K_ATP channel gain-of-function leads to increased myocardial L-type Ca²⁺ current and contractility in Cantu syndrome

Significance

ATP-sensitive potassium (K_ATP) channels are present in cardiac and smooth muscle; when activated, they relax blood vessels and decrease cardiac action potential duration, reducing cardiac contractility. Cantu syndrome (CS) is caused by mutations in K_ATP genes that result in overactive channels. Contrary to prediction, we show that the myocardium in both CS patients and in animal models with overactive K_ATP channels is hypercontractile. We also show that this results from a compensatory increase in calcium channel activity, paralleled by specific alterations in phosphorylation of the calcium channel itself. These findings have implications for the way the heart compensates for decreased excitability and volume load in general and for the basis of, and potential therapies for, CS specifically.

Abstract

Cantu syndrome (CS) is caused by gain-of-function (GOF) mutations in genes encoding pore-forming (Kir6.1, KCNJ8) and accessory (SUR2, ABCC9) K_ATP channel subunits. We show that patients with CS, as well as mice with constitutive (cGOF) or tamoxifen-induced (icGOF) cardiac-specific Kir6.1 GOF subunit expression, have enlarged hearts, with increased ejection fraction and increased contractility. Whole-cell voltage-clamp recordings from cGOF or icGOF ventricular myocytes (VM) show increased basal L-type Ca²⁺ current (LTCC), comparable to that seen in WT VM treated with isoproterenol. Mice with vascular-specific expression (vGOF) show left ventricular dilation as well as less-markedly increased LTCC. Increased LTCC in K_ATP GOF models is paralleled by changes in phosphorylation of the pore-forming α₁ subunit of the cardiac voltage-gated calcium channel Cav1.2 at Ser1928, suggesting enhanced protein kinase activity as a potential link between increased K_ATP current and CS cardiac pathophysiology.

Cantu syndrome (CS), characterized by hypertrichosis, osteochondrodysplasia, and multiple cardiovascular abnormalities (1), is caused by gain-of-function (GOF) mutations in the genes encoding the pore-forming (Kir6.1, KCNJ8) and regulatory (SUR2, ABCC9) subunits of the predominantly cardiovascular isoforms of the K_ATP channel (2–5). Because the same disease features arise from mutations in either of these subunits, it is concluded that CS arises from increased K_ATP channel activity, as opposed to any nonelectrophysiologic function of either subunit.

However, this conclusion does not provide immediate explanation for many CS features. In the myocardium, for example, acute activation of K_ATP channels results in shortening of the action potential (AP), with concomitant reduction of both calcium entry and contractility (6). The naïve prediction in CS would therefore be that K_ATP GOF mutations should shorten the AP, reduce contractility, and reduce cardiac output. We previously reported high cardiac output with low systemic vascular resistance in CS (7). Cantu syndrome cardiac pathology is therefore opposite to prediction, and also unlike classical hypertrophic or dilated cardiomyopathies, in that the ventricle is dilated, but there is increased cardiac output. Here we characterize CS cardiac pathology in patients, and explore the mechanistic basis using mice that express K_ATP GOF mutant subunits in the heart and vasculature.

Results

Low Blood Pressure in CS Patients.

Eleven CS individuals (five male, six female, aged 17 mo to 47 y), all harboring ABCC9 mutations (Table S1), participated in CS research clinics at St. Louis Children’s Hospital. Five had been previously followed at this institution (7) and the remainder were enrolled via the CS Interest Group (www.cantu-syndrome.org). Patient demographic data, genotype, and available cardiac historical, physical, and test information are summarized in Table S1. Most patients had no recalled cardiac symptomatology, although prior office notes revealed episodes of chest pain, fatigue, shortness of breath, and exercise intolerance associated with pericardial effusion. Additionally, three patients (cs002, cs004, cs005) reported palpitations and exercise intolerance. One of these (cs004) also had symptoms with orthopnea, resulting from “idiopathic” high-output state and atrial fibrillation. Five patients had patent ductus arteriosus that required surgical ligation or catheter-based closure, two had significant pericardial effusions, three had been diagnosed with pulmonary hypertension, and five had histories of lower extremity edema. All patients had full but noncollapsing peripheral pulses. All but one patient (cs004, who was on several medications; Table S1) had supine systolic and diastolic blood pressure (BP) that was well below mean for age (Table S1) [mean age: 16.6 ± 13.5; systolic BP: 90.5 ± 12.8 mmHg; diastolic BP: 58.2 ± 6.2 mmHg; heart rate (HR): 85 ± 17 beats per minute (bpm)] (8). Despite these low BP values, no patient demonstrated orthostatic HR or BP changes. There were relatively few cardiac findings on physical examination, with the exception of one patient with a diastolic murmur (cs004) at the apex. Electrocardiograms revealed first-degree atrioventricular (AV) block in four patients, fascicular block in two, and T-wave abnormalities (T-wave axis 180° displaced from QRS axis and morphologic abnormalities) in seven patients, but no evidence of QT shortening or correct QT (QTc) prolongation (Table S2).

Table S1.

Patient genotype and history

	cs001	cs002	cs003	cs004	cs005	cs006	cs007	cs008	cs009	cs010	cs011
Age (y), gender	12.3 M	17.9 F	23.4 F	47.0 F	18.1 F	12.8 M	32.0 F	1.5 F	4.0 M	6.0 M	8.0 M
Height (cm)	152.1	155.3	163.7	181.2	157.8	139.8	177	81.5	109	121.2	143.3
Height (%) for age	75	10	Adult	Adult	10–25	<5	Adult	50–75	90–95	75–90	>95
Weight (kg)	45.6	45.4	57.2	100.1	68.3	30.7	87.6	10.48	21.3	23.4	39.7
Amino acid substitution	R1154W	R1154Q	R1154Q	R1154Q	R1154W	H1005L	R1116C	R1116C	R1116C	R1116C	R1116H
History
Orthopnea	—	—	—	Yes	—	—	—	—	—	—	—
Exercise intolerance	—	—	—	Yes	—	—	—	—	—	—	—
Symptoms with exertion	—	—	—	Yes	Occasionally	—	—	—	—	—	—
Palpitations	—	Yes	—	Yes	Yes	—	—	—	—	—	—
Presyncope	—	—	—	—	—	—	—	—	—	—	—
Syncope	—	—	—	—	—	—	—	—	—	—	—
Fatigue	—	—	—	—	—	—	—	—	—	—	—
Headaches		Yes	Yes	Yes	No	No	Yes	—	—	—	—
Rhythm abnormalities	No	No	No	AF	No	No	No	No	No	No	No
PDA	Yes	No	No	No	Yes	No	Yes	Yes	No	Yes	Yes
Pericardial effusion	No	Yes	No	Yes	No	No	No	No	No	No	No
Pulmonary hypertension	No	Yes	No	No	No	No	No	Yes	No	No	Yes
Lower extremity edema	No	Yes	Yes	Yes	Yes	No	Yes	No	No	No	No
Medications				Atenolol, Cardizem Lasix, spironolactone		Enalapril, Focalin		Sildenafil
Physical
HR (bpm)
Supine	78	72	74	92	66	100	75	126	90	90	72
Upright	84	88	88	92	76	104	83	120	102	112	96
HR norms for age (10–90%)	62–96	62–96	60–80	60–80	62–96	62–96	60–80	98–135	86–123	81–117	74–111
BP (mmHg)
Supine	92/62	88/52	92/70	116/60	88/50	74/56	106/64	72/50	84/56	85/60	98/60
Upright	96/66	92/58	88/62	108/70	98/62	74/50	102/60	94/64	74/56	72/54	100/62
Orthostatic changes: BP	No	No	No	No	No	No	No	No	No	No	No
Orthostatic changes: HR	No	No	No	No	No	No	No	No	No	No	Yes
BP norms for age (50–99%)	108–133/63–90	109–133/65–90	120–140/80–90	120–140/80–90	120–140/80–90	101–126/59–86	120–140/80–90	90–114/42–67	95–120/51–78	98–123/53–80	102–127/61–88
Age at which systolic BP represents 50th percentile	4-y-old boy	4-y-old girl	6-y-old girl	Normal adult	4-y-old girl	<1-y-old boy	11-y-old girl	1-y-old girl	<1-y-old boy	<1-y-old boy	8-y-old boy
CV findings	None	None	None	Chronic venous changes	None	None	Chronic venous changes	None	None	None	None

