Sex, Age and Regional Differences in L-Type Calcium Current are important determinants of arrhythmia phenotype in Rabbit Hearts with drug induced Long QT type 2

Abstract

In congenital and acquired long QT type 2 (LQT2), women are more vulnerable than men to Torsade de Pointes (TdP) but in pre-pubertal rabbits (and children), the arrhythmia phenotype is reversed; yet females still have longer action potential durations (APDs) than males. Thus, sex-differences in K⁺ channels and APDs alone cannot account for sex-dependent arrhythmia phenotypes. The L-type calcium current (I_Ca,L) is another determinant of APD, Ca²⁺-overload, early-afterdepolarizations (EADs) and TdP. Therefore, sex, age and regional-differences in I_Ca,L density and in EAD susceptibility were analyzed in epicardial left ventricular myocytes isolated from the apex and base of pre-pubertal and adult rabbit hearts.

In pre-pubertal rabbits, peak I_Ca,L at the base was 22% higher in males than females (6.4±0.5 vs.5.0±0.2 pA/pF, p<0.03) and higher than at apex (6.4±0.5 vs. 5.0±0.3 pA/pF, p<0.02) myocytes. Sex differences were reversed in adults; I_Ca,L at the base was 32% higher in females than males (9.5±0.7 vs.6.4±0.6 pA/pF, p<0.002) and 28% higher than the apex (9.5±0.7 vs. 6.9±0.5 pA/pF, p<0.01). Apex-base differences in I_Ca,L were not significant in adult male and pre-pubertal female hearts. Western blot analysis showed that Cav1.2α levels varied with sex, maturity and apex-base with differences similar to variations in I_Ca,L and optical mapping revealed that the earliest EADs fired at the base. Single myocyte experiments and Luo-Rudy simulations concur that I_Ca,L elevation promotes EADs and is an important determinant of LQT2 arrhythmia phenotype, most likely by reducing repolarization reserve and by enhancing Ca²⁺-overload and the propensity for I_Ca,L reactivation.

Introduction

Women have longer rate corrected QT (QTc) intervals and are especially prone to QT prolongation and Torsade de Pointes (TdP) after treatment with drugs that inhibit K⁺ channels.^{1, 2} A number of studies have shown an increase of TdP in women versus men following an exposure to agents known to block the K⁺ channel HERG and inhibit the rapid component of the delayed rectifying current, I_Kr^{1, 3-5} The increase in vulnerability to sudden death in women has been reported for cardiac^{1 5}, as well as non-cardiac drugs.⁶ These sex differences result most likely from the regulation of ionic channel expression by sex steroids.⁷ In the congenital form of LQT2, the underlying genetic defects of HERG reduces I_Kr (loss of function) and may be asymptomatic in some conditions but in the presence of a mild block of I_Kr tend to precipitate TdP in women more frequently than in men.⁸

In rabbit models of drug-induced LQT2, adult females had significantly lower I_Kr and perhaps I_Kir (inward rectifying K⁺ current) which contributed to their longer QT interval and greater arrhythmia vulnerability compared to their male counterpart.⁹ The current consensus is that normal female hearts express fewer functional K⁺ channels resulting in longer action potential durations (APDs) and when treated with agents that inhibit I_Kr, adult females have a greater vulnerability to early afterdepolarizations (EADs) and Torsade de Pointes (TdP). The concept of ‘repolarization reserve’ (RR) emerged to explain the greater vulnerability of women to TdP where K⁺ channel inhibition prolong APDs more markedly in females than males.

In pre-pubertal rabbit hearts with drug-induced LQT2, we showed that sex differences in arrhythmia phenotype are reversed with males being highly vulnerable to I_Kr blockade compared to females. In pre-pubertal (before the surge of sex hormones) rabbits (< 42 days old), female hearts had longer APDs than males yet the potent I_Kr blocker E4031 failed to elicit EADs and TdP despite a marked prolongation of APDs of over 1 s. Findings in pre-pubertal rabbit hearts seemed to differ from human data from children with congenital forms of LQT2.^{10, 11} Analysis of human registry data revealed that adult females with congenital LQT2 had a significantly higher risk of cardiac events (syncope, aborted cardiac arrest) and that in pre-pubertal children (<14 years old), girls had an equal likelihood of cardiac events as in boys.¹⁰ However, a closer scrutiny of the data revealed that boys had a 3-fold greater likelihood of a lethal arrhythmia.¹⁰ Thus, the lethality of LQT2 arrhythmias in boys trumps the number of cardiac events and indicates that the arrhythmia phenotype is reversed in children compared to adults. Thus, the arrhythmia phenotype found in adult and pre-pubertal rabbit hearts with drug-induced LQT2 are congruent to that found for LQT2 in humans.

Interestingly, APDs were longer in pre-pubertal female than male rabbits; yet E4031 elicited TdP within minutes in male hearts but merely prolonged APDs in female hearts.¹¹ Thus, additional factors, other than K⁺ currents and APD prolongation, must be considered to predict the arrhythmia phenotype; namely factors that the propensity to early afterdepolarizations (EADs).

The L-type Ca²⁺ channel is a major regulator of cardiac Ca²⁺ homeostasis and has been implicated in the genesis of EADs and TdP.¹² The classic hypothesis of EAD genesis suggests they arise from reactivation of I_Ca,L.^{13, 14} Evidence for this mechanism has come from experimental reactivation of I_Ca,L with Bay K4864¹³ and a theoretical model.¹⁵ Another hypothesis of EAD formation proposes that APD prolongation promotes cellular Ca²⁺ overload, triggering spontaneous Ca²⁺ release from the sarcoplasmic reticulum (SR),¹⁶ enhancing the turnover rate of the Na⁺/Ca²⁺ exchanger and its depolarizing current, I_NCX ^{12, 17, 18} which may reactivate I_Ca,L. In the classic hypothesis, the EAD voltage depolarization precedes the rise of intracellular free Ca²⁺, [Ca²⁺]_i whereas [Ca²⁺]_i precedes EADs in the alternative mechanism. Compelling support for the second hypothesis comes from simultaneous maps of APs and [Ca²⁺]_i where E4031-induced EADs generated a rise of [Ca²⁺]_i of such magnitude and kinetics that it was most likely produced by spontaneous SR Ca²⁺ release.¹⁷ Nevertheless, both mechanisms implicate I_Ca,L as a trigger of EADs.

Studies of the genomic effects of estrogen on the expression of cardiac Ca²⁺ channels and I_Ca,L have yielded contradictory results. In papillary muscles of female rabbits, ovariectomy (OVX) increased and estrogen replacement (7 days) decreased isometric force. Estrogen reduced ³H-nitrendipine binding in plasma membrane preparations compared to OVX and control groups; yet, peak L-type calcium currents (I_Ca,L) was not significantly different for the three treatment groups.¹⁹ In contrast, Pham et al., reported higher I_Ca,L density on the epicardium of adult female rabbit hearts compared to males and no sex differences on the endocardium such that female hearts but they did not examine apex-base differences in I_Ca,L.²⁰ In rat hearts, Western blots indicated that females had higher levels of ryanodine receptor, Cav1.2 (the α subunit of the L-type Ca²⁺ channel protein) and sodium-calcium exchange (NCX) proteins yet their mRNA levels were lower than males.²¹

New Zealand rabbits offer significant advantages as a model of human LQT2 and to investigate sex differences in arrhythmia phenotype: a) Rabbit cardiac APs and ionic currents (in particular K⁺ currents: I_K1, I_to, I_Kr and I_Ks) are similar to human APs, with similar responses to blockers of K⁺ currents.^{22, 23} b) Sex differences in arrhythmia phenotype are similar in rabbits and men.¹¹ c) Numerous studies have used rabbit models of drug induced LQT to investigate the factors that precipitate TdP.^24-26 d) Rabbits are ‘reflex ovulators’ with estrogen levels that remain elevated until mating ²⁷ which avoids estrogen oscillations that occur in most mammals during the estrus cycle and thereby minimizes estrogen-dependent genomic variations of ion channel expression.

