Overexpression of HCN-encoded pacemaker current silences bioartificial pacemakers

Overexpression of HCN-encoded pacemaker current silences bioartificial pacemakers Deborah K. Lieu, PhD,*†‡ Yau Chi Chan, MPhil,¶ Chu Pak Lau, MD,¶ Hung Fat Tse, MD, PhD,¶ Chung Wah Siu, MBBS,*†¶ Ronald A. Li, PhD*†¶§ From the *Stem Cell Program and †Department of Cell Biology and Human Anatomy, University of California, Davis, California, ‡NSF Center for Biophotonics Science and Technology, Sacramento, California, ¶Division of Cardiology, Department of Medicine, Stem Cell Program and Heart, Brain, Hormone & Healthy Aging Center, University of Hong Kong, Hong Kong, and §Institute of Pediatric Regenerative Medicine, Shriners Hospital for Children of North America, Sacramento, California. BACKGROUND Current strategies of engineering bioartificial pacemakers from otherwise silent yet excitable adult atrial and ventricular cardiomyocytes primarily rely on either maximizing the hyperpolarization-activated If or on minimizing its presumptive opponent, the inwardly rectifying potassium current IK1. OBJECTIVE The purpose of this study was to determine quantitatively the relative current densities of If and IK1 necessary to induce automaticity in adult atrial cardiomyocytes. METHODS Automaticity of adult guinea pig atrial cardiomyocytes was induced by adenovirus (Ad)-mediated overexpression of the gating-engineered HCN1 construct HCN1-⌬⌬⌬ with the S3-S4 linker residues EVY235-7 deleted to favor channel opening. RESULTS Whereas control atrial cardiomyocytes remained electrically quiescent and had no If, 18% of Ad-CMV-GFP-IRES-HCN1⌬⌬⌬ (Ad-CGI-HCN1-⌬⌬⌬)–transduced cells demonstrated automaticity (240 ⫾ 14 bpm) with gradual phase 4 depolarization (143 ⫾ 28 mV/s), a depolarized maximal diastolic potential (⫺45.3 ⫾ 2.2 mV),

Introduction Heart rhythms originate from the sinoatrial node, which consists of a few thousand pacemaker cells. Malfunction of the pacemaker cells often necessitates implantation of an electronic pacemaker. If, which is encoded by the hyperpolarization-activated cyclic nucleotide-modulated (HCN) channel gene family, plays a pivotal role in cardiac pacing by acting as an intrinsic oscillator that drives diastolic deThis work was supported by grants from the National Institutes of Health (R01-HL72857) to Dr. Li, from the Stem Cell Program of the University of California to Dr. Li, and from the Hong Kong Research Grant Council (HKU 7459/04M) to Drs. Lau, Tse, and Li. Dr. Chan was supported by a postgraduate studentship from the University of Hong Kong. Dr. Lieu was supported by a fellowship from the Shriners Hospital for Children. Dr. Siu was supported by a postdoctoral fellowship award from the Croucher Foundation. Drs. Lieu, Chan, and Siu contributed equally to this work. Address reprint requests and correspondence: Dr. Ronald Li, University of California, Davis, Room 650, Shriners Hospital, 2425 Stockton Boulevard, Sacramento, California 95817. E-mail address: ronaldli@ ucdavis.edu. (Received December 12, 2007; accepted May 10, 2008.)

and substantial If at ⫺140 mV (If,⫺140 mV ⫽ ⫺9.32 ⫾ 1.84 pA/pF). In the remaining quiescent Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes, two distinct immediate phenotypes were observed: (1) 13% had a hyperpolarized resting membrane potential (⫺56.7 ⫾ 1.3 mV) with If,⫺140 mV of ⫺4.85 ⫾ 0.97 pA/pF; and (2) the remaining 69% displayed a depolarized resting membrane potential (⫺27.6 ⫾ 1.3 mV) with If,⫺140 mV of ⫺23.0 ⫾ 3.71 pA/pF. Upon electrical stimulation, both quiescent groups elicited a single action potential with incomplete phase 4 depolarization that was never seen in controls. Further electrophysiologic analysis indicates that an intricate balance of IK1 and If is necessary for induction of atrial automaticity. CONCLUSION Optimized pacing induction and modulation can be better achieved by engineering the If/IK1 ratio rather than the individual currents. KEYWORDS HCN channel; Gene transfer; Pacemaker (Heart Rhythm 2008;5:1310 –1317) © 2008 Heart Rhythm Society. All rights reserved.

