Allan J. Levi

Cultured adult cardiac myocytes : Future applications, culture methods, morphological and electrophysiological properties

Isolated adult cardiac myocytes maintained in primary culture have been used as a model of the adult myocardium for 20 years. With the recent advances and current interest in using molecular biological techniques to investigate cardiac physiology, culturing myocytes is becoming an increasingly important technique. Acutely isolated myocytes do not remain viable for the time needed for the changes in gene expression to occur, and therefore it is necessary to maintain myocytes in culture. The aims of this review are: (1) To describe a method for isolating and culturing myocytes in serum-free medium. This section is targeted at new researchers in the field, with particular emphasis on aspects of the isolation procedure which are important for optimising myocyte culture. (2) To review current knowledge of how contractile, electrophysiological and morphological properties of adult myocytes are preserved in culture. Over the past 5 to 10 years significant advances have been made in developing novel techniques which help maintain the in-vivo properties of myocytes in culture. Efficient methods for transporting exogenous genes and anti-sense oligonucleotides into adult myocytes are now available. We anticipate that in future these advances will make cultured myocytes more attractive for use in biophysical and molecular investigations of cardiac physiology.

Time course and voltage dependence of expressed HERG current compared with native ”rapid” delayed rectifier K current during the cardiac ventricular action potential

Abstract It is widely believed that HERG (human ether-a-go-go-related gene) encodes the major subunit of the cardiac ”rapid” delayed rectifier K channel. The aims of the present study were threefold: (1) to record directly the time course and voltage dependence of expressed HERG current in a mammalian cell line, during an imposed ventricular action potential (AP); (2) to compare this with native rapid delayed rectifier current (IKr) elicited by applying an AP command to isolated guinea-pig ventricular myocytes; (3) to provide mechanistic information regarding the profile of HERG/IKr during the AP. We used the AP clamp technique and conventional whole-cell patch-clamp recordings at 32–34°C. HERG was transiently expressed in Chinese hamster ovary (CHO) cells. There was an outward current in transfected CHO cells, which developed progressively during the AP plateau and slow repolarisation phase. The instantaneous current-voltage (I-V) relation for both leak-subtracted HERG current (n=10) and E-4031-sensitive current (n=6) during AP repolarisation was maximal between –30 mV and –40 mV. The conductance-voltage (G-V) relation was maximal at potentials between –60 and –75 mV. A similar voltage dependence for HERG current was observed during a descending ramp from +60 to –80 mV (n=5), but not during either an ascending ramp (n=5), or a reversed AP waveform (n=8). These data suggest that instantaneous HERG current during the AP does not depend on the instantaneous command voltage alone, but upon the previous voltages during the applied waveform. The time course of activation of HERG current at potentials near the AP plateau was rapid. Tail currents recorded on premature repolarisation at different time points in the AP showed directly that HERG also activates rapidly during the AP. The I-V profiles of fully activated HERG and of current during the AP were very similar. IKr from guinea-pig ventricular myocytes was measured as E-4031-sensitive current during the AP clamp command. The current had a similar I-V and G-V profile to HERG current in CHO cells. These data indicate that HERG current and native IKr are similar during an applied AP waveform. Activation of HERG is rapid during the AP. However, due to rapid inactivation relatively little current flows until the potential becomes less positive than 0 mV. The removal of inactivation then allows more current to flow, giving rise to the distinct instantaneous I-V profile during the AP. The correlation between the voltage dependence of HERG during the AP and the fully activated I-V relation indicates that the contribution of HERG/IKr to AP repolarisation is more significantly determined by the open-channel I-V relation, than the precise activation time course of the current.

ACTION POTENTIALS, ION CHANNEL CURRENTS AND TRANSVERSE TUBULE DENSITY IN ADULT RABBIT VENTRICULAR MYOCYTES MAINTAINED FOR 6 DAYS IN CELL CULTURE

Adult rabbit ventricular myocytes were cultured in a basic medium (Medium 199) for up to 6 days to assess preservation of morphology and ion channel currents. In culture, cells remained rod shaped and striated but their ends became progressively rounded. Cell cross-sectional area declined slightly (by 14%) over the first 24 h, in contrast, whole-cell capacitance (which reflects external surface membrane plus membrane infoldings) decreased by 42% over the same time. Using whole-cell patch-clamp, we observed that the typical “N” shape steady-state current-voltage (I-V) relation became flattened after 24 h in culture. L-type Ca channel density was assessed as barium flux (IBa,L) via the channel. IBa,L (normalised to cell capacitance) declined by 50% after 24 h and recovered partially by days 4 and 6. The density of inward rectifier K current declined by 54% after 24 h and showed no recovery subsequently. In contrast, there was no significant decline in the density of transient outward K current after 24 h, but it declined subsequently by 65% after 6 days. We speculate that the time course of change in each ion channel density may reflect a change in pattern of ion channel expression, or differential membrane loss since the density of transverse tubules decreased by 57% after 6 days in culture. These results suggest that even by 24 h in culture, ion channel density in myocytes has changed substantially from the acutely isolated state.