Here, we investigated sex, age and regional differences in voltage-gated Ca²⁺ channels by measuring I_Ca,L density using the whole-cell voltage-clamp technique, by analyzing Cav1.2α protein levels by Western blots and mRNA levels by real time PCR, by correlating the regional elevation of I_Ca,L to the origin of the earliest EADs and to the LQT2 arrhythmia phenotype by optical mapping, and by showing that adult female and pre-pubertal male myocytes were more prone to fire EADs using experimental and simulation techniques. These findings provide new insight on the mechanisms underlying the firing of EADs and on sex and age differences in arrhythmia phenotype in LQT2.

Materials and Methods

Arrhythmia Phenotype in Langendorff Model of Drug-Induced LQT2

New Zealand white rabbits were anesthetized with pentobarbital (50 mg/kg) and injected with heparin (200 U/kg, intra venous). Hearts were excised and perfused in a Langendorff apparatus with a Tyrode’s solution containing (in mM): 130 NaCl, 24 NaHCO₃, 1.0 MgCl₂, 1.2 NaHPO₄, 4.0 KCl, 50 Dextrose, 1.25 CaCl₂ gassed with 95% O₂ and 5% CO₂ (pH 7.4). Perfusion pressure was adjusted to 60-70 mmHg by controlling the flow rate of the perfusion. Hearts were placed in a specially designed chamber to reduce movement artifacts and control the temperature in the medium bathing the heart using a feedback control device to maintain temperature at 37.0±0.2°C.²⁸ Hearts were stained with the voltage sensitive dye, di-4-ANEPPS (25 μl of 1 mg/ml dimethyl sulfoxide, DMSO) (Molecular Probes, Eugene, OR, USA) by injecting the dye through a port in the bubble trap (or a compliance chamber) located above the aortic cannula to the heart.¹⁸ Hearts were then perfused with the I_Kr blocking agent E4031 ((1-[2-(6-methyl-2-pyridyl)-ethyl]-4-(4- methylsulfonylaminobenzoyl) piperidine = 0.5 μM) to produce a drug-induced LQT2 and allowed to beat at their intrinsic rate, as previously described.¹¹ Four groups of rabbits were tested: adults (A) males (n=8) and (B) females (n=8) 3-4 month old and prepubertal (C) males (n=10) and (D) females (n=18) six weeks old weighing ∼ 1.5 kg. For each Langendorff heart, the arrhythmia phenotype was determined by treating the heart with E4031 and tracking the emergence of EADs and TdP which typically occurred within 5 min or failed to occur for over 30 min. Thus, the protocol using E4041 at 0.5 μM provided ‘yes’ or ‘no’ assay of arrhythmia phenotype.

Regional Distribution of the Earliest EADs

Upon perfusion with E4031, APs and EADs were monitored by optical mapping to identify the locations on the heart that fired the earliest EADs that progressed to TdP. In cases where EADs appeared at several sites on the epicardium, the earliest EAD was identified from the temporal delays between all sites that fired an EAD. Precautions were taken to insure that EADs signals represented electrical events and not motion artifacts. The earliest EAD had to occur synchronously with a voltage change measured by surface EKG recordings and had to propagate to adjacent regions of the heart for at least ∼3 mm or 3 pixels. In most cases, the first EADs appeared at the base exclusively and propagated out but did not reach the apex regions. In other cases, EADs appeared first at the base and propagated to the apex; in those cases, an activation map was generated to ascertain the origin of the EAD wave. Cumulative plots of the sites that fired the earliest EADs were generated separately for adult female and pre-pubertal male rabbit hearts to determine if EADs were more likely to start from basal or apical regions of the ventricles. A one-tailed ‘binomial test’ was used as a test for the statistical significance of deviation from the null hypothesis (p = 0.5, or 50% probability) that EADs were equally likely to occur at the base or the apex. Statistical significance for EADs firing from a preferential location is reached when the ‘binomial test’ rejects the null hypothesis with p ≤ 2%. The probability of firing the earliest EAD at the apex is given by:

$$\mathrm{p}=\Sigma \left({{\mathrm{C}}_{\mathrm{N}}}^{\mathrm{j}}\right)\times (1∕2{)}^{\mathrm{N}}$$

where the sum is taken from j = 0 to a, a is the number of experiments where EADs fire first at the apex, N is the total number of experiments where EADs were measured, and C_N^a is the combination of ‘a’ out of ‘N’.

Cell Isolation

Ventricular myocytes were isolated from either pre-pubertal (30-49 days old) or adult (3 month old) male and female New Zealand white rabbits by a modification of a previously described method.²⁹ Briefly, rabbits were anesthetized with pentobarbital (50 mg/kg) and injected with heparin (200 U/kg iv). The hearts were excised and perfused via the aorta with a physiological salt solution (PSS) containing (in mM): 140 NaCl, 5.4 KCl, 1.5 CaCl₂, 2.5 MgCl₂, 11 glucose, and 5.5 HEPES (pH 7.4). Hearts were then perfused with Ca²⁺-containing PSS for 5 min followed by perfusion with nominally Ca²⁺-free PSS for 10 min, after which collagenase type 2 (Worthington, at 0.60 mg/ml) was added to Ca²⁺ free PSS for 15 min of digestion at 35 °C. The ventricles were removed and placed in a high potassium buffer (KB) containing (in mM) 110 KGlutamate, 10 KH₂PO₄, 25 KCl, 2 MgSO₄, 20 taurine, 5 creatine, 0.5 EGTA, 20 glucose, and 5 HEPES (pH 7.4). Sections of epicardium approximately 1 mm in depth were surgically removed from apex and base regions of the left ventricle and cell isolation was performed separately for each region.²⁰ Myocytes from the apex were taken from 3-6 mm from the very bottom of the heart, those from the base were taken from 1-4 mm below the left atrium and no cells were studied from a 3-4 mm region in the middle of the heart. The tissues were minced and single myocytes were obtained by filtering through 100 μm nylon mesh. Cells were allowed to settle, the supernatant was aspirated, and the pellet was re-suspended in KB. Experiments were performed on the day of cell isolation and 4-8 myocytes were studied from each heart.

The methods and protocols used in the study were all in accordance with the University of Pittsburgh Institutional Animal Care and Use Committee and complied with the Guide for the Care and Use of Laboratory Animals as adopted by the National Institutes of Health.

