP channel activation is beneficial in mammals raises the possibility that a similar response in aquatic ectotherms could promote tolerance of environmental hypoxia. SarcK_ATP channel currents have been recorded in isolated myocytes from the hearts of several teleost species(Ganim et al., 1998; Paajanen and Vornanen, 2002),and there is evidence for the existence of cardiac mitoK_ATPchannels in fish (MacCormack and Driedzic,2002). In addition, nitric oxide synthase (NOS) has been localized in the hearts of goldfish (Bruning et al.,1996), and NO has been shown to play a critical role, viacGMP, in regulating the heart and peripheral vasculature in a variety of teleost species (Imbrogno et al.,2003; Jennings et al.,2004; Pellegrino et al.,2003). In goldfish, NO synthesized in cardiac myocytes plays a role in sarcK_ATP channel activation during moderate hypoxia(Cameron et al., 2003). Although it has been proposed that hypoxia-induced K_ATP channel activation, whether in the heart or in the brain, contributes to the enhanced tolerance of environmental oxygen depletion exhibited by many ectothermic vertebrates (Cameron and Baghdady,1994; Pek-Scott and Lutz,1998), this possibility has not been directly tested in fish.

The purpose of the present study, then, was to determine whether sarcK_ATP or mitoK_ATP channel activation exerts a cardioprotective effect during hypoxia in goldfish. This species is highly tolerant of depleted environmental oxygen; indeed, the congeneric crucian carp Carassius carassius retains normal cardiac function for 5 days of complete anoxia (Stecyk et al.,2004). A cellular model of environmental hypoxia was used to examine the effects of altered channel activity on isolated myocytes. Our hypothesis was that the NO- and cGMP-dependent activation of cardiac K_ATP channels would promote survival of isolated cells, thereby contributing to a suite of mechanisms that dramatically enhance tolerance of hypoxia in these fish. Preliminary results of this work have appeared in abstract form (Zhu et al.,2004).

Goldfish Carassius auratus L. were obtained from local suppliers and acclimated at 21°C in filtered, aerated 200 liter aquaria under a 12 h:12 h light:dark cycle. On the day of an experiment, goldfish of either sex were anesthetized by immersion in a solution of MS-222 (tricaine methanesulfonate, 0.2 g l^–1; pH 7.6). The heart was rapidly excised and prepared for standard intracellular recording using the intact ventricle, or for enzymatic dispersion of individual myocytes.

Animals used for conventional intracellular recording ranged in size from 9.5–28.1 g (mean ± s.e.m.,18.1±4.8 g). Procedures have been previously described(Cameron et al., 2003);briefly, the entire heart was mounted in a tissue bath and superfused with normoxic saline solution at a rate of 15 ml min^–1. This solution contained (in mmol l^–1) 150 NaCl, 5 KCl, 1.2 NaH₂PO₄, 1.2 MgSO₄, 1.8 CaCl₂, 10 Hepes and 10 glucose; pH was maintained at 7.6 and temperature held at 21±1°C.

Standard glass microelectrodes filled with 3 mol l^–1 KCl were connected to a differential amplifier (WPI, Sarasota, FL, USA), and spontaneous intracellular action potentials (APs) were recorded from the most superficial layer of ventricular muscle. APs were digitized (PowerLab;ADInstruments, Mountain View, CA USA) and analyzed using appropriate software(Chart, Peak Parameters; ADI). Parameters measured were action potential amplitude, resting membrane potential, slope of action potential upstroke phase and action potential duration at various levels of repolarization.

Oxygen partial pressure in the tissue bath was monitored using an oxygen meter (Cameron Instruments, Port Aransas, TX, USA) while moderate hypoxia(6.1±0.2 kPa) was induced by switching the superfusate from air-bubbled saline (20 kPa) to a glucose-free solution gassed with 100% N₂. The new condition was reproducible, reversible, and achieved in the tissue bath within 1 min. Electrophysiological responses to the acute onset of hypoxia and to reoxygenation were first assessed in the absence of any drug. Recordings were collected after 10–30 min exposure to hypoxia, and again 20 min after a return to normoxic solution.

