Dynamics of the inward rectifier K+ current during the action

potential of guinea pig ventricular myocytes

Jose Ibarra, Gregory E. Morley, and Mario Delmar

Department of Pharmacology, SUNY Health Science Center. Syracuse, New York 13210 USA

ABSTRACT The potassium selective, inward rectifier current (/K1) is known to be responsible for maintaining the resting membrane
potential of quiescent ventricular myocytes. However, the contribution of this current to the different phases of the cardiac action
potential has not been adequately established. In the present study, we have used the action potential clamp (APC) technique to
characterize the dynamic changes of a cesium-sensitive (i.e., /K1) current which occur during the action potential. Our results show
that (a) /K1 is present during depolarization, as well as in the final phase of repolarization of the cardiac action potential. (b) The
current reaches the zone of inward-going rectification before the regenerative action potential ensues. (c) The maximal outward
current amplitude during repolarization is significantly lower than during depolarization, which supports the hypothesis that in adult
guinea pig ventricular myocytes, /Kj rectification is accentuated during the action potential plateau. Our results stress the
importance of /K1 in the modulation of cell excitability in the ventricular myocyte.

INTRODUCTION
It is widely accepted that the potassium selective, inward APC technique (Starzak and Starzak, 1976, 1978; Cole,
rectifier current (IK1) is the major determinant of the 1980; deHaas and Vogel, 1989; Doerr et al., 1989; Doerr
resting potential in the mammalian ventricular myocyte et al., 1990) to analyze the behavior of IK1 during the
(Pennefather and Cohen, 1990). Results obtained in this cardiac cycle of adult guinea pig ventricular myocytes.
and other laboratories suggest that IKI also plays an Our results reveal the specific contribution of IK1 to the
active role during action potential depolarization (Tour- different phases of the cardiac action potential. The data
neur, 1986; Delmar et al., 1989a), and contributes to the also show that the maximal outward IK1 during repolariza-
final phase of action potential repolarization (Tourneur tion is significantly less than during subthreshold depo-
et al., 1987). Such results were obtained using conven- larization, just before activation of the cell.
tional current and voltage clamp techniques, and the
inferences derived from them await confirmation in a
system in which a more direct analysis of current METHODS
dynamics may be performed.
Recent evidence obtained in chick embryo cardiac Cell dissociation and recording
cells (Mazzanti and DeFelice, 1990) also suggests that, procedures
although the inward rectifier channels are essentially Experiments were carried out in freshly-dissociated guinea pig ventric-
time-independent in the voltage region of the action ular myocytes. Details of the dissociation procedure are given in a
potential (Tourneur et al., 1987), the inward-going previous publication from our laboratory (Delmar et al., 1989a). The
composition of the Tyrode solution was (in mM): NaCl, 150; KCI, 5.4;
rectification of the channel can be accentuated by the CaCl2, 1.8; MgCl2, 1.0; NaHCO3, 5.8; NaH2PO4, 0.4; Glucose, 5.5;
cellular events which occur during the action potential HEPES, 5.0; pH was balanced to 7.4. Temperature was maintained at
plateau. The latter observation has not been tested for 350C.
the whole-cell IKI of adult, mammalian myocytes main- Recordings were obtained using sylgard-coated suction pipettes
tained at physiological concentrations of extracellular filled with an internal pipette solution containing (in mM): KCI, 150;
MgCl2, 1.0; EGTA, 5; HEPES, 5; 13-OH-butyric acid, 2.0; ATP
potassium. (disodium salt), 5.0; phosphocreatine (disodium salt), 5.0. The pH was
The action potential clamp (APC) techniques (Fisch- balanced to 7.2. Electrodes were coupled to an Axoclamp 2A amplifier
meister et al., 1984; deHaas and Vogel, 1989) offer a operating in the current clamp or voltage clamp mode.
valuable approach to the study of the dynamics of
specific currents as they actually occur during the action Action potential clamp
potential. This paper describes the use of the whole cell The whole-cell APC technique consists of voltage clamping the cell
membrane to its own action potential (deHaas and Vogel, 1989; Doerr
et al., 1990). Membrane potential recordings were obtained from
Address correspondence to Dr. Delmar. cardiac myocytes paced repetitively at a constant cycle length of

