Notch signaling modulates the electrical behavior of cardiomyocytes

Abstract

Notch receptor signaling is active during cardiac development and silenced in myocytes after birth. Conversely, outward K⁺ Kv currents progressively appear in postnatal myocytes leading to shortening of the action potential (AP) and acquisition of the mature electrical phenotype. In the present study, we tested the possibility that Notch signaling modulates the electrical behavior of cardiomyocytes by interfering with Kv currents. For this purpose, the effects of Notch receptor activity on electrophysiological properties of myocytes were evaluated using transgenic mice with inducible expression of the Notch1 intracellular domain (NICD), the functional fragment of the activated Notch receptor, and in neonatal myocytes after inhibition of the Notch transduction pathway. By patch clamp, NICD-overexpressing cells presented prolonged AP duration and reduced upstroke amplitude, properties that were coupled with reduced rapidly activating Kv and fast Na⁺ currents, compared with cells obtained from wild-type mice. In cultured neonatal myocytes, inhibition of the proteolitic release of NICD with a γ-secretase antagonist increased transcript levels of the Kv channel-interacting proteins 2 (KChIP2) and enhanced the density of Kv currents. Collectively, these results indicate that Notch signaling represents an important regulator of the electrophysiological behavior of developing and adult myocytes by repressing, at least in part, repolarizing Kv currents.

NEW & NOTEWORTHY We investigated the effects of Notch receptor signaling on the electrical properties of cardiomyocytes. Our results indicate that the Notch transduction pathway interferes with outward K⁺ Kv currents, critical determinants of the electrical repolarization of myocytes.

INTRODUCTION

Outward K⁺ Kv-mediated currents are important determinant of the repolarization phase of the action potential (AP) of cardiomyocytes (57). Alterations in the expression of Kv channels and regulatory proteins during development, physiological, and pathological circumstances affect the profile of the AP with consequences on myocardial excitation-contraction coupling and electrical stability of the heart (68).

The Kv channel-interacting protein 2 (KChIP2), which is part of the Kv channel multiprotein complex assembly, has emerged as a key modulator of Kv currents and AP duration (17, 46, 74). KChIP2 levels are attenuated in the heart during early phases of development (35, 79) and in response to diseased conditions (46, 59, 62), scenarios that are associated with weaker repolarizing Kv currents and prolonged duration of the AP (4, 11, 19, 34, 52, 54, 69, 76, 78, 80). However, molecular cues governing the expression of Kv, KChIP2, and other ion channel subunits dictating AP properties in myocytes in homeostatic and pathological conditions remain largely unknown.

Notch receptor signaling is active during embryonic development and critical in cardiogenesis by regulating trabeculation, myocyte proliferation, and valve formation (8, 12, 20, 30, 60, 65). Notch signaling is progressively silenced in the postnatal heart, but it is partly restored in the adult myocardium after ischemic injury (22, 44), hypertrophy (10), or heart failure (40), a factor that appears to preserve the structural and functional integrity of the organ by promoting prosurvival and proliferating signaling (10, 21, 22, 44, 67). Activation of the Notch transduction system involves cell-to-cell contact and binding of the extracellular domain of the Notch receptor to delta-like and/or Jagged families of membrane-bound ligands. After receptor-ligand interaction, the multisubunit protease complex γ-secretase cleaves the intracellular domain of the Notch receptor (NICD) (6, 14, 42), which, in turn, translocates to the nucleus. NICD then interacts with transcription regulators of the CSL family, promoting the activation of genes comprising the hairy and enhancer of split (HES) and hairy-related transcription factor (HRT) families (31, 33, 55) involved in cell fate determination. But whether the Notch receptor signaling axis ultimately interferes with the electrophysiological properties of myocytes in the developing and adult heart remains to be clarified.

In the present study, we examined the possibility that NICD modulates the density of ionic currents in cardiomyocytes. We found that overexpression of NICD in adult cells alters the repolarization phase of the AP by decreasing Kv currents, whereas suppression of NICD signaling in neonatal cells promotes the appearance of Kv currents and enhances KChIP2 gene expression. Thus, Notch signaling represents an important regulator of the electrophysiological properties of neonatal and adult cardiomyocytes.

METHODS

Transgenic animals.

C57Bl/6 neonatal mice at postnatal day 2 (n = 8 litters) or postnatal days 5–7 (n = 4) and adult transgenic mice of either sex (n = 54) and C57Bl/6 adult mice (n = 8) were used in this study. Experiments were performed in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals and approved by the local animal care committees (Institutional Animal Care and Use Committee at New York Medical College and at Brigham and Women's Hospital, Harvard Medical School). To activate Notch1 signaling in cardiomyocytes, a Cre-conditional NICD expression mouse model was used (49). This was achieved by crossing homozygous mice for tamoxifen-inducible Cre recombinase under the cardiac-specific α-myosin heavy chain promoter (αMHC-MerCreMer, MCM^+/+ mice; stock no. 5657, The Jackson Laboratory) (71) with hemizygous ZEG-NICD mice, in which the transgenic construct uses a loxP-flanked STOP sequence between a cytomegalovirus enhancer chicken β-actin promoter and the coding sequence for intracellular portion of the Notch1 protein (NICD), with an internal ribosomal entry site (IRES) linked the enhanced green fluorescence protein (GFP) gene (NICD-GFP^+/− mice, stock no. 6850, The Jackson Laboratory) (49). The progeny consisted of double transgenic (αMHC-MerCreMer and NICD-GFP, MCM-NICD-GFP) and single transgenic [αMHC-MerCreMer, MCM-wild type (WT)] mice born with a normal Mendelian ratio of ~1:1.