polarization to the action potential (AP) threshold after each excitation cycle.1–3 Expression of If and the absence of the Kir2-encoded inwardly rectifying potassium current IK1, a resting membrane potential (RMP) stabilizer,4 are signatures of sinoatrial nodal pacemaker cells. Conversely, IK1 but not If is robustly expressed in the silent yet excitable adult atrial and ventricular cardiomyocytes.5,6 Based on these differences, two genetic approaches—IK1 suppression7,8 and If overexpression1,9 –11— have been used in independent experiments to convert normally quiescent cardiomyocytes into spontaneously AP-firing cells as bioartificial pacemakers.12 However, previous cellular and electrophysiologic experiments performed to test these strategies were largely limited to ventricular cardiomyocytes rather than atrial cardiomyocytes, although the latter are the more likely recipient for future gene-based therapies for sick sinus syndrome. Given the different ion channel expression profiles of atrial cardiomyocytes and ventricular cardiomyocytes, further investigations of atrial cardiomyocytes are war-

1547-5271/$ -see front matter © 2008 Heart Rhythm Society. All rights reserved.

doi:10.1016/j.hrthm.2008.05.010

Lieu et al

If and IK1 in Induction of Atrial Automaticity

ranted. Using a combination of computational modeling and somatic gene transfer techniques, we explored in detail the biophysical basis and interactions of If and IK1 in the induction of atrial automaticity.

Materials and methods Mathematical formulation Our mathematical model for atrial cardiomyocytes was formulated based on a previous guinea pig ventricular model13 and modified by decreasing IK1 and INa, with ion concentrations fixed for simplicity. If was introduced and adjusted as previously described,14,15 and its conductance magnitude was varied to test for automaticity dependency on If. Numerical results were obtained from MATLAB using a variable order ordinary differential equation solver (ode15) plus a built-in backward-difference method, with a relative tolerance of 10⫺8 and an absolute tolerance of 10⫺4.

Molecular biology Polymerase chain reaction– based mutagenesis of mouse HCN1 (generously provided by Dr. Steve Segalbaum, Columbia University) of the bicistronic adenovirus shuttle vector pAdCMV-GFP-IRES (pAdCGI) was performed with overlapping oligos as described in our previous publications.16,17 The internal ribosomal entry site (IRES) allows simultaneous translation of two transgenes— green fluorescent protein (GFP) and an engineered-HCN1 construct— with a single transcript. Mouse HCN1 was chosen because its biophysical properties were most extensively characterized in our previous reports,1,14,16,18 –24 making it the most suitable candidate for gene transfer experiments. Although the heart rates of mouse and guinea pig significantly differ (cycle length ⬃150 and 300 ms, respectively), their phase 4 slopes are relatively comparable (human ⬃49 mV/s, mouse ⬃50 mV/s, guinea pig ⬃45 mV/s). The heart rate differences can be attributed to other factors, such as AP duration, which in turn are dependent on the particular species-specific ion channel composition.25 Because a gating-engineered construct was used to compensate for any contextdependent effects, the specific species or isoform chosen is not as important as the gating property per se. Adenoviruses were generated by Cre-lox recombination of purified ⌿5 viral DNA and shuttle vector DNA using Cre4 cells.26 The recombinant products were plaque purified, amplified, and purified again by CsCl2 gradients, yielding concentrations on the order of 1010 plaque-forming units per milliliter.

Adenovirus-mediated gene transfer and isolation of atrial cardiomyocytes Adult female guinea pigs (weight ⬃250 g) were euthanized by intraperitoneal injection of pentobarbital 80 mg/kg. The hearts were quickly excised, followed by perfusion with 200 U/mL collagenase II using a customized Langendorff apparatus.27 Atrial cardiomyocytes were plated at 5 ⫻ 105 per laminin-coated glass coverslip in medium containing 5 mM carnitine, 5 mM creatine, 5 mM taurine, 100 ␮g/mL penicillin-streptomycin, and 10% fetal bovine serum in Medium

1311 199 (Invitrogen Corp., Carlsbad, CA, USA) at 37°C with 5% CO2 for 2 hours. For transduction, atrial cardiomyocytes were incubated for 1 hour in serum-free medium containing adenoviral particles at a concentration of ⬃2 ⫻ 109 plaqueforming units per milliliter, then refreshed with normal culture medium. A transduction efficiency of ⬃70% to 80% typically could be achieved with this protocol. The same in vitro transduction system of adult guinea pig cardiomyocytes has been used previously by us1,28 and by others.29 Because cardiomyocytes isolated from the hearts of animals that underwent in vivo intracardiac injection of adenoviruses showed identical data trends with the in vitro transduction system as we reported previously,30 we switched to the latter to reduce the need for sacrificing animals without compromising experimental efficiency.