Data Acquisition and Analysis

L-type Ca²⁺ currents were studied using the conventional whole cell configuration of the patch clamp technique.³⁰ Patch pipettes had resistances of 1-2.5 mΩ when filled with (in mM): 130 CsCl, 20 tetraethylammonium chloride (TEA-Cl), 5 MgATP, 5 EGTA, 0.1 TRIS-GTP, and 5 HEPES (pH 7.2). Cells were bathed in K⁺-free solution containing (in mM): 140 NaCl, 5.4 CsCl, 2.5 CaCl₂, 0.5 MgCl₂, 11 glucose, and 5.5 HEPES (pH 7.4). Currents recorded using an Axopatch 200 amplifier were filtered at 5 kHz and sampled at 10 kHz using a Digidata 1200a interface and pCLAMP (v 9.2) software (Axon Instruments). The magnitude of the peak inward current I_Ca,L was measured during 100 ms voltage-clamp steps to 0 mV applied following a 50 ms pre-pulse to -30 mV from a holding potential of -80 mV every 6 seconds. All recordings were made 3-5 minutes after gaining whole cell access and after I_Ca,L had stabilized.³¹ Series resistance was partially compensated to achieve values of ≤ 3.0 mΩ to prevent large voltage errors when measuring larger (1.5 nA) whole cell I_Ca,L. Capacitance measurements were obtained from membrane test parameters using Axon software. Capacitance of male and female pre-pubertal myocytes was 67.2±2.1 and 69.9±3.7 pF, respectively. Adult male and female myocyte capacitance was 110.7±7.3 and 109.3±8.7 pF, respectively. I_Ca,L was isolated by blocking K⁺ channels with Cs⁺ and TEA⁺, inactivating Na⁺ channels with a voltage clamp pre-pulse step to -30mV and eliminating the driving force for Cl^- currents by measuring I_Ca,L close to the predicted Cl^- equilibrium potential (0 mV).

The voltage dependence of Ca²⁺ channel activation and inactivation was determined as described previously.³² Parameters for the voltage dependence of activation were obtained from the least squares fit of data points to the equation: g/g_max=1/(1+exp-((V_T—V_0.5)/b)), where g/g_max represents normalized Ca²⁺ conductance, V_T represents test potentials from -30 to 30 mV, V_0.5 is the potential at half maximal activation, and b is the slope. Parameters for the voltage dependence of inactivation were obtained from the equation: I=I_ir+(1-I_ir)/(1+exp-((V_c-V_0.5)/b)), where I is the normalized magnitude of the peak inward current measured during a test pulse to 0 mV following a 5 s conditioning pulse (V_c) to Vm between -90 and 30 mV, I_ir is the inactivation resistant current, V_0.5 is the potential at which inactivation was half maximal, and b is the slope. The current elicited during the test pulse was normalized to the magnitude of the current recorded during a pretest pulse to 0 mV, which preceded each conditioning pulse. This corrected for changes in current magnitude due to rundown.³¹ Time to half inactivation of I_Ca,L (t_1/2) was determined by fitting the inactivating component of the I_Ca,L trace (defined as the region between the peak Ca²⁺ current and the end of the depolarizing pulse to 0 mV) to the bi-exponential curve fitting function of Clampfit (Axon Instruments). The larger of the two exponential components (97% of the inactivation curve) was used to measure t_1/2. Results are reported as the mean ± SE of at least three or more independent experiments. Statistical comparisons between two groups of experimental data were performed using the Student’s two tailed t test.

Action Potentials

APs were recorded using the current-clamp mode as previously described ³³ with an internal solution containing (in mM): 150 KCl, 5 MgATP, 5 EGTA, 0.1 TRIS-GTP, and 5 HEPES (pH 7.2). Cells were bathed in an extracellular solution containing (in mM): 140 NaCl, 5.4 KCl, 2.5 CaCl₂, 0.5 MgCl₂, 11 glucose, and 5.5 HEPES (pH 7.4), at 35 °C. APs were elicited by injecting current pulses (1-2 nA) of 5 ms duration through the patch pipette at frequencies of 0.33 or 1 Hz. Once stable AP recordings were obtained, myocytes were exposed to 5 μM E4031 to completely block I_Kr ^{33, 34} and recordings were continued to monitor AP prolongation and the incidence of EADs. “Fisher’s exact test” was used to test the null hypothesis of equal probability (p) of EADs between male and female myocytes (see http://en.wikipedia.org/wiki/Fisher%27s_exact_test). A probability of less than 2% (p< 2%) is considered to be a rejection of the null hypothesis and indicates a statistically significant difference in the likelihood of firing an EAD between male and female myocytes.

Quantitative Assays of Protein and mRNA

Tissues from the apex or base of rabbit hearts were dissected from exactly the same regions as described for the isolation of ventricular myocytes for voltage-clamp studies. The tissues were disrupted using a PowerGen model 125 homogenizer (setting 5, 30 sec) in 1 ml Enhanced Lysis Buffer (ELB; 250 mM NaCl, 0.1% NP-40, 50 mM HEPES [pH 7.0], 5 mM EDTA, 0.5 mM DTT) supplemented with a cocktail of protease and phosphatase inhibitors. The extract was rocked for 15 min at 4°C, cellular debris was removed by centrifugation (12 000 × g, 5 min) and supernatant containing the protein extract was stored at -80°C. Protein concentrations were determined by the Bradford method (Bio-Rad protein assay, Bio-Rad). Cell lysates (50 μg protein/lane) were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride (PVDF) membranes (Immobilon-P, Bedford, MA, USA), and incubated overnight at 4°C with a rabbit antibody (Affinity Bioreagents, Golden, CO, USA) that was directed against Cav (1.2) (diluted 1:1000) or mouse β-actin (Sigma, St. Louis, MO, USA, diluted 1:20,000), followed by horseradish peroxidaseconjugated second antibody (Sigma). The antigen-antibody complex was visualized with Millipore Immobilon^™ Western Chemiluminescent HRP substrate. Digitized fluorograms were quantified by using NIH Image 1.6 Software.

mRNA Extraction and real time PCR

Left ventricle epicardial tissue from the apex or base of rabbit hearts was disrupted as described for protein extracts but in guanadinine isothiocyanate buffer and mRNA was purified by centrifugation through cesium chloride (Chomcznski 1987). The resulting RNA (200 ng) was subjected to reverse transcription (RT) in 100 μl of Geneamp PCR buffer (Applied Biosystems) containing 1mM dNTPs, 2.25 mM random hexamers, 7.5mM MgCl₂ and Superscript II (Invitrogen). The reaction was carried out in a thermocycler at 25°C for 5 min, 48°C for 30 min, and 95°C for 5 min. The real time PCR was performed using 2 μl of the RT reaction products in a total volume of 50 μl containing 25 μl ABsolute SYBR Green mix (Thermo Scientific) and 20 nM primers. The reaction was carried out in an Applied Biosystems 7900HT thermocycler. Cav (1.2) expression was normalized to that of GAPDH. The primers used included Cav (1.2) forward — 5′-CATTGGGAACATCGTGATTGTC -3′, Cav( 1.2) reverse 5′ CAGGCGAACATGAACTGCAG —3′, GAPDH forward 5′ — TCCTGGGCTACACCGAGG —3′ and GAPDH reverse 5′ — TGGCACTGTTGAAGTCGCAG —3′. Relative quantitation was carried out using the 2^{- ΔΔCt} method

Action Potential Modeling of Cardiac Myocytes

Left ventricular epicardial action potentials were simulated using the Luo-Rudy model of the mammalian ventricular AP.^{35, 36} The model incorporates the transient outward K⁺ current, I_to,³⁷ and ionic currents are mathematically represented by the Hodgkin-Huxley formulation. The computational model includes simulations for ion transporters and pumps that regulate Na⁺, K⁺, Ca²⁺ concentrations across the sarcolemmal and sarcoplasmic reticulum (SR) membranes. The model tracks the dynamic changes in intracellular Ca²⁺ by incorporating Ca²⁺ release and uptake from the SR network, a delay for Ca²⁺ diffusion from the longitudinal to junctional SR, Na⁺/Ca²⁺ exchange, and Ca²⁺ buffering by calmodulin, troponin (in the cytoplasm) and calsequestrin (in the SR). Experimentally determined gender, age and regional differences in current densities and voltage-dependent parameters for I_CaL (Table I and II) were incorporated into the AP model by modifying the equations representing I_CaL. A Table in Supplementary Materials is provided to define the simulations parameters used to model APs from the base of pre-pubertal and adult male and female myocytes. To mimic LQT2, the rapid component of the delayed rectifier K⁺ current, I_Kr was suppressed by either 50% or 100%.¹⁵ The last AP from a train of 50 APs with a simulated cycle length of 1 s displayed was used to evaluate the effects of altered I_Ca,L on APs with suppressed I_Kr. Simulations were repeated to examine the influence of the Na/Ca exchanger (NCX) and its current, I_NCX on APD and EAD generation during 50 or 100 % I_Kr block with and without an increase of I_Ca,L The LRd model has been extensively used in several studies during the past decade to understand the mechanisms of arrhythmias arising due to ion channel mutations and/or drug block.^38-40

Table I.