To account for the potential off-target effects of tamoxifen treatment and expression of Cre and GFP in myocytes (41), mice with tamoxifen-inducible Cre recombinase (MCM^+/+ mice) were crossed with transgenic mice with inducible cell membrane-localized green fluorescence in Cre recombinase-expressing cells (stock no. 7676, The Jackson Laboratory) (53). The progeny consisted of double transgenic animals (αMHC-MerCreMer and GFP, MCM-GFP mice).

Cre recombinase was induced by administration of tamoxifen (Sigma-Aldrich) for 4 consecutive days (0.1 g·kg body wt⁻¹·day⁻¹ ip) (29). Age- and sex-matched MCM-NICD-GFP and MCM-WT mice were treated with tamoxifen, and cardiomyocytes were isolated 19–58 days later.

Histological analysis.

Mouse hearts were fixed in phosphate-buffered formalin (10%, Sigma) and embedded in paraffin (9). Thin sections of the left ventricle (LV) were indirectly immunolabeled for the detection of GFP (Anti-GFP Antibody, catalog no. ab5450, Abcam) and KChIP2 (KCNIP2 Polyclonal Antibody, catalog no. PA5-41075, ThermoFisher Scientific). Nuclei were counterstained with DAPI (Sigma-Aldrich). Secondary antibodies included donkey anti-goat [Alexa Fluor 488 AffiniPure Donkey Anti-Goat IgG (H⁺L), catalog no. 705-545-147, Jackson ImmunoReasearch Laboratories] and goat anti-rabbit [Alexa Fluor 647 AffiniPure Goat Anti-Rabbit IgG (H⁺L), catalog no. 111-605-003, Jackson ImmunoReasearch Laboratories]. Images were acquired using a Zeiss LSM 710 confocal microscope with a ×63 objective. For comparison of KChIP2 expression levels in the myocardium of MCM-WT and MCM-NICD-GFP mice, digital images obtained from immunofluorescent-stained sections were analyzed using ImageJ software. Mean fluorescence intensity in the channel of KChIP2 labeling was measured in areas corresponding to cardiomyocytes, as determined by cell shape and GFP expression.

Isolation and culture of neonatal mouse cardiomyocytes.

Neonatal mouse cardiomyocytes were prepared from hearts of 2-day-old pups. Briefly, mice were anesthetized by hypothermia and euthanized by decapitation with sharp surgical scissors. Hearts were explanted, and ventricles were dissected, minced, incubated in trypsin dissolved in HBSS medium (1 mg/ml, Sigma-Aldrich) at 4°C for ~12 h, and then serially digested with collagenase type II (1 mg/ml, 285 U/mg, Worthington Biochemical) at 37°C. Alternatively, minced myocardial ventricles were digested in a solution containing the following (in mM): 116 NaCl, 4.7 KCl, 0.8 MgSO₄, 20 HEPES, 5.6 glucose, 20 taurine, 0.005 creatine, 0.005 Na pyruvate, and 0.8 NaH₂PO₄ (pH 7.4), in which 0.4 mg/ml of collagenase type II (285 U/mg, Worthington Biochemical) and 0.6 mg/ml of pancreatin from the porcine pancreas (Sigma-Aldrich) were added. Collagenase type IV (Worthington Biochemical) was used for one isolation. After dissociation, cells were pooled and preplated in two consecutive steps of 30–45 min to allow fibroblasts to adhere. The myocyte-enriched population remaining in suspension was plated in culture dishes or coverglasses precoated with 0.1% gelatin (Sigma-Aldrich) for 1 h. Myocytes were cultured in DMEM (Life Technologies) containing 10% FBS (GIBCO) and the antibiotics penicillin-streptomycin (100 U/ml penicillin and 0.1 mg/ml streptomycin, Sigma-Aldrich). After 24 h, the medium was changed and cells were cultured in the presence of 20 µM γ-secretase inhibitor (DAPT; Sigma-Aldrich) or vehicle only (DMSO; Sigma-Aldrich). The treatment was repeated every 48 h.

Isolation of adult cardiomyocytes.

With the animal under deep anesthesia (isoflurane), thoracotomy was performed, the heart was excised, and LV myocytes were enzymatically dissociated as previously reported (52, 69, 70, 72). Briefly, the heart was connected to a plastic cannula for retrograde perfusion through the aorta in a Langendorff system (Radnoti) at 37°C. The perfusate consisted of a Ca²⁺-free solution gassed with 85% O₂-15% N₂. After 5 min, 0.1 mM CaCl₂, 274 U/ml collagenase (type 2; Worthington Biochemical), and 0.57 U/ml protease (type XIV, Sigma-Aldrich) were added to solution that contained the following (in mM): 126 NaCl, 4.4 KCl, 5 MgCl₂, 20 HEPES, 22 glucose, 20 taurine, 5 creatine, 5 Na pyruvate, and 5 NaH₂PO₄ (pH 7.4). At completion of digestion, the atria and right ventricle were dissected and the LV was cut in small pieces; these fragments were shaken in resuspension solution and filtered using a 200-µm nylon mesh (Spectrum Laboratories). For electrophysiological experiments, only rod-shaped myocytes exhibiting cross striations and showing no spontaneous contractions or contractures were selected; cells were used within 8 h after enzymatic digestion.

Whole cell patch-clamp experiments.

Isolated adult LV myocytes or neonatal myocytes plated on a coverglass were placed in a bath on the stage of an Olympus IX51, Olympus IX71, or Zeiss Axiovert 200 microscopes for patch-clamp measurements (52, 66, 69, 70, 73). Data were acquired by means of the whole cell patch-clamp technique in voltage- and current-clamp modes using Multiclamp 700A or 700B amplifiers (Molecular Devices). Electrical signals were digitized using 250-kHz 16-bit resolution analog-to-digital converters (Digidata 1322, 1440A, or 1550B, Molecular Devices) and recorded using pCLAMP 9.0 or 10 software (Molecular Devices) with low-pass filtering at 2 kHz. Membrane capacitance (C_m) was measured in voltage-clamp mode using a 5-mV voltage step and pCLAMP software algorithm; this parameter was used to normalize transmembrane currents (69, 70). Pipettes were pulled by means of a vertical (PB-7, Narishige) or horizontal (P-97, Sutter Instruments) glass microelectrode puller; when filled with intracellular solution, pipettes had a resistance of 1–3 MΩ.