Electrophysiology and data analysis All electrophysiologic experiments were performed at 37°C to simulate physiological conditions. Whole-cell patch clamp technique used an Axopatch 200B amplifier and pClamp 9.2 software (Axon Instruments Inc., Sunnyvale, CA, USA). A xenon arc lamp was used to view GFP fluorescence at 488/530 nm (excitation/emission). Successfully transduced cells showed green epifluorescence. Patch pipettes were prepared from 1.5-mm thin-walled borosilicate glass tubes using a Sutter micropipette puller P-97 and had typical resistances of 3 to 5 M⍀ when filled with an internal solution containing the following (in mM): 110 K-aspartate, 20 KCl, 1 MgCl2, 0.1 Na-GTP, 5 Mg-ATP, 5 Na2-phosphocreatine, 1 EGTA, and 10 HEPES, pH adjusted to 7.3 with KOH. The external Tyrode’s bath solution consisted of the following (in mM): 140 NaCl, 5 KCl, 1 MgCl2, 1 CaCl2, 10 glucose, and 10 HEPES, pH adjusted to 7.4 with NaOH. To optimize the time window for electrical recording and to avoid contamination of automaticity due to time-dependent IK1 reduction of atrial cardiomyocytes in culture, we first performed a detailed comparison of I K1 and RMP of freshly isolated cardiomyocytes and those after 24 to 36 hours in culture. During this time window, no significant changes in either RMP (⫺59.2 ⫾ 0.9 mV vs ⫺58.5 mV ⫾ 1 mV, P ⬎.05) or IK1 density at ⫺100 mV (⫺10.6 ⫾ 1.0 pA/pF vs ⫺11.8 ⫾ 2.9 pA/pF, P ⬎.05) were observed between freshly isolated and cultured atrial cardiomyocytes. More importantly, absolutely no cultured atrial cardiomyocytes recorded during this period exhibited automaticity. All subsequent recordings were performed within 24 to 36 hours. Upon membrane rupturing, AP recording was always performed first. Spontaneous electrical activity was measured in atrial cardiomyocytes that were left passive without current input, whereas stimulation of 0.1 to 1 nA for 5 ms was given to silent atrial cardiomyocytes for eliciting an AP. No obvious rundown was observed as gauged by reproducible AP phenotype, firing or not. Measurements reported were not corrected for the experimentally determined liquid junction potential of ⫺20.1 ⫾ 0.9 mV. After determining the AP phenotype, voltage clamp experiments were per-

1312

Heart Rhythm, Vol 5, No 9, September 2008

formed to characterize IK1 and If expressed in the same cells. To elicit inward currents, atrial cardiomyocytes were held at ⫺30 mV and pulsed from 0 to ⫺140 mV in 10-mV increments for 2 seconds, followed by a 1-second, ⫺100 mV pulse. If was defined as ZD7288-sensitive but Ba2⫹insensitive time-dependent currents and IK1 as Ba2⫹-sensitive current. The steady-state current/voltage (I-V) relationship was determined by plotting the currents measured at the end of a 2-second test pulse of the protocol mentioned against the test potentials. The voltage dependence of HCN channel activation was assessed by plotting time-dependent tail currents at ⫺100 mV measured immediately after the 2-second test pulse (0 to ⫺140 mV) as a function of the test pulse voltage. Currents were normalized to the maximum tail current recorded. These recordings were made in the presence of Ba2⫹. Data are reported as mean ⫾ SEM. P ⬍.05 was considered significant as determined by unpaired t-test.

tential of K⫹ (Figure 1A, left). Upon receiving a current stimulus, a typical adult AP could be elicited (Figure 1B). With gradual increase in the maximum conductance for If (from GIf ⫽ 0 pS) while keeping a constant conductance for IK1, spontaneous firing was achieved, consistent with what we previously demonstrated mathematically14 and experimentally.1,30 Increasing the conductance of If (from GIf ⫽ 5 to 7.5 nS) accelerated the firing frequency, which was associated with reduced AP amplitude, depolarized maximal diastolic potential, and diastolic phase 4 depolarizing phase similar to those of genuine adult nodal pacemaker cells (Figure 1A, middle). Interestingly, rather than accelerating the firing frequency, further increase in If (GIf ⱖ20 nS) significantly depolarized RMP and subsequently led to cessation of automaticity (Figure 1A, right). However, unlike the normal ventricular phenotype, a single electrical stimulus gave rise to damped oscillations of the membrane potential (Figure 1C).

