Effects of the Renin-Angiotensin System on the Current Ito in Epicardial and Endocardial Ventricular Myocytes From the Canine Heart
Abstract—The Ca2+-independent portion of transient outward K+ current (Ito) exhibits a transmural gradient in ventricle. To investigate control mechanisms for this gradient, we studied canine epicardial and endocardial ventricular myocytes with use of the whole-cell patch-clamp technique. Ito was larger in amplitude, had a more negative voltage threshold for activation, and had a more negative midpoint of inactivation in epicardium. Recovery from inactivation was >10-fold slower in endocardium. Incubation of epicardial myocytes with angiotensin II for 2 to 52 hours altered Ito to resemble unincubated endocardium and reduced the amplitude of the phase 1 notch of the action potential. In contrast, incubation of endocardial myocytes with losartan for 2 to 52 hours altered Ito to resemble unincubated epicardium and induced a phase 1 notch in the action potential. With RNase protection assays, we determined that incubations with angiotensin II or losartan did not alter mRNA levels for either Kv4.3 or Kv1.4; thus, a change in the α subunit for Ito is unlikely to be responsible. To test whether posttranslational modification produced the effects of angiotensin II, we coexpressed Kv4.3 and the angiotensin II type 1a receptor in Xenopus oocytes. Incubation with angiotensin II increased the time constant for recovery from inactivation of the expressed current by 2-fold with an incubation time constant of 3.7 hours. No effect on activation or inactivation voltage dependence was observed. These results demonstrate that the properties of Ito in endocardium and epicardium are plastic and likely under the tonic-differing influence of the renin-angiotensin system.
Differences in the action potential duration of myocytes from the endocardium (ENDO) and the epicardium (EPI) of the canine heart contribute to the morphology and polarity of the T wave of the ECG.1 2 The ionic currents responsible for these differences are determinants of the T wave. One major difference between EPI and ENDO myocyte ion currents is in the Ca2+-independent transient outward current (referred to here as Ito). Previous investigations have demonstrated a lower density and slower recovery from inactivation in ENDO compared with in EPI.3 4 We recently identified the molecular basis of Ito in canine ventricle as a member of the Shaker family of K channels, Kv4.3.5
Among the potential modulators of the T wave is the renin-angiotensin system and its active hormone, angiotensin II (Ang II).6 Ang II has direct actions on cardiac membrane currents, increasing chloride current7 and decreasing Na+/K+ exchange current8 in rabbit ventricular myocytes. These actions occur through the Ang II type 1 (AT1) receptor, which is prevalent in cardiac tissues of many species.9 10 Not only is the AT1 receptor present in heart, but also an autocrine renin-angiotensin system is present in rat ventricular myocytes.11 12 13 This suggests that mammalian cardiac myocytes or nonmyocyte elements in mammalian hearts can produce this hormone, which modulates repolarizing currents.
In the present study, we first describe the properties of Ito in canine EPI and ENDO, outlining differences in both current density and gating properties. We then demonstrate that the differences between these 2 tissues are entirely plastic; that is, long-term exposure of EPI to Ang II or of ENDO to the AT1 receptor blocker losartan can convert properties of Ito of either region into those of the other, thereby providing a basis for an altered transmural gradient of repolarization in the ventricle and resultant Twave changes.
Materials and Methods
Adult dogs of either sex were euthanized with an injection of sodium pentobarbital. EPI and ENDO ventricular myocytes were dissociated according to the trituration method developed in our laboratory.14 Cells were stored in KB solution (which contained [in mmol/L] KCl 83, K2HPO4 30, MgSO4 5, Na-pyruvic acid 5, β-OH butyric acid [Na-salt] 5, taurine 20, glucose 10, EGTA 0.5, HEPES 5, Na2ATP 5, and creatine 5, titrated to pH 7.2 with KOH)15 at room temperature for at least 1 hour before electrophysiological experiments. Cells were incubated in KB solution at room temperature the first day of experiments (2 to 5 hours) and then at 4°C for experimentation on days 2 and 3 (6 to 52 hours). Ang II and saralasin were purchased from Sigma Chemical Co, and losartan was kindly provided by Merck.