Drugs, including the nitric oxide (NO) donor S-nitroso-N-acetylpenicillamine (SNAP; 100 μmol l^–1), the nitric oxide synthase (NOS) inhibitor Nω-Nitro-l-arginine methyl ester (l-NAME;50 μmol l^–1), the K_ATP channel antagonist glibenclamide (5 μmol l^–1), the mitoK_ATPchannel agonist 3-methyl-7-chloro-1,2,4-benzothiadiazine-1,1-dioxide(diazoxide; 100 μmol l^–1), the stable and cell-permeable cGMP analog 8-bromoguanosine 3′:5′-cyclic monophosphate(8-Br-cGMP; 100 μmol l^–1), and the mitoK_ATPchannel inhibitor 5-hydroxydecanoic acid (5-HD; 100 μmol l^–1), were purchased from Sigma (St Louis, MO, USA). Diazoxide was first dissolved in a drop of DMSO. Following the initial period of equilibration in normoxic saline, the effects of SNAP, 8-Br-cGMP and diazoxide on ventricular action potential parameters were recorded after 10 min exposure. In other experiments, glibenclamide or l-NAME was made up in hypoxic solution and applied to the tissue bath as described above. Dosages were chosen from among those recently employed in comparable studies.

Animals used for this procedure ranged in size from 23.7 to 64.9 g (mean± s.e.m., 43.3±12.4 g). Cell isolation procedures were modified from those of Karttunen and Tirri(1986), as previously described (Cameron et al.,2003). Upon removal, the heart was bathed in a chilled Ca²⁺-containing saline buffer (μmol l^–1): 137 NaCl, 4.6 KCl, 3.5 NaH₂PO₄, 1.2 MgCl₂, 11 glucose, 10 Hepes and 0.025 CaCl₂. The atrium was tied off and an olive-tipped cannula (27 gauge) was inserted into the ventricle through the opening of the bulbus arteriosus. The ventricle was then perfused with an identical but nominally Ca²⁺-free saline solution for 5 min,followed by 25 minperfusion with an enzyme buffer containing 50 ml Ca²⁺-free saline, 35 mg collagenase (Type I), 25 mg trypsin and 25 mg bovine serum albumin (Fraction V). The heart was removed from the perfusion apparatus and carefully torn into small pieces. After 5 min of swirling agitation followed by 5 min of gentle trituration with a Pasteur pipette, the supernatant containing single free-floating cells was removed. Fresh buffer was added, and the procedure repeated to obtain a second and third fraction of cells. This procedure produced a high yield (80–90%) of viable,elongated myocytes.

The method used to assess cellular viability in hypoxia was derived from the in vitro model of cellular ischemia developed by Vander Heide et al. (1990), as modified by Sato et al. (2000). Portions of the myocyte suspension (0.5 ml) were pipetted into 5 ml test tubes, and 0.5 ml of saline buffer or drug solution was added. The hypoxic condition was induced in some tubes by layering an additional 0.5 ml of mineral oil on top of the myocyte suspension, preventing gaseous diffusion of oxygen. After 60 min incubation, 0.5 ml of myocytes were sampled from each tube and mixed with 0.5 ml of staining solution containing 0.5% gluteraldehyde and 0.5% Trypan Blue in saline buffer. The resulting ml of suspended cells was placed on a slide and examined by phase contrast microscopy at 100×. Elongated myocytes with a normal morphological appearance were included in the analysis,while dead, `rounded-up' cells were excluded(Vander Heide et al., 1990). These dead cells were always blue, and were presumed to have been damaged during the enzymatic isolation procedure. Viability of the elongated cells was assessed by their capacity to exclude Trypan Blue(Fig. 1); the percentage staining clearly after 3 min exposure to the staining solution was recorded for each experimental condition. Myocytes staining dark blue were considered irreversibly damaged. Drugs and concentrations were identical to those used in the intracellular experiments; they were made up in saline buffer on the day of an experiment and were present throughout the period of in vitrohypoxia.

Data obtained from the intracellular and cellular hypoxia experiments were transferred to Microsoft Excel spreadsheets; t-tests and analysis of variance for repeated measures (ANOVA with post-hoc Fisher PLSD and Scheffe F-test; Statview) were used to determine statistical significance (P<0.05) among group means; results were expressed as means ± s.e.m.

Transmembrane intracellular recording techniques were used to monitor the effects of moderate hypoxia (6.1±0.2 kPa) on the configuration of ventricular action potentials in goldfish heart. Cardiac action potential duration (APD) was decreased in all experiments (N=5). As previously observed (Cameron et al.,2003), there was a significant (P<0.01) reduction in APD at 90% of full repolarization (APD₉₀) after 10 min of exposure to substrate-free hypoxia (Fig. 2). Mean APD₉₀ under normoxic conditions was 575±24 ms; after exposure to hypoxia, this value fell to 488±12 ms, a decrease of 15.1%. APD at 50% of full repolarization (APD₅₀)was not significantly reduced. There were no significant effects of hypoxia on any other electrophysiological parameters, including resting membrane potential, action potential amplitude or slope of action potential upstroke.