Results

If overexpression in atrial cardiomyocytes resulted in distinct electrical phenotypes

If overexpression silences instead of inducing automaticity In the absence of If but in the presence of IK1, our baseline simulations reproduced the quiescent phenotype of adult atrial cardiomyocytes with RMP close to the reversal po-

Our modeling results suggest that although If alteration can be an effective strategy for inducing and modulating automaticity, an optimal operating range exists. Outside this range, pacing cannot be induced at best or arrhythmias result at

Figure 1 Mathematical model results of If-induced pacing in cardiomyocytes. A: Effect of increasing If on spontaneous generation of action potential (AP). Modeling results for membrane potential (Em), IK1, and If simulated for increasing magnitudes of conductance are shown. Without If, resting membrane potential is close to the reversal potential of K⫹, so pacing is not possible. With addition of If, spontaneous firing occurs and frequency increases with increasing If until the introduced If is markedly greater than IK1, when the quiescent phenotype returns due to a depolarized resting membrane potential. AP can be elicited by a current stimulus both in quiescent cardiomyocytes with no If (B) and in quiescent cardiomyocytes due to large If (C). Phase 4 depolarization is present in a quiescent cardiomyocyte with large If but is absent in quiescent cell without If.

Lieu et al

If and IK1 in Induction of Atrial Automaticity

1313

Figure 2 IK1 and If tracings of control and Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes. Control cells displayed IK1 (A) but no If (B), whereas transduced atrial cardiomyocytes displayed comparable IK1 (C) and sizable If (D).

worst. To validate our computational predictions and explore in detail the basis of atrial automaticity induction, we introduced If into atrial cardiomyocytes by transduction with the recombinant adenovirus Ad-CGI-HCN1-⌬⌬⌬. The engineered HCN1-⌬⌬⌬ construct, whose S3-S4 linker has been shortened by deleting residues 235–237 to favor channel opening, was chosen to reproduce native If without the need for considering such poorly defined factors as accessory subunits and cellular context as we previously described.16,17,30 Control (untransduced) atrial cardiomyocytes had IK1 but no If (Figures 2A and 2B), whereas the transduced atrial cardiomyocytes displayed robust IK1 and If (Figures 2C and 2D). Detailed current/voltage relationships are shown in Figure 3. As anticipated, all control atrial cardiomyocytes were electrically quiescent but capable of generating a single AP upon injection of a depolarizing current stimulus (0.1–1.0 nA for 5 ms; Figure 3A). Ad-CGI-HCN1-⌬⌬⌬ transduction of atrial cardiomyocytes resulted in two distinct electrical phenotypes: an intrinsically firing and a quiescent-yet-excitable type. The first, accounting for 18% of all Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes recorded (n ⫽ 113), exhibited rhythmic firing even in the complete absence of any external stimulus (Figure 3B). The spontaneous APs had a rate of 240 ⫾ 14 bpm, upstroke velocity of 121 ⫾ 34 mV/s, and gradual phase 4 depolarization slope of 143 ⫾ 28 mV/s. Of note, their maximal diastolic potential (⫺45.3 ⫾ 2.2 mV) was significantly depolarized relative to the RMP of control atrial cardiomyocytes (⫺58.5 ⫾ 1.0 mV, P ⬍.01). The remaining 82% of the Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes were completely quiescent. Detailed analysis of these electrically silent atrial cardiomyocytes posttransduction enables us to further categorize them into two subgroups: one with a mean hyperpolarized RMP of ⫺56.7 ⫾ 1.3 mV (n ⫽ 15 of 93; Figure 3C) and another

with a significantly more depolarized RMP of ⫺27.6 ⫾ 1.3 mV (n ⫽ 78 of 93; Figure 3D). Of note, the latter phenotype was not observed in our previous report using adult ventricular cardiomyocytes.1 Electrical stimulation of both groups elicited single APs displaying an incomplete “phase 4-like” depolarization that did not lead to subsequent firing (arrows). Such phase 4-like depolarization was more prominent in the group that had depolarized RMPs, showing a depolarization slope of 68 ⫾ 7 mV/s (vs 27 ⫾ 5 mV/s of transduced atrial cardiomyocytes with hyperpolarized RMP). RMP, maximal diastolic potential, and phase 4 slope are summarized as bar graphs in Figures 4A and 4B. In addition, their mean AP duration to 90% repolarization (APD90 ⫽ 58.5 ⫾ 4.8 ms) was shorter than that of the hyperpolarized RMP group (APD90 ⫽ 96.4 ⫾ 23.6 ms), although this difference did not reach statistical significance (P ⫽ .13).

