Sarcolemmal ion currents and sarcoplasmic reticulum Ca²⁺content in ventricular myocytes from the cold stenothermic fish, the burbot(Lota lota)

Ectotherms are able to acclimatize to different temperatures, adjusting physiological and biochemical processes to meet the demands of their environment. Acclimatization of the heart is especially important as cardiac output must match the changes in activity level, metabolic rate and blood viscosity that occur with changes in temperature. The effects of temperature acclimation on cardiac contractility in ectotherms are well documented(Keen et al., 1994; Shiels and Farrell, 1997; Aho and Vornanen, 1999) (for a review, see Driedzic and Gesser,1994). More recent work has focused on elucidating adaptations in cardiac myocyte excitability (Vornanen et al., 2002a; Paajanen and Vornanen, 2004), cellular Ca²⁺ cycling(Hove-Madsen and Tort, 1998; Hove-Madsen et al., 1998; Harwood et al., 2000, Shiels et al., 2000; Shiels et al., 2002a; Shiels et al., 2002b; Hove-Madsen et al., 2003) and protein structure (Yang et al.,2000; Gillis et al.,2000; Gillis et al.,2003) associated with maintained cardiac viability in the cold(for reviews, see Vornanen et al.,2002b; Gillis and Tibbits,2002). Although some ectotherms cope with cold temperatures via cold-torpor and resultant reductions in metabolic rate and heart rate, there are several compensative changes often associated with ectotherms that remain active in the cold. These include an increase in relative ventricular mass, possibly to offset increased blood viscosity(Goolish, 1987; Graham and Farrell, 1989), a proliferation of the sarcoplasmic reticulum (SR)(Bowler and Tirri, 1990)suggesting increased reliance on intracellular Ca²⁺ cycling during excitation-contraction coupling (Keen et al., 1994; Shiels and Farrell,1997; Aho and Vornanen,1999; Tiitu and Vornanen,2002b) and changes in K⁺ channel conductances that decrease action potential duration (APD) ensuring myocyte excitability in the cold (Vornanen et al., 2002a; Paajanen and Vornanen,2004).

Some ectothermic animals do not tolerate large seasonal increases in temperature, and therefore inhabit a cold stenothermic environment. Fish species such as the burbot Lota lota, are cold stenotherms spending most of their life at temperatures between 1°C and 7°C, and are rarely found in waters above 13°C (Carl,1995). Burbot are benthic, and are sluggish swimmers but are cold-active, spawning in winter under ice-covered lakes(Pääkkönen and Marjomäki, 2000). Examination of burbot heart morphology and contractility suggests that the changes normally associated with cold-acclimation in active species such as rainbow trout (Oncorhynchus mykiss), may be routine for cold-adapted species. The relative ventricular mass of the burbot (∼0.15% body mass) is elevated in comparison with most eurythermal species (∼0.08% body mass)(Tiitu and Vornanen, 2002a). Both atrial and ventricular muscle isolated from burbot heart exhibit increased ryanodine-sensitivity of contraction, suggesting that the SR may be routinely involved in delivering Ca²⁺ to the myofilaments during force development (Tiitu and Vornanen,2002b). Furthermore, [³H]ryanodine binding to cardiac preparations from burbot and rat show similar Ca²⁺-dependent activation of the SR Ca²⁺ release channel, suggesting that Ca²⁺-induced Ca²⁺-release (CICR) may operate during excitation-contraction coupling in this species(Vornanen, 2006). Collectively, these results suggest that many of the subcellular changes that are required to maintain cardiac function during cold acclimation may also play a role in long-term cold adaptation. However, at present no studies have examined excitation-contraction coupling at the level of the myocyte in the burbot or any cold stenothermic species.