AF, atrial fibrillation; PDA, patent ductus arteriosus.

Table S2.

Human electrocardiological findings

	cs001	cs002	cs003	cs004	cs005	cs006	cs007	cs008	cs009	cs010	cs011
Age (y)	12.3	17.9	23.4	47	18.1	12.8	32	1.5	4	6	8
HR (bpm)	90	73	71	70	73	87	75	114	132	n.d.	91
Rhythm	NSR	NSR	NSR	NSR	NSR	NSR	NSR	NSR	NSR		NSR
P axis (°)	76	46	45	37	36	40	37	43	44		56
QRS axis (°)	81	110	51	−36	68	109	20	10	69		113
T axis (°)	77	25	−42	−42	21	81	30	58	9		76
PR interval (ms)	170	190	148	208	154	108	210	150	150		170
QRS interval (ms)	96	82	96	114	94	82	96	86	78		96
RR interval (ms)	690	816	808	848	826	612	810	560	440		690
QT interval (ms)	330	390	356	400	388	366	400	320	290		330
QTc interval (ms)	404	430	387	432	427	440	447	440	430		440
T-wave abnormality	Yes	Yes	Yes	Yes	Yes	No	Yes (mild flattening)	Yes (right precordial)	No		No
Conduction delay	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes		Yes
Early repolarization	No	No	Yes	No	No	No	No	No	Yes		Yes
QRS axis norm for age	9–128	9–128	−10–95	−31–90	9–128	9–128	−25–91	89–152	10–139		6–116
QRS interval norm for age	40–90	69–107	69–107	69–107	69–107	40–90	69–107	30–80	30–70

HR normative values from Fleming et al. (62). Blood pressure norms from National Heart Lung and Blood Institute (NHLBI) normative blood pressure values (8). Height percentiles based on Centers for Disease Control and Prevention published growth charts. Normative ECG interval and axes values for age were from Moss and Allen (25). Normal QTc <450 ms; borderline QTc 450–470 (1). Values listed represent the second and 98th percentile norms for age. ND, not determined; NSR, normal sinus rhythm.

Cardiovascular Structural and Functional Characteristics in CS Patients.

Two-dimensional color Doppler echocardiographic interrogation and strain imaging (Fig. 1, Fig. S1, and Table S3) revealed normal segmental anatomy, but markedly enlarged hearts (Fig. 1), with enhanced cardiac output and contractility, in CS patients. Left ventricular (LV) chambers were significantly dilated in CS patients, and cardiac output was markedly increased (Fig. 1). LV mass was increased but, interestingly, LV posterior wall thickness (LVPWD) was not different from controls (Fig. 1). This constellation of features is distinct from both typical hypertrophic cardiomyopathies (where LVPWD is increased) and typical dilated cardiomyopathies (where there is chamber dilation with diminished cardiac function). In part, this dilated hypercontractile phenotype could be a secondary response to chronically elevated blood volume as a result of vasodilation—a directly predicted consequence of vascular K_ATP GOF (9). Pulsed-wave velocity testing offers a noninvasive correlate of vascular tone. Consistent with BP measurements, there was diminished pulsed-wave velocity in CS patients (control 5.6 ± 0.9 ms⁻¹, n = 8; CS 4.5 U+ 0.8 m/s, n = 7, P = 0.003), implying diminished vascular tone.

Fig. 1.

Enhanced cardiac volume and contractility in CS patients. Comparisons of pertinent echocardiographic measures from control and CS patients.

Fig. S1.

Human echocardiography. (Upper) Two-dimensional four-chamber view images of hearts from CS patient (cs001) and control individual at both end systole and end diastole. Note multicolor tag markings along the LV chamber for computing strain (deformation) measurements. (Lower) CS patient shows markedly greater strain (−22.3) than control (−16.8). Comparisons of pertinent echocardiographic measures from control and CS patients.

Table S3.

Human echocardiographic findings

	Control (n = 17)		Cantu (n = 11)		P value
	Average	SD	Average	SD
BMI	19.52	0.44	21.49	1.57	NS
Weight (kg)	50.94	16.42	48.74	27.45	NS
Height (cm)	157.75	30.77	143.64	29.63	NS
BSA (DuBois method)	1.5	0.41	1.36	0.52	NS
Systolic BP	108.29	12.08	93.64	13.43	0.00585
Diastolic BP	67.76	5.61	58.09	8.77	0.0014
HR	82.82	3.92	90.00	5.62	NS
LVDD	4.62	0.95	5.32	1.28	NS
LVDD Z-score	1.15	0.98	2.77	1.92	0.0064
LVPW	0.6429	0.1358	0.7383	0.1386	NS
RWT	0.28	0.09	0.28	0.09	NS
LVMI-MMODE	27.55	5.57	58.33	19.84	<0.00001
LVM-2D	90.04	29.71	143.85	83.00	0.0208
LVEDV	93.00	25.35	139.09	80.57	0.0357
LVEDV Z-score	−0.29	1.49	2.53	0.97	<0.00001
LVESV	36.5	12.04	56.05	39.99	NS (0.0662)
LVMI-2D (g/m2.7)	23.6	3.2	60.79	20.06	<0.00001
Remodeling	0.0	0.0	1.82	0.6	<0.00001
FS (%)	37.0	2.6	42.18	4.95	0.0014

Circadian Abnormalities and Low Vagal Function in CS Patients.