Successful automaticity induction depends on the relative magnitude of If to IK1 To explore the ionic basis underlying the different electrical phenotypes for Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes, we measured If and IK1 (Figures 3E–3H). Given the observed heterogeneity of AP profiles, voltage and current clamp recordings were simultaneously performed on the same cells to directly correlate If and IK1 magnitudes to the resultant AP phenotype. Quiescent control atrial cardiomyocytes robustly expressed Ba2⫹-sensitive IK1 with the typical inwardly rectifying behavior but no If (Figure 3E, n ⫽ 12). Conversely, hyperpolarization-activated time-dependent If could be seen in all Ad-CGI-HCN1⌬⌬⌬–transduced atrial cardiomyocytes. The range of magnitudes varied due to the sporadic nature of Ad-mediated transgene expression (Figures 3F–3H). Despite these differences in If, there was no statistically significant IK1 density (at ⫺140 mV) difference relative to control (⫺22.0 ⫾ 2.1

1314

Heart Rhythm, Vol 5, No 9, September 2008

Figure 3 Ad-CGI-HCN1-⌬⌬⌬ transduction resulted in three electrophysiologic phenotypes. Action potential (AP) and normalized current/voltage relationships of If and IK1 for control (A and E, n ⫽ 12) and Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes with spontaneously firing APs (B and F, n ⫽ 7), quiescent with hyperpolarized resting membrane potential (RMP; C and G, n ⫽ 11), and quiescent with depolarized RMP (D and H, n ⫽ 17) are shown. Arrows indicate phase 4 depolarization. MDP ⫽ maximal diastolic potential.

pA/pF) among spontaneously firing atrial cardiomyocytes (⫺23.3 ⫾ 6.5 pA/pF, P ⫽ .85), quiescent cells with depolarized RMP (⫺26.9 ⫾ 4.4 pA/pF, P ⫽ .34), and quiescent cells with hyperpolarized RMP (⫺30.6 ⫾ 3.8 pA/pF, P ⫽ .07). Of note, spontaneously firing Ad-CGI-HCN1-⌬⌬⌬– transduced atrial cardiomyocytes expressed a sizable If, but its magnitude was only 64% ⫾ 7% and 50% ⫾ 17% of IK1

when assessed at ⫺40 mV and ⫺140 mV, respectively. The voltages of ⫺40 and ⫺140 mV were chosen for analysis to reflect their physiologic relevance and conductances (when the open probability was 100% at ⫺140 mV), respectively. Interestingly, transduced but electrically silent atrial cardiomyocytes with hyperpolarized RMPs expressed smaller If than IK1, whereas those with depolarized

Lieu et al

If and IK1 in Induction of Atrial Automaticity

1315 transduced atrial cardiomyocytes with depolarized RMP (If/IK1 ⬃ 1 at ⫺140 mV; Figure 4D), If suppression (to lower the ratio) with 20 ␮M ZD7288, a specific HCN channel blocker, shifted RMP from ⫺29.2 ⫾ 0.8 mV to ⫺47.6 ⫾ 1.2 mV and subsequently triggered rhythmic firing (Figure 5C, n ⫽ 3). Collectively, our experimental results were consistent with our computational predictions and further indicate that an optimal balance of If and IK1, rather than one-sided If overexpression or IK1 suppression, is necessary for automaticity induction and modulation in atrial cardiomyocytes.

Discussion

Figure 4 Action potential characteristics of control and Ad-CGI-HCN1⌬⌬⌬–transduced atrial cardiomyocytes. A: Resulting membrane potential (RMP) or maximal diastolic potential (MDP) for control, transduced pacing atrial cardiomyocytes, and transduced quiescent atrial cardiomyocytes with depolarized or hyperpolarized RMP. B: Phase 4 slope for transduced atrial cardiomyocytes. C: Effect of If to IK1 ratio (|If/IK1|) at ⫺40 mV on electrophysiologic phenotypes of atrial cardiomyocytes. D: Effect of If to IK1 ratio (|If/IK1|) at ⫺140 mV on electrophysiologic phenotypes of atrial cardiomyocytes. Shaded bars represent Ad-CGI-HCN1-⌬⌬⌬–transduced cells.