For AP measurements, cells were stimulated with current pulses 1–1.5 times threshold (52, 66, 69, 70, 72, 73). Myocytes were bathed with Tyrode solution containing the following (in mM): 140 NaCl, 5.4 KCl, 1 MgCl₂, 5 HEPES, 5.5 glucose, and 1 CaCl₂ (pH 7.4 adjusted with NaOH). The composition of the pipette solution was as follows (in mM): 10 NaCl, 113 KCl, 0.5 MgCl₂, 5 K₂-ATP, 5.5 glucose, 5 EGTA, and 10 HEPES (pH 7.2 with KOH). AP measurements were conducted at 37°C. Data were analyzed using Clampfit (pCLAMP) software.

The fast Na⁺ current (I_Na) was measured in adult myocytes with a previously reported protocol (69). Cells were bathed with modified Tyrode solution of the following composition (in mM): 5 NaCl, 1 MgCl₂, 1 CaCl₂, 0.1, CdCl₂, 20 HEPES, 11 glucose, and 132.5 CsCl (pH 7.4 with CsOH). The composition of the pipette solution was as follows (in mM): 5 NaCl, 135 CsF, 10 EGTA, 5 Mg-ATP, and 5 HEPES (pH 7.2 with CsOH). Current-voltage (I–V) relations were determined by applying depolarizing steps 200 ms in duration from a holding potential (V_h) of −120 mV in 5-mV increments. The interpulse interval was 3 s. I_Na amplitude was measured as the current difference between the inward peak current and the current at the end of the 200-ms step. I_Na was normalized by C_m.

At each membrance potential (V_m) tested, Na⁺ conductance (g) was calculated as follows: g = I_Na/(V_m − E_Na) (1), where I_Na is the amplitude of the Na⁺ current at V_m and E_Na is the reversal potential, which was defined by the intercept of the I–V relation to the zero-current axis.

Conductance-voltage relations were plotted and fitted with the following Boltzmann function: g = g_max/{1 + exp[(V_1/2G − V_m)/k_G]} (2), where g_max is the maximal conductance, V_1/2G is the potential at which conductance is halfway between 0 and g_max, and k_G is the slope of the conductance curve.

Values of normalized Na⁺ conductance (G = g/g_max) were plotted to obtain activation curves; half-maximal activation potential (V_1/2G) and the slope of the activation curve (k_G) were obtained by fitting the data with the following Boltzmann equation: G = {1 + exp[(V_1/2G − V_m)/k_G]}⁻¹ (3).

A two-pulse protocol was used to assess the voltage dependence of steady-state inactivation of I_Na (69). Prepulses were introduced to depolarize the cell to different membrane voltages starting from −120 mV for 300 ms in 5-mV increments. Each prepulse was followed by a single 30-ms test pulse, which depolarized the cell to −40 mV. Values of normalized Na⁺ conductance (G = g/g_max) were plotted with V_m relative to the preconditioning steps to compute the steady-state inactivation curves, which were fitted with a Boltzmann equation, as indicated above (3).

I_Na reactivation was studied using a two-pulse protocol with a variable interpulse duration (69) ranging from 1 to 50 ms in increments of 1 ms. With each pulse, I_Na was activated by depolarizing cells from a V_h of −120 to −40 mV for 30 ms. The double pulse protocol was repeated every 3 s. For each interpulse duration (t), I_Na reactivation was calculated by dividing the amplitude of the current measured during the second pulse [I_(t)] by the amplitude of the current measured during the first pulse (I_max). Values of I_Na reactivation were plotted to obtain reactivation curves.

In adult cardiomyocytes, voltage-gated outward K⁺ Kv currents were assessed with a previously reported protocol (48, 69) to dissect K⁺ currents with rapid activation and fast (I_to), intermediate (I_K,slow1), and slow (I_K,slow2) kinetics of inactivation, together with a sustained (noninactivating) component (I_ss) (48, 57). Cells were bathed with modified Tyrode solution containing the following (in mM): 140 NaCl, 4 KCl, 1 MgCl₂, 10 HEPES, 10 glucose, 1.2 CaCl₂, and 0.3 mM CdCl₂ to block L-type Ca²⁺ current (pH 7.4 adjusted with NaOH) (48, 69). The composition of the pipette solution was as follows (in mM): 5 NaCl, 20, KCl, 120 K-aspartate, 1 MgCl₂, 5 Mg-ATP, 10 EGTA, and 10 HEPES (pH 7.2 with KOH) (48, 69). After C_m and series resistance compensation (≥70%), Kv currents were activated using depolarizing voltage steps from a V_h of −80 to +60 mV for 25 s in duration. By this protocol, loss of voltage control was minimized due to the proximity to the reversal potential of Na⁺. Interpulse interval was 60 s. I_to, I_K,slow1, I_K,slow2, and I_ss amplitudes and time constants were obtained by fitting the current decay during the depolarizing step with a triexponential function using Clampfit 10 software (48, 69). For cells lacking I_to, a biexponential function was used for fitting. Currents (expressed as pA/pF) were normalized by C_m. A representative multiexponential fitting curve for total outward currents and monoexponential fitting curve for each Kv current component are reported.