Measurement of Ito, Action Potentials, and Data Analysis
Isolated cells were maintained at 32° to 35°C (±0.5°C in each experiment).16 Ito and action potentials were recorded with use of the whole-cell patch-clamp technique. Ito was defined as the difference between the peak value and the current level at the end of a 300- or 400-ms pulse. The threshold was the most negative test voltage that elicited an Ito of >10 pA. The pipettes were filled with solution containing (in mmol/L) NaCl 6, K-aspartate 130, MgCl2 2, CaCl2 5, EGTA 11, Na2ATP 2, Na-GTP 0.1, Na-cAMP 0.2, and HEPES 10 (pH adjusted to 7.2 with KOH). The external solution contained (in mmol/L) NaCl 137.7, NaOH 2.3, MgCl2 1, glucose 10, HEPES 5, KCl 5.4, CaCl2 1.8, MnCl2 2, and CdCl2 0.2, pH 7.4. The liquid junction potential of −11 mV between the electrode tip and the cell interior (cell interior negative) was not corrected.17 Mn2+ and Cd2+ were used to block Ca2+ currents, which can obscure Ito activation.
RNase Protection Assays
ENDO and EPI myocytes were incubated for 24 hours either with or without 2 μmol/L Ang II (EPI) or 2 μmol/L losartan (ENDO). mRNA was isolated with the use of paramagnetic poly(dT) beads (Dynal). RNA probes for Kv1.4 and Kv4.3 were prepared as described previously.5 RNase protection assays were performed and quantified as described previously.5
Heterologous Expression in Oocytes
Oocytes were prepared from Xenopus laevis as previously described.5 Oocytes were injected with 50 nL Kv4.3 mRNA or an equimolar ratio of Kv4.3 and AT1a mRNA. Injected oocytes were incubated at 18°C for 24 to 48 hours. In the study of Ang II treatment, oocytes were then incubated in OR35 solution containing 4 μmol/L Ang II for 25 to 32 hours at 18°C before testing.
Oocytes were voltage-clamped with the use of a 2-microelectrode voltage clamp. The extracellular recording solution was OR2.5
Data are presented as mean±SEM. Statistical significance was tested with Student’s t tests. P<0.05 was considered statistically significant.
An expanded Materials and Methods section is available online at http://www.circresaha.org.
Gating Properties of Ito in ENDO Differ From Those in EPI
We began our investigation by characterizing Ito in both ENDO and EPI ventricular myocytes. We first examined the voltage dependence of activation. We held the cells at −65 mV and first briefly depolarized (5 ms) to −45 mV to inactivate INa and then depolarized to voltages from −40 to +50 mV in 10-mV increments for 300 ms. The average threshold for first observation of transient outward current (see Materials and Methods for definition) was −25±2 mV (n=24) in EPI myocytes and 4±3 mV (n=15) in ENDO myocytes. As previously reported,3 the amplitude of transient outward current is also larger in the EPI myocyte. The larger Ito was determined in the Table⇓ by measuring the dependence of the transient outward conductance on voltage, which increases almost 4-fold more rapidly in EPI than in ENDO. Our results are reported normalized to capacitance. The average capacitance was 125±26 pF (n=20) in ENDO myocytes and 104±8 pF (n=26) in EPI myocytes.
We also examined the voltage dependence of inactivation of Ito. We held the membrane at −65 mV and used a 3-pulse protocol. We first depolarized for 5 ms to −45 mV to inactivate INa and then used a first (conditioning) pulse, which either depolarized or hyperpolarized for 2 seconds to bring the membrane to a new starting value for inactivation; with the second (test) pulse, we depolarized to +10 mV for 400 ms (see Figure 1A⇓). Ito amplitude normalized to Ito for the most negative conditioning pulse is plotted against each conditioning voltage in Figure 1B⇓, along with the fits of a Boltzmann 2-state model. Ito exhibited a more positive midpoint of inactivation in ENDO myocytes and a reduction in steepness of the inactivation curve.