To examine the possibility that APD shortening in hypoxia was caused by the activation of K_ATP channels, the effects of the channel antagonist glibenclamide (5 μmol l^–1) on hypoxia-induced shortening were monitored in isolated hearts. Simultaneous exposure to both glibenclamide and hypoxia abolished the reduction in APD characteristic of hypoxia alone(N=3; Fig. 3).

To determine whether NO plays a role in cellular responses to hypoxia in goldfish, the effects of the NOS inhibitor l-NAME were studied in isolated hearts and in ventricular myocytes. Hypoxia-induced shortening of APD₉₀ was eliminated by previous exposure to 50 μmol l^–1l-NAME (N=4; Fig. 3).

To determine whether NO and cGMP could mimic cellular responses to hypoxia,the effects on action potential parameters of the NO donor SNAP and the stable cGMP analog 8-Br-cGMP were monitored under normoxic conditions. At a concentration of 100 μmol l^–1, SNAP significantly(P<0.05) reduced APD₉₀ after 10 min exposure(N=4; Fig. 4). With prolonged exposure to SNAP, APD was further reduced; at 60 min, both APD₅₀ and APD₉₀ were significantly decreased. There were no significant effects of SNAP on any other action potential parameter.

At a concentration of 200 μmol l^–1, 8-Br-cGMP also shortened ventricular APD under normoxic conditions (P<0.05). APD₉₀ was reduced by 17.6% (N=3; Fig. 4), although APD₅₀ remained unchanged. At a concentration of 100 μmol l^–1, diazoxide had no significant effect on ventricular APD₉₀ (N=6; Fig. 4).

A cellular model of environmental hypoxia was used to assess the potential cardioprotective effects of K_ATP channel activation. Under normoxic conditions, 37.8±4.0% of elongated, isolated ventricular myocytes from goldfish were unable to exclude Trypan Blue, suggesting cellular injury(N=6; Fig. 5). After 60 min hypoxia, however, the incidence of Trypan Blue-positive cells was significantly (P<0.001) increased compared to that observed in normoxia. The percentage of stained cells in hypoxia was 81.2±2.6, a 115% increase. More than 500 myocytes were individually evaluated.

Following simultaneous exposure of isolated myocytes to l-NAME and 60 min hypoxia, the percentage of Trypan Blue-stained cells was moderately increased relative to that seen with hypoxia alone(Fig. 6). After 60 min exposure of isolated myocytes to hypoxia plus SNAP, however, the percentage of Trypan Blue-stained cells was markedly reduced compared to that observed with hypoxia alone. To determine whether this cardioprotective action of SNAP was linked to the hypoxia-induced production of NO in muscle cells, the effect of SNAP in combination with l-NAME also was tested. Blockade of NO synthesis by l-NAME entirely eliminated the reduction in stained cells observed with SNAP and hypoxia alone (Fig. 6).

In isolated cells exposed to hypoxia, diazoxide exerted a marked cardioprotective influence. Isolated myocytes exposed to diazoxide during 60 min hypoxia survived in greater numbers than did those affected by hypoxia alone (Fig. 7).

In mammals, diazoxide and 5-HD have been shown to be selective for mitoK_ATP over sarcK_ATP channels — the former as an activator and the latter as an inhibitor(Mannhold, 2004). In goldfish,the beneficial effects of diazoxide in preserving isolated myocytes was reduced, but not eliminated, by 5-HD (Fig. 7). In contrast, 5-HD did not alter the cardioprotective effect of 8-Br-cGMP; the reduction in the percentage of stained cells in hypoxia that was produced by 8-Br-cGMP alone was not affected when this agent was given in combination with the presumed mitoK_ATP blocker.

The present findings support the hypothesis that hypoxia-induced activation of cardiac sarcolemmal or mitochondrial K_ATP channels in goldfish protects muscle cells during metabolic stress, possibly enhancing whole-animal tolerance of environmental hypoxia. The cardioprotective activation of K_ATP channels in this species, whether in whole hearts or isolated myocytes, is linked to the production of nitric oxide, which acts through a cGMP-dependent mechanism.