To further dissect the two components of I_to with fast (I_to,f) and slow (I_to,s) inactivation kinetics (7, 83), the decay phases of outward currents evoked by 4.5-s depolarization steps from a V_h of −80 to +60 mV were analyzed with a triexponential function to evaluate amplitude of I_to,f, I_to,s, and I_K,slow based on their distinct time constants of decay that differ by an order of magnitude (7, 23, 25, 83). For cells lacking I_to,f, I_to,s, or both components, biexponential or monoexponential functions were used. The interpulse interval was 5.5 s. Currents (expressed as pA/pF) were normalized by C_m. The representative multiexponential fitting curves for total outward currents and monoexponential fitting curve for each Kv current component are reported. In addition, the Kv channel inhibitor 4-aminopyridine was used at a high concentration (5 mM) (48, 83) to evaluate the nature of Kv currents affected by Notch signaling. For the same cell, the 4-aminopyridine-sensitive current was assessed as the difference current obtained by subtracting traces recorded in 4-aminopyridine from those measured in Tyrode solution for the same cell.

The inward rectifier K⁺ current (I_K1) was evaluated at 37°C using a depolarizing ramp (18, 66) from −110 to +80 mV for 10 s in duration using solutions comparable to those used for AP measurements.

Rapidly activating outward K⁺ Kv currents were assessed in neonatal myocytes with Tyrode solution containing 0.3 mM CdCl₂ to block L-type Ca²⁺ current. The composition of the pipette solution was as follows (in mM): 10 NaCl, 113 KCl, 0.5 MgCl₂, 5 K₂-ATP, 5.5 glucose, 5 EGTA, and 10 HEPES (79). I–V relations were determined applying depolarizing steps 500 ms in duration from a V_h of −70 mV in 10-mV increments. Rapidly activating K⁺ current amplitude was measured as the difference between the peak outward current at the beginning of the step and I_ss, the current at the end of the 500-ms step (79), using Clampfit 10 software.

Quantitative RT-PCR.

Total cellular RNA was prepared from cultured neonatal mouse ventricular myocytes or myocardial tissue using RNeasy Plus Micro kit (Qiagen) or Quick-RNA MiniPrep kits (Zymo Research) following the manufacturer's instructions. The kit contains gDNA Eliminator Mini Spin Columns to avoid genomic DNA contaminations. cDNA was obtained from total RNA using the MultiScribe reverse transcriptase kit (Applied Biosystems) or Moloney murine leukemia virus reverse transcriptase (Promega). Real-time RT-PCR was performed with the primers shown in Table 1. To avoid the influence of genomic contamination, forward and reverse primers for each gene were located in different exons. The StepOnePlus Real-Time PCR system (Applied Biosystems) was used for quantitative RT-PCR. In each case, cDNA was combined with Power SYBR Green Master Mix (Applied Biosystems) or Perfecta SYBR green FastMix Rox qPCR master mix (Quanta BioSciences) in a 20-μl reaction. Cycling conditions were as follows: 95°C for 10 min followed by 40 cycles of amplification (95°C denaturation for 10–15 s, 60°C annealing and extension for 1 min). The melting curve was then obtained. Threshold cycle (C_t) values were normalized with respect to β₂-microglobulin or hypoxanthine guanine phosphoribosyl transferase.

Table 1.

Primers for PCR analysis

Gene	Primers
GFP
Forward	5′-GCACGACTTCTTCAAGTCCGCCATGCC-3′
Reverse	5′-GCGGATCTTGAAGTTCACCTTGATGCC-3′
Mouse Kcna4
Forward	5′-AAAGGGGAAACAAATCACCG-3′
Reverse	5′-GCAGGAAATGAAGAGCATCC-3′
Mouse Kcna5
Forward	5′-ATGAGGCCCATCACTGTAGG-3′
Reverse	5′-AAAATTGGAGACGATGACGG-3′
Mouse Kcnd2
Forward	5′-TCAGCAAGCAAGTTCACCAG-3′
Reverse	5′-TTCCCTGCTATGGTTTTTGG-3′
Mouse Kcnd3
Forward	5′-TCTGCCAGCAAGTTCACAAG-3′
Reverse	5′-TTCCCTGCGATTGTCTTAGG-3′
Mouse Kcnip2
Forward	5′-TGAGGGTCTGGAACAACTCC-3′
Reverse	5′-GGGACATTCGTTCTTGAAGC-3′
Mouse Knip2a
Forward	5′-CTCCAGGGCCCAGTAAAAA-3′
Reverse	5′-TGGGGCAGCTAATGTTTCAC-3′
Mouse Knip2b
Forward	5′-TCAGTGAAAACAGCGTGGAG-3′
Reverse	5′-TGCGTGTGAACTTGGTTTGT-3′
Mouse Knip2c
Forward	5′-AGCTTACGGACAGCGTGGA-3′
Reverse	5′-ACTCTCTGCGTGTGAACTTGG-3′
Mouse Notch1
Forward	5′-AAGTGGGACCTGCCTGAAT-3′
Reverse	5′-CTGAGGCAAGGATTGGAGTC-3′
Mouse Hes1
Forward	5′-GGAAATGACTGTGAAGCACCT-3′
Reverse	5′-GTCACCTCGTTCATGCACTC-3′
Mouse Hey1
Forward	5′-CACCTGAAAATGCTGCACAC-3′
Reverse	5′-ATGCTCAGATAACGGGCAAC-3′
Mouse Hprt
Forward	5′-AGCCCCAAAATGGTTAAGGT-3′
Reverse	5′-CAAGGGCATATCCAACAACA-3′
Mouse β2m
Forward	5′-CTCGGTGACCCTGGTCTTTCTG-3′
Reverse	5′-ATGTGAGGCGGGTGGAACTG-3′

GFP, green fluorescent protein; Hprt, hypoxanthine guanine phosphoribosyl transferase; β2m, β₂-microglobulin.

Data analysis.