The time constant of inactivation of Ito did not differ between ENDO and EPI. We then used a 2-pulse protocol to investigate the kinetics of recovery from inactivation. The cycle length was 8 seconds. The first depolarizing pulse of 300 ms in duration was applied from the holding potential of −65 mV to the test potential of +5 mV EPI, or +15 mV ENDO, to activate a measurable Ito; then, a variable interval at the holding potential was allowed for recovery from inactivation, followed by a second identical test pulse. Data are provided in Figures 1C⇑ and 1E⇑ for EPI and ENDO myocytes, respectively. Ito recovered from inactivation at a much slower rate in ENDO than in EPI myocytes. This difference in kinetics is quantified in Figures 1D⇑ and 1F⇑. For this example, recovery from inactivation of EPI Ito is fit with a time constant of 30 ms, whereas the recovery from inactivation for ENDO Ito is fit by a constant of 950 ms. The average results summarized in the Table⇑ suggest that Ito in EPI is more dense, activates more negatively, inactivates more negatively, and recovers faster from inactivation than does the same current in ENDO.
Incubation of EPI Myocytes With Ang II Alters Ito so That Its Gating Properties Resemble Those Found in ENDO
Superfusion of EPI myocytes with 1 μmol/L Ang II for 5 to 20 minutes induced no change in the activation or density of Ito (n=8). Recovery from inactivation was also studied in 4 myocytes, where no change was observed. We then studied the effects of chronic exposure of EPI myocytes to Ang II by storing myocytes for a period of 2 to 52 hours in KB medium containing 0.5 to 2 μmol/L Ang II, after which electrophysiological studies of Ito were performed in the absence of Ang II at 32° to 35°C. We also incubated EPI myocytes from the same animals in the same storage solution and for the same time period but without Ang II. These myocytes showed no significant difference from the control values for EPI myocytes studied acutely. Figure 2A⇓ provides sample data from EPI myocytes incubated in control solution or with Ang II for 24 hours (Figure 2B⇓). Figure 2C⇓ clearly shows the current density is reduced by Ang II and that the activation has been shifted to more positive potentials (see Table⇑).
We also examined the steady-state voltage dependence of inactivation and the recovery from inactivation in the Ang II–incubated myocytes; the results are presented in Figure 3⇓. Steady-state inactivation is shifted to more positive potentials, and the recovery from inactivation is dramatically slowed by incubation with Ang II.
The Table⇑ shows a comparison of the properties of Ito in Ang II–incubated EPI ventricular myocytes with those from unincubated EPI and ENDO. It is clear that the gating properties of Ito and its density have been altered by incubation with Ang II. Most of the properties of Ito in EPI myocytes incubated with Ang II now more closely resemble those of ENDO Ito.
The effects of Ang II incubation on EPI Ito were prevented when the AT1 receptor blocker losartan (1 μmol/L) was included with Ang II in the incubate.
Incubation of ENDO Myocytes With the AT1 Receptor Blocker Losartan Alters Ito Such That Its Properties Resemble Those of EPI
Given the effects of incubation with Ang II on EPI myocytes and the existence of an autocrine renin-angiotensin system, we explored the possibility that blockade of the normal Ang II pathway via the AT1 receptor might modulate Ito in ENDO. We first acutely exposed ENDO myocytes to losartan (1 μmol/L). No acute effects on Ito activation or density (5 myocytes) or recovery from inactivation (3 myocytes) were observed (data not shown).
We next incubated ENDO myocytes for 2 to 52 hours with 1 to 2 μmol/L losartan in the same incubation solution as was used for EPI myocytes. We also examined ENDO myocytes stored in the same incubate without added losartan. We observed no change in the properties of Ito from those of unincubated ENDO myocytes. The effects of losartan on Ito amplitude and the voltage dependence of activation are presented in Figure 4⇓. Ito is increased and the voltage dependence of activation is shifted to more negative potentials (see the Table⇑), similar to control EPI.
We next examined the voltage dependence of inactivation and the time course of recovery from inactivation in the losartan-incubated ENDO myocytes. Figure 5⇓ illustrates the results. Steady-state inactivation shifts to more negative potentials, and the time constant for recovery from inactivation becomes much faster.
The Table⇑ compares properties of Ito in ENDO myocytes incubated with losartan with unincubated ENDO and EPI myocytes. Losartan incubation converts most properties of Ito in ENDO myocytes to resemble those of unincubated EPI myocytes.