The electrophysiological characteristics of hypertrophied ventricular myocytes from the spontaneously hypertensive rat.

Objectives: Previous studies on multicellular preparations have shown that hypertrophied cardiac muscle from the spontaneously hypertensive rat (SHR) has a prolonged action potential. The first aim of the present study was to determine whether the action potential of isolated left ventricular myocytes was similarly prolonged and to study the underlying membrane currents that might be responsible. The second aim was to evaluate the L-type calcium current amplitude of SHR myocytes, as we have recently shown that they have an increased contraction and an increase in the calcium trigger entering via the L-type calcium channel might be one possible mechanism for this. Methods: The electrophysiological characteristics of left ventricular myocytes isolated from the SHR were compared with those from normotensive control rats. Action potentials were recorded with microelectrodes. Cells were voltage-clamped and the membrane currents elicited by steps to different potentials were analysed. Blockers of potassium and calcium currents were used to reveal the contribution made by these currents to net membrane currents. Results: SHR myocytes had prolonged action potentials. The action potential duration of SHR myocytes at 90% repolarization was found to be longer, although at 20% and 50% repolarization no difference was found. There was no difference in the resting membrane potential between SHR and control myocytes. Using a voltage clamp we studied the L-type calcium current and potassium currents. The major change in SHR myocytes was a decrease in the magnitude (normalized to the membrane capacitance) of the inward rectifier potassium current elicited by negative potentials. There was no detectable difference in either the transient outward or delayed rectifier potassium currents. We also found no difference in the magnitude, time course or voltage dependence of L-type calcium current in hypertrophied SHR myocytes. Conclusions: First, the action potential of SHR myocytes was prolonged compared with control myocytes. Secondly, the main change in SHR myocytes was that pulses to negative potentials elicited a lower inward rectifier potassium current. A reduction in the density of inward rectifier channels might play a role in prolonging the SHR action potential, since a lower outward repolarizing current will flow through inward rectifier potassium channels during the SHR action potential repolarization. Thirdly, there was no difference in L-type calcium current density or time course between SHR and control myocytes. Thus, a change in L-type calcium current probably plays no role in causing the prolonged SHR action potential or the increased contraction of hypertrophied SHR ventricular myocytes. Finally, the prolonged action potential in SHR myocytes may itself be one factor responsible for the increased contraction of these cells.

A method for making rapid changes of superfusate whilst maintaining temperature at 37°C

We have developed a technique for making a rapid solution change, whilst at the same time maintaining the temperature of the preparation at 37°C. It is technically difficult to use rapid solution changes when experiments are performed at normal mammalian body temperature. As a solution is heated from room temperature to 37°C, gas bubbles form in the rapid-flowing solution streams, and these disturb a cell or attached recording pipettes. We describe a system that has been developed to eliminate these problems. We show how to construct the different components of the system, and we have designed an electronic circuit to control solution changes. We have performed tests to characterise the function of this system. Solution flow out of the nozzle of the device (0.88 ml min−1, linear flow velocity 11.6 cm s−1) caused a fall in the steady-state temperature at the experimental preparation of only 0.3°C. The device which takes between 0.5 and 1 s to completely change the superfusate of a single cell, was used routinely with five different experimental solutions. This system may be valuable in studies which require rapid solution changes to be performed at a normal mammalian body temperature.

Role of Intracellular Sodium Overload in the Genesis of Cardiac Arrhythmias

Intracellular Na and Arrhythmia. A number of clinical cardiac disorders may be associated with a rise of the intracellular Na concentration (Nai) in heart muscle. A clear example is digitalis toxicity, in which excessive inhibition of the Na/K pump causes the Nai concentration to become raised above the normal level. Especially in digitalis toxicity, but also in many other situations, the rise of Nai may be an important (or contributory) cause of increased cardiac arrhythmias. In this review, we consider the mechanisms by which a raised Nai may cause cardiac arrhythmias. First, we describe the factors that regulate Nai, and we demonstrate that the equilibrium level of Nai is determined by a balance between Na entry into the cell, and Na extrusion from the cell. A numher of mechanisms are responsible for Na entry into the cell, whereas the Na/K pump appears to be the main mechanism for Na extrusion. We then consider the processes by which an increased level of Nai might contrihute to cardiac arrhythmias. A rise of Nai is well known to result in an increase of intracellular Ca, via the important and influential Na/Ca exchange mechanism in the cell membrane of cardiac muscle cells. A rise of intracellular Ca modulates the activity of a numher of sarcolemmal ion channels and aflects release of intracellular Ca from the sarcoplasmic reticulum, all of which might be involved in causing arrhythmia. It is possible that the increase in contractile force that results from the rise of intracellular Ca may initiate or exacerbate arrhythmia, since this will increase wall stress and energy demands in the ventricle, and an increase in wall stress may be arrhythmogenic. In addition, the rise of Nai is anticipated to modulate directly a number of ion channels and to affect the regulation of intracellular pH, which also may be involved in causing arrhythmia. We also present experiments in tbis review, carried out on the working rat heart preparation, which suggest that a rise of Nai causes an increase of wall stress‐induced arrhythmia in this model. In addition, we have investigated the effect on wall stress‐induced arrhythmia of maneuvers that might be anticipated to change intracellular Ca, and this has allowed identification of some of the factors involved in causing arrhythmia in the working rat heart.