Data are presented as means ± SE or means ± SE and scatterplots. Statistical analysis was performed using SigmaPlot 11.0. Data were initially tested for normality (Shapiro-Wilk) and equal variance for assignment to parametric or nonparametric analysis. Parametric tests included Student’s t-test or one-way repeated ANOVA followed by the Bonferroni test for two groups or multiple comparisons, respectively. Nonparametric tests included the Mann-Whitney rank sum test or Kruskal-Wallis one-way ANOVA on ranks followed by Dunn’s method for two groups or multiple comparisons, respectively (69, 73). P < 0.05 was considered significant.

RESULTS

Notch1 alters the AP in adult myocytes.

To establish the effects of Notch1 signaling on the electrical properties of cardiomyocytes, mice with inducible myocyte-specific expression of NICD were used. Specifically, the MerCreMer system allowed tamoxifen-inducible synthesis of Cre recombinase in cardiomyocytes, promoting the expression of NICD and GFP (MCM-NICD-GFP). By RT-PCR, transcripts for GFP were apparent in isolated cell preparations obtained from MCM-NICD-GFP mice, together with significant increase in levels of the Notch downstream target Hes1, with respect to cells from WT mice (Fig. 1), which is consistent with NICD overexpression.

Fig. 1.

Activation of Notch signaling in MCM-NICD-GFP myocytes. Transcript levels for green fluorescence protein (GFP) and downstream targets of Notch1–Notch4 signaling Hes1 and Hey1 in isolated myocyte preparations from C57Bl/6 [wild type (WT); n = 4] and MCM-NICD-GFP mice [Notch1 intracellular domain (NICD); n = 4] are shown. *P < 0.001 vs. WT.

For electrophysiological experiments, LVs were enzymatically digested and GFP-positive myocytes from MCM-NICD-GFP animals were compared with GFP-negative cells obtained from single transgenic mice with tamoxifen-inducible MerCreMer transgene (MCM-WT). By patch-clamp analysis, MCM-NICD-GFP myocytes presented substantial (P < 0.001) prolongation of the early repolarization phase of the AP and reduction of the maximal velocity of repolarization (dV/dt_min) with respect to MCM-WT cells (Fig. 2). Conversely, resting V_m was similar in the two groups, but upstroke amplitude and velocity were attenuated in MCM-NICD-GFP cells. Thus, overexpression of NICD in myocytes is coupled with altered electrical properties.

Fig. 2.

The Notch1 intracellular domain (NICD) prolongs the duration of the action potential (AP) of adult myocytes. A: images for a left ventricular (LV) cardiomyocyte from a tamoxifen-treated MCM-NICD-GFP mouse obtained in bright-field (left) and fluorescent light (right). B: APs recorded in a green fluorescent protein (GFP)-negative myocytes isolated from a wild-type (WT) mouse (MCM-WT) and a GFP-positive cell obtained from a transgenic mouse with NICD expression (MCM-NICD-GFP). The derivative of the transmembrane potential signal for the two cells is shown at the bottom. C: AP properties of GFP-negative myocytes from MCM-WT mice (WT, n = 23; from 9 animals) and GFP-positive myocytes from MCM-NICD-GFP mice (NICD, n = 17; from 8 animals) are shown as means ± SE and scatterplots. RMP, resting membrane potential; APD, AP duration. *P < 0.05 vs. WT; **P < 0.001 vs. WT.

Notch1 attenuates fast I_Na in adult myocytes.

To define the basis for the attenuated depolarization of the AP in myocytes overexpressing NICD, fast I_Na, which is responsible for the AP upstroke, was evaluated in the two groups of cells. With respect MCM-WT myocytes, peak I_Na density and channel conductance were reduced, respectively, by 33% and 32% in MCM-NICD-GFP cells (Fig. 3). I_Na activation and reactivation properties were comparable in the two groups, but steady-state inactivation was shifted toward more positive potentials in cells overexpressing NICD. Thus, reduced I_Na density may account, at least in part, for the slower depolarization rate and attenuated AP amplitude in cells with active Notch signaling.

Fig. 3.

The Notch1 intracellular domain (NICD) reduces fast Na⁺ current (I_Na) in adult myocytes. A: whole cell voltage-gated I_Na recorded in voltage clamp in a green fluorescent protein (GFP)-negative myocyte isolated from a wild-type (WT) mouse (MCM-WT) and a GFP-positive cell obtained from a transgenic mouse with NICD expression (MCM-NICD-GFP). The voltage-clamp protocol consisted of a family of depolarizing steps in 5-mV increments. B: current-voltage (I–V) relations, voltage dependency of conductance, steady-state activation, steady inactivation, and time dependency of reactivation for I_Na in GFP-negative myocytes from MCM-WT mice (WT, n = 14–16; from 3 animals) and GFP-positive myocytes from MCM-NICD-GFP mice (NICD, n = 11–14; from 4 animals) are shown as means ± SE. g_max, maximal conductance. *P < 0.05 vs. WT. Fitting parameters are shown in Table 2.

Table 2.

Fitting parameters for I_Na properties

		gmax, mS/µF	V1/2G, mV	kG, mV	R2	P
INa conductance
WT		0.54 ± 0.02	−43.9 ± 0.7	3.61 ± 0.6	0.81	<0.0001
NICD		0.38 ± 0.01	−42.4 ± 0.8	3.98 ± 0.7	0.79
INa activation
WT			−42.8 ± 0.3	4.1 ± 0.3	0.94	0.559
NICD			−42.0 ± 0.4	4.3 ± 0.4	0.89
INa steady-state inactivation
WT			−74.3 ± 0.2	4.8 ± 0.2	0.98	0.0419
NICD			−70.9 ± 0.2	4.4 ± 0.2	0.98

Values are means ± SE. I_Na, fast Na⁺ current; g_max, maximal conductance; V_1/2G, potential at which conductance is halfway between 0 and g_max; k_G, slope of the conductance curve; WT, wild type; NICD, Notch1 intracellular domain.

Notch1 reduces Kv currents in adult myocytes.