1 534 0006-3495/91 /12/1534/06 $2.00 Biophys. J. e Biophysical Society

Volume 60 December 1991 1534-1539
700-1,000 ms with current pulses (duration, 40 ms) of just-threshold
amplitude, applied through a patch pipette in the current clamp
configuration. Long pulse durations were used to induce prolonged
foot potentials, similar to those recorded from well-polarized cells
during impulse propagation across a heterogeneous conduction path-
way (Jalife, 1983; Delmar and Jalife, 1990; Hill et al., 1990; Davidenko
150 mV
et al., 1991 [in press]). A computer (Zenith Z248) was used to store AP

(via a 12 bit analog/digital converter; DAS-20; Metrabyte; sampling

rate 10 KHz) several action potentials. The amplifier was then
switched to single electrode voltage clamp mode and the recorded
trace was played back (via a 12-bit digital/analog converter) to act as 0
Ip
the voltage command for the preparation.
We have often observed that the resting potential of the myocyte
shifts slightly (within 2 mV) during control recordings, particularly
when the cell is paced repetitively at a relatively short ( < 1 s) cycle
length (Lorente et al., 1991). Thus, for the control runs, we always
adjusted the holding potential at the onset of recording to match
(within 1 mV) the value of resting potential that was measured IK
immediately before switching to voltage clamp mode. Consequently,
holding current during control conditions was equal to zero.
Cesium chloride (15 mM) was used as an IKI blocker (Tourneur et
al., 1987; Oliva et al., 1990; see Discussion). Active currents were 100 msec
stored on a videotape-based system (Neurocorder Model DR 484) and
later analyzed with the computer using the CLAMPEX and
CLAMPFIT routines of the pCLAMP system. Dynamic IKI current- FIGURE 1 Action potential clamp recordings of IKI Top trace depicts
voltage (IV) relations were constructed by correlating the value of the action potential (AP) used as voltage command for the same cell. Ip
membrane potential wth the value of IK1 at each sample point. and ICS represent, respectively, single-current traces obtained under
A variable gain amplifier was included in the action potential clamp APC before and after addition of 15 mM cesium to the superfusate. IKI
circuit to ensure that the voltage signal delivered to the cell as represents the Cs-sensitive current. Bottom vertical calibration: 500
command potential had the same amplitude (within 1 mV at the action pA for trace Ip and 300 pA for traces Ic, and IK1. Membrane outward
potential plateau) as the one generated by the cell during current currents are represented as downward deflections. The horizontal
clamp recording (Doerr et al., 1990). In our initial series of experi- lines across the Ip, ICS and IKI traces indicate zero current level.
ments (n = 11), the variable gain amplifier was interposed between the
output of the D/A converter and the external command input of the
patch clamp amplifier operating in the voltage clamp mode. In these after addition of 15 mM cesium to the superfusate. In this case, the
experiments, 10 sequential recordings were averaged to obtain the voltage clamp amplifier was forced to compensate for both the absent
pertinent current and voltage data to generate the IKI IVrelations. In a pulse and the Cs-sensitive (i.e., IKI) currents. By digitally subtracting
second series of experiments (n = 9), the variable gain amplifier was trace Ip from ICS we obtained a trace of IKI throughout the different
interposed between the output of the patch clamp amplifier operating phases of the cardiac cycle (trace labelled IKI). Note that, in the APC
in current clamp mode and the input of the A/D converter (during technique, the amplifier compensates for absent currents, and thus,
action potential acquisition). This modification allowed us to take the conventional polarity of the current is inverted (deHass and Vogel,
advantage of most of the voltage range of the A/D board, thus 1989; Doerr et al., 1990).
enhancing the resolution of the voltage signal and reducing the A plot of voltage (AP) versus current (IKI) at each point in time
amplitude of the individual voltage steps. This procedure led to a yielded the dynamic IKI IV relation (Mazzanti and DeFelice, 1990),
drastic reduction in the noise amplitude of the current traces during shown in Fig. 2. In fact, this technique allowed us to obtain an IV
action potential clamp. Noise was further reduced by placing a Bessel relation for IKI both during depolarization (open squares) and during
filter (1-2 KHz cut-off frequency) between the output of the D/A repolarization (crosses). Data obtained during the action potential
converter and the external command input of the voltage clamp
upstroke (between -54 mV and 50 mV) were affected by the lack of
amplifier. Given the reduction in noise amplitude, dynamic IKI IV voltage control during the rapid depolarization (Doerr et al., 1990)
relations could be generated from single traces. Similar results were and thus were not included in the plot (see Discussion).
obtained regardless of whether averaged or single traces were ana- In Figs. 1 and 2, the holding current for the Ic, and IKI traces was
lyzed. slightly outward during diastole. This has been previously shown for
APC experiments in which all (or most) outward conductances were
blocked (see Fig. 2 D of Doerr et al., 1990) and it probably reflects the
RESULTS fact that the resting membrane potential is not identical to the
equilibrium potential for IKI As shown in Fig. 2, at the subthreshold
Dynamics Of 'Ki in the cardiac cycle level of potentials, the inward rectifier increased progressively during
An example of our results is presented in Fig. 1. In A, the trace the pulse and reached the zone of inward-going rectifcation before the
labelled AP illustrates an action potential acquired from a guinea pig fast action potential upstroke ensued. Current was almost negligible
ventricular myocyte using the whole-cell current clamp method. After during the action potential plateau but increased again as the cell
switching to the voltage clamp mode, the same action potential was repolarized beyond -40 mV; in contrast with what would be expected
used as the voltage command for the cell membrane, eliciting the from a time-independent current, the IV relation during depolariza-
current labelled Ip. This current trace mimics the original stimulating tion was clearly distinct from that during repolarization. The following
pulse, except for a small offset from the baseline observed during the characteristics were consistently observed in our experiments (n = 20):
action potential plateau (see Discussion). Trace ICS was recorded 3 min (a) the IV curves intersected 10-15 mV above the resting potential