In this study, we investigated the electrophysiological properties of burbot ventricular myocytes at 4°C, which is the typical habitat temperature of this species for most of the year. We also examined electrophysiological parameters after acutely warming the myocytes to 11°C because we were interested in how excitation-contraction coupling in the stenothermic heart is modulated during acute temperature change and because 11°C is approaching the upper temperature at which this species is found(Carl, 1995; Pääkkönen and Marjomäki, 2000). We first set out to measure APs and the major sarcolemmal (SL) ion currents involved in regulating myocyte excitability and maintaining electrical stability in the cold. Next, to investigate possible changes in cellular Ca²⁺ dynamics in a cold-adapted species we examined SR Ca²⁺ accumulation and release using caffeine. We assumed that if the changes found in rainbow trout and other active teleosts under cold-acclimation were adaptive, then, through evolutionary processes, similar changes may be permanent in the genome of cold stenothermic fish. First, we hypothesized that reorganisation of K⁺currents would occur with a shift from the dominance of the inward rectifier K current (I_K1) in favour of the delayed rectifier K⁺ current (I_Kr)(Vornanen et al., 2002a). Second, we expected that intracellular Ca²⁺ stores of the SR would make a significant contribution to excitation-contraction coupling(Tiitu and Vornanen, 2002b)and appear as accelerated decay of the l-type Ca²⁺current (I_Ca) (Shiels et al., 2002b). In agreement with the first hypothesis, I_K1 was small and I_Kr large in ventricular myocytes of the burbot heart. However, we did not find evidence of increased SR Ca²⁺ involvement in our measurements of I_Ca inactivation. Rather, we report an increase in SL Na⁺-Ca²⁺ exchange (NCX) activity, which we suggest is the primary pathway for SL Ca²⁺ influx in this species.

Sexually mature burbot Lota lota L. (body mass 224.6±12.7 g, N=22) of both sexes were caught during spawning time from Lake Orivesi (62°30′N) in Finland. In the laboratory fish were held in 500 liter stainless steel tanks at 4°C with continuous circulation(approximately 0.5 l min^-1) of aerated groundwater. Fish were fed with dead vendace (Coregonus albula) three times a week. Photoperiod was 15 h:9 h dark:light.

All procedures were in accordance with local animal handling protocols. A detailed description of myocyte preparation has been previously published for other fish species (Vornanen,1997; Shiels et al.,2000). Briefly, fish were stunned with a blow to the head, the spine was cut just behind the brain and the heart was excised. The heart was then perfused first with an isolating solution for 8-10 min, and then with a proteolytic enzyme solution for 15 min at ∼15°C. After enzymatic treatment, the ventricle was placed in isolating solution, cut into small pieces with scissors and then triturated through the opening of a Pasteur pipette to free individual myocytes. The myocytes were stored in fresh isolating solution at 4°C and used within 8 h.

The isolating solution contained (mmol l^-1): NaCl 100, KCl 10,KH₂PO₄ 1.2, MgSO₄ 4, taurine 50, glucose 20,and Hepes 10, adjusted to pH 6.9 with KOH. For enzymatic digestion,collagenase (Type IA from Sigma, St Louis, MI, USA; 0.75 mg ml^-1),trypsin (Type IX from Sigma; 0.5 mg ml^-1) and fatty acid-free bovine serum albumin (BSA, from Sigma; 0.75 mg ml^-1) were added to this solution.

The external solution used for measuring ventricular action potentials (AP)contained (mmol l^-1): NaCl 150, KCl 3, MgSO₄ 1.2,NaH₂PO₄ 1.2, CaCl₂1.8, glucose 10 and Hepes 10, adjusted to pH 7.6 with NaOH. The external solution used for measuring K⁺ currents contained (mmol l^-1): NaCl 150, KCl 5.4,CaCl₂ 1.8, MgCl₂ 1.2, glucose 10 and Hepes 10, adjusted to pH 7.6 with NaOH at 20°C. Specific inhibition of the rapid component of the delayed rectifier K⁺ current (I_Kr) was accomplished with E-4031 (1 μmol l^-1; Alomone Labs Ltd,Jerusalem, Israel). The external solution used for measuring Na⁺-Ca²⁺ exchange current (I_NCX) and l-type Ca²⁺ current (I_Ca) contained(mmol l^-1): NaCl 150, CsCl 5.4, MgSO₄ 1.5,NaH₂PO₄ 0.4, CaCl₂ 1.8, glucose 10 and Hepes 10, adjusted to pH 7.6 with CsOH. Unless otherwise stated, 0.5 μmol l^-1 TTX (Tocris Cookson, Bristol, UK), 10 μmol l^-1nifedipine and 100 μmol l^-1 ouabain (both from Sigma) were included to block Na⁺ channels, l-type Ca²⁺channels and Na⁺/K⁺ ATPase, respectively, when recording I_NCX. Nifedipine and ouabain were omitted when recording I_Ca.