Voltage-gated outward K⁺ Kv currents are the primary determinants of AP repolarization (57, 69), and, therefore, voltage-clamp experiments were performed to test the influence of Notch signaling on Kv current densities. Specifically, the amplitudes of I_to, I_K,slow1, I_K,slow2, and I_ss were studied by fitting the decay phase of the total outward K⁺ current elicited by a 25 s-depolarizing step (17, 48, 69). The analysis was performed in 1) GFP-negative myocytes from MCM-WT mice, 2) GFP-NICD-positive myocytes from MCM-NICD-GFP mice, and 3) GFP-positive myocytes from animals with tamoxifen-inducible MerCreMer transgene-driving myocyte-restricted expression of GFP (MCM-GFP). The latter group of cells was introduced to evaluate the potential effects of the Cre recombinase system on K⁺ currents. Quantitatively, I_to and I_K,slow1 components of Kv currents were significantly reduced in NICD-expressing myocytes, whereas I_K,slow2 and I_ss were comparable in the three groups of cells (Fig. 4).

Fig. 4.

The Notch1 intracellular domain (NICD) reduces Kv current components of adult myocytes. A: whole cell voltage-gated K⁺ currents recorded in voltage clamp in a green fluorescent protein (GFP)-negative myocyte isolated from a wild-type (WT) mouse (MCM-WT) and a GFP-positive cell obtained from a transgenic mouse with NICD expression (MCM-NICD-GFP). The voltage-clamp protocol consisting of a 25 s-depolarizing step is shown at the bottom. B: fitted curves of current traces shown in A with multiexponential function (black curves) and monoexponential fitted components (shades of gray curves). For the MCM-NICD-GFP myocyte, a biexponential function was used because this cell lacked transiet outward K+ current (I_to). C: quantitative data for the amplitude of Kv current components in GFP-negative myocytes from MCM-WT mice (WT, n = 17; from 3 animals), GFP-positive myocytes from a MCM-GFP mouse (GFP, n = 11), and GFP-positive myocytes from MCM-NICD-GFP mice (NICD, n = 18; from 4 animals) are shown as means ± SE. I_to, K⁺ currents with rapid activation; I_K,slow1 and I_K,slow2, intermediate and slow kinetics of inactivation; I_ss, sustained (noninactivating) component. *P < 0.05 vs. WT; **P < 0.05 vs. GFP.

The transient outward current (I_to) can be further separated in components with fast (I_to,f) and slow (I_to,s) inactivation kinetics (17, 23, 25, 48, 83). Thus, we implemented additional tests to determine which of these currents were affected by the activation of Notch signaling. For this purpose, we studied cells obtained from the intraventricular septum, the anatomic region of the heart where myocytes present both I_to,f and I_to,s (17, 83). Specifically, the decay phases of outward currents evoked by a 4.5 s-depolarization steps were analyzed with a triexponential function to dissect I_to,f, I_to,s, and I_K,slow based on their distinct time constants of decay, which differ by an order of magnitude (23, 25, 83). Both I_to,f and I_to,s were substantially reduced in NICD-expressing myocytes compared with WT myocytes, contributing in a similar manner to the overall reduction of I_to. Additionally, NICD-expressing cells presented diminished I_K,slow, whereas I_ss was not affected (Fig. 5).

Fig. 5.

The Notch1 intracellular domain (NICD) affects fast and slow components of the transient outward current (I_to). A: whole cell voltage-gated K⁺ currents recorded in voltage clamp in a green fluorescent protein (GFP)-negative myocyte isolated from the intraventricular septum of a wild-type (WT) mouse (MCM-WT) and a GFP-positive cell from the intraventricular septum of a transgenic mouse with NICD expression (MCM-NICD-GFP). The voltage-clamp protocol consisting of a 4.5 s-depolarizing step is shown at the bottom. B: fitted curves of current traces shown in A with triexponential function (black curves) and relative monoexponential fitted components (shades of gray curves). C: quantitative data for the amplitude of Kv current components in GFP-negative myocytes from the ventricular septum of MCM-WT mice (WT, n = 18; from 3 animals) and GFP-positive myocytes from the ventricular septum of MCM-NICD-GFP mice (NICD, n = 17; from 5 animals) are shown as means ± SE. *P < 0.01 vs. WT.

To better define the nature of K⁺ currents altered by Notch signaling activation, myocytes from WT and NICD-overexpressing mice were exposed to 4-aminopyridine, a blocker of Kv1 and Kv4 channel subunits (48, 83). K⁺ currents were elicited using 4.5-s depolarization steps before and after application of 5 mM 4-aminopyridine; the amplitude of the 4-aminopyridine-sensitive current was evaluated for the two groups of cells by subtracting currents in the presence of the drug from currents obtained in baseline condition (Fig. 6, A–C). For both cell types, exposure to 4-aminopyridine eliminated I_to,f and I_to,s and reduced I_K,slow, whereas no major changes were observed for I_ss (Fig. 6D). In the presence of the drug, the component of I_K,slow not affected by 4-aminopyridine remained larger in WT myocytes with respect to NICD myocytes. Importantly, quantification of the difference current before and after exposure to the blocker indicated that the 4-aminopyridine-sensitive components of I_to,f, I_to,s, and I_K,slow were reduced in myocytes with active NICD (Fig. 6E). Thus, Notch signaling activation reduces Kv currents that are sensitive to 4-aminopyridine (i.e., I_to,f, I_to,s, and part of I_K,slow) or resistant to this blocker (i.e., a component of I_K,slow).

Fig. 6.