Parameters of voltage-dependent I_Ca,L activation and inactivation in pre-pubertal myocytes

	Activation				Inactivation
Pre- pubertal	Vr (mV)	V0.5 mV)	Slope	n	V0.5 inact (mV)	Slope	% Ir	n	t1/2 (ms)	n
Male
Apex	54±2.0	-4.7±0.8	7.9±0.8 †	30	-30±0.5 *	-5.3±0.2	5.7±1.0	6	24±0.6 ‡	30
Base	55±2.1	-4.6±2.2	7.1±0.7 †	25	-29±0.6	-5.3±0.2	4.1±0.9	7	21±0.5 ‡	25
F e m a l e
Apex	55±1.8	-5.3±2.3	8.0±0.7**	18	-33±0.6 *	-5.9±0.1	4.6±0.5	7	22±0.6 §	18
Base	55±1.7	-5.4±2.7	7.4±0.4**	17	-32±0.6	-5.8±0.2	5.0±0.4	7	22±0.7	17

Cumulative data were analyzed as in Methods and expressed as mean ± S.E. V_r, reversal potential

^*the voltage of half maximal inactivation (V_0.5) was significantly shifted to hyperpolarized potentials in female compared to male myocytes (p<0.03)

^†the slope factor in male and female apex myocytes was significantly different from base myocytes (p<0.001); I_r, inactivation resistant current

^‡the time to half inactivation of I_Ca,L (t_1/2) was significantly different between male apex and base (p<0.0001)

^§male and female apex (p<0.05).

Table II.

Parameters of voltage dependent I_Ca,L activation and inactivation in adult myocytes

	Activation				Inactivation
Adult	Vr (mV)	V0.5 (mV)	Slope	n	V0.5 inact (mV)	Slope	% Ir	n	t1/2 (ms)	n
Male
Apex	55±1.2	-5.6±1.4	6.3±0.09	7	-29±1.20	-5.0±0.44	4.5±0.58	6	17.2±0.50 *	7
Base	55±1.3	-4.6±1.1	6.3±0.09	10	-28±1.84	-5.6±0.16	3.3±0.58	6	17.3±0.64 *	10
Female
		-4.6± 0.78	6.7±0.18
Apex	56±1.3			9	-28±1.36	-5.3±0.30	5.7±0.75	5	18.8±1.1 *	8
		-6.6±0.87	6.2±0.14
Base	54±0.75			10	-26±1.95	-5.1±0.36	4.6±0.33	6	17.0±0.9 *	11

Cumulative data were analyzed as in Methods and expressed as mean ± S.E. V_r, reversal potential; V_0.5, the voltage which gave half maximal activation or inactivation; I_r, inactivation resistant current

^*time to half inactivation of I_Ca,L in adult myocytes was significantly different from corresponding sex and regions of pre-pubertal myocytes.

Results

Sex and Age Differences in Arrhythmia Phenotype

Adult rabbit hearts exhibited the expected sex differences in arrhythmia phenotype, as that reported for clinical drug-induced LQT2. As shown in Figure 1, perfusion of adult male hearts with E4031 produced a marked prolongation of APDs (> 1 s) yet failed to develop TdP (1 out of 8 hearts had an arrhythmia) (panel a). In contrast, female hearts treated with the same concentration of E4031 consistently developed TdP in 7 out of 8 hearts (panel b). The arrhythmia phenotype was opposite in hearts isolated from pre-pubertal rabbits where perfusion with E4031 elicited TdP in male hearts (7 out of 10 hearts, as in panel c) but failed to elicit TdP in pre-pubertal female hearts (2 out of 18 developed TdP, panel d). Note the drug E4031 was effective in all cases and caused a marked APD prolongation in adult males and pre-pubertal female hearts; yet no TdP. The highly reproducible sex differences of arrhythmia phenotype in adults and the reverse in pre-puberty did not correlate with sex differences in APDs where pre-pubertal female hearts had longer APDs than their male counterpart.¹¹

Figure 1. Sex Differences in Arrhythmia Phenotype in Rabbit Hearts.

Each panel illustrates a control action potential measured optically using the voltage-sensitive dye, di4-ANEPPS before treatment with E4031. The arrhythmia phenotype of male (left panels) and female (right panels) upon perfusion with E4031 (0.5 μM) is illustrated for adults (panels a & b) and pre-pubertal rabbit hearts (panels c & d). E4031 elicited Torsade de Pointes (TdP) in 5-10 min in adult female hearts (b) and elicited markedly prolonged APDs in adult male hearts without progressing to TdP, VT or VF (a). In contrast, E4031 elicited TdP in pre-pubertal male (c) but not female hearts (d). Electrical activity was recorded optically for a minimum of 30 min after perfusion with E4031 to verify that adult male and pre-pubertal female hearts did not develop EADs and TdP.

Sex and Regional Comparisons of Epicardial I_Ca,L in Pre-pubertal Rabbit hearts

In pre-pubertal hearts, peak whole-cell Ca²⁺ currents (normalized to cell capacitance) were significantly higher in male compared to female epicardial myocytes isolated from the base of the left ventricles (Figure 2). I_Ca,L measured at 0 mV from the base of the heart was higher in absolute magnitude in male (-6.4±0.5 pA/pF, n=26 cells, H=7 hearts) compared to female myocytes (-5.0±0.2 pA/pF n=17, H=4, p<0.03). Representative individual current traces were superimposed (Fig. 2A) to demonstrate the differences in I_Ca,L between the sexes. Current to voltage (I/V) relationships were plotted for test potentials between -30 and +60 mV (Fig. 2B). I/V plots were bell-shaped for both sexes, reached a single maximum value at 10 mV and had identical reversal potentials (V_r) (see Table 1).

Figure 2. Sex differences in epicardial I_Ca,L in pre-pubertal rabbits.

A) Representative I_Ca,L traces from male and female myocytes from the base of the left ventricle; the myocytes were chosen for their similar sizes or membrane capacitance (± 1 pA/pF), the inward current was evoked by 100 ms depolarizing pulses to 0 mV.

B) Current to voltage (I/V) relationships in male and female myocytes from the base of the left ventricle.

C) As for (A) representative traces from male and female myocytes from the apex.

D) I/V relationships in male and female apex myocytes.

E) Cumulative peak I_Ca,L from male and female, base and apex myocytes. At the apex, differences in peak I_Ca,L were not significantly different (n = 18 to 30, p < 0.70). At the base, cumulative differences in peak I_Ca,L between male and female myocytes *, were statistically significant (n = 17 to 26, p < 0.04).