The pipette solution used during K⁺ current experiments contained (mmol l^-1): KCl 140, MgATP 4, MgCl₂ 1, EGTA 5 and Hepes 10, adjusted to pH 7.2 with KOH. Pipette solutions for measurement of I_NCX contained (mmol l^-1): CsCl 140,MgCl₂ 1, CaCl₂ 9, BAPTA 20, Na₂ATP 5,Na₂GTP 0.03 and Hepes 10, adjusted to pH 7.2 with CsOH at 20°C. The free intracellular Ca²⁺ concentration of this solution was calculated (MaxChelator) to be 179.5 and 186.6 nmol l^-1 at 4°and 11°C, respectively. Under these conditions, intracellular calcium is buffered to a diastolic level. In some experiments, we investigated the effect of using a lower level of Ca²⁺ buffering on I_NCX by replacing BAPTA with 0.025 mmol l^-1EGTA. I_Ca was initially characterised with pipettes containing (mmol l^-1): CsCl 130, MgATP 5, tetraethylammonium chloride (TEA) 15, MgCl₂ 1, oxaloacetate 5, EGTA 5,Na₂GTP 0.03 and Hepes 10 adjusted to pH 7.2 with CsOH. In latter experiments I_Ca and SR Ca²⁺ loading were assessed using the same pipette solution except that EGTA concentration was decreased from 5 mmol l^-1 to 0.025 mmol l^-1 to better mimic in vivo cytosol Ca²⁺ buffering(Hove-Madsen and Tort, 1998). Steady-state kinetics parameters of I_Ca were obtained by fitting activation and inactivation data to Boltzman functions to determine the half-activating, half-inactivating potentials (V_h) and the slope (k) of activation and inactivation, as previously described(Vornanen, 1998).

Intracellular APs were measured from spontaneously beating whole-heart preparations at 4°C and 11°C as described previously(Vornanen, 1996). Briefly, the excised ventricle was medially opened, spread and secured on the bottom of a 10 ml tissue chamber filled with oxygenated saline. Ventricular APs were recorded using high-resistance microelectrodes (30-60 MΩ when filled with 3 mol l^-1 KCl) fabricated from borosilicate glass (World Precision Instruments, 1BBL, Sarasota, FL, USA) with a two-stage horizontal puller (Campden Instruments Ltd, UK). Microelectrode signals were recorded using a high-impedance amplifier (KS-700, WPI, Sarasota, FL, USA), digitized(DigiData 1200, Axon Instruments, Foster City, CA, USA) and stored to a computer using Axotape 2.2 acquisition software and then analysed offline(Clampfit, Axon Instruments). The time course of contraction was recorded simultaneously with APs by attaching one corner of the ventricle to a force transducer (FT03 Grass Instruments, West Warwick, RI, USA) by a small metal hook and braided silk thread. Muscle was slightly tensioned and the force signal was amplified by a Grass 7D polygraph amplifier and fed through the digitizer to the computer for later off-line analysis.

Stimulation, acquisition and analysis of ventricular myocyte whole-cell voltage and current signals was achieved using established methods(Vornanen, 1997; Shiels et al., 2000; Paajanen and Vornanen, 2002)on either an Axopatch 1D amplifier in conjunction with pClamp 8.2 and Clampfit software, or on an EPC-9 amplifier in conjunction with Pulse 6.3 and Pulsefit software (Heka, Lambrecht, Germany). Myocytes (capacitance 23.7±0.5 pF, N=179 cells) were placed in the recording chamber (RCP-10T, Dagan,Maryland, MI, USA, volume 500 μl or RC-26, Warner Instruments Corp. Brunswick, Handen, CT, USA, volume 150 μl) and were superfused continuously with external saline at the rate of 1.5-2 ml min^-1. The temperature of the saline was regulated at either 4±1°C or 11±1°C by circulating water baths or a Peltier device. Bath temperature was continuously monitored by thermocouples positioned no less than 5 mm from the cell under investigation. Patch electrodes were pulled from borosilicate glass (Garner F-78045, Claremont, CA, USA) with a two-stage vertical puller (L/M-3P-A, List Medical, Darmstadt, Germany). The resistance of the electrodes was 2-4 MΩ when filled with pipette solutions. Pipette and whole cell capacitances were routinely compensated, and access resistance was recorded but not compensated. Currents were filtered at 2.0 kHz using either a 4-pole or a 8-pole Bessel filter.