The effect of internal sodium and caesium on phasic contraction of patch-clamped rabbit ventricular myocytes.

1. The voltage dependence of phasic contraction was assessed in rabbit ventricular myocytes. Phasic contraction at all potentials was abolished by exposure to ryanodine‐thapsigargin, showing that it was due primarily to Ca2+ release from the sarcoplasmic reticulum (SR). Experiments were performed at 35 degrees C, cells were whole‐cell patch clamped and contraction was measured optically as unloaded shortening. Cells were held at ‐40 mV to inactivate the Na+ current (INa) and T‐type Ca2+ current. A standard cellular Ca2+ load was established by applying a train of conditioning pulses at 0.5 Hz before each test pulse. The effect of replacing K+ with Cs+ in the dialysing pipette solution, and the effect of altering dialysing [Na+] between 0 and 20 mM, was assessed on contraction. 2. Cells dialysed with a K(+)‐based, Na(+)‐free solution exhibited a ‘bell‐shaped’ voltage dependence of the L‐type Ca2+ channel current (ICa,L), with a maximum ICa,L at +10 mV. Replacing internal K+ with Cs+, or altering pipette [Na+], did not affect the voltage dependence of ICa,L. 3. The voltage dependence of phasic contraction in cells dialysed with a K(+)‐based solution was modulated by pipette [Na+]. The voltage dependence of phasic contraction was bell‐shaped with 0 Na+, became much loss bell‐shaped with 10 mM Na+ and with 20 mM Na+ the phasic contraction elicited at +100 mV was 1.6‐fold larger than that at +10 mV. 4. Replacing 80% of K+ with Cs+ in the pipette dialysis solution led to a significant reduction in contraction amplitude and a more rapid decline in contraction amplitude after beginning the dialysis of the cell. 5. Cells dialysed with a Cs(+)‐based solution displayed a voltage dependence of phasic contraction which was more bell‐shaped (i.e. more similar to that of ICa,L) than that obtained with the corresponding K(+)‐based dialysis solution. The level of pipette [Na+] still modulated the voltage dependence of phasic contraction in cells dialysed with a Cs(+)‐based solution. 6. Time‐to‐peak contraction (tpk) also displayed voltage dependence; it had a minimum value between 0 and +20 mV (the voltage range for maximum ICa,L), but increased at more negative and positive potentials. Alteration of tpk contraction is discussed in relation to the stochastic behaviour of L‐type Ca2+ channels and SR Ca2+ release channels. 7. The shape of the voltage dependence of contraction in rabbit myocytes at 35 degrees C is modulated by dialysing [Na+] over the tested range, 0‐20 mM. Modulation of voltage dependence of contraction by dialysing [Na+] is consistent with an influence of reverse Na(+)‐Ca2+ exchange in triggering intracellular Ca2+ release, in addition to the trigger Ca2+ which enters via ICa,L. 8. The marked effect of dialysing Cs+ on contraction amplitude, and on the voltage dependence of phasic contraction, does not appear to have been reported previously. Internal dialysis with Cs+ is a commonly used technique for blocking interfering outward K+ currents, in order to measure ICa,L more selectively. The present study suggests that Cs+ might also interfere with processes involved in excitation‐contraction coupling and indicates that it might be wise to exercise caution with the use of internal Cs+ in experiments investigating excitation‐contraction coupling.

Characteristics of the delayed rectifier K current compared in myocytes isolated from the atrioventricular node and ventricle of the rabbit heart.