At the apex, no significant differences were found in peak I_Ca,L between pre-pubertal male (-5.0±0.3 pA/pF, n=30, H=9, p<0.02) and female (-4.8±0.3 pA/pF, n=18, H=5) myocytes. The superposition of current traces from male and female myocytes (Fig. 2C) demonstrated that at the apex, I_Ca,L was similar for both sexes. Averaged I/V relationships (Fig. 2D) and cumulative data for I_Ca,L measured at 0 mV (Fig. 2E) demonstrated that the normalized current magnitudes were comparable in both sexes for myocytes isolated from the apex of the hearts.

The apex-base distribution of I_Ca,L in pre-pubertal male and female ventricles were analyzed because previous reports implicated regional heterogeneities in current distribution as contributors to dispersion of repolarization and arrhythmia vulnerability.^{11, 20, 33} Individual current traces and I/V plots (Fig. 2, A-E) showed that male epicardial cells from the base had significantly higher peak I_Ca,L (-6.4±0.5 pA/pF) than those from the apex (-5.0±0.3 pA/pF). In female pre-pubertal rabbit hearts, apex-base differences in I_Ca,L (-4.8±0.3 pA/pF vs. -5.0±0.2 pA/pF, respectively) were not statistically significant (Fig. 2E).

Voltage Dependence of I_Ca,L in Pre-pubertal Rabbit Hearts

The voltage dependence of I_Ca,L activation and inactivation was measured to determine whether sex and regional differences exist in these channel properties. I_Ca,L activation curves for male and female apex and base epicardial myocytes are presented in Fig 3A. Although gender differences in I_Ca,L activation were not observed, there were significant regional differences in the slope factor for I_Ca,L activation in both sexes (Table 1). In males, the slope factor for current activation at apex and base was 7.9±0.84 and 7.1±0.67, respectively (p<0.001). The slope factor for current activation measured in female apex and base myocytes was 8.0±0.67 and 7.4±0.42, respectively (p<0.01). The higher slope factor implied that during an AP the time-course of Ca²⁺ entry via voltage gated Ca²⁺ channels might be faster at the apex than the base of the heart.

Figure 3. Voltage-dependence of I_Ca,L activation and inactivation in pre-pubertal rabbit myocytes.

A) I_Ca,L activation curves from male and female apex and base myocytes. B) Steady state I_Ca,L inactivation curves from male and female, apex and base myocytes.

As shown in Figure 3B, I_Ca,L inactivation occurred at more negative potentials in female compared to male myocytes. In female apex and base epicardial myocytes, the voltage at half maximal inactivation (V_0.5) was -33±0.6 and -32±0.6 mV, respectively. The V_0.5 of current inactivation in male apex and base cells occurred at significantly more positive potentials of -30±0.5 and -29±0.6 mV (p<0.03), respectively.

Analysis of I_Ca,L inactivation kinetics revealed that the time to half maximal inactivation, t_1/2 was significantly longer in male (24.0±0.56 ms, n=30, H=9) compared to female (22.2±0.58 ms, n=18, H=5, p<0.05) myocytes at the apex. No gender differences were observed at the base. In addition, there were marked apex-base differences in t_1/2 in male hearts; t_1/2 was significantly longer at the apex than the base (24.0±0.56 vs.20.6±0.45 ms, n=26, H=7, p<0.0001); whereas, regional differences in t_1/2 were not observed in female hearts.

Sex and Regional Comparisons of I_Ca,L in Adult Rabbits

In contrast to the findings in pre-pubertal hearts, peak I_Ca,L was significantly higher at the base of adult female compared to male myocytes. Representative current traces from myocytes with nearly identical membrane capacitance (Fig. 4A), averaged I/V relationships (Fig. 4B) and cumulative data (Fig. 4E) demonstrated the differences in I_Ca,L between the sexes. I_Ca,L at the base in females was 9.5±0.7 pA/pF (n=11, H=5) compared to 6.4±0.6 pA/pF, (n=11, H=5, p<0.002) for males. No differences in I_Ca,L were found in adult apex myocytes. I_Ca,L in female apex myocytes was 6.9±0.5 pA/pF (n=9, H=3) and in male myocytes from the same region was 7.3±0.4 pA/pF (n=7, H=4, p<0.5). The magnitude of representative currents (Fig. 4C), I/V relationships (Fig. 4D) and cumulative data for I_Ca,L (Fig. 4E), all demonstrated that I_Ca,L was similar at the apex for both sexes.

Figure 4. Sex Differences in I_Ca,L are reversed in Adult Rabbits.

A) Representative I_Ca,L traces from adult female and male base myocytes with similar sizes or capacitance (± 3 pA/pF).

B) I/V relationships from adult female and male base myocytes.

C) Representative traces from female and male base myocytes of similar capacitance (± 3 pA/pF).

D) I/V relationships in female and male apex myocytes.

E) Cumulative differences in peak I_Ca,L between female and male base myocytes. *, the magnitude of peak I_Ca,L between sexes was significantly different at the base (n = 9 to 11, p < 0.01) but not at the apex; ns, not significantly different (n = 7 to 9, p < 0.50).

Regional comparisons revealed that peak I_Ca,L was 28% higher in female base (9.5±0.7 pA/pF, n=11, H=5) compared to apex (6.9±0.5 pA/pF, n=9, H=3, p<0.01) myocytes. In contrast, there were no significant trans-epicardial differences in I_Ca,L in adult male ventricular myocytes, with values of 6.4±0.5 (n=11 H=5) and 7.3±0.4 pA/pF (n=7, H=4, p<0.3) respectively for base and apex myocytes.

Voltage Dependence of I_Ca,L in Adult Rabbit Hearts

Evaluation of channel properties in adult ventricles revealed no significant sex or regional differences in I_Ca,L activation or inactivation (Fig 5A-B, Table 2). However, significant differences in the rate of I_Ca,L inactivation were observed between adult and pre-pubertal hearts. The t_1/2 for I_Ca,L at the base of adult hearts was 17.3±0.64 and 17.0±0.87 ms for male and females, respectively. The t_1/2 values in male and female apex myocytes of adult hearts were 17.2±0.50 and 18.8±1.1 ms, respectively. These figures were statistically significant (p<0.01) for t_1/2 values of pre-pubertal hearts (Tables 1 and 2).

Figure 5. Voltage-dependence of I_Ca,L activation and inactivation in adult rabbit myocytes.

A) I_Ca,L activation curves from adult female and male apex and base myocytes. B) Steady state I_Ca,L inactivation curves from female and male apex and base myocytes.

Sex, Age and Regional distribution of Cav1.2α

A reasonable explanation for sex differences in I_Ca,L is that sex steroids modulate the expression levels of the Ca²⁺ channel protein, Cav1.2α. Quantitative Western blot analysis of total protein supported the notion that differences in the current density of I_Ca,L was due to differences in protein levels. Hearts from adult and pre-pubertal rabbits of each sex (n=4 per group) were rapidly isolated flash frozen in liquid nitrogen and segments of tissue were dissected for protein and mRNA analysis. The tissues were dissected from exactly the same regions of epicardium as described for the isolation of myocytes for I_Ca,L. In adult rabbits, Cav1.2α was expressed at a statistically significant higher level at the base of female hearts compared to at the apex and was higher than at the base and apex of male hearts (Figure 6, a and c). In pre-pubertal rabbits, Cav1.2α was higher at the base of male hearts compared to the apex and was higher than at the base and apex of female hearts but the differences were not statistically significant (Figure 6, b and d). Cav1.2α mRNA levels exposed statistically significant higher levels at the base than apex of adult female hearts but were otherwise not statistically significant in either adult or pre-pubertal comparisons of mRNA levels (Figure 6, e and f).

Figure 6. Protein and Message Levels for Cav1.2α from the Base and Apex of Adult and Pre-pubertal Female and Male Left Ventricular Tissues.