Several patients received 24-h ambulatory ECG monitoring. None demonstrated atrial or ventricular arrhythmia during the recording period. The data were subsequently used for heart rate variability analysis (HRV). Fig. 2 displays HR and high-frequency (HF) power, a marker for vagal activity, as a function of time of day, for a representative CS patient and age/sex-matched control. The control 18-y-old female shows typical circadian variation: in general, higher HR during waking hours, and significantly lower HR during sleeping hours; HF also has a circadian rhythm that is inversely related to this HR trend, rising during sleeping hours, and low during waking hours. Interestingly, all CS patients demonstrated markedly diminished HF power (Fig. 2 and Fig. S2) and generally high HR for age that failed to lower appropriately during sleeping hours (Figs. S2 and S3). These latter findings suggest a relatively elevated sympathetic activity but diminished vagal activity.

Fig. 2.

Cardiovascular control in CS patients. CS patient (cs005, 18-y-old female) and control (18-y-old female) heart rate (color) and high-frequency power (right axis, black) plotted over 24 h.

Fig. S2.

Cantu patient circadian rhythm and high-frequency power graphs.

Fig. S3.

Cantu patient circadian rhythm and high-frequency power graphs.

Enhanced Contractility in Hearts Expressing K_ATP GOF Mutations.

The predicted effect of K_ATP GOF in the myocardium itself is AP shortening and reduced contractility, whereas the predicted effect in the vasculature is reduced peripheral resistance. The above clinical evaluation of CS patients reveals physiologically low BPs, consistent with the latter prediction. However, myocardial hypercontractility and no evidence of K current-induced QT shortening on ECG are grossly opposite to prediction. We hypothesize that these features are secondary consequences of the primary predictions. Specifically, we propose that increased K_ATP current in either ventricular or smooth muscle myocytes will lead to lower cardiac output and decreased peripheral resistance, respectively; both of these will result in diminished tissue perfusion, which in turn will induce a systemic feedback to increase cardiac output (see, for example, Fig. 7C).

Fig. 7.

Cellular and molecular basis of Cantu syndrome. (A) Quantitation of Western blot analysis of isolated total Cav1.2a subunit protein, distal C terminus, and processed C-terminal fragments, normalized to WT protein levels in cGOF, Kir6.1[AAA], and vGOF cardiac samples, as well as relative levels of phospho-S1700 and phospho-S1928 residues. (B) Kaplan–Meier survival curve for male icGOF (n = 7) and littermate control mice (n = 6) transplanted with slow-release propranolol pellets (5 mg/21-d release) on day 0, and then induced (both icGOF and control) with tamoxifen starting on days 2–7 (gray bar). (C) Proposed mechanistic basis of Cantu syndrome. GOF mutations in either the pore-forming (KCNJ8) or accessory (ABCC9) K_ATP subunits will directly cause action potential (AP) shortening and reduced heart rate in cardiomyocytes and reduced excitability in smooth muscle myocytes, which will result in diminished calcium uptake, decreased contractility, and decreased cardiac output, as well as decreased peripheral resistance. Combined, these reactions would reduce blood pressure, stimulating baroreceptors and triggering PKA- or PKC-dependent stress-signaling pathways in the heart and potentially in smooth muscle. These pathways lead to phosphorylation of LTCC, specifically the Cav1.2 subunit, resulting markedly enhanced basal activity of the LTCC, enhanced contractility of the myocyte, and restored APD. In vascular smooth muscle cell, the diminished peripheral resistance will result in diminished effective tissue perfusion, giving rise to long-standing volume load on the heart, and chamber dilation.

To examine these hypotheses in a tractable system, we first evaluated the cardiac consequences of Kir6.1 GOF mutations expressed in the myocardium under α-myosin heavy chain (αMHC) control in mice [constitutive GOF (cGOF) mice]. cGOF mice exhibit prolonged PR intervals, as well as episodes of junctional rhythm, and diminished AV nodal conduction, but no QT shortening (10). Echocardiography under light anesthesia (11) (Fig. 3A and Table S4) revealed normal chamber wall dimensions, but increased contractility and ejection fraction in cGOF hearts, again counter to naïve prediction, but consistent with the hypercontractile phenotype of CS patients.

Fig. 3.

Enhanced contractility in cGOF mice. (A) Echocardiographic analysis showing increased ejection fraction and strain rate in cGOF mice compared with controls. (B) Contractility parameters for isolated ventricular myocytes from cGOF and control sex-matched littermate hearts.

Table S4.

Mouse K_IR 6.1 GOF (cGOF) echocardiographic findings

	Control (n = 5)		cGOF (n = 6)		P value
	Mean	SD	Mean	SD
Body weight (mg)	26.7	4.2	27.8	3.7	NS
HR (bpm)	620.2	44.6	608.7	36.2	NS
LVPWd (mm)	0.842	0.059	0.820	0.068	NS
IVSd (mm)	0.846	0.065	0.813	0.056	NS
LVIDd (mm)	3.37	0.47	3.60	0.19	NS
LVPWs (mm)	1.44	0.09	1.41	0.08	NS
IVSs (mm)	1.50	0.06	1.46	0.16	NS
LVIDs (mm)	1.61	0.39	1.77	0.17	NS
LVM (mg)	96.8	28.7	101.5	17.7	NS
LVMI	3.60	0.77	3.64	0.28	NS
RWT	0.506	0.054	0.453	0.023	NS (0.057)
FS	52.6	5.0	50.9	3.7	NS
EDV	37.9	8.6	39.0	12.4	NS
ESV	10.3	2.5	9.1	3.2	NS
EF	72.9	2.9	76.9	1.6	0.017
dV/dt-s	1.0	0.2	1.1	0.3	NS
dV/dt-d	0.9	0.2	1.1	0.3	NS
LVMd	95.4	15.5	92.9	21.5	NS
LVMdI	3.6	0.3	3.3	0.5	NS
SV	27.53	6.43	29.83	9.33	NS
CO	16.45	3.26	17.76	5.52	NS
S-rad	36.753	4.193	41.504	6.865	NS
S-long	21.539	2.760	21.522	4.936	NS
SR-rad-s	9.964	0.876	10.456	1.267	NS
SR-rad-d	11.320	1.229	13.890	1.280	0.008
SR-long-s	8.524	0.751	8.566	1.783	NS
SR-long-d	12.254	2.461	11.352	3.200	NS
	Control (n = 5)		cGOF (n = 6)
	Mean	SD	Mean	SD	P value
Body weight	26.7	4.2	27.8	3.7	0.641
HR	620.2	44.6	608.7	36.2	0.647
LVPWd	0.842	0.059	0.820	0.068	0.585
IVSd	0.846	0.065	0.813	0.056	0.394
LVIDd	3.37	0.47	3.60	0.19	0.287
LVPWs	1.44	0.09	1.41	0.08	0.630
IVSs	1.50	0.06	1.46	0.16	0.582
LVIDs	1.61	0.39	1.77	0.17	0.396
LVM	96.8	28.7	101.5	17.7	0.746
LVMI	3.60	0.77	3.64	0.28	0.908
RWT	0.506	0.054	0.453	0.023	0.057
FS	52.6	5.0	50.9	3.7	0.529
EDV	37.9	8.6	39.0	12.4	0.871
ESV	10.3	2.5	9.1	3.2	0.512
EF	72.9	2.9	76.9	1.6	0.017
dV/dt-s	1.0	0.2	1.1	0.3	0.793
dV/dt-d	0.9	0.2	1.1	0.3	0.376
LVMd	95.4	15.5	92.9	21.5	0.833
LVMdI	3.6	0.3	3.3	0.5	0.338
SV	27.53	6.43	29.83	9.33	0.653
CO	16.45	3.26	17.76	5.52	0.653
S-rad	36.753	4.193	41.504	6.865	0.211
S-long	21.539	2.760	21.522	4.936	0.995
SR-rad-s	9.964	0.876	10.456	1.267	0.483
SR-rad-d	11.320	1.229	13.890	1.280	0.008
SR-long-s	8.524	0.751	8.566	1.783	0.961
SR-long-d	12.254	2.461	11.352	3.200	0.619