Ibarra
lbarra at al
et al. Dynamica of 'KI
Dynamics 'Ki in the Cardiac
Cardian Cycle
Cycle 1535
1 535
 I (pA) currents that are difficult to identify at the single-
channel level. For example, in the case of IK,, single-
channel outward currents are impossible to detect at a
normal extracellular concentration of potassium ([K]O),
making it necessary to raise [K]o to unphysiological levels
(11.3 mM or above; Mazzanti and DeFelice, 1988, 1990).
This procedure not only alters channel conductance
(Sakmann and Trube, 1984; Matsuda, 1991) but it also
shifts the reversal potential of the current and, conse-
quently, alters the voltage gradient at which the channel
operates. The latter may be important since the kinetic
lI I
properties of the channel are known to be voltage
-90 -80 -70 -60 -50 -40 0 dependent (Ishihara et al., 1989; Oliva et al., 1990). In
Voltage (mV) addition, the shift of reversal potential due to high [K]o
would prevent the study of the dynamics of outward IK1
in the voltage range of the subthreshold response. In
FIGURE 2 Dynamic IK1 IV relation. Data points were obtained during summary, patch-APC and whole-cell APC techniques
either subthreshold depolarization (open squares) or during phase-3 complement each other, and they both represent excel-
repolarization (crosses) of the action potential shown in Fig. 1. lent tools in the study of the dynamic properties of
Following the conventional representation, outward currents are
considered to be of positive sign. membrane ionic currents.