Ventricular tissues from adult and pre-pubertal male and female hearts and female rabbit hearts were isolated, flash frozen and processed to extract ‘total protein’ or mRNA from the base and apex of each heart, as described in Methods. Protein samples from the base and apex of each heart were loaded on 16 lanes and run on the same gel to compare their concentrations of Cav1.2 α compared to β-actin.

A) Data from adult rabbits shows a Western blot for Cav1.2α obtained from the apex and base of 4 adult male (Hearts: 1-4) and 4 adult female hearts (Hearts: 5-8) (panel a). Cumulative density histograms for Cav1.2α (normalized with respect to β-actin) (panel c) and relative mRNA levels from real time PCR measurements (panel e) are summarized for 4 adult male and 4 adult female rabbit hearts. Protein and message levels were statistically higher at the base compared to the apex of adult female hearts (p< 0.01).

B) As for (A) but from pre-pubertal rabbits. Western blot for Cav1.2α obtained from the apex and base of 4 pre-pubertal female (Hearts: 1-4) and 4 pre-pubertal male hearts (Hearts: 5-8) (panel b). Cumulative density histograms for Cav1.2α (normalized with respect to β-actin) (panel d) and relative mRNA levels (panel f) are summarized for 4 pre-pubertal male and 4 pre-pubertal female rabbit hearts. Protein levels were statistically higher at the base compared to the apex of pre-pubertal male hearts (p< 0.01). There were no statistical differences in message levels. For panels C-F, the ordinates represent are in an arbitrary scale derived from densitometry measurements. Labels are F: female, M: male; A: Apex and B: Base.

Spatial Distribution of the Earliest EADs

The correlation between the arrhythmia phenotype and the enhanced I_Ca,L at the base of the hearts suggests that the earliest EADs that capture and progress to TdP should also occur at the base of adult female and pre-pubertal male hearts. More precisely, a higher Ca²⁺ current density was measured from myocytes isolated from the top 1/3^rd of the base just below the left atrium compared to the bottom 1/3^rd of the apex.

To measure the spatial distribution of the earliest EADs on the epicardium, E4031 was added to the perfusate and maps of optical APs were recorded to detect the earliest EADs that appeared on the epicardium. As shown in figure 7, the location of the earliest EADs clustered around the base of the heart in both adult female (panel A; n=9/9 hearts) and pre-pubertal male hearts (panel B; n=8/9 hearts). For these measurements, only hearts that developed E4031-induced TdP were considered in the analysis and only one pre-pubertal male heart had an early EAD that occurred below the midline (red horizontal line, Fig. 7 A and B). A one-tailed binomial test was used to test the hypothesis that EADs occur with equal probability at the base and the apex. For adult female and pre-pubertal males, the p-values were 0.001953 and 0.017578, respectively. Thus, the clustering of EADs around the base of the heart was statistically significant with p < 0.002.

Figure 7. Epicardial Distribution of the Earliest EADs in adult female and pre-pubertal male hearts with LQT2.

A: Membrane potential (V_m) was optically mapped from 256 sites from the anterior surface of adult female hearts (N=9). Upon perfusion with E4031 (0.5 μM), APs became prolonged and began to fire EADs. The earliest site to fire an EAD was labeled with a red cross (X) to identify the regions of myocardium that are most susceptible to fire an EAD. In 9 hearts, EADs clustered at the base and no EADs appeared first at the apex of heart. The superposition of 2 APs recorded from a pixel at the base and apex illustrate an example of EADs that appeared first at the base and did not propagate to the apex (Heart 1, Middle Panel) and EADs that captured at the base and propagated to the apex (Heart 2, Right Panel). The propensity of EADs to fire first at the base was statistically significant, p< 0.02 based on a one-tailed binomial test.

B: As for (A) except that pre-pubertal male hearts were mapped to detect the regions of the heart that fired the earliest EADs. The earliest EADs occurred preferentially at the base (left panel, n=8/9) and the superposition of APs recorded from the base and apex illustrate an example of an EAD from the base that failed to propagate all the way to the apex (middle panel) and an example of an EAD that propagated from the base to elicit a smaller, delayed depolarization at the apex (right panel)

Sex differences on the Incidence of EADs in Isolated Myocytes

APs were recorded from ventricular myocytes isolated from the base of pre-pubertal and adult, male and female hearts then treatment with E4031 revealed a sex-dependent propensity to fire EADs. As shown in Figure 8, pre-pubertal male (A: n=4 cells, H = 3 hearts) and adult female (D: n=6, H=4) myocytes fired EADs whereas pre-pubertal female (B: n=4, H=3) and adult male (C: n=6, H=4) myocytes failed to fire EADs when treated with E4031. In all myocytes, E4031 produced a marked APD prolongation but EADs could only be observed in pre-pubertal male and adult female myocytes. To observe EADs, the external Ca²⁺ concentration had to be raised to 2.5 mM and the myocytes had to be paced for 10-20 beats at 1 Hz for pre-pubertal myocytes and at 0.33 Hz for adult myocytes. Fisher’s exact test was applied to test the null hypothesis null hypothesis that both males and females have equal likelihood of having EADs. The null hypothesis is rejected because its probability, p = 1.43% and 0.76% for pre-pubertal and adult myocytes respectively. With p< 2%, statistical significance is achieved for the greater incidence of EADs in pre-pubertal males compared to female and for adult female compared to adult male myocytes.

Figure 8. EAD Susceptibility in isolated Ventricular Myocytes from the base of the heart.

Myocytes were isolated from the base of rabbit hearts as described in Methods and tested for their susceptibility to fire EADs once treated with E4031. In myocytes from pre-pubertal hearts, EADs occurred spontaneously in male (A) (n=4/4; H=3 hearts) but not in female (B) (n=0/4; H=3) ventricular cells. In myocytes isolated from adult hearts, there was a reversal of sex-differences; EADs occurred in female (D) (n=5/6; H=4) but no in male (n=0/6; H=4) myocytes. Note that treatment with E4031 elicited a marked depolarization or EAD (→) in adult female and pre-pubertal male myocytes compared to their adult male and pre-pubertal female counterpart. “Fisher’s exact test” rejects the null hypothesis of equal probability of EADs between male and female myocytes with p< 2%, such that statistical significance is reached to predict that EADs are more likely pre-pubertal male than female myocytes and more likely with adult female compared to male myoctes.

AP Simulations: Influence of elevated I_Ca,L on EAD induction

Simulations of the cardiac AP based on a modified version of the Luo-Rudy model were used to evaluate the role of enhanced I_Ca,L density as a predictor of the propensity to EADs in drug-induced LQT2. AP simulations for pre-pubertal and adult myocytes from the base of the heart were generated by incorporating the experimentally determined current densities and voltage-dependent parameters for I_CaL (Figure 2-5, Table I and II) by modifying the equations representing I_CaL. To mimic LQT2, the rapid component of the delayed rectifier K⁺ current, I_Kr was suppressed by either 50% or 100%. Figure 9 a and b illustrate simulations of control APs in pre-pubertal male and female myocytes, respectively; Although experimental differences in I_Ca,L properties (see Table I) were incorporated in the simulations of control APs, there were no discernable differences in the shape and time course of pre-pubertal APs. However, the subsequent 50% block of I_Kr resulted in the firing of EADs in the male (higher I_Ca,L) but not the female model of a myocyte. In adults, female and male myocytes were modeled according experimental differences in their I_Ca,L properties (Table II) and again there were no discernible differences in control APs (Figure 9 c and d). When a 50% inhibition of I_Kr was inserted, female myocytes fired EADs while male myocytes did not (Fig. 9 c and d). The theoretical analysis confirmed the experimentally recorded APs (Figure 8) using E4031 to suppress I_Kr and mimic the propensity to fire EADs which in turn are consistent with the arrhythmia phenotype recorded in intact perfused hearts. The simulations support the hypothesis that a 25-30% increase of I_Ca,L was alone sufficient to promote EADs in myocytes with reduced I_Kr.