The Notch1 intracellular domain (NICD) affects 4-aminopyridine (4-AP)-sensitive currents. A: whole cell voltage-gated K⁺ currents recorded in voltage clamp in a green fluorescent protein (GFP)-negative myocyte isolated from a wild-type (WT) mouse (MCM-WT) and a GFP-positive cell obtained from a transgenic mouse with NICD expression (MCM-NICD-GFP) before (Baseline) and after progressive exposure to 5 mM 4-AP. B: 4-AP-sensitive (Difference) current obtained by subtracting current traces before and after exposure to 4-AP shown in A. C: fitted curves of difference current traces shown in B with triexponential function (black curves) and relative monoexponential fitted components (shades of gray curves). D: quantitative data for the amplitude of Kv current components before (left) and after (right) exposure to 5 mM 4-AP for GFP-negative myocytes from MCM-WT mice (WT, n = 14; from 3 animals) and GFP-positive myocytes from MCM-NICD-GFP mice (NICD, n = 9; from 2 animals) are shown as means ± SE. *P < 0.05 vs. WT. E: quantitative data for the amplitude of 4-AP-sensitive (Difference) currents for myocytes reported in D. F: current-voltage (I–V) relations obtained by 10 s-depolarizing ramps from −110 to +80 mV. Data for MCM-WT mice (WT, n = 16; from 5 animals) and GFP-positive myocytes from MCM-NICD-GFP mice (NICD, n = 20; from 5 animals) are shown as means ± SE. Inward current at negative potential reflects inward rectifier K⁺ current.

To test whether activation of Notch signaling had any consequences on I_K1, which controls resting V_m in ventricular myocytes, slow depolarizing ramps were used (18, 66). No differences were observed between WT and NICD myocytes (Fig. 6F) in the range of potential where I_K1 is operative. Thus, Notch signaling reduces Kv current components with fast and intermediate inactivating kinetics, contributing, at least in part, to the protracted repolarization phase of the AP.

Notch interferes with Kv currents and KChIP2 expression in neonatal myocytes.

The expression of Notch receptors is progressively reduced in postnatal hearts (22), whereas outward K⁺ currents gradually appear in maturing cardiomyocytes, resulting in a progressive shortening of the AP duration (19, 34, 78). To test the possibility of a causative link between Notch activation and electrophysiological properties of developing myocytes, Notch signaling was perturbed using an antagonist of γ-secretase, the enzyme responsible for the cleavage of the Notch receptor and subsequent translocation of active NICD to the nucleus (14, 16). The γ-secretase inhibitor DAPT was used based on the ability of this compound to interfere with Notch signaling in various cell types under in vitro conditions (1, 38, 45, 77, 84). As previously reported (22), transcripts for the Notch1 receptor and downstream targets of Notch signaling were highly expressed in neonatal hearts compared with adult hearts (Fig. 7A). Thus, neonatal cells obtained from C57Bl/6 WT mice at postnatal day 2 were cultured and treated with DAPT or vehicle alone (DMSO). Transcripts of the Notch1 receptor and Hey1 gradually declined when neonatal myocytes were cultured for 1 or 5 days in the presence of DAPT or DMSO, but the γ-secretase inhibitor further decreased expression of Hey1, indicating that this compound effectively interferes with Notch signaling (Fig. 7B). Thus, we used this approach to modulate Notch signaling in cardiomyocytes and test the effects of this intervention on the expression of Kv channel subunit and Kv currents. Quantitatively, compared with vehicle-treated cells, levels of Kcna4, Kcna5, Kcnd2, and Kcnd3, which encode the K⁺ channel subunits Kv1.4, Kv1.5, Kv4.2, and Kv4.3, respectively, were not affected by DAPT treatment. Conversely, levels of Kcnip2 and relative isoforms Kcnip2a, Kcnip2b, and Kcnip2c, which encode KChIP2, modulators of I_to (35, 59, 79), were increased in the group of cells with perturbed Notch signaling (Fig. 7C).

Fig. 7.

Inhibition of Notch1 intracellular domain (NICD) formation in neonatal myocytes alters their gene expression profile. A: transcript levels, by RT-PCR, for the Notch1 receptor and downstream target genes Hes1 and Hey1 in the myocardium of C57Bl/6 mice at 5–7 days (Neonate, n = 4) and 3.8 mo (Adult, n = 4) of age. *P < 0.05 vs. neonate. B: transcript levels for the Notch1 receptor and its downstream target gene Hey1 relative to a preparation of isolated neonatal cardiomyocytes before [control (Ctrl)] and after culture with a γ-secretase inhibitor (DAPT) or vehicle (DMSO) for 1 or 5 days. C: average changes for the expression of transcripts in cultured neonatal myocytes treated with the γ-secretase inhibitor (DAPT) relative to vehicle-treated cells (dashed line) obtained from the same isolations (n = 11 cell cultures for each condition). Cells were cultured for 1–6 days after the initial treatment. Data are shown as means ± SE. *P < 0.05 DAPT- vs. vehicle-treated cells.

The effects of Notch signaling inhibition on Kv currents were established at 1, 2, and 3−7 days after treatment with DAPT. By patch clamp, the fraction of neonatal myocytes presenting fast activating and rapid inactivating K⁺ currents was progressively elevated with culture time and was twofold greater in myocytes treated with DAPT compared with vehicle-treated myocytes (Fig. 8, A and B). Additionally, overall Kv current density was severalfold larger in cells exposed to the γ-secretase inhibitor relative to control, but the amplitude of the sustained component (I_ss) was comparable (Fig. 8C). These results point to a positive correlation between activated KChIP2 transcription and enhanced I_to densities in myocytes with reduced Notch signaling.

Fig. 8.