The delayed rectifier potassium current (IK) is known to be important in action potential repolarisation and may contribute to the diastolic pacemaker depolarisation in pacemaker cells from the heart. In this study, using whole-cell patch clamp, we investigated the characteristics ofIK in morphologically normal cells from the atrioventricular node (AVN) and ventricle of the rabbit heart. Cells were held at −40 mV and 5 μM external nifedipine was used to block L-type calcium current (ICa,L). SignificantIK was observed with pulses to potentials more positive than −30 mV. The steady-state activation curve in both cell types showed maximal activation at between + 10 and + 20 mV. Half-maximal activation ofIK occurred at −4.9 and −4.1 mV with slope factors of 8.3 and 12.4 mV in ventricular and AVN cells, respectively. Using pulses of increasing duration, significantIK tails after repolarisation from + 40 mV were observed with pulses of 20 ms and increased with pulses up to 100–120 ms in both cell types. Pulses of longer duration did not activate furtherIK and this suggested that only the rapid component ofIK, calledIKr, was present in either cell type. Moreover,IK tails after pulses to all potentials were blocked completely by E-4031, a selective blocker ofIKr. The reversal potential ofIK varied with the concentration of external K. Superfusion of AVN cells with medium containing 4, 15 and 40 mM [K+]o resulted in reversal potentials of −81, −56 and −32 mV respectively, which are close to values predicted if theIK channel were highly selective for K. The time constants for deactivation ofIK in ventricle and AVN on return to −40 mV after a 500-ms activating pulse to + 60 mV were 480 ms and 230 ms, respectively. The faster deactivation ofIK in AVN cells was a distinguishing feature and suggests that there may be differences in theIKr channel protein between ventricular and AVN cells.

Effect of streptomycin on wall-stress-induced arrhythmias in the working rat heart

OBJECTIVE To assess whether streptomycin, an inhibitor of mechano-sensitive cation channels, has an effect on arrhythmias-induced by an increase of ventricular wall stress in the rat heart. METHODS The isolated working rat heart preparation was used. Arrhythmias were induced by increasing the afterload (i.e., aortic pressure) against which the left ventricle (LV) pumped for 20 s. This led to an increase of LV pressure, stretch of the LV and an increase in LV wall stress. The number of ventricular premature beats induced by each afterload step was compared in the absence and presence of streptomycin, a compound known to block mechano-sensitive cation channels in the heart. RESULTS Perfusion with 200 microM streptomycin caused a significant reduction in wall-stress-induced arrhythmias. The effect of streptomycin on arrhythmias reached steady-state within 10 min of application. In the presence of streptomycin, arrhythmias elicited by a 40 mmHg afterload increase were reduced to 38% of control. Arrhythmias induced by an 80 mmHg afterload increase were reduced to 61% of control. Complex arrhythmias (ventricular tachycardia) induced by an afterload increase were also reduced in the presence of 200 microM streptomycin. There was no change in inotropic state with streptomycin, as assessed either by cardiac output or by maximum developed LV pressure. Streptomycin 50 microM (a typical therapeutic plasma concentration in patients) had no effect on wall-stress-induced arrhythmias. CONCLUSIONS The results were inconsistent with streptomycin acting by modulating inositol phosphate production, or altering the level of intracellular calcium or inotropic state. The anti-arrhythmic effect of streptomycin appears more consistent with inhibition of mechano-sensitive cation channels, suggesting that these ion channels might be involved in causing wall-stress-induced arrhythmias.

Intracellular calcium transients recorded with Fura-2 in spontaneously active myocytes isolated from the atrioventricular node of the rabbit heart

We have used the fluorescent Ca indicator Fura-2 to assess the changes in intracellular calcium (Ca1) in single spontaneously active myocytes isolated from the rabbit atrioventricular node (avn). Simultaneous recordings of membrane potential and the Fura-2 ratio signal (which reflects Ca1) showed that a transient rise of Ca1 occurred with each spontaneous action potential (ap). The ap upstroke preceded the rise in Ca1 and repolarization of the ap occurred faster than the decline of Ca1. The level of Ca1 remained raised and progressively declined towards a baseline diastolic level during the subsequent pacemaker depolarization. The Fura-2 (Ca1) transient in spontaneously active avn cells had a time-to-peak of 49.2±5.4 ms (mean + s.e.m.; n= 7) and declined with a single exponential time course (time constant = 139.8±23.9 ms; n = 7). Application of 10 µM ryanodine completely and irreversibly abolished the Ca1 transient, identifying the sarcoplasmic reticulum (sr) as the major source of releasable Ca. Both removal of external Ca and block of L-type Ca channels (with 2 µM nifedipine) also abolished Ca1 transients, suggesting that Ca entry via L-type Ca-channels is involved in triggering the sr Ca release underlying the Ca1 transient. Removal of external Na (in the presence of 20 µM nifedipine to block L-type Ca channels) caused a reversible increase in Ca1, showing that Na/Ca exchange is present in avn cells and that it is involved in Ca1 regulation. Spontaneous Ca1 transients were abolished by 1 µM acetylcholine, and this was associated with a hyperpolarization of membrane potential and cessation of action potentials. Isoprenaline (1 µM) increased the rate and amplitude of spontaneous Ca1 transients; this corresponded to an increase in the rate and a change in the shape of spontaneous action potentials observed in patch-clamped avn myocytes.