Figure 9. LRd Simulations of APs from the base of adult and pre-pubertal hearts.

Left panels show APs from male base myocytes and right panels show APs from female base myocytes. Top row (panels a, b) illustrates Luo-Rudy simulations of APs from pre-pubertal myocytes; bottom row (panels c, d) are AP simulations of adult myocytes. The simulated AP shown in traces a-d represent the 50^th AP from a train of APs stimulated at a cycle length of 1 s, either in the presence or absence of 100% or 50% I_Kr block (to mimic LQT2) in the simulated conditions. I_Kr suppression leads to EAD development in the pre-pubertal male myocytes (panel a) but not in pre-pubertal female myocyte (panel b). In contrast, I_Kr suppression leads to EAD development in the adult female myocytes (panel d) but not in the adult male myocytes (panel c). The male vs. female myocytes were simulated according to the experimentally determined activation and inactivation parameters shown in Tables 1 and 2 and experimental differences in I_Ca,L.

Influence of I_NCX

A is reasonable concern is that the up-regulation of I_Ca,L will increase Ca²⁺ influx on a beat-to-beat basis which is likely to be accompanied by an increase in Na/Ca exchange current, I_NCX to increase Ca²⁺ efflux and balance influx to efflux. Based on simulations of the AP, Figure 10 shows that in pre-pubertal male (panel A) and adult female (panel B) myocytes, an increase in I_NCX (30%) had no discernible effect on APD and during 100% or 50% I_Kr block, the higher I_NCX does not inhibit the generation of EADs. Nevertheless, a 30% increase in I_NCX decreased intracellular free Ca²⁺ in the cytosol (panel D) and the Ca²⁺ concentration in the lumen of the sarcoplasmic reticulum (panel E). Moreover, during I_Kr blockade increasing I_NCX alone did not elicit EADs (panel C), whereas increasing I_Ca,L alone was sufficient to elicit EADs (Figure 9) which highlights the importance of I_Ca,L as an important determinant of EADs susceptibility and of arrhythmia phenotype.

Figure 10. Influence of I_NCX on EAD susceptibility.

Simulated APs from the base of pre-pubertal male and adult female myocytes were simulated as in Figure 9 but with a 30% increase in I_NCX. A: APs from pre-pubertal male myocytes; B: APs from adult female base myocytes. A higher I_NCX has no discernible effect on APD and does not inhibit the induction of EADs when I_Ca,L is elevated and I_Kr is inhibited. C: In adult male and pre-pubertal female myocytes (i.e. normal I_Ca,L) and increase in I_NCX alone did not elicit EADs after imposing an I_Kr block. Simulated free [Ca²⁺] in the cytosol and the lumen of the SR. Changes in cytoplasmic free Ca²⁺ during an AP (panel D) and SR free Ca²⁺ (panel E) without and with a 30% increase in I_Ca,L density and 100% I_Kr block. Ca²⁺ in control conditions (light traces), with higher I_NCX (bold traces) and AP (dotted traces). A 30% increase in I_NCX caused a slight decrease in free Ca²⁺ in both the cytosol and the SR lumen but did not inhibit EADs. The simulated APs shown represent the 50^th AP from a train of APs stimulated at a cycle length of 1000 ms, either in the presence of 100% or 50% I_Kr block.

Discussion

Our main findings are that I_Ca,L is elevated at the base of hearts that are prone to EADs and TdP in E4031-induced LQT2 and that the first EADs that elicit TdP originate from the base of the heart. Protein and mRNA levels suggest that these sex differences in I_Ca,L are predominantly due to sex-differences in ion channel expression at the base but not the apex of the left ventricle. More precisely, we found that I_Ca,L at the base is higher in pre-pubertal male than female myocytes, that this sex difference is unique to the base of the heart and not the apex resulting in a gradient of I_Ca,L in pre-pubertal male but not in female hearts. Moreover, sex differences in I_Ca,L are reversed in adult rabbit hearts such that the adult female myocytes now have higher I_Ca,L at the base compared to adult males. Current to voltage relationships and the kinetics of I_{Ca,L for} all 4 groups of rabbits suggest that these sex differences in I_Ca,L are due to genomic changes in the density of functional L-type channels rather than alterations in channel properties. Besides changes in peak I_Ca,L, differences in the V_0.5 suggest that inactivation of I_Ca,L occurs later during the AP of adult female myocytes which would tend to contribute to a slight increase in APD and higher Ca²⁺ influx per AP. The other statistically different parameter was the slower inactivation kinetics of apical versus basal myocytes from pre-pubertal male hearts. The physiological consequences of slower inactivation remain unclear but suggest that if all else remains unchanged then the Ca²⁺-dependent inactivation of I_Ca,L is delayed at the apex, perhaps increasing Ca²⁺ influx during the AP plateau. Western blots of the Cav1.2α subunit of Ca²⁺ channel proteins revealed that protein levels are higher at the base of adult female hearts compared to the apex of female hearts and are higher than protein levels at the base and apex of adult male hearts. In adult hearts, differences in Cav1.2α were statistically significant and were consistent with functional measurements of whole-cell current densities. The mRNA coding for Cav1.2α was statistically (2.5X) higher at the base than the apex of female hearts, but there were no other differences in mRNA between male and female hearts. Thus, the whole-cell current, protein and mRNA analysis support the interpretation that there are sex differences in the expression of voltage gated L-type Ca²⁺ channels at the base but not the apex of the heart. In pre-pubertal hearts, Cav1.2α protein levels were statistically higher at the base of male hearts compared to the apex and had a tendency to be greater than at the base and apex of female hearts without reaching statistical significance. Similarly, message levels for Cav1.2α were not statistically different in pre-pubertal hearts.

The correlation of the arrhythmia phenotype with a) higher I_Ca,L , b) higher protein levels, c) the firing of EADs in single cells and in simulations and c) a statistically higher incidence of EADs that originate first at the base of the heart, together provide compelling evidence that I_Ca,L is an important determinant of the arrhythmia phenotype. A higher I_Ca,L reduces the repolarization reserve and during I_Kr inhibition can promote EADs by one of two possible mechanisms: a) spontaneous reactivation of I_Ca,L during the long AP plateau or b) the reactivation of I_Ca,L triggered by an inward I_NCX which is in turn elicited by spontaneous SR Ca²⁺ release. A 20-30% increase in I_Ca,L is sufficiently large to enhance a) Ca²⁺ influx per AP and intracellular Ca²⁺ load, b) luminal Ca²⁺ in the sarcoplasmic reticulum (SR), c) spontaneous SR Ca²⁺ release and I_NCX during I_Kr inhibition, and d) thus initiate EADs that progress to TdP. Optical mapping of membrane potential changes showed that in adult female hearts treated with E4031, EADs originated preferentially from the base. Similarly, AP recordings from adult ventricular myocytes isolated from the base of the heart showed that I_Kr blockade with E4031 elicited EADs in female but not in male myocytes. Thus, single cell properties are consistent with the arrhythmia phenotype of intact hearts. AP simulations confirmed that I_Kr inhibition prolonged APDs without eliciting EADs and that an elevation of I_Ca,L was necessary and sufficient to elicit EADs. These data do not exclude sex-differences in the expression of other Ca²⁺ channels and transporters, namely a) I_NCX, b) cardiac ryanodine receptor (RyR2) and/or c) SERCA2, sarcoplasmic reticulum Ca²⁺ pumps.