RMPs expressed If larger than and comparable to IK1 at ⫺40 and ⫺140 mV, respectively. The If/IK1 ratios for all AP are summarized in Figures 4C and 4D. They displayed the same rank orders at ⫺40 and ⫺140 mV. Because our computational and experimental results strongly suggest that automaticity occurs only within a finite range of If/IK1 ratios, we proceeded to obtain experimental proof to the “counterintuitive” notions that (1) IK1 suppression can cease spontaneous firing and (2) If suppression can induce automaticity. Figure 5A shows that, in typical spontaneously firing HCN1-⌬⌬⌬–transduced atrial cardiomyocytes (n ⫽ 4), IK1 suppression with Ba2⫹ to increase the If/IK1 ratio paradoxically terminated the rhythmic firing and shifted RMP to ⫺0.8 ⫾ 0.2 mV. Indeed, application of Ba2⫹ to HCN1-⌬⌬⌬–transduced atrial cardiomyocytes with hyperpolarized RMP, whose If/IK1 ratio was smaller than the spontaneously firing HCN1-⌬⌬⌬–transduced atrial cardiomyocytes (Figures 4C and 4D), induced transient automaticity as their RMP depolarized and achieved an optimal If/IK1 (n ⫽ 9). Interestingly, in quiescent HCN1-⌬⌬⌬–

In recent years, efforts to create bioartificial pacemakers as potential alternatives or supplements to electronic devices for treatment of cardiac arrhythmias have been experimented exclusively by either maximizing If11,30 or by minimizing its “antagonizing” IK1,7 to convert nonpacing cardiomyocytes into spontaneously firing pacemaker-like cells. Through in silico and in vitro HCN1 gene transfer atrial cardiomyocyte experiments, we demonstrated in the present study that engineering the balance between If and IK1, rather than a one-sided manipulation, is a more flexible approach for automaticity induction and its modulation. Mechanistically, for any given current density of IK1, the introduced If could induce automaticity only when its maximal current density at ⫺140 mV is approximately 50% of IK1 (i.e., If/IK1 ⫽ 0.5; Figure 4C). Even for a cell with small or no If (e.g., adult ventricular cardiomyocytes), automaticity still could be induced by IK1 suppression to unleash the latent pacing ability as demonstrated presently for the case of HCN1-⌬⌬⌬–transduced atrial cardiomyocytes with hyperpolarized RMP in the presence of Ba2⫹ and previously with genetic suppression of IK1.7 At the other extreme, excess If that overwhelms IK1 (i.e., If/IK1 ⬎0.5 at ⫺140 mV) would depolarize RMP to such an extent that automaticity cannot be induced. Based on our mathematical model, spontaneous AP generation of atrial cardiomyocytes occurred only when an optimal balance between If and IK1 was established. Therefore, perturbation of this balance by changing either current would terminate pacing. The necessity of achieving this optimal balance, however, was not apparent in our previous report using guinea pig ventricular cardiomyocytes,1 whose intense expression of intrinsic IK1 relative to the induced If magnitude limited the experimental exploration to this facade of the overall picture. The present experiment in guinea pig atrial cardiomyocytes with less intrinsic IK1 provides important experimental observations supporting the notions that (1) the firing activity of pacing Ad-CGI-HCN1⌬⌬⌬–transduced atrial cardiomyocytes (with If/IK1 ⫽ 0.5) could be ceased after IK1 blockade (which previously was shown to unleash the latent pacemaker activity of ventricular cardiomyocytes7) and (2) silent Ad-CGI-HCN1-⌬⌬⌬– transduced cells with hyperpolarized (If/IK1 ⬍0.5) and depolarized (If/IK1 ⬎0.5) RMP could be rendered electrically active by inhibiting IK1 and If (to increase and decrease If/IK1), respectively. Thus, If/IK1 is crucial for inducing

1316

Heart Rhythm, Vol 5, No 9, September 2008

Figure 5 Effects of IK1 blocker Ba2⫹ and If blocker ZD7288 on Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes. A: Spontaneous action potential (AP) in Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes was abolished by addition of 3 mM Ba2⫹ (n ⫽ 4). B: Quiescent Ad-CGI-HCN1-⌬⌬⌬– transduced atrial cardiomyocytes with hyperpolarized resting membrane potential (RMP) generated spontaneous AP, with 3 mM Ba2⫹ blocking excess IK1 (n ⫽ 9). C: Quiescent Ad-CGI-HCN1-⌬⌬⌬–transduced atrial cardiomyocytes with depolarized resting membrane potential from overexpression of If generated AP with 20 ␮M ZD7288 (n ⫽ 3).