During SR Ca²⁺ loading and release experiments, rapid (∼50 ms) application of caffeine and/or channel inhibitors was achieved by switching between temperature controlled barrels of a rapid solution changer(RS200, Biologic, Claix, France). CdCl₂ (100 μmol l^-1) or a combination of CdCl₂ (30 μmol l^-1) and verapamil (10 μmol l^-1) was used to rapidly block I_Ca.. NiCl₂ (10 mmol l^-1) was used to inhibit the NCX. SR Ca²⁺ content was assessed by the application of caffeine (10 mmol l^-1), which induces the release of Ca²⁺ from the SR. This Ca²⁺ is then extruded from the cell via the NCX generating an inward current (see Results), which is directly proportional to the Ca²⁺ released from the SR(Varro et al., 1993). The time integral of this caffeine-induced I_NCX current was used to calculate the SR Ca²⁺ content (in pC) at the time of caffeine application. This value was expressed per unit capacitance (pC pF^-1). SR Ca²⁺ content was also expressed in μmol Ca²⁺ l^-1 non-myofibrillar cell volume [40% as determined previously (Vornanen, 1998)]. Cell volume was calculated from cell surface area, obtained by measurements of cell capacitance (pF) and assuming a specific membrane capacitance of 1.59μF cm^-2, and a surface-to-volume ratio of 1.15.

Details of the voltage clamp waveforms and protocols used to study the electrophysiological properties of whole-cell currents are provided in the results and the figures. When values are presented as means, the number of observations (N) and statistical significance are provided in the text or appropriate figure legend.

At 4°C the beating frequency of the burbot heart was 17.98±0.36 contractions min^-1 (N=10) (0.3 Hz). The duration of contraction and AP closely matched and were slightly less than 2 s at 4°C(Fig. 1). The close match between AP duration (APD) and the duration of contraction suggest that the activation and relaxation of cardiac twitch are governed by membrane potential. Peak ventricular force occurred when the AP had repolarised to about -20 mV. The resting membrane potential (RMP) and AP overshoot were-70.4±1.1 mV and 14.4±0.7 mV, respectively (N=10). AP duration at plateau (0 mV), and at 50, 90 and 100% repolarisation was 292±48, 1108±28, 1544±29 and 1830±46 ms,respectively (N=10). At 11°C, RMP and AP overshoot were-75.0±1.8 and 15.8±1.2 mV, respectively, whereas corresponding values of AP duration were 250±44, 771±39, 1159±61 and 1313±72 ms at 0 mV, and at 50, 90 and 100% repolarisation, respectively(N=12). The APD was significantly shorter at 11°C than 4°C with APD at 90% repolarization (APD₉₀) values of 1091±179 and 1594±16 ms, respectively (P<0.0002). Absolute force decreases with increasing temperature (Fig. 1). This was a regular finding and has been quantitatively reported previously (see Tiitu and Vornanen, 2002b).

The two main K⁺ currents in burbot ventricular myocytes are the background inward rectifier (I_K1) and the rapid component of the delayed rectifier current (I_Kr). The conductance of I_K1, which is the major K⁺ current in ventricular myocytes of most vertebrate species, was surprisingly low in burbot myocytes (39.2±5.4 pS pF^-1 at 4°C; Fig. 2A). Even at 11°C, I_K1 conductance of the burbot ventricular myocyte was only 71.4±1.7 pS pF^-1. In contrast, the size of I_Kr was large. At 4°C, the maximum density of the E4031-sensitive (1 μmol l^-1) tail current was 2.9±0.3 pA pF^-1 (Fig. 2B). When temperature was increased to 11°C, current density increased to 3.8±1.1 pA pF^-1 (P=0.04) and caused a 15 mV shift(P=0.04) of the current-voltage relation to hyperpolarising voltages. The slope conductance of I_Kr was 199±27 and 320±8 pS pF^-1 at 4°C and 11°C, respectively, i.e. 4.5-5 times that of the I_K1.

We investigated the efficacy of the NCX at two different levels of intracellular Ca²⁺ buffering. In the first series of experiments(Fig. 3A) we measured the NCX under conditions that held intracellular Ca²⁺ at diastolic levels(see Materials and methods). I_NCX was elicited at 4 s intervals from the calculated reversal potential of the exchanger (-26.5 mV)by ramp pulses (Fig. 3A,inset). I_NCX was measured as the Ni²⁺-sensitive current during the hyperpolarizing phase of the ramp. At 4°C, the conductance of I_NCX was 35.9±6.3 pS pF^-1between -40 and 20 mV and it increased with a Q₁₀ of 2.49±0.29 to 58.6±8.4 pS pF^-1 when temperature was increased to 11°C (Fig. 3A). The measured reversal potential of I_NCXwas -23.5±0.1 and -23.7±0.8 mV at 4°C and 11°C,respectively, which is close to theoretical equilibrium potential. Neither 2μmol l^-1 isoprenaline nor 10 mmol l^-1 caffeine had any effect on burbot I_NCX (not shown). Isoprenaline (10μmol l^-1) caused a small but non-significant (P=0.142, N=5) increase in I_NCX (not shown).