Two additional control experiments were performed. In 1 experiment (n=10), EPI was incubated for 2 to 52 hours with losartan, and in the other (n=5), ENDO was incubated with Ang II for the same time period. No effect was observed of either incubation on Ito (data not shown).
Action Potential Notch Is Influenced by Ang II
Ito is responsible for phase 1 rapid repolarization. Given the effects of Ang II and of losartan on EPI and ENDO Ito, respectively, changes in the action potential contour might be expected. We recorded action potentials from 11 control EPI and 10 control ENDO myocytes as well as from 5 EPI myocytes incubated for 24 hours with 1 μmol/L Ang II and from 6 ENDO myocytes incubated for 24 hours with 1 μmol/L losartan. Representative results are illustrated in Figure 6⇓. Control EPI action potentials always demonstrated a notch, whereas control ENDO action potentials did not (a difference noted previously by others1 ). EPI myocyte incubation with Ang II resulted in loss of the notch, whereas ENDO myocyte incubation with losartan produced a notch. These changes are consistent with the changes in Ito that we described earlier.
mRNA levels for Kv4.3 and Kv1.4
The patch-clamp results revealed changes in Ito induced by incubation with either Ang II (EPI) or losartan (ENDO). However, they do not provide a mechanism for the observed effects. One possibility is a change in the molecular correlate of the current. We previously demonstrated that Kv4.3 underlies Ito in the canine ventricle. However, Kv1.4 is present in canine myocytes and recovers from inactivation much more slowly. We therefore examined the effects of incubation of EPI myocytes with 2 μmol/L Ang II for 24 hours and the effects of incubation of ENDO myocytes with 2 μmol/L losartan for the same period on the mRNA level for these K+ channel subunits. The mRNA levels were quantified from 3 samples with thee use of RNase protection assays. For EPI, (EPI+Ang II)/EPI=1.00±0.07 for Kv4.3 and 0.98±0.09 for Kv1.4. For ENDO, (ENDO+losartan)/ENDO=1.01±0.04 for Kv4.3 and 1.02±0.10 for Kv1.4. Neither incubation resulted in any change in mRNA levels for either transcript.
Heterologous Expression of Kv4.3 With the AT1 Receptor in X laevis Oocytes
If a change in the molecular correlate of Ito does not occur, possibly the changes could be induced by posttranslational modification of the existing protein. To test this hypothesis, we heterologously expressed the dominant molecular correlate of canine Ito, Kv4.3, in X laevis oocytes along with the AT1a receptor.
Figure 7⇓ shows the results of our study of the effects of incubation with 4 μmol/L Ang II for up to 32 hours on recovery from inactivation with the following protocol. Membrane potential was depolarized to −20 mV for 600 to 900 ms from a holding potential of −90 mV to completely inactivate Kv4.3 current. The recovery potential was −100 mV followed by test steps to −20 mV at various intervals. The recovery kinetics were slower for the treated oocyte (τ=361 ms) compared with the control oocyte (τ=141 ms) (Figures 7A⇓ and 7B⇓). The average ≈2-fold slowing of the time constant for recovery from inactivation is provided in Figure 7C⇓ for all oocytes studied. Figure 7C⇓ also illustrates that this effect is not observed in the absence of either Ang II or the AT1a receptor. Figure 7D⇓ illustrates the time course of the change in recovery from inactivation after incubation with Ang II. The time constant is 3.7 hours.
Although an effect is observed on recovery from inactivation, there was no effect of incubation with Ang II on either the voltage dependence of inactivation or activation (data not shown).
In agreement with previous studies,3 4 we found a smaller density and a slower recovery from inactivation of Ito in ENDO than in EPI. We also observed a more positive threshold for Ito activation in ENDO. The smaller current density, more positive activation, and slower recovery from inactivation all contribute to the absence of a notch in the ENDO action potential during steady-state ventricular beating. One additional difference between these 2 tissue types is the voltage dependence of Ito inactivation, which is more positive in ENDO.