CO, cardiac output mL/min; +dV/dt, peak rate of LV volume increase in diastole; ); −dV/dt, peak rate of LV volume decrease in systole; EDV, end-diastolic LV volume (μL) based on long-axis images; EF, ejection fraction [(EDV − ESV)/EDV]; ESV, end-systolic LV volume (μL); IVS, interventricular septum; LV FS, left ventricular fractional shortening; LLViDs, left ventricular internal dimension in systole; LViDd, left ventricular internal dimension in diastole; VPWd, left ventricular posterior wall width in diastole; LVPWs, left ventricular posterior wall width in diastole; S-long, average peak longitudinal strain (%); SR-long-d, average peak diastolic longitudinal strain rate (1/s); SR-long-s, average peak systolic longitudinal strain rate (1/s); S-rad, average peak radial strain (%); SR-rad-d, average peak diastolic radial strain rate (1/s); SR-rad-s, average peak systolic radial strain rate (1/s); SV, stroke volume [μLLVM, LV mass] (mg). Data are presented as the mean ± SD, and tests of statistical difference were computed using two-tailed Student’s t test for all parameters.

Isolated cGOF myocytes displayed diminished resting sarcomere length, significantly increased fractional shortening and increased rates of shortening and relaxation (Fig. 3B and Table S5). These features are similar to those seen in WT myocytes following exposure to isoproterenol (ISO). The subsequent effect of ISO was reduced in cGOF myocytes (Fig. 3B), consistent with the idea that cGOF myocytes are effectively “prestimulated” (12) via a pathway that is at least convergent with that responding to β-adrenergic signaling. To confirm that this prestimulation effect was not an artefactual effect of increased Kir6.1 protein expression per se, mice expressing dominant-negative Kir6.1[AAA] subunits under αMHC control (AAA) were studied (13). No significant differences in contractile parameters were seen between AAA and littermate control myocytes (Table S6).

Table S5.

Kir6.1 cGOF contractility data

Baseline measures	Control, n = 39		cGOF, n = 45		P value
	Mean	SD	Mean	SD
Sarcomere length	1.799	0.043	1.770	0.050	0.0063
Contraction rate	−1.932	0.886	−2.825	1.468	0.0014
Contraction time	0.291	0.017	0.289	0.012	NS
End-systolic length	1.664	0.063	1.585	0.083	0.000006
Peak contraction amplitude	0.135	0.051	0.185	0.071	0.0004
Fractional shortening	7.473	2.810	10.429	3.976	0.0002
Time to peak contraction	0.398	0.031	0.414	0.026	0.0115
Relaxation rate	1.768	1.019	2.578	1.486	0.0052
Relaxation time	0.462	0.046	0.486	0.042	0.0151
Isoproterenol	Control, n = 30		Kir 6.1 GOF, n = 41
	Mean	SD	Mean	SD
Sarcomere length	1.678	0.059	1.713	0.063	0.0214
Contraction rate	−7.322	1.604	−7.926	2.232	NS
Contraction time	0.280	0.008	0.279	0.010	NS
End-systolic length	1.368	0.072	1.386	0.055	NS
Peak contraction amplitude	0.309	0.046	0.327	0.052	NS
Fractional shortening	18.462	2.814	19.051	2.724	NS
Time to peak contraction	0.388	0.018	0.387	0.019	NS
Relaxation rate	4.937	1.193	5.421	1.046	NS (0.0736)
Relaxation time	0.456	0.038	0.454	0.034	NS

All times in milliseconds, all lengths in μm, and all rates in μm/milliseconds.

Table S6.

Kir6.1 AAA contractility data

Baseline measures	Control, n = 34		Kir 6.1 AAA, n = 35		P value
	Mean	SD	Mean	SD
Sarcomere length	1.815	0.025	1.822	0.037	NS
Contraction rate	−2.001	0.909	−2.499	1.313	NS
Contraction time	−0.094	0.018	−0.098	0.016	NS
End-systolic length	1.692	0.057	1.674	0.086	NS
Peak contraction amplitude	0.124	0.047	0.148	0.076	NS
Fractional shortening	6.826	2.577	8.109	4.175	NS
Time to peak contraction	0.133	0.015	0.134	0.013	NS
Relaxation rate	1.739	1.065	2.153	1.576	NS
Relaxation time	0.057	0.013	0.056	0.016	NS
Isoproterenol	Control, n = 20		Kir 6.1 AAA, n = 27
	Mean	SD	Mean	SD
Sarcomere length	1.752	0.048	1.729	0.066	NS
Contraction rate	−8.130	1.190	−8.210	1.409	NS
Contraction time	−0.096	0.012	−0.089	0.010	0.040
End-systolic length	1.425	0.044	1.395	0.067	NS
Peak contraction amplitude	0.327	0.043	0.334	0.037	NS
Fractional shortening	18.654	2.206	19.346	2.127	NS
Time to peak contraction	0.128	0.015	0.123	0.015	NS
Relaxation rate	6.188	1.225	6.444	1.184	NS
Relaxation time	0.058	0.011	0.049	0.008	0.003

All times in milliseconds, all lengths in μm, and all rates in μm/milliseconds.