In the second series of experiments we reduced intracellular Ca²⁺ buffering to a more physiological level by replacing 20 mmol l^-1 BAPTA with 0.025 mmol l^-1 EGTA. Repolarising ramps from 30 to -100 mV were applied at every 4th second from a holding potential of -80 mV (see inset Fig. 3B). These experiments were preformed at 4°C only. The outward NCX current is significantly smaller in the weakly buffered pipette solution than in the heavily buffered solution given in Fig. 3A, with a mean conductance of 13.8±2.3 pS pF^-1at 10 mV. The mean reversal potential of the NCX current was -8.88±1.6 mV (N=9).

_Ca was elicited from a holding potential of -80 mV to voltages between -70 and +60 mV for 1 s, in the absence and presence of a saturating concentration of isoprenaline (10 μmol l^-1)(Fig. 4). The density of I_Ca measured using 5 mmol l^-1 EGTA, which will augment current amplitude, was only 0.81±0.13 pA pF^-1 at 4°C, and increased to 1.35±0.18 pA pF^-1 at 11°C(Q₁₀=2.08; P=0.01). The charge density(Q_Ca) of I_Ca was 0.24±0.043 pC pF^-1 and 0.21±0.034 pC pF^-1 at 4°C and 11°C, respectively. Temperature does not affect charge density(P=0.6), primarily due to the slowing of current decay at the colder temperature. Single exponential equations fit to the decay of the I_Ca under physiological buffering conditions (i.e. 25μmol l^-1 EGTA, Fig. 6C) provide a time constant (τ) of 211.9±12.9 ms and 155.8±19.7 ms, at 4°C and 11°C, respectively(P<0.001).

I_Ca was blocked by both CdCl₂ (100 μmol l^-1) and a combination of CdCl₂ (30 μmol l^-1) and verapamil (10 μmol l^-1) (see below). Isoprenaline increased peak I_Ca in the burbot ventricular myocytes by 65% and 95% at 4°C and 11°C, respectively(Fig. 4B,C).

The slow inactivation time constants and the long AP duration, especially at 4°C, suggest a prominent role for the I_Ca window current in this species. Steady-state activation and inactivation(Fig. 4D) and the I_Ca window current(Fig. 4E) was measured in burbot myocytes at 4°C using 5 mmol l^-1 EGTA in the pipette. The voltage at which inactivation was halfmaximal (V_h) and slope (k) that describes the Boltzman fit to inactivation are-11.95±1.37 mV and -11.14±0.89 (mean ± s.e.m., N=9), respectively. Corresponding values for steady-state activation are -10.89±1.20 mV and 10.40±1.43 mV, respectively. These results indicate slow transition between activation and inactivation at 4°C and result in the large window current given in Fig. 4E.

The Ca²⁺ stores of the burbot SR were first released with caffeine so that all myocytes started with a negligible SR Ca²⁺content. SR Ca²⁺ was then replenished with a series of stimulus pulses which, under control conditions, consisted of 25 square pulses from -80 to +10 mV for 600 ms at a frequency of 0.2 Hz. SR Ca²⁺ accumulation was assessed by recording the NCX current generated upon the re-application of caffeine and calculating its time integral. Representative recordings of I_NCX and its time integral at 4°C and 11°C are given in Fig. 5A,B. The mean values for SR Ca²⁺ content in burbot myocytes at 4°C and 11°C are given in Fig. 5Cand are expressed as charge (pC) normalized to myocyte capacitance (pF). The ability to load Ca²⁺ into the SR is not significantly affected by acute warming with Ca²⁺ content being 124±23 μmol l^-1 at 4°C and 165±33 μmol l^-1 at 11°C. This small steady-state SR Ca²⁺ content in burbot cells was not the result of incomplete Ca²⁺ release during the 3 s caffeine pulse as longer single caffeine applications of up to 10 s did not result in a greater SR Ca²⁺ release, nor did repeated shorter duration single caffeine applications (not shown).