The critical findings in our study were that, first, Ang II converts many of the properties of EPI Ito to those normally seen in ENDO, and second, losartan converts many of the properties of ENDO Ito to those normally seen in EPI. Important to these observations is that neither exposure of ENDO to agonist nor exposure of EPI to antagonist had any effect and that exposure of EPI to agonist plus antagonist also had no effect. Moreover, none of the effects on EPI or ENDO were acute: incubation was required. Such incubation of EPI in Ang II (1) decreases the Ito current density, (2) shifts the threshold for Ito activation to more positive voltages, (3) shifts the voltage dependence of Ito inactivation to more positive voltages, and (4) slows Ito recovery from inactivation by >1 order of magnitude. Although the acute actions of Ang II on cardiac membrane currents in other species are mediated by activation of protein kinase C18 and may include mitogen-activated protein kinase in the signaling cascade,19 the signaling pathway that underlies the chronic action of Ang II on Ito remains to be determined.
Chronic incubation of ENDO with losartan induces (1) an increase in Ito density, (2) a more negative Ito activation, (3) a more negative voltage dependence of Ito inactivation, and (4) a dramatic speeding of Ito recovery from inactivation. The resultant ENDO Ito resembles that observed in normal EPI. This observation is all the more remarkable because it occurs in the absence of applied agonist. One plausible explanation is that tonic activation of the AT1 receptor of normal ENDO myocytes occurs both in vivo and in the incubate and that the source of this tonic activation is an autocrine renin-angiotensin system present in the canine ventricular myocytes or in nonmyocyte cells that remain in the incubate.
If an autocrine renin-angiotensin system contributes to modulation of the T wave, then ENDO is likely exposed to a larger tonic influence of Ang II than EPI. This supposition is based on the observation that the conversion of ENDO Ito to EPI Ito involves AT1 receptor blockade, whereas the converse action requires AT1 receptor activation. This raises the possibility that these 2 cell types are normally under a chronically differing influence of the renin-angiotensin system. An autocrine renin-angiotensin system has already been reported in rat ventricular myocytes.11 12 13 Either Ang II production, AT1 receptor density, or coupling of the AT1 receptor to its second messenger pathway might be more efficient in the ENDO, leading to a larger tonic effect. A higher activity of ACE in ENDO than in EPI has been reported in rat ventricle.20 Also, angiotensin mRNA is higher in ENDO than in EPI in human ventricle.21 Preliminary experiments with ACE inhibition in dogs have demonstrated alterations in the T wave vector.22
The changes we observed in EPI and ENDO Ito could have been caused by at least 3 alternatives: (1) a change in the molecular correlate of Ito, (2) posttranslational modification of the channel protein, and (3) an auxiliary (β) subunit. We consider each in turn. (1) Regarding a change in the molecular correlate of Ito, in canine myocytes, Ito has been identified as Kv4.3, although in control conditions, there also is some expression of Kv1.4. Because Kv1.4 recovers from inactivation much more slowly than Kv4.3, an isoform switch had to be given serious consideration. Our RNase protection assays demonstrated no change in transcript level for either channel with chronic incubation. Barring a dramatic difference between message and protein level, a change in the α subunit of Ito is a remote possibility. (2) Regarding posttransitional modification of the channel protein, our results in X laevis oocytes demonstrated that after incubation with Ang II, Kv4.3 exhibited a 2-fold slowing of recovery from inactivation. However, no change in the voltage dependence of activation or inactivation was observed. Possibly, the oocyte environment cannot reproduce that of the ventricular myocyte, and mammalian cells may allow for a more complete reconstruction of the observed effects. (3) Regarding an auxiliary (β) subunit, it is known that β subunits can alter the gating properties of ion channels. In cardiac myocytes, IKs is a heteromultimer of KCNQ1 and minK.23 KCNQ1 gates markedly more rapidly in the absence of minK. Recently, a β subunit has been identified for the Kv4 family of ion channel subunits.24 The channel recovers from inactivation more rapidly than in its absence. If this β subunit were to underlie our results, its density should be higher in EPI than in ENDO, and this gradient should be altered by our long-term incubations.
One area in which our observations are relevant is cardiac memory, in which ECG T wave changes are associated with an altered transmural gradient for repolarization6 and phenotypic expression is blocked by ACE inhibition or Ang II receptor blockade.22 Another setting in which our observations may play a role is myocardial hypertrophy, whether pathological (eg, hypertension) or physiological (eg, postnatal left ventricular hypertrophy). Nonetheless, the importance of our findings to the evolution of repolarization in the heart and the morphology of the T wave remains to be established.