K_ATP GOF Myocytes Demonstrate Increased LTCC.

Cardiac excitation–contraction (EC) coupling depends on Ca entry via the L-type Ca²⁺ current (LTCC), which is modulated by phosphorylation in response to β-adrenergic and other neurohumoral inputs. Prestimulated cGOF myocyte contractility is consistent with enhancement of such modulation. Cell size, assessed by cell capacitance, was normal in cGOF myocytes, but baseline LTCC amplitude and density were double those of control myocytes (Fig. 4B). Commensurate with the idea that cGOF myocytes were prestimulated, LTCC amplitudes and densities were markedly increased by ISO in control myocytes but much less so in cGOF myocytes (Fig. 4). There was a slight shift in voltage dependence in the baseline cGOF current–voltage (I/V) relationship, again similar to that in control I/V curves after ISO exposure.

Fig. 4.

Enhanced LTCCs in cGOF myocytes. (A) Representative families of Ca²⁺ current obtained from control (Left) or cGOF (Right) mice. Na⁺ and T-type Ca²⁺ channels were inactivated by slow voltage ramp from holding potential of −70 to −45 mV, and then voltage was stepped to +55 mV in 10-mV increments. The identical protocol was used to elicit I_Ca following exposure to ISO (1 mM). (B) Peak I_Ca amplitude and I_Ca density as a function of voltage (mean ± SEM). (Inset) Mean capacitance in each case. (Scale bars, 500 pA and 20 ms.) Solid lines here and in following figures indicate currents in baseline, dotted lines in ISO; asterisks indicate significance of cGOF vs. control in each condition (two-way ANOVA, P < 0.001). In this figure only, daggers indicate significance of ISO vs. baseline for control (black) and cGOF (red).

Because the Kir6.1 GOF transgene is constitutively activated in cGOF hearts, the above changes in LTCC properties might be developmentally induced. To examine the temporal link between expression of K_ATP GOF and consequent increases in LTCC, we assessed cardiomyocytes in which GOF was induced postnatally, rather than developmentally [tamoxifen-induced GOF (icGOF)] by crossing Kir6.1[G343D] and tamoxifen-inducible Myh6-Cre mice (14). Double-transgenic icGOF and littermate controls were treated with tamoxifen for 5 d; ventricular myocytes were isolated, and LTCC recorded, on day 6. Recorded icGOF myocytes were larger than control myocytes but, even controlling for this, icGOF LTCC density was still dramatically increased at baseline (Fig. 5). Again, as in cGOF myocytes, ISO significantly stimulated LTCC in control but not icGOF myocytes. These experiments further confirm remodeling of LTCC in response to Kir6.1 GOF, and demonstrate that this can occur in adult animals and rapidly (within hours or days) following transgene induction, implying that such responses are physiologically, not developmentally, mediated.

Fig. 5.

Enhanced LTCCs in inducible icGOF ventricular myocytes. (A) Representative families of Ca²⁺ current obtained from either control (Left) or icGOF (Right) at baseline and in presence of ISO. Protocols as in Fig. 4. (B) Peak I_Ca amplitude and I_Ca density plotted as a function of voltage (mean ± SEM). (Inset) Mean capacitance in each case. (Scale bars, 500 pA and 10 ms.)

Vascular K_ATP GOF Expression Results in Enlarged Hearts and Increased LTCC.

Kir6.1 and SUR2 are prominently expressed in vascular smooth muscle (VSM). VSM expression of K_ATP GOF under smooth muscle (SM)-MHC promoter control reduces vascular contractility, resulting in a chronically vasodilated state with low systolic and diastolic BP (9). Young vGOF mice, induced with tamoxifen at 2 mo and studied by echocardiography 1 mo later, revealed no significant changes in cardiac structural parameters, but did exhibit enhanced cardiac contractility (Fig. 6A). A second cohort of vascular-specific expression (vGOF) mice in which expression had been induced for more than 6 mo (old vGOF) showed significant chamber dilation and now showed diminished contractility (Fig. 6A and Table S7). Consistent with these older vGOF echocardiographic findings, baseline LTCC in vGOF myocytes was elevated, although not as dramatically as in cGOF or icGOF mice, and both vGOF and control LTCC increased following ISO (Fig. 6B). Interestingly, the baseline I–V relationship was left-shifted by ∼10 mV (Fig. 6B) in vGOF myocytes.

Fig. 6.

Dilated heart and enhanced LTCCs in vGOF cardiac muscle. (A) Echocardiographic features of vGOF mice at 3 mo and >6 mo of age. (B) Representative Ca²⁺ currents obtained from either control (Left) or vGOF (Right) cardiomyocytes. Protocols as in Fig. 4. Peak I_Ca amplitude and I_Ca density plotted as a function of voltage (mean ± SEM). (Inset) Mean capacitance in each case. (Scale bars, 500 pA and 20 ms.)

Table S7.

Comparison of echocardiographic parameters in old cohort of vGOF mice

	Control (n = 4)		vGOF (n = 6)		P value
Age (d)	276.5	109.1	308.0	97.6	0.6454
HR	635.3	23.6	646.2	12.2	0.3586
LVPWd	0.85	0.10	0.87	0.11	0.8104
IVSd	0.90	0.10	0.89	0.10	0.8350
LVIDd	3.17	0.28	3.31	0.21	0.3958
LVPWs	1.498	0.192	1.422	0.230	0.6023
IVSs	1.510	0.153	1.478	0.135	0.7391
LVIDs	1.50	0.16	1.71	0.19	0.1124
LVM	90.58	13.37	96.84	8.68	0.3909
LVMI	3.06	0.88	3.12	0.39	0.8967
RWT	0.557	0.103	0.534	0.092	0.7186
FS	52.7	1.8	48.5	3.8	0.0736
EDV	28.75	2.87	32.61	2.04	0.0367
ESV	6.63	1.70	8.94	1.01	0.0255
EF	76.3	3.6	72.3	1.9	0.0519
dV/dt-s	0.843	0.070	0.890	0.063	0.3066
dV/dt-d	0.869	0.166	0.962	0.131	0.3518
LVMd	76.1	9.0	79.7	10.5	0.5984
LVMdI	2.90	0.33	2.91	0.23	0.9720
SV	22.13	2.06	23.67	1.17	0.1648
CO	13.83	1.54	15.10	0.99	0.1482
S-rad	40.88	3.63	37.13	1.40	0.0476
S-long	22.81	0.75	21.74	2.18	0.3811
SR-rad-s	11.9	1.7	10.6	1.0	0.1775
SR-rad-d	14.6	2.1	14.3	1.5	0.7943
SR-long-s	11.5	0.4	9.1	1.8	0.0325
SR-long-d	12.9	1.8	12.5	1.9	0.7874

Increased Cav1.2 Protein Phosphorylation in Kir6.1 GOF Hearts.