In the present study we confirm our previous observation(Cameron et al., 2003) that transmembrane action potential duration is significantly shortened during moderate hypoxia in intact, isolated goldfish heart(Fig. 2). In mammals, cardiac APD shortening in hypoxia is thought to arise through the activation of sarcK_ATP channels (Flagg et al.,2004; Shigematsu and Arita,1997; Suzuki et al.,2002). Patch clamp studies also have revealed a hypoxia-induced increase in the activity of sarcK_ATP channels in both mammals(Budas et al., 2004; Liu et al., 2001) and goldfish(Cameron et al., 2003). We now show that in fish, the APD shortening can be eliminated by simultaneous exposure of the tissue glibenclamide (Fig. 3), an antagonist of both sarc- and mitoK_ATP channels. The hypoxia-induced reduction in APD was also abolished by l-NAME,an inhibitor of NO synthase, suggesting an upstream role for NO in the activation of sarcK_ATP channels(Fig. 3). Further, APD shortening can be similarly induced by the NO donor SNAP, as well as by a stable analog of cGMP (Fig. 4). NO has previously been shown to activate sarcK_ATP channels in mammalian cardiac muscle (Chen et al.,2000; Moncada et al.,2000), and this response has been linked to cGMP(Chen et al., 2000; Han et al., 2001). In contrast, the mitoK_ATP activator diazoxide had no significant effect on APD.

In an attempt to demonstrate that hypoxia-induced and NO-dependent activation of K_ATP channels promotes the survival of ventricular myocytes, we used an in vitro model of environmental hypoxia. Similar methods have been used to simulate the effects of ischemia on isolated mammalian cells, and to demonstrate, for example, the cardioprotective activation of mitoK_ATP channels by diazoxide(Sato et al., 2000) or volatile anesthetics (Zaugg et al.,2002). In each study, the viability of apparently intact,elongated myocytes was assessed by their capacity to exclude Trypan Blue dye.

The present results indicate that ventricular myocytes from goldfish are damaged after 60 min of in vitro hypoxia under mineral oil. As illustrated in Fig. 5, cells exposed to the low-oxygen environment were significantly more often positive for Trypan Blue than those maintained under normoxic conditions. However, the cellular damage caused by hypoxia could be reduced by agents thought to activate K_ATP channels. Cells exposed to hypoxia in the presence of SNAP or 8-Br-cGMP, both of which reduced APD in the intact heart, were much more likely to survive and exclude the dye than were cells exposed to hypoxia alone (Figs 6, 7). In contrast, the NOS inhibitor l-NAME, which prevented hypoxia-induced APD shortening,increased the numbers of stained, damaged cells. Perhaps surprisingly, l-NAME also eliminated the cardioprotective influence of SNAP,suggesting that SNAP promotes the endogenous synthesis and release of NO from cardiac muscle cells. Confounding this analysis is the fact that hypoxia itself may increase the NO concentration in cardiac muscle through activation of NOS isoforms (which vary from one species to the next), or through NOS-independent pathways (Schulz et al.,2004). SNAP has been shown to increase the concentration of cGMP in mammalian myocardium (Qin et al.,2004).

The present data strongly suggest that NO, in its capacity to activate sarcK_ATP channels in goldfish, promotes survival of cardiac muscle cells during hypoxia. The cardioprotective, cGMP- and K_ATP-dependent influence of NO in hypoxia has been recently reviewed (Kolar and Ostadal,2004; Schulz et al.,2004). In mammalian myocytes, NOS is upregulated during ischemia(Wang et al., 2002), and a similar enhancement of NOS activity occurs in response to hypoxia in the vasculature of an elasmobranch fish(Renshaw and Dyson, 1999).

While a cardioprotective role for K_ATP channel activation has been described by many investigators, there remains some question as to the relative roles of sarcolemmal vs mitochondrial channels(Gross and Peart, 2003). In the present study, the cellular damage induced by hypoxia could be reduced by agents thought to specifically activate either sarcK_ATP or mitoK_ATP channels. SNAP and 8-Br-cGMP, which reduce APD through activation of sarcK_ATP channels, also promote tolerance of hypoxia in isolated cells (Figs 6, 7). Diazoxide, which is thought to act primarily at mitoK_ATP, also decreased the numbers of stained, damaged myocytes (Fig. 7). In the present study, diazoxide did not affect APD(Fig. 4), supporting the view that its effect is primarily at the mitochondrial channel. However, a recent study involving rat ventricular myocytes during metabolic inhibition indicates that this agent can open sarcK_ATP channels as well, probably by an indirect mechanism (Rodrigo et al.,2004). Finally, 5-HD, a specific inhibitor of mitoK_ATPin mammals, was found to reduce (but not eliminate) the cardioprotective influence of diazoxide, but to have no effect on the beneficial response to 8-Br-cGMP.