This work was supported by National Institutes of Health grants HL-20558, HL-53956, HL-28958, and NS-29755. Drs Wymore and Yu are supported by Scientist Development Awards of the American Heart Association. Dr Wang holds a postdoctoral fellowship from the American Heart Association, Northeast Affiliate.
- Received February 4, 2000.
- Accepted March 30, 2000.
- © 2000 American Heart Association, Inc.
Sicouri S, Antzelevitch C. A subpopulation of cells with unique electrophysiological properties in the deep subepicardium of the canine ventricle. Circ Res. 1991;68:1729–1741.
Anyukhovsky EP, Sosunov EA, Rosen MR. Regional differences in electrophysiologic properties of epicardium, midmyocardium and endocardium: in vitro and in vivo correlations. Circulation. 1996;94:1981–1988.
Liu D-W, Gintant GA, Antzelevitch C. Ionic bases for electrophysiological distinctions among epicardial, midmyocardial, and endocardial myocytes from the free wall of the canine left ventricle. Circ Res. 1993;72:671–687.
Nabauer M, Beuckelmann DJ, Uberfuhr P, Steinbeck G. Regional differences in current density and rate-dependent properties of transient outward current in subepicardial and subendocardial myocytes of human left ventricle. Circulation. 1996;93:168–177.
Dixon JE, Shi W, Wang H-S, McDonald C, Yu H, Wymore RS, Cohen IS, McKinnon D. Role of the Kv4.3 K+ channel in ventricular muscle: a molecular correlate for the transient outward current. Circ Res. 1996;79:659–668.
Rosen MR, Cohen IS, Danilo P Jr, Steinberg SF. The heart remembers. Cardiovasc Res. 1998;40:469–482.
Hool LC, Gray DF, Robinson BG, Rasmussen HH. Angiotensin-converting enzyme inhibitors regulate the Na+-K+ pump via effects on angiotensin metabolism. Am J Physiol. 1996;271:C172–C180.
Tsuchiya K, Horie M, Watanuki M, Albrecht CA, Obayashi K, Fujiwara H, Sasayama S. Functional compartmentalization of ATP is involved in angiotensin II–mediated closure of cardiac ATP-sensitive K+ channels. Circulation. 1997;96:3129–3135.
Dostal DE, Rothblum KN, Chernin MI, Cooper GR, Baker KM. Intracardiac detection of angiotensinogen and renin: a localized renin-angiotensin system in neonatal rat heart. Am J Physiol. 1992;263:C838–C850.
Zhang, X., Dostal, DE, Reiss K, Cheng W, Kajstura J, Li P, Huang H, Sonnenblick EH, Meggs LG, Baker KM, Anversa P. Identification and activation of autocrine renin-angiotensin system in adult ventricular myocytes. Am J Physiol. 1995;269:H1791–H1802.
Clement-Chomienne O, Walsh MP, Cole WC. Angiotensin II activation of protein kinase C decreases delayed rectifier K+ current in rabbit vascular myocytes. J Physiol (Lond). 1996;495:3:689–700.
Yamazaki T, Komuro I, Zou Y, Kudoh S, Mizuno T, Hiroi Y, Shiojima I, Takano H, Kinugawa K, Kohmoto O, Takahashi T, Yazaki Y. Protein kinase A and protein kinase C synergistically activate the Raf-1 kinase/mitogen-activated protein kinase cascade in neonatal rat cardiomyocytes. J Mol Cell Cardiol. 1997;29:2491–2501.
Yamada H, Fabris B, Allen AM, Jackson B, Johnston CI, Mendelsohn FAO. Localization of angiotensin converting enzyme in rat heart. Circ Res. 1991;68:141–149.
Mattsson KI, Monaghan MM, Strassle BW, An WF, Bowlby MR, Trimmer JS, Buchwalder L, Rhodes KJ. Colocalization and coassociation of a novel family of neuronal calcium sensor homologs with Kv4-family K+ channels in mammalian brain. Soc Neurosci. 1999;25:983. Abstract.