The LTCC in mature ventricular myocytes is conducted by Cav1.2 channels, composed of a pore-forming α₁ subunit in association with α₂δ, β, and possibly γ subunits (15). In the heart, the large C-terminal domain of the 250-kDa α1 subunit is proteolytically processed, resulting in a complex of the core Cav1.2 protein plus its noncovalently bound distal C terminus (16), a potent autoinhibitor of channel activity (17). β-Adrenergic stimulated phosphorylation of the Cav1.2 channel relieves this autoinhibition and enhances LTCC and myocyte contractility, in response to sympathetic signaling (18–20). We found no change in the level of full-length Cav1.2 channel protein, proteolytically processed Cav1.2 protein, or distal C-terminal protein, between cGOF and control hearts (Fig. 7A). We investigated two phosphorylation sites on Cav1.2 channels that have been identified in intact ventricular myocytes using phosphospecific antibodies (Ser1700 at the interface between the distal and proximal C-terminal domains and Ser1928 in the distal C-terminal domain). Phosphorylation of Ser1700 by PKA is directly implicated in β-adrenergic regulation of Cav1.2 channels (18, 19), whereas phosphorylation of Ser1928 may correlate with PKA or PKC pathway activity (18, 21). We found no significant change in phosphorylation of Ser1700 in cGOF, but Ser1928 phosphorylation was increased approximately twofold in cGOF mice (Fig. 7A), and ∼1.25-fold in vGOF mice. Because Ser1928 is also a substrate for phosphorylation by PKC (22), phosphorylation of this site by either PKA or PKC signaling pathways may be involved in persistent increase in basal activity, and consequent loss of β-adrenergic stimulation of Cav1.2 channels in CS.

Intolerance of β-Blockade in icGOF Animals.

To further explore the mechanistic basis of Cantu disease and directly assess whether β-adrenergic signaling is required for adaptation to K_ATP GOF in the heart, we chemically ablated sympathetic signaling in icGOF and littermate control mice by implanting adult (12- to 32-wk-old) animals with slow-release propranolol pellets, and then initiating transgene expression with tamoxifen 2–7 d later. The majority of icGOF mice treated with propranolol pellets died within 2 wk of transgene induction, but no deaths were observed in littermate controls (Fig. 7B). Though further mechanistic details remain to be elucidated, these data are a striking indication that adrenergic signaling is required for adaptation to K_ATP GOF induction.

Discussion

CS and Cardiovascular Disease.

K_ATP channels are heterooctameric complexes of pore forming Kir6.1 or Kir6.2 subunits and regulatory SUR1 or SUR2 subunits (23). ABCC9 (SUR2) and KCNJ8 (Kir6.1) are prominently expressed in cardiac myocytes, VSM, and vascular endothelial cells (23), suggesting that cardiovascular features will predominate in CS. We show that CS patients have dilated LVs, increased cardiac output and ejection fraction, and increased myocardial contractility. LV mass is increased compared with controls, but LV wall thicknesses are normal. Such findings are distinct from hallmarks of either hypertrophic cardiomyopathy (HCM) or dilated cardiomyopathy; the former typically demonstrates thickened LV chamber walls with hypercontractile function, whereas the latter demonstrates a dilated LV chamber with diminished cardiac function. Distinguishing between these cardiac phenotypes is critical, because standard therapies for HCM, for example, might actually worsen a Cantu patient’s clinical status.

High cardiac output states such as we observe can arise from chronic vasodilation, and can lead to LV volume overload and subsequent chamber dilation (24–27). Vital statistic and echocardiographic data support the hypothesis that CS patients are vasodilated, but in a compensated high cardiac output state. Though it is controversial to term BP as clinically low in an asymptomatic patient, systolic and diastolic BPs were greatly diminished compared with mean for age.

Feedback Control via L-Type Ca Current in Compensated Cardiac Function.

Gain of K⁺ channel function in blood vessels would be predicted to cause reduced contractility and vasodilation, and in the heart to shorten APs, reducing Ca²⁺ entry and contractility. Available clinical data regarding ECG phenotype in CS patients is limited, but there is no evidence for AP shortening (10). The high cardiac output state evident by echocardiography is also not naively predicted and leads us to postulate that feedback response to both the heart and vasculature remodels EC coupling to produce the unexpected hypercontractile function (Fig. 7). In mice, echocardiograms and isolated myocyte contractility studies reveal that such compensation does indeed occur: cardiac contractility is increased, concomitant with increased LTCC. Isolated cGOF myocytes reveal basal hypercontractility and elevated LTCC, but similar maximal contractility and LTCC after ISO exposure to control, whereas old vGOF myocytes also show some increase in baseline LTCC. Direct analysis of LTCC in hearts of CS patients is not feasible, but there are indirect suggestions of similarly enhanced basal LTCC in the ambulatory ECG data: higher than normal heart rates, as well as blunted heart rate variability and T-wave abnormalities (Fig. 2).

In the presence of ISO, LTCC densities in cGOF or icGOF myocytes are not significantly different from control (Figs. 4 and 5) (12). This lack of ISO response is consistent with chronic in vivo activation of signaling pathways that result in prestimulation. This prestimulation must be a consequence of the initial defect; that is, the gain of K_ATP conductance, which we suggest will indeed be a reduced cardiac output, but that this results in activation of adrenergic or parallel signaling pathways. We previously showed that basal and ISO-stimulated cAMP concentrations are not altered in transgenic Kir6.2 GOF hearts that also show prestimulation (12). Cardiac LTCC can be increased by PKA activation (28–30), but also by PKC in response to α-adrenergic and endothelin-1–dependent signaling (31–35). PKC overexpression may result in diminished LTCC current in some circumstances (36), although dialysis of critical LTCC-associated proteins may confound results in whole-cell patch-clamp experiments (32). Both PKA and PKC modulation of LTCC (22, 37) are mediated by phosphorylation of Ca_V1.2 serine residues on the pore or accessory subunits (18–20, 38). The identity of relevant residues phosphorylated in response to adrenergic signaling remains an active area of investigation. Ser1700 and Ser1928 phosphorylation of Cavα1.2 have both been demonstrated, but Ser1928 may also be phosphorylated by PKC (35, 39), and Ser1700 may also be a target of CaMKII (18, 40). Our analysis implicates Ser1928 phosphorylation as a marker of the response to K_ATP GOF but not Ser1700. Though Ser1928 phosphorylation may not itself be directly involved in the enhancement of LTCC, it is suggestive that PKC or PKA pathways may ultimately be activated.

Feedback Response to Vascular Defects in CS.