automaticity of atrial cardiomyocytes by modulating RMP, which subsequently affects voltage-gated Na⫹ and L-type Ca2⫹ currents that are involved in AP depolarization. To explore the mechanisms of cardiac automaticity, it is important to study native sinoatrial nodal cells. However, the main scope of the current study was to determine the relative current density of If and its presumptive “opponent” IK1 that is necessary to induce automaticity in normally quiescent yet excitable atrial cardiac muscle cells, which are the likely target recipients of gene therapy for constructing bioartificial pacemakers (e.g., in sick sinus syndrome). Given the significantly different ion channel panoplies in sinoatrial nodal cells and atrial cardiomyocytes (e.g., the former have virtually no IK1, whereas the latter express ⬃1/6 of the IK1 in ventricular cardiomyocytes), we chose to focus both our computational and experimental efforts on the conversion of atrial cardiomyocytes. The results of the present study have two major implications for future efforts on the generation of bioartificial pacemaker. The first concerns the successful conversion of quiescent ventricular or atrial cardiomyocytes into pacemaker-like cells. Because an optimal operating ratio of If to IK1 exists, the dosage of If to be administered should be dependent on intrinsic IK1 expression, which varies among chamber-specific cells. For example, right atrial cardiomyocytes (for sinoatrial node reconstruction) that have an intrinsically smaller IK1 density than that of ventricular cardiomyocytes would require less If expression for conversion into pacing cells. Second, our work suggests the need for safety consideration in converted pacing cells for clinical

application. If and IK1 mismatch not only fails to induce automaticity but will result in heterogeneous RMP and AP duration in atrial cardiomyocytes, which may become a substrate for arrhythmias, thus further highlighting the importance of accurate and customized If dosing. This study has several limitations. First, adenovirus-mediated expression of If was sporadic. However, it was this sporadic nature that enabled us to explore the effects of If/IK1 in the first place. Second, the cellular effects of autonomic influences on HCN-induced bioartificial pacemaker have not been examined. Presumably, activation of adrenergic receptor and IKACh in vivo likely affects AP parameters such as phase 4 slope, RMP, maximal diastolic potential, and AP duration. Further experiments for defining the optimal dose of If in relation to autonomic modulations are warranted. Finally, although Ba2⫹ is commonly used to block IK1, it is known to influence ionic channels such as If, IKr, and IKs. Collectively, we have shown that a fine balance between If and IK1 is the key for successful induction of automaticity and, more importantly, its modulation in atrial cardiomyocytes. A qualitative approach of If overexpression or IK1 suppression may be replaced with a more quantitative approach in titrating the different ionic components, designed specifically for the target cells of interest to engineer a functionally viable bioartificial pacemaker.

References 1. Xue T, Siu CW, Lieu DK, et al. Mechanistic role of I(f) revealed by induction of ventricular automaticity by somatic gene transfer of gating-engineered pacemaker (HCN) channels. Circulation 2007;115:1839 –1850.