Sex Differences in I_Ca,L

Several studies have investigated sex differences in I_Ca,L in various mammalian species but the findings remain inconclusive and no general consensus has thus far been achieved. In 50-60 days old rabbits, Pham et al reported a transmural dispersion of I_Ca,L (higher on the epicardium than endocardium) at the base of female hearts that was absent in male hearts.²⁰ In contrast to rabbit hearts, a study on mongrel dogs found uniformly higher levels of I_Ca,L in female than male hearts across the left ventricular wall.⁴¹ In guinea pig hearts, the opposite result was obtained where I_Ca,L was significantly higher in males than females even when the female current density was measured at different phases of the estrus cycle.⁴² Moreover, in mice and rat hearts no significant differences were found in I_Ca,L between males and females.^{43, 44} In human midmyocardial left ventricular myocytes from patients with end-stage heart failure, I_Ca,L was found to be higher (∼10%) in female than male hearts but the difference did not reach statistical significance.⁴⁵ Nevertheless, simulations and experiments showed that at long cycle lengths myocytes from women were prone to EADs whereas myocytes from men rarely fired an EAD.⁴⁵ In the absence of similar studies in ‘healthy’ human myocytes, Verkerk et al., pointed out that the properties of myocytes from failing hearts were consistent with those obtained from healthy human hearts in terms of QTc, EAD susceptibility and sex differences in APDs and thus, proposed that the differences in I_Ca,L between female and male hearts represent a characteristic of normal human hearts.⁴⁵

Previous studies were attentive to transmural differences in I_Ca,L but neglected apex-base differences or differences in pre- versus post-puberty. The current findings of higher I_Ca,L in adult female base myocytes are in agreement with previous studies on rabbit ²⁰ and human⁴⁵ hearts but extend the data to reveal I_Ca,L differences between apex and base and between pre-and post puberty. It may be that once apex-base heterogeneities are included in the analysis, I_Ca,L differences between men and women will be statistically significant and will expose larger I_Ca,L differences between the sexes. In rat and guinea pig hearts, sex differences in I_Ca,L were not detected perhaps because the myocytes were isolated from random regions of the left ventricle, and regional heterogeneities of I_Ca,L might conceal differences of 25-30% in current density.⁴⁴ In another study, no apex-base differences in I_Ca,L were detected in mongrel dogs and human hearts but the study was not attentive to possible sex differences.⁴⁶

More intriguing is the finding that higher I_Ca,L densities in adult and pre-pubertal rabbits correlate with the propensity to EADs at the base of the heart and the vulnerability to TdP in E4031-induced LQT2.¹¹ Sex differences in arrhythmia phenotype (as defined by E4031 in perfused hearts) arise from the properties of ventricular myocytes since E4031 elicited EADs in freshly isolated adult female but not male myocytes (Figure 8 e and f). The significance of I_Ca,L differences in rabbit hearts is amplified by the remarkable similarity between rabbit (Fig. 8e,f) and human (Fig. 1D),⁴⁵ recordings of male and female APs and the firing of EADs in human female but not male ventricular myocytes.

Regional Elevation of I_Ca,L as a Predictor of Arrhythmia Phenotype

Numerous studies showed that enhanced dispersion of repolarization is pro-arrhythmic and contributes to the initiation of TdP induced by drugs that prolong the AP.^{11, 17, 20, 33, 46-49} In congenital, drug-induced LQT2 and in animal models of LQT2, APD prolongation does not automatically result in TdP and additional factors are needed to elicit EADs and TdP; sex being a major risk factor. Our findings of higher I_Ca,L at the base in pre-pubertal males and adult females correlate with sex and age-related differences in arrhythmia phenotype are consistent with the hypothesis that in the setting of prolonged APDs, the severity of Ca²⁺ overload is a critical determinant of EADs and TdP. More precisely, the propensity to TdP in LQT2 is determined by I_Ca,L, the higher the current, the greater the severity of Ca²⁺ overload and the greater propensity that EADs originate from the base and progress to TdP.

Study Limitations

A comprehensive analysis of the cellular determinants of LQT-mediated arrhythmias requires us to examine all channels and transporters involved in Ca²⁺ handling. One would reasonably expect that an increase in I_Ca,L would be matched with an increase in Ca²⁺ pumps at the cellular membrane and/or Na⁺/Ca²⁺ exchanger (NCX) to balance the influx to the efflux of Ca²⁺ during an AP. Likewise, Ca²⁺-handling proteins like the Ca²⁺ release channels (or ryanodine receptors, RyR) and Ca²⁺,Mg²⁺-ATPase (SERCA2) on the SR network may exhibit sex differences and thereby contribute to the arrhythmia phenotype by altering SR-Ca²⁺ overload, spontaneous SR Ca²⁺ release and the initiation of EADs. In rat hearts, a recent study reported higher protein levels of Ca(v)1.2, RyR and NCX in females yet paradoxically Ca(v)1.2 mRNA levels were higher in males.²¹

In adult and pre-pubertal rabbit hearts, we found sex differences in NCX protein that were similar to Cav1.2 α with higher protein levels in adult female and pre-pubertal males based on Western blot analysis of pre-pubertal and adult rabbit left ventricles (not shown). Similarly, the density of I_NCX was higher in pre-pubertal male compared to age-matched female base myocytes (not shown). Further studies that probe differences in Ca²⁺ handling proteins are needed to obtain a better understanding of all determinants of the arrhythmia phenotype.

Optical mapping of membrane potential on the epicardium of rabbit hearts revealed at statistically significant propensity of EADs to appear first at the base of the heart. The 2-dimensional nature of optical mapping does not resolve the exact origins of such EADs but does identify the breakthrough sites that appear first on the epicardium. Until a high speed 3-dimensional technique is developed, one cannot completely exclude the possibility that EADs originate from deeper layers at the base of the ventricles. However, it is highly unlikely that EADs originate from deeper layers near the apex which then propagate inside the ventricular wall to breakthrough at the base before the apex.

The density of I_Ca,L depends on the number of functional channels and on the modulation of channel activity by regulatory peptides and multiple phosphorylation sites through β-adrenergic activity. Sex dependent regulation of channel activity presents another level of complexity which has yet to be analyzed in a comprehensive fashion. The findings raise important questions regarding genomic regulation of ion channel expression by sex steroids. What mechanisms produce sex differences in ion channel expression in pre-puberty before the surge of estrogen and testosterone? What cues produce regional differences in ion channel expression?

Nevertheless, the role of I_Ca,L as a determinant of arrhythmia phenotype in drug-induced LQT2 may be fundamental to our ability to evaluate the safety of new drugs that produce small but measurable QT or APD prolongation. Female sex is well known to be a risk factor to lethal TdP but the current data supports the more precise notion that sex-differences in I_Ca,L is a critical factor is the assessment of arrhythmogenic risk. It is interesting to speculate that I_Kr inhibition may pose less of an arrhythmogenic risk if it is combined with I_Ca,L and/or I_NCX inhibition. Similarly, congenital LQT2 may be asymptomatic well into adulthood then enhanced I_Ca,L and/or I_NCX through a genomic regulation can precipitate a shift in arrhythmia phenotype. Thus, sex steroids, heart failure and cardiac hypertrophy may alter the LQT2 arrhythmia phenotype through genomic regulation of Ca²⁺ channels.