Limitations of the action potential

(indicated by an arrow in Fig. 2). (b) For voltage values between the clamp technique
intersecting point and activation threshold (-54 mV in the case shown
in Figs. 1 and 2), the outward current obtained during depolarization A common problem that we have faced in our experi-
was larger than that measured following the action potential plateau. ments is the loss of voltage clamp control during the
(c) Repolarization current was larger than depolarization current action potential upstroke, probably due to the high
between resting potential and the intersecting point. (d) The voltage
value corresponding to the maximal outward current during repolariza- frequency and large amplitude of the command signal
tion was more negative than the one obtained during depolarization. during that phase of the action potential (Doerr et al.,
(e) The maximal amplitude of outward current reached during 1990). To avoid errors of interpretation, we systemati-
subthreshold depolarization (703 + 306.8 pA. X SD; n = 20) was
±
cally discarded currents obtained during the action
larger than that measured during repolarization (514 247 pA). The
±

difference between the two groups was highly significant (two tailedp potential upstroke. Consequently, we could not deter-
value = 0.0001) as determined by a paired t test. The relative differ- mine the exact shape of the negative slope region during
ence varied widely from one experiment to another, the average being depolarization.
27.9 + 11.6%. These differences strongly suggest that events occurring A small current (which we shall refer to as "offset
during the action potential plateau significantly alter the conductance current") was often observed during the action potential
of the IK1 channel (see also Mazzanti and DeFelice, 1990).
plateau in normal Tyrode solution (see Ip trace, Fig. 1).
A small offset from the baseline was also found by
deHaas and Vogel (1989) in their APC experiments of
DISCUSSION single nerve axon. In some experiments we repeated the
APC protocol after Cs washout. In those instances, the
The action potential clamp technique offers a valuable offset current was very similar to the one initially found
approach to the study of the dynamics of specific for the Ip trace.
currents as they actually occur during the action poten- The presence of an offset current could be explained
tial. This can be carried out by either allowing the cell to by at least two mechanisms. First, it is possible that this
clamp to its own action potential while recording the current results from temporal variations in the back-
membrane channel activity with a cell-attached pipette ground conductance (Belles et al., 1987). However, it is
(i.e., action potential patch clamp, or patch-APC; unlikely that an increase in background conductance
Fischmeister et al., 1984) or by digitizing the action would occur in our experimental conditions (pipette
potential and then using that action potential as the [ATP] = 5 mM) in the brief time (<3 min) elapsed
voltage command for the same cell (whole cell-APC; between current clamp recording and the moment of
Starzak and Starzak, 1976, 1978; Cole, 1980; deHaas and acquisition of Ip. A second possibility is that the offset
Vogel, 1989; Doerr et al., 1989, 1990). current results from a small (<1 mV) mismatch be-
Whole-cell APC recordings are useful for the study of tween the original action potential (as recorded from the