Stimulating burbot myocytes with long square depolarizing pulses (1-4 s) to high voltages (+50 mV) did not significantly increase SR Ca²⁺content compared with that obtained under the control loading conditions given in Fig. 5C. Furthermore,applying loading pulses (25-75 pulses, to either +10 or +50 mV, for either 1 s or 4 s) in bathing solution without TTX, in an attempt to augment reverse-mode NCX, did not result in greater SR Ca²⁺ accumulation upon application of caffeine (not shown). The amount of Ca²⁺ accumulated by the SR did not significantly increase when stimulating pulses (either control pulses or 1 s pulses to +50 mV) were applied in the presence of 1 or 10 μmol l^-1 isoprenaline, although I_Ca was augmented. Thus, SR Ca²⁺ content in burbot ventricular myocytes was at a steady-state between 100 and 300 μmol l^-1 Ca²⁺under the conditions of our study. It should be noted, however, that a few cells (11 out of 119 at 4°C and 2 out of 20 at 11°C) had an SR Ca²⁺ content in excess of 1000 μmol l^-1 upon the first application of caffeine. Although these cells were Ca²⁺overloaded, it does suggest that the maximal Ca²⁺ storage capacity of the burbot SR can be large.

SR Ca²⁺ levels were considerably reduced (∼70%) when control loading pulses were applied during blockade of I_Ca (with 100 μmol l^-1 CdCl₂) or I_NCX (with 10 mmol l^-1 NiCl). For example, SR Ca²⁺ content decreased from 244±72 μmol l^-1 (N=13) under control conditions, to 73±20 and 75±41 μmol l^-1when the 25 stimulation pulses were applied in the presence of CdCl₂ (100 μmol l^-1) or NiCl₂ (10 mmol l^-1), respectively. This suggests equal SR Ca²⁺ loading capabilities of these two Ca²⁺ influx pathways. However, because CdCl₂ at a concentration of 100 μmol l^-1 can potentially impact NCX activity, we examined SR Ca²⁺ accumulation in the presence of 30 μmol l^-1 CdCl₂ and 10 μmol l^-1 verapamil, which has been shown to quickly and effectively block l-type Ca²⁺ channel currents in crucian carp(Carassius carassius) myocytes without inhibiting the NCX(Vornanen, 1999). We found less of a reduction (∼56%) in SR Ca²⁺ content under these conditions, suggesting that reverse-mode NCX contributes a greater amount of Ca²⁺ to SR stores than I_Ca in burbot ventricular myocytes. It is important to note that these experiments were conducted with 0.025 mmol l^-1 EGTA in the pipette to better simulate in vivo cytosolic Ca²⁺ buffering(Hove-Madsen and Tort, 1998)and thus peak I_Ca was reduced by ∼65% (to 0.28±0.02 and 0.48±0.04 pA pF^-1 at 4°C and 11°C, respectively) compared with the values presented in Fig. 4, possibly reducing I_Ca contribution to SR Ca²⁺ loading. On the other hand, the lack of an increase in SR Ca²⁺ content during isoprenaline stimulation, despite augmented I_Ca,emphasizes the limited role of I_Ca and the potential importance of Ca²⁺ influx via NCX during excitation-contraction coupling in this species.

Inactivation kinetics of I_Ca were examined to assess the impact of SR Ca²⁺ release on excitation-contraction coupling in burbot myocytes at 4°C and 11°C. I_Ca records initiated immediately after depletion of SR Ca²⁺ by caffeine allowed the effects of subsequent progressive accumulation and release of SR Ca²⁺ on I_Ca to be monitored. Single exponential fits (tau, τ) to the decaying portion of I_Ca revealed no change in inactivation kinetics as Ca²⁺ was loaded into the SR(Fig. 6), indicating a lack of SR-Ca²⁺-release dependent inactivation of I_Caat either temperature.

In general, the RMP and AP of the burbot ventricle are similar to those recorded from other vertebrate hearts(Jaeger, 1965; Anderson et al., 1977; Morad et al., 1983; Venditti et al., 1996). However, there are several features of SL ion currents in the burbot ventricle that are not especially typical for vertebrate heart. These characteristics include (i) very small I_K1, (ii) large I_Kr, (iii) small I_Ca and (iv) large I_NCX.