Two distinct effects were observed on cGOF and icGOF LTCC: left-shift in activation and increased current density. We also observed a left-shift of voltage dependence of LTCC in vGOF, but only a mild increase in current density, and a nonsignificant increase in Ser1928 phosphorylation. In contrast, older vGOF animals exhibited increased diastolic volume, as observed in CS patients. Vascular K_ATP GOF results in increased smooth muscle K_ATP current and diminished BP (9), providing a potentially direct explanation for the markedly low-for-age BPs in CS patients; this also provides a systemic mechanism that may link the primary vGOF phenotype to secondary consequences in the heart: vascular K_ATP GOF expression results in vasodilation that will ultimately result in a long-standing volume load, evidenced by LV chamber dilation. Early after transgene induction (6 wk), vGOF hearts exhibit increased contractility and increased ejection fraction (Fig. 6), but end diastolic volume is increased in older vGOF animals, potentially reflecting dilation manifesting after prolonged exposure to volume overload (26).

Conclusions and Implications.

CS can arise from GOF in either Kir6.1 or SUR2 proteins of the cardiovascular K_ATP channel (2–5). A distinct CS cardiac pathology is characterized by high output state with enhanced cardiac contractility and enhanced chamber volume, associated with decreased vascular pulse wave velocity and low BP. Our animal studies suggest that CS cardiac pathology emerges from combined cardiac and vascular K_ATP GOF mutation expression (Fig. 7C). The vascular findings are readily explained by the expected molecular consequences, but K_ATP GOF mutations in the myocardium will tend to reduce cardiac action potential duration (APD) and decrease pacemaker activity, both of which would reduce cardiac contractility and output. The counter observation of enhanced contractility and maintained APD is explained by enhanced LTCC, with left-shifted activation and increased basal conductance, associated with enhanced phosphorylation of the PKA/PKC target residue Ser1928. These findings are consistent with compensatory chronic signaling, potentially involving adrenergic stimulation, through pathways that converge on the LTCC.

This consistent explanation for CS cardiac features raises the question of whether or how to treat them. The dramatic difference in response of icGOF and control animals to β-blockade (Fig. 7B) is consistent with adrenergic signaling being involved in at least the early compensation to K_ATP GOF, further suggesting that β-blockade could be a dangerous approach to treating Cantu patients. Alternately, appropriate therapies should target K_ATP channels directly, and the success of sulphonylurea drugs in treating neonatal diabetes, which results from GOF in the pancreatic K_ATP isoforms, gives promise that similar or more selective K_ATP antagonists may reverse some or all CS disease features.

Methods

Human studies were carried out on CS patients recruited to an annual research clinic at St. Louis Children’s Hospital. Written informed consent was provided by all patients. The study was approved by the Human Research Protection Office of Washington University School of Medicine and performed at St. Louis Children’s Hospital in St. Louis. Echocardiographic and electrocardiographic studies were performed. All animal studies complied with the standards for the care and use of animal subjects as stated in the NIH Guide for the Care and Use of Laboratory Animals (41) and were reviewed and approved by the Washington University Institutional Animal Care and Use Committee. Mouse strains used included cGOF (10), αMHC-Cre (42), icGOF (tamoxifen-inducible Kir6.1[G343D] transgenic), Mer-Cre-Mer-α-MHC (43), and vGOF (9). Transgene expression was induced by 5× daily injections of 10 mg/kg tamoxifen (44). Isolated myocyte studies and isolated tissue Western blot analyses were carried out as previously described (45–48). Detailed methods are available in SI Methods.

SI Methods

Human Study Design.

Patients were recruited to an annual research clinic through invitations from cardiology and genetic outpatient clinics at St. Louis Children’s Hospital, and through the CS interest group website (www.cantu-syndrome.org). After a detailed explanation of the study, and before enrollment, written informed consent was obtained from each subject or, if under 18, their parents. We included healthy age, gender, and body mass index (BMI)-matched subjects from a separate study (49). The study was approved by the Human Research Protection Office of Washington University School of Medicine and performed at St. Louis Children’s Hospital.

Standard clinical histories and physical examinations, including height, weight, and head circumference measurements, were made. BP was measured by manual sphygmomanometer with appropriate cuff size and heart rate, in the resting supine position and after 2 min in the upright position, to ascertain orthostatic vital sign changes. BP measurements were repeated three times to confirm measurement accuracy, and were subsequently compared with published normative values by gender and age (www.nhlbi.nih.gov/files/docs/guidelines/child_tbl.pdf).

Ventricular structure.

Transthoracic complete M-mode, 2D, and color Doppler echocardiographic examinations were performed with a commercially available ultrasound imaging system with appropriate size-phased array transducer (Vivid 7 and 9; General Electric Medical Systems). M-mode imaging of the LV in the parasternal short axis view was used to measure LV end-diastolic diameter (LVEDD), LV end-systolic diameter (LVESD), end diastolic posterior wall thickness (PWT), relative wall thickness (RWT = 2 × PWT/LVEDD) (50) and LV mass. LV mass was calculated and indexed to height (LVMI) using the Devereux equation (51). To define ventricular structural phenotype, the 95th percentiles for LVMI for children older than 9 y of age (>40 g/height^2.7 in females and >45 g/height^2.7 in males) and the 95th percentile value for RWT for normal children and adolescents (RWT > 0.41) were used as cutoff values to categorize abnormal LV geometry (52, 53). Children categorized as having normal LV geometry had LVMI and RWT below the 95th percentiles. Concentric remodeling was defined as normal LVMI and elevated RWT, eccentric hypertrophy as elevated LVMI and normal RWT, and concentric hypertrophy as elevated LVMI and RWT (54).

Ventricular function.

Cardiac output was calculated from measured aortic valve annular diameter (D), time velocity integral (TVI) of pulse-wave Doppler velocity in LV outflow tract, and HR (beats per minute) using equation: π × (D/2)² × TVI × HR × 100 (l per min), averaged for three consecutive cardiac cycles and indexed against body surface area (BSA) (l⋅min⁻¹⋅m²) (55). In addition, M-mode imaging was used to measure LV fractional shortening [FS = (LVEDD – LVESD)/LVEDD], and 2D imaging was used to measure LV ejection fraction using the modified Simpson method. Myocardial mechanics were evaluated by measurement of strain (dimensionless measure of myocardial deformation) and strain rate (time derivative of strain), which correlates with LV peak elastance, reflecting contractile performance (56). Strain and strain rate were analyzed by 2D speckle-tracking echocardiography. This technique is noninvasive, minimally angle-dependent, and has been validated against sonomicrometry and tagged MRI (57, 58). LV global longitudinal strain (GLS) and rates (GLSR) were measured from grayscale 2D apical images of the LV (four-, three-, and two-chamber views) (57). The peak LV GLS, systolic GLSR, early diastolic GLSR, and late diastolic GLSR values were calculated using EchoPAC (General Electric Medical Systems) software, by averaging the GLS and GLSR measurements from the three apical views.