1536
1 536 Biophysical Journal
Biophysical Journal Volume 60
Volume 60 December
1991
December 1991
cell) and the command signal delivered to the cell during digital subtraction of Cs-sensitive currents (see, for
APC. We routinely adjusted for possible differences (see example, Tourneur et al., 1987; Oliva et al., 1990) are
Methods). However, small unresolved mismatches may quite similar to those observed by measuring the differ-
have remained. ence between records obtained at 14 mM [K]o and those
Regardless of its underlying mechanism, the character- obtained in K-free solution, using the oil-gap technique
istics of the offset current were such that it would not be (see, e.g., Ishihara et al., 1989).
expected to significantly alter the shape of the IK1 IV In summary, our results as well as those of others
relation. Indeed, the offset current was still present after strongly suggest that most (if not all) of the Cs-sensitive
Cs washout and is therefore possible that it was also outward current that was recorded in our experiments
present during Cs superfusion; in that case, the offset between -40 mV and resting potential indeed corre-
current would be cancelled out during the subtraction sponds to the whole-cell IK1 previously identified through
procedure. Moreover, even if the offset current was not more conventional voltage clamp techniques (Pennefa-
completely eliminated by the subtraction, it would only ther and Cohen, 1990).
alter the amplitude of the subtracted currents recorded
in the plateau range of potentials. Yet, in this paper, we
only analyze the time and voltage dependence of IK1 for Role of 'Kl in cell depolarization
voltages more negative than - 40 mV. Finally, we should Our results demonstrate that, when just-threshold cur-
point out that in two experiments, no offset current was rent pulses are used to activate the cell, the membrane
detected. Results from the latter experiments were potential enters the negative slope region of the IK1 IV
similar to those illustrated in Figs. 1 and 2. relation before the rapid upstroke ensues (see Fig. 2).
The above limitations notwithstanding, the data pre- These data conclusively demonstrate the active role of
sented in this paper do show the dynamic contribution of IKi in determining the amplitude and shape of the
a Cs-sensitive current to the different phases of the subthreshold depolarization (Delmar et al., 1989a), and
cardiac action potential. As discussed below, the evi- strongly support the hypothesis that the negative slope
dence available in the literature strongly supports the region of IK1 contributes to the overall depolarization of
hypothesis that such a Cs-sensitive current is largely the cell (Tourneur, 1986).
carried through inward rectifier, IK1 channels. Cell excitability has generally been associated with the
ability of inward currents to generate an action potential
Cesium selectivity upstroke. Yet, even in the presence of normal sodium
channels, the cell will not generate an action potential if
The applicability of the APC technique to the study of it fails to depolarize from its resting level to threshold. It
individual ionic currents is highly dependent on the is in this regard that input resistance may play a role in
specificity of the channel blocker being used. An excel- the excitability of cardiac cells.
lent discussion on the nature of Cs-sensitive currents can Under normal circumstances, the propagating current
be found in Oliva et al. (1990). These authors provide is much larger than the minimum current required for
strong arguments in support of the premise that, for excitation and, consequently, the characteristics of IK1
voltages more negative than -20 mV, the cesium- would be irrelevant to the overall process of activation.
sensitive current is largely (if not exclusively) IK1. A However, when propagation is impaired, electrotonic
similar conclusion was drawn by Tourneur et al. (1987) depolarizations of long duration can precede the devel-
when studying guinea pig ventricular myocytes. opment of the active response (see Jalife, 1983; Delmar
In the studies cited above, Cs-sensitive currents were and Jalife, 1990, for review). In those instances, the
recorded in the presence of calcium channel blockers. In dynamic properties of IK1 play a critical role in determin-
our case, calcium currents were still functional. How- ing the amplitude and shape of the subthreshold re-
ever, it is unlikely that calcium currents were affected by sponse (Tourneur, 1986; Delmar et al., 1989a, b; Lorente
cesium because no sizable inward currents were de- et al., [1991]) and consequently, determine the success
tected at any time in the subtracted trace. In addition, or failure of activation.
Ca-channel permeability studies have shown that cesium
is the least permeable of all cations (Hess et al., 1986).
Moreover, we are not aware of any studies indicating a /Ki conductance is reduced by events
blocking effect of cesium on calcium currents. In fact, occurring during the action potential
cesium-containing solutions are commonly used to study plateau
the functional properties of calcium currents in cardiac We consistently observed IV relations which showed a
cells (see e.g., Tseng, 1988; Hartzell and White, 1989). more pronounced inward-going rectification during ac-
The characteristics of the IK1 IV curves obtained by tion potential repolarization. This observation corre-