Lieu et al

If and IK1 in Induction of Atrial Automaticity

2. Irisawa H, Brown HF, Giles W. Cardiac pacemaking in the sinoatrial node. Physiol Rev 1993;73:197–227. 3. DiFrancesco D. Pacemaker mechanisms in cardiac tissue. Annu Rev Physiol 1993;55:455– 472. 4. Boyett MR, Honjo H, Kodama I. The sinoatrial node, a heterogeneous pacemaker structure. Cardiovasc Res 2000;47:658 – 687. 5. Schram G, Pourrier M, Melnyk P, et al. Differential distribution of cardiac ion channel expression as a basis for regional specialization in electrical function. Circ Res 2002;90:939 –950. 6. Yu H, Chang F, Cohen IS. Pacemaker current exists in ventricular myocytes. Circ Res 1993;72:232–236. 7. Miake J, Marban E, Nuss HB. Gene therapy: biological pacemaker created by gene transfer. Nature 2002;419:132–133. 8. Milanesi R, Baruscotti M, Gnecchi-Ruscone T, et al. Familial sinus bradycardia associated with a mutation in the cardiac pacemaker channel. N Engl J Med 2006;354:151–157. 9. Qu J, Barbuti A, Protas L, et al. HCN2 overexpression in newborn and adult ventricular myocytes: distinct effects on gating and excitability. Circ Res 2001; 89:E8 –E14. 10. Qu J, Plotnikov AN, Danilo P, et al. Expression and function of a biological pacemaker in canine heart. Circulation 2003;107:1106 –1109. 11. Bucchi A, Plotnikov AN, Shlapakova I, et al. Wild-type and mutant HCN channels in a tandem biological-electronic cardiac pacemaker. Circulation 2006; 114:992–999. 12. Cowan DB, McGowan FX Jr. A paradigm shift in cardiac pacing therapy? Circulation 2006;114:986 –988. 13. Matsuoka S, Sarai N, Kuratomi S, et al. Role of individual ionic current systems in ventricular cells hypothesized by a model study. Jpn J Physiol 2003;53:105–123. 14. Azene EM, Xue T, Marban E, et al. Non-equilibrium behavior of HCN channels: insights into the role of HCN channels in native and engineered pacemakers. Cardiovasc Res 2005;67:263–273. 15. Kurata Y, Hisatome I, Imanishi S, et al. Dynamical description of sinoatrial node pacemaking: improved mathematical model for primary pacemaker cell. Am J Physiol Heart Circ Physiol 2002;283:H2074 –H2101. 16. Lesso H, Li RA. Helical secondary structure of the external S3-S4 linker of pacemaker (HCN) channels revealed by site-dependent perturbations of activation phenotype. J Biol Chem 2003;278:22290 –22297. 17. Tsang SY, Lesso H, Li RA. Critical intra-linker interactions of HCN1-encoded pacemaker channels revealed by interchange of S3-S4 determinants. Biochem Biophys Res Commun 2004;322:652– 658.

1317 18. Tsang SY, Lesso H, Li RA. Dissecting the structural and functional roles of the S3-S4 linker of pacemaker (hyperpolarization-activated cyclic nucleotide-modulated) channels by systematic length alterations. J Biol Chem 2004;279:43752– 43759. 19. Azene EM, Sang D, Tsang SY, et al. Pore-to-gate coupling of HCN channels revealed by a pore variant that contributes to gating but not permeation. Biochem Biophys Res Commun 2005;327:1131–1142. 20. Xue T, Li RA. An external determinant in the S5-P linker of the pacemaker (HCN) channel identified by sulfhydryl modification. J Biol Chem 2002;277: 46233– 46242. 21. Xue T, Marban E, Li RA. Dominant-negative suppression of HCN1- and HCN2-encoded pacemaker currents by an engineered HCN1 construct: insights into structure-function relationships and multimerization. Circ Res 2002;90: 1267–1273. 22. Azene EM, Xue T, Li RA. Molecular basis of the effect of potassium on heterologously expressed pacemaker (HCN) channels. J Physiol 2003;547(Pt 2):349 –356. 23. Henrikson CA, Xue T, Dong P, et al. Identification of a surface charged residue in the S3-S4 linker of the pacemaker (HCN) channel that influences activation gating. J Biol Chem 2003;278:13647–13654. 24. Au KW, Siu CW, Lau CP, et al. Structural and functional determinants in the S5-P region of HCN-encoded pacemaker channels revealed by cysteine-scanning substitutions. Am J Physiol 2008;294:C136 –C144. 25. Satoh H. Sino-atrial nodal cells of mammalian hearts: ionic currents and gene expression of pacemaker ionic channels. J Smooth Muscle Res 2003;39:175– 193. 26. Hardy S, Kitamura M, Harris-Stansil T, et al. Construction of adenovirus vectors through Cre-lox recombination. J Virol 1997;71:1842–1849. 27. Mitra R, Morad M. Two types of calcium channels in guinea pig ventricular myocytes. Proc Natl Acad Sci U S A 1986;83:5340 –5344. 28. Ennis IL, Li RA, Murphy AM, et al. Dual gene therapy with SERCA1 and Kir2.1 abbreviates excitation without suppressing contractility. J Clin Invest 2002;109:393– 400. 29. Alseikhan BA, DeMaria CD, Colecraft HM, et al. Engineered calmodulins reveal the unexpected eminence of Ca2⫹ channel inactivation in controlling heart excitation. Proc Natl Acad Sci U S A 2002;99:17185–17190. 30. Tse HF, Xue T, Lau CP, et al. Bioartifical sinus node constructed via in vivo gene transfer of an engineered pacemaker HCN channel reduces the dependence on electronic pacemaker in a sick-sinus syndrom model. Circulation 2006;114: 1000 –1011.