Data arising from mammalian studies of hypoxia and ischemic preconditioning also suggest that activation of either sarcK_ATP or mitoK_ATP channels can be cardioprotective(Sanada et al., 2001; Tanno et al., 2001; Toyoda et al., 2000). Studies of transgenic mice expressing a mutant K_ATP channel with reduced ATP-sensitivity suggest that the sarcolemmal channel, at least in mammals, is required for optimal response to ischemia(Rajashree et al., 2002). In pharmacological studies, sarcK_ATP activation protects against ischemic injury (Suzuki et al.,2002), and the present data suggest a similar action against hypoxia in fish (Fig. 6). In contrast, Sato et al. used an in vitro model of cellular ischemia,very similar to that employed in the present study, to show that activation of mitoK_ATP channels, and not sarcK_ATP channels, was responsible for ischemic cardioprotection in rabbit myocytes(Sato et al., 2000). Indeed,many recent studies, most involving diazoxide, 5-HD and other agents presumed to be selective for cardiac sarcK_ATPvsmitoK_ATP channels, implicate the latter subtype in protection against ischemia (Murata et al.,2001; Oldenburg et al.,2002; Uchiyama et al.,2003).

There has been considerable speculation as to the specific benefits provided to the myocardium by sarcK_ATP or mitoK_ATPchannel activation (Gross and Fryer,1999; Oldenburg et al.,2002). Recent evidence suggests that NO/cGMP-dependent cardioprotection involves mitoK_ATP channels, not as an end-effector, but as an element in a cascade of events leading to the production of reactive oxygen species (ROS; Lebuffe et al., 2003; Qin et al., 2004; Xu et al., 2004).

One factor that hinders the interpretation of data in studies such as these is that many of the pharmacological agents typically used, including several of those employed in the present study, may protect the heart by means unrelated to ion channel activation. It has been reported that the cardioprotective influence of SNAP, for example, may involve the mitochondria and be unrelated to its capacity to affect sarcK_ATP channels(Rakhit et al., 2001). Both diazoxide and 5-HD can have beneficial but channel-independent effects on mitochondrial metabolism (Dzeja et al.,2003; Hanley et al.,2002), leading at least one group to challenge the existence of diazoxide- and 5-HD-sensitive K_ATP channels in mammalian mitochondria (Das et al.,2003). In addition, preliminary data from our laboratory suggest that 5-HD eliminates the reduction in APD characteristic of hypoxia,suggesting that this agent is not specific to mitoK_ATP channels in fish, or that other channels are affected. The present data cannot exclude the possibility that diazoxide benefits goldfish myocytes by means other than mitoK_ATP channel activation. This could underlie our observation that 5-HD reduced, but did not completely eliminate, the cardioprotective effects of diazoxide (Fig. 7).

It is now widely accepted that the NO- and cGMP-dependent activation of sarcK_ATP and/or mitoK_ATP channels is cardioprotective under conditions of metabolic stress in mammals, and that channel activity underlies the phenomenon of ischemic preconditioning(Gross and Peart, 2003; Kolar and Ostadal, 2004). Aquatic ectotherms may be regularly exposed to acute or chronic environmental hypoxia. Moreover, both sarcK_ATP(Cameron et al., 2003; Paajanen and Vornanen, 2002)and mitoK_ATP (MacCormack and Driedzic, 2002; MacCormack et al., 2003) channels exist in the hearts of teleost fish, and the former are specifically activated in hypoxia by a NO-dependent mechanism(Cameron et al., 2003). The present findings support the hypothesis that sarcK_ATP or mitoK_ATP channel activation promotes the survival of ventricular myocytes exposed to conditions of depleted oxygen in vitro. A similar channel activation in critical tissues in vivo may extend whole-animal tolerance of hypoxia.

The authors gratefully acknowledge the assistance of Lilian G. Perez and Natasha K. Soodoo in conducting some experiments. This project was funded in part by a grant from the National Science Foundation (DBI-0097499) and by Brachman Hoffman and other grants from Wellesley College.