Ibarra at al.
lbarra et Dynamics of I,K in the
Dynamics the Cardiac
Cardiac Cycle
Cycle 1537
1537
lates well with that of Mazzanti and DeFelice (1990), IK1 outward conductance is also a function of time. The
showing that IK1 rectification is accentuated following implications that this temporal component may have on
the action potential plateau. the origin and maintenance of normal and abnormal
We may speculate that the increased rectification cardiac rhythms, remains to be determined.
observed after active depolarization occurs as a conse-
quence of the voltage-dependent blocking effect of We wish to thank Dr. Jose Jalife for his advice and support throughout
cytosolic free calcium ions on the conductance of the IK1 this project. The technical assistance of Ms. Wanda Coombs and Ms.
JoAnne Getchonis, and the secretarial skills of Ms. La Verne Gilbert
channels (Mazzanti and DiFrancesco, 1989). An argu- are also appreciated.
ment against this possibility is the fact that our experi-
ments were carried out in the presence of 5 mM EGTA This work was supported by grants R01-HL 40923 and P01-HL 39707
from the Heart, Lung and Blood Institute, National Institutes of
in the pipette solution, which would presumably chelate Health.
Cat. Yet, we can not be certain as to whether the EGTA
was able to diffuse into the intracellular space and
abolish the systolic Ca transients. Further experiments Received for publication 31 January 1991 and in final form 16
in which the cell is actively dialyzed with a known August 1991.
concentration of Ca2" would be necessary to test this
hypothesis. Moreover, it is possible that, as shown by
Mazzanti and DeFelice (1990) for chick embryo myo- REFERENCES
cytes, IK1 channel block during the action potential
Belles, B., J. Hescheler, and G. Trube. 1987. Changes of membrane
plateau is caused by the same calcium ions entering the currents in cardiac cells induced by long whole-cell recordings and
cell during the action potential; in that case, it is possible tolbutamide. PfluegersArch. (Eur. J. Physiol.) 409:582-588.
that calcium ions could escape EGTA chelation because Cole, K. S. 1980. Digital operations can be deceptive. Biophys. J.
they are confined to the restricted subsarcolemal space. 31:433-434.
Another possibility is that the reduced IK1 is a conse- Davidenko, J. M., A. V. Pertsov, R. Salomonsz, W. Baxter, and J.
quence of the time- and voltage dependence of Mg- Jalife. 1991. Saptio-temporal irregularities of spiral wave activity in
mediated open channel block (Matsuda, 1988) and/or a isolated ventricular muscle. J. Electrocardiol. In press.
result of the voltage- and time-dependent gating proper- deHaas, V., and W. Vogel. 1989. Sodium and potassium currents
recorded during an action potential. Eur. Biophys. J. 17:49-51.
ties of the channel (Kurachi, 1985). Indeed, the combina-
Delmar, M., D. C. Michaels, and J. Jalife. 1989a. Slow recovery of
tion of particle block and gate closing may lead to excitability and the Wenckebach phenomenon in the single guinea
further reduction in open channel probability as the pig ventricular myocytes. Circ. Res. 65:761-774.
membrane depolarizes into the voltage levels of the Delmar, M., L. Glass, D. C. Michaels, and J. Jalife. 1989b. Ionic basis
action potential plateau (Ishihara et al., 1989), with a and analytical solution of the Wenckebach phenomenon in guinea
consequent reduction in whole-cell outward current pig ventricular myocytes. Circ. Res. 65:775-788.
during repolarization (Ishihara et al., 1989; Oliva et al., Delmar, M., and J. Jalife. 1990. Wenckebach periodicity: from deduc-
1990). Yet, partial blockade of open channels may tive electrocardiographic analysis to ionic mechanisms. In Cardiac
Electrophysiology: From Cell to Bedside. D. P. Zipes and J. Jalife,
explain the paradoxical increase in outward current editors. W. B. Saunders Company, Philadephia, PA. 128-138.
during repolarization at the most negative levels of Doerr, T., R. Denger, A. Doerr, and W. Trautwein. 1990. Ionic
membrane potential (Oliva et al., 1990). Further imple- currents contributing to the action potential in single ventricular
mentation of kinetic models in which both particle block myocytes of the guinea pig studied with action potential clamp.
and channel gating are taken into account (such as those PfluegersArch. (Eur. J. Physiol.) 416:230-237.
of Ishihara et al., 1989, or Oliva et al., 1990) will be Doerr, T., R. Denger, and W. Trautwein. 1989. Calcium currents in
necessary to advance our understanding of the dynamic single SA nodal cells of the rabbit heart studied with action potential
behavior of IK1 during the cardiac action potential. clamp. PfluegersArch. (Eur. J. Physiol.) 413:599-603.
Fischmeister, R., L. J. DeFelice, R. K. Ayer, R. Levi, and R. L.
DeHaan. 1984. Channel currents during spontaneous action poten-
tials in embryonic chick heart cells. The action potential patch
CONCLUSIONS clamp. Biophys. J. 46:267-271.
Hartzell, H. C., and R. E. White. 1989. Effects of magnesium on
Our results show that IK1 dynamically contributes to the inactivation of the voltage-gated calcium current in cardiac myo-
electrophysiological events of the cardiac cycle. Indeed, cytes. J. Gen. Physiol. 94:745-767.
IKi participates in the regenerative depolarization pro- Hess, P., J. B. Lansman, and R. W. Tsien. 1986. Calcium channel
selectivity for divalent and monovalent cations. Voltage and concen-
cess and, therefore, may play a role in cell excitability. It tration dependence of single channel curret in ventricular heart
also provides outward current during the final phase of cells. J. Gen. Physiol. 88:293-319.
action potential repolarization. Finally, our results show Hill, B. C., K. R. Courtney, and M. Delmar. 1990. Laser scanner
that, in addition to its well-known voltage dependence, observations of heterogeneous conduction in rabbit epicardium