In cardiac myocytes, I_K1 maintains the negative RMP and contributes to final phase 3 repolarisation(Christie, 1995; Barry and Nerbonne, 1996). I_K1 of the burbot ventricle is small, only one third of those found in cold-acclimated (4°C) trout and even less than the values measured in warm-acclimated (18°C) trout and crucian carp(Paajanen and Vornanen, 2002; Paajanen and Vornanen, 2004; Vornanen et al., 2002a), which explains its 7-10 mV less negative RMP in comparison with these species. Because I_K1 is small, it cannot contribute very much to the repolarisation of the AP. It is perhaps not unexpected then, that the other repolarising current, I_Kr, is about five times larger than I_K1 in burbot ventricular myocytes. The relative sizes of K⁺ currents in burbot ventricular myocytes resemble those of cold-acclimated trout but are in fact more extreme, with burbot ventricular K⁺ currents showing an electrical excitability phenotype similar to atrial K⁺ currents in other fish species(Vornanen et al., 2002a). The similarity of K⁺ currents, i.e. small I_K1 and large I_Kr, in cold-acclimated trout and burbot suggest that this might be a physiologically significant mechanism by which excitability is maintained at low temperatures. Studies on other cold-adapted species are necessary to assess the universality of this strategy and its influence on factors such as maintenance of RMP, AP duration and the prevention of cardiac arrhythmias.

Fish cardiac myocytes have a large surface area-to-volume ratio, increasing the efficacy of SL ion exchange in cytosolic Ca²⁺ management. This is especially true in myocytes of the burbot heart, which are 30% smaller in both length and width than those from similarly sized rainbow trout, which should result in even smaller diffusion distances between the SL and the myofilaments (Tiitu and Vornanen,2002a). In the physiological voltage range, the NCX avidly transports Ca²⁺ in both directions across the SL, while l-type Ca²⁺ channels provide an entry pathway for extracellular Ca²⁺ and maintain the long AP duration. The density of I_Ca is very low in burbot myocytes(Fig. 4) being half of that of crucian carp myocytes under identical conditions (not shown). However, the temperature sensitivity of I_Ca in the burbot ventricle(Q₁₀∼2) is similar to that of other fish(Shiels et al., 2000). The low I_Ca density may be related to the low density of the l-type Ca²⁺ channels on the burbot SL as dihydropyridine(DHPR) binding studies have indicated significantly lower B_max compared with that of trout and carp(Tiitu and Vornanen, 2003).β-adrenergic stimulation increases the size of I_Ca,but at present the adrenergic tonus on the burbot heart and its impact on excitation-contraction coupling are unknown.

Despite the slow inactivation of I_Ca, the charge carried (Q_Ca) was still very small. Indeed, the values for Q_Ca after a 1 s square pulse in the burbot were similar to the values of the rainbow trout ventricular myocytes for 0.5 s pulse and 58-75% of the values of the crucian carp ventricular myocytes for 0.5 s pulse(Vornanen, 1997; Vornanen, 1998; Shiels et al., 2000) at similar temperatures and identical conditions of cytosolic Ca²⁺buffering. Thus, I_Ca and Q_Ca in the burbot heart are smaller than in either rainbow trout or crucian carp. However, the slow inactivation of I_Ca may have another important role in cytosolic Ca²⁺ management. It increases calcium influx via the I_Ca-window current, which may play an important role in the long APD of burbot myocytes. We have previously shown that acute cold temperature increases the I_Ca-window current in trout myocytes, increasing the relative importance of SL calcium influx (Shiels et al., 2000). Here we show that the size the I_Ca window current in burbot ventricular myocytes at 4°C is twofold greater than in rainbow trout at 7°C (Shiels et al.,2000). Despite this, the peak density of I_Cais so small in the burbot that the integrated charge transfer during the duration of an AP (∼1 s at 4°C) remains smaller in burbot than trout under similar experimental conditions.

I_NCX density in burbot ventricular myocytes is approximately double that observed in crucian carp under identical conditions(Vornanen, 1999). I_NCX and I_Ca have been measured previously in crucian carp ventricular myocytes and it was estimated they contribute almost equally to SL Ca²⁺ entry(Vornanen, 1999). The present findings show that the density of I_NCX is clearly larger and the density of I_Ca much smaller in burbot than in carp myocytes. For example, the mean current density of the I_NCX is 0.2615 pA pF^-1 at +10 mV at 4°C(see Fig. 3B), and if this current is integrated for 1000 ms, it results in 0.2615 pC pF^-1 of charge transferred in 1 s. Because only one charge is carried by the NCX for each Ca²⁺ atom, this charge density corresponds to 0.523 pC pF^-1 if it were carried by Ca²⁺ channels. This is substantially more than that transported by l-type Ca²⁺channels at peak current (0.24±0.043 pC pF^-1 at the same voltage and temperature). This clearly indicates that in burbot ventricular myocytes, the NCX is much more important than l-type Ca²⁺ channels in trans-sarcolemmal Ca²⁺ influx. This is especially true when one considers that the charge transferred by the l-type Ca²⁺ channels given above was measured with 5 mmol l^-1 EGTA in the pipette solution, which augments Ca²⁺ influx.