Characterization of heart rate and rhythm.

Cardiac rhythm and conduction were assessed by 15-lead surface ECG and rhythm strip, and two-lead 24-h ambulatory ECG recording. Data were loaded onto a MARS PC Holter Analyzer (GE Medical Systems) and analyzed using standard research Holter techniques. Time-domain HRV was used to measure the statistical properties of NN interval variation over various time scales (from beat-to-beat to 24 h). Frequency domain HRV using power spectral analysis (fast Fourier transforms) was used to quantify variance (power) of heart rate as a function of underlying physiological oscillations, reflecting different aspects of the regulatory action of the autonomic system on the heart (59).

Animal Studies.

Generation and strains of transgenic mice.

All procedures complied with the standards for the care and use of animal subjects as stated in the NIH Guide for the Care and Use of Laboratory Animals (41) and were reviewed and approved by the Washington University Institutional Animal Care and Use Committee. Animals used were as follows::

cGOF.

Mice harboring a Kir6.1[G343D] mutation downstream of the β-chicken actin promoter fragment knocked into the ROSA26 locus as previously described (9) were crossed with tissue- and time-specific Cre-driver mice to yield experimental models. Germ-line cardiac-specific K_ATP GOF was achieved by crossing with αMHC-Cre mice, which express Cre recombinase under direction of an α-myosin heavy-chain promoter fragment (60).

icGOF.

Kir6.1[G343D] mice were crossed with Mer-Cre-Mer–regulated α-MHC promoter fragment mice (43) and transgene expression was induced in double-transgenic (TG) animals by 5× daily injections of tamoxifen (10 mg/kg) as detailed previously (44). Control single-TG littermates were treated the same way.

vGOF.

Kir6.1[G343D] mice were crossed with animals expressing CRE recombinase under control of a smooth muscle-myosin heavy-chain promoter fragment (61) (SM-MHC-Cre). Transgene expression was induced in double-TG animals by 5× daily injections of 10 mg/kg tamoxifen (44). Control single-TG littermates were treated the same way.

Kir6.1[AAA].

TG mice carrying an inducible dominant-negative mutation in the pore-forming subunit of Kir6.1 were described previously (13). Double-TG animals were obtained by crossing with αMHC-Cre driver mice as described previously.

Isolation of cardiac myocytes.

Mice were anesthetized by subcutaneous injection with 2.5% (vol/vol) Avertin (10 mL/kg, i.p.). The heart was rapidly excised, and retrogradely perfused through the aorta with Ca²⁺-free solution for 5 min followed by perfusion with digestion solution containing 0.8 mg/mL collagenase (Type 2; Worthington Biochemical). LV myocytes were dispersed by manual trituration. Cells were stored at room temperature in WIM solution.

Cellular electrophysiology.

Isolated myocytes were transferred into a recording chamber containing Tyrode solution (below; 5.4 mM CsCl replaces KCl) and TTX (10 μM). ISO (1 μM) was added where indicated. Patch-clamp electrodes (1–3 MΩ when filled with electrode solution) were fabricated from soda lime glass microhematocrit tubes (Kimble 73813). Cell capacitance and series resistance were determined using 5–10 mV hyperpolarizing square pulse from a holding potential of –70 mV following establishment of whole-cell recording configuration. Series resistance was electronically compensated by 60–80%. pCLAMP 9.0 software and Digidata 1200 converter were used to generate command pulses and collect data. Data were filtered at 1 kHz and recorded at 3 kHz.

Single-cell contraction measurement.

To assess contractility, unloaded cell length was measured using a video edge detection system (IonOptix). Myocytes were bathed with normal Tyrode solution and stimulated with bipolar-stimulating electrodes placed near the cell.

Solutions.

Wittenberg isolation medium.

NaCl, 116 mM; KCl, 5.3 mM; NaH₂PO₄, 1.2 mM; glucose, 11.6 mM; MgCl₂, 3.7 mM; Hepes, 20 mM; l-glutamine, 2.0 mM; NaHCO₃, 4.4 mM; KH₂PO₄, 1.5 mM; 1× essential vitamins (GIBCO catalog no. 12473-013); 1× amino acids (GIBCO catalog no. 11120-052); pH 7.3–7.4.

Tyrode solution with CsCl substituted for KCl.

NaCl, 137 mM; CsCl, 5.4 mM; NaH₂PO₄, 0.16 mM; glucose, 10 mM; CaCl₂, 1.8 mM; MgCl_2, 0.5 mM; Hepes, 5.0 mM; NaHCO₃, 3.0 mM; pH 7.3–7.4.

Krebs–Henseleit.

NaCl, 118 mM; KCl, 4.7 mM; NaHCO₃, 25 mM; NaH₂PO₄, 1.2 mM; MgSO₄, 1.2 mM; CaCl₂, 1.2 mM; glucose, 10 mM; saturated with a 95% O₂–5% CO₂ gas mixture; pH 7.3–7.4.

K_INT.

KCl, 140 mM; K₂ATP, 5 mM; Hepes, 10 mM; EGTA, 10 mM; pH 7.3–7.4.

Cs_INT.

CsCl, 120 mM; TEA-Cl, 20 mM; K₂ATP, 5 mM; Hepes, 10 mM-; pH 7.3–7.4.

Western blot analysis.

Hearts were homogenized in buffer containing 50 mM Tris (pH 8.5), 5 mM EDTA, 150 mM NaCl, 10 mM KCl, 1% Triton X-100, 5 mM NaF, 5 mM β-glycerophosphate, 1 mM sodium-orthovanadate, plus protease inhibitors and membranes were solubilized in homogenization buffer for 2 h at 4 °C. Insoluble material was removed by centrifugation at 16,000 × g for 20 min. Proteins were separated by SDS/PAGE, transferred to nitrocellulose, and analyzed by immunoblotting. Polyclonal anti-CNC1 was generated against amino acid sequences in the intracellular loop between domains II and III of Cav1.2 (821–838) and detects both the long and short forms of Cav1.2 channels (45, 46). Polyclonal CH2 antibodies were generated against peptides corresponding to residues 2051–2066 in the distal C terminus of Cav1.2 (16) and were used to detect the distal C terminus and the processed C terminus. Anti-CH3P was generated against residues 1923–1932 and recognizes phosphorylated S1928 (47). Anti-Cav1.2-pS1700 phosphospecific antibody was generated against residues 1694–7109 in the proximal C terminus (48).

Data analysis.

Unless indicated, data were analyzed using Clampfit, Microsoft Excel, and GraphPad Prism6 (GraphPad) software. Results are presented as mean ± SEM. Statistical tests and P values are denoted in figures.