1538 Biophysical Journal Volume 60 December 1991

 induced by cellular uncoupling with heptanol. J. Cardiovasc. Electro- K-inward rectification in cardiac myocytes. Pfluegers Arch. (Eur. J.
physiol. 1:426-436. Physiol.) 413:322-324.
Ishihara, K., T. Mitsuiye, A. Noma, and M. Takano. 1989. The Mg2+ Oliva, C., I. S. Cohen, and P. Pennefather. 1990. The mechanism of
block and intrinsic gating underlying inward rectification of the K' rectification of IKI in canine Purkinje myocytes. J. Gen. Physiol.
current in guinea-pig cardiac myocytes. J. Physiol. (Lond.). 419:297- 96:299-318.
320. Pennefather, P., and I. S. Cohen. 1990. Molecular mechanisms of
Jalife, J. 1983. The sucrose gap preparation as a model of AV nodal cardiac K-channel regulation. In Cardiac Electrophysiology: From
transmission: are dual pathways necessary for reciprocation and AV Cell to Bedside. D. P. Zipes and J. Jalife, editors. W. B. Saunders
nodal echoes? PACE (Pacing Clin. Electrophysiol.). 6:1106-1122. Company, Philadelphia, PA. 17-28.
Kurachi, Y. 1985. Voltage-dependent activation of the inward-rectifier Sakmann, B., and G. Trube. 1984. Conductance properties of a single
potassium channel in the ventricular cell membrane of guinea-pig inwardly-rectifying K channels in ventricle guinea-pig. J. PhysioL
heart. J. Physiol. (Lond.). 366:365-385. (Lond.). 347:641-657.
Lorente, P., C. Delgado, M. Delmar, D. Henzel, and J. Jalife. 1991. Starzak, M. E., and R. J. Starzak. 1976. Voltage clamp currents from
Hysteresis in the excitability of isolated guinea pig ventricular squid giant axons clamped with their action potentials. Bio. Bull.
myocytes. Circ. Res. 69:1301-1315. 151:430431.
Matsuda, H. 1988. Open-state substructure of inwardly rectifying Starzak, M. E., and R. J. Starzak. 1978. An action potential clamp to
potassium channels revealed by magnesium block in guinea-pig probe the effectiveness of space clamp in axons. IEEE Trans.
heart cells. J. Physiol. (Lond.). 397:237-258. Biomed. Eng. 25:201-204.
Matsuda, H. 1991. Effects of external and internal K' ions on Tourneur, Y. 1986. Action potential-like responses due to the inward
magnesium block of inwardly rectifying K' channels in guinea-pig rectifying potassium channel on mammalian heart cells. J. Membr.
heart cells. J. PhysioL (Lond.). 435:83-99. Bio. 90:115-122.
Mazzanti, M., and L. J. DeFelice. 1988. K channel kinetics during the Tourneur Y., R. Mitra, M. Morad, and 0. Rougier. 1987. Activation
spontaneous heart beat in embryonic chick ventricle cells. Biophys. J. properties of the inward-rectifying potassium channel on mamma-
54:1139-1148. lian heart cells. J. Membr. Bio. 97:127-135.
Mazzanti, M., and L. J. DeFelice. 1990. Ca modulates outward current Tseng, G.-N. 1988. Calcium current restitution in mammalian ventric-
through IK1 channels. J. Membr. Biol. 116:41-45. ular myocytes is modulated by intracellular calcium. Circ. Res.
Mazzanti, M., and D. DiFrancesco. 1989. Intracellular Ca modulates 63:468482.

Ibarra et al. Dynamics of 'K1 in the Cardiac Cycle 1539