Studies using non-teleost expression systems demonstrate that the trout NCX temperature-sensitivity is quite low (Q₁₀∼1.1)(Elias et al., 2001), whereas the NCX of a tropical fish species, the tilapia Oreochromis mossambicus, shows a similar temperature-sensitivity to mammals(Marshall et al., 2005). Surprisingly, the burbot I_NCX shows a relatively strong temperature dependence (Q₁₀=2.47) between 4°C and 11°C when recorded in native myocytes. This is interesting in light of the cold stenothermic environment in which the burbot lives and the predominance of the NCX in mediating SL Ca²⁺ flux. It should be noted, however, that regardless of the Q₁₀-effect, the activity of the NCX at 4°C is still high in burbot in comparison with other fish species. Further studies are necessary to assess the impact of this temperature sensitivity on excitation-contraction coupling.

We found the SR Ca²⁺ content in burbot ventricular myocytes was at a steady state between 100 and 300 μmol l^-1, which is comparable to that observed in mammals [100-150 μmol l^-1(Bassani et al., 1995; Negretti et al., 1995)] but smaller than that reported for rainbow trout myocytes (>500 μmol l^-1) (Hove-Madsen et al.,1998; Shiels et al.,2002b).

In the present study we examined the effect of SR Ca²⁺ content on the inactivation kinetics of I_Ca as an indirect means of assessing Ca²⁺-induced Ca²⁺-release (CICR). Despite the modest total SR Ca²⁺ content, we had expected to see evidence of CICR in burbot myocytes given that (i) isolated muscle experiments show that SR Ca²⁺ contributes significantly to contractility(Tiitu and Vornanen, 2002b),(ii) electron micrographs of burbot ventricle show that peripheral couplings between the SR and the SL are bridged by distinct foot particles akin to ryanodine receptors, indicating the structural organization necessary for CICR(Tiitu and Vornanen, 2002a),and (iii) [³H]ryanodine binding studies in cardiac vesicles indicate that the burbot ventricle has a substantial number of ryanodine receptors (65% of the value of the rat heart) whose opening is very Ca²⁺ sensitive (Vornanen,2006). However, we saw no effect of SR Ca²⁺ content on the inactivation kinetics of I_Ca, which suggests limited CICR in burbot ventricular myocytes. This apparent contradiction may be related to the low density of the l-type Ca²⁺ channels,which may preclude sufficient trigger signal for propagative CICR (see Shiels and White, 2005). Alternatively, as the NCX can provide the trigger Ca²⁺ in both mammal and fish hearts (Vornanen et al.,1994; Hove-Madsen et al.,2003), it is possible that the large I_NCX of the burbot ventricular myocytes overwhelms I_Ca as a trigger in the dyadic junction and, at the same time, masks the effect of CICR on I_Ca inactivation. According to this proposal the relative roles of I_Ca and I_NCX in triggering SR Ca²⁺ release would be quite different in fish and mammalian hearts. Obviously further studies examining the time course of intracellular Ca²⁺ transients in burbot ventricle with and without SR inhibition would provide valuable insight into the physiological role and mechanism of SR Ca²⁺ cycling during excitation-contraction coupling in this species.

Electrical excitation of burbot ventricular myocytes suggest that cold stenothermy and compensative cold-acclimation involve many of the same subcellular mechanisms. In particular, burbot K⁺ currents are organised similarly to those of cold-acclimated active species, demonstrating a small delayed rectifier K⁺ current (I_Kr) and a large inward rectifier K⁺ current (I_K1). This data strongly suggests that K⁺ current reorganisation may be necessary for AP regulation in cold-adapted species. Based on inactivation of I_Ca, we found no evidence of an upregulation of SR Ca²⁺ flux pathways in burbot ventricular myocytes, which is contrary to previous findings from other cold-acclimated fish. This may be related to the fluminous Ca²⁺ influx through the NCX, which probably provides the major part of the contractile Ca²⁺.

List of abbreviations
 
• AP

  action potential

 
• APD

  action potential duration

 
• CICR

  Ca²⁺-induced Ca²⁺-release

 
• NCX

  Na²⁺-Ca²⁺ exchange

 
• RMP

  resting membrane potential

 
• SL

  sarcolemma

 
• SR

  sarcoplasmic reticulum

We thank Jaakko Haverinen for his help with some of the experiments and acknowledge the support of a Journal of Experimental Biology Travel Fellowship to H.A.S. and by the Academy of Finland to V.P. and M.V.