Inhibition of Notch1 intracellular domain (NICD) formation in neonatal myocytes promotes Kv currents. A: whole cell voltage-gated K⁺ currents recorded in voltage clamp in neonatal myocytes in cultures treated with vehicle alone (DMSO) or a γ-secretase inhibitor (DAPT). Currents were elicited by a family of depolarizing steps from a holding potential (V_h) of −70 mV. B: fraction of myocytes presenting outward K⁺ Kv components in the groups of cells treated with vehicle (DMSO, n = 20 at 1–2 days; n = 19 at 3–7 days) or γ-secretase inhibitor (DAPT, n = 17 at 1–2 days; n = 27 at 3–7 days). The numbers of cells expressing or negative for Kv currents are shown in the graphs. *P < 0.05 vs. DMSO. C: quantitative data for the rapidly activating outward Kv currents in cells reported in B. *P < 0.05 vs. DMSO. D: quantitative data for the sustained current (I_ss) measured during the last 50 ms of depolarizing steps in cells shown in B and C.

Notch interferes with KChIP2 expression in adult myocytes.

Based on in vitro findings, expression of KChIP2 was evaluated in the heart of transgenic mice with active NICD to substantiate the possibility of a causative link between Notch1 activation and modulation of KChIP2 expression. By immunohistochemistry, KChIP2 was homogeneously distributed in the myocardium of MCM-WT mice, whereas levels of this protein were variably localized in the ventricular tissue of MCM-NICD-GFP mice. Specifically, levels of KChIP2 were reduced in GFP-positive myocytes with respect to GFP-negative myocytes (Fig. 9). Thus, Notch signaling activation represses KChIP2 expression.

Fig. 9.

Notch signaling activation reduces Kv channel-interacting protein 2 (KChIP2) expression in cardiomyocytes. A: KChIP2 (white) expression in the left ventricular (LV) myocardium of a MCM-WT mouse. Nuclei are stained by DAPI. Autofluorescence (AF) in the green channel was collected to visualize the structure of myocardial tissue. The green channel is omitted at right. B: green fluorescent protein (GFP; green) and KChIP2 (white) expression in the LV myocardium of a MCM-NICD-GFP mouse. Nuclei are stained by DAPI. The green channel is omitted at right. C: quantification of KChIP2 levels in myocytes in LV tissue of MCM-WT mice [wild type (WT), n = 2 mice] and MCM-NICD-GFP mice [Notch1 intracellular domain (NICD), n = 3 mice]. For each acquired image, the intensity of the fluorescent signal (F) from individual myocytes was normalized with respect to the signal of the myocyte presenting the highest fluorescent intensity (F_max). Normalized data are presented as F/F_max. *P < 0.001 vs. WT. D: quantification of relative KChIP2 levels in NICD-GFP-negative and NICD-GFP-positive myocytes in the myocardium of the MCM-NICD-GFP mice (NICD) shown in C. *P < 0.001 vs. NICD-GFP negative.

DISCUSSION

The results of the present study indicate that Notch signaling controls the electrophysiological properties of cardiomyocytes by negatively regulating Kv and I_Na densities. Also, our findings support the possibility that the modulatory action of NICD on the electrical behavior of neonatal and adult cells involves, at least in part, repression of KChIP2 gene expression.

The ability of cleaved NICD to modify the cellular electrophysiology of mouse ventricular myocytes has been previously reported (63, 64), but the ionic basis for the protracted electrical recovery found in cells with Notch activation was not clarified. A close resemblance of the AP properties of Notch-positive cells and myocytes of the cardiac conduction system was noted, together with the identification of a conduction-like phenotype of the transcriptional profile of newborn myocytes with ectopic Notch expression (64). More recently, additional studies have provided novel information on the mechanisms of action of Notch signaling on the electrophysiological aspects of ventricular and atrial myocytes, which include the regulation of Kv currents (40, 61). Our results complement these findings and strengthen the notion that Notch signaling has the ability to modulate the electrical behavior of cardiomyocytes during development and adult life.

Protracted electrical repolarization is a typical feature of neonatal cells (78, 81), which is also observed in myocytes of the postinfarcted, hypertrophied, failing, and aged heart (2–4, 27, 28, 32, 37, 58, 69, 73, 76). Thus, common gene regulatory networks active in the fetal and diseased myocardium may account for these phenotypic similarities. The finding that Notch levels decline postnatally and are partly restored after ischemic insults (22, 44), hypertrophy (10), or heart failure (40) favors the possibility that this receptor contributes to the electrophysiological adaptations of cardiomyocytes during development and with intervening pathologies. Thus, additional studies are necessary to define a causative relationship between Notch expression and electrical remodeling in myocytes under diseased conditions. Moreover, abundance of Notch ligands may represent an important variable in the electrophysiological alterations occurring in the damaged heart and this aspect also requires further clarification.

Our data indicate that Notch signaling maintains neonatal myocytes in an immature electrophysiological state and has the capability to remodel the AP of adult cells. Prolongation of the AP under conditions of increased workload of the myocardium represents an adaptive mechanism of inotropic support (5, 39, 68, 69, 80), whereas long AP in myocytes during the perinatal period seems the necessary condition to sustain sarcolemmal Ca²⁺ entry for cell contraction, in view of the immature t-tubular/sarcoplasmic reticulum systems of developing cells (43). Thus, by modulating the electrical properties of myocytes, Notch signaling may have important consequences on myocardial mechanical function.

Outward Kv currents in cardiomyocytes comprise components that differ in biophysical and pharmacological properties: I_to,f (encoded by subunits Kv4.2 and Kv4.3), I_to,s (encoded by Kv1.4), I_K,slow (encoded by Kv1.5 and Kv2.1), and I_ss (based on the two-pore domain K⁺ channel TASK1) (26, 48, 50, 51, 74, 82). The analysis of the decay of K⁺ currents and the use of 4-aminopyridine in WT and NICD myocytes allowed us to identify that I_to,f, I_to,s, and I_K,slow are critically reduced by Notch1 signaling. We found that KChIP2 levels declined in NICD-positive myocytes, constituting, at least in part, the molecular basis for the decrease of I_to,f. In fact, KChIP2 was reported to coimmunoprecipate (24, 47), stabilize Kv4.2 and Kv4.3 proteins, and generate I_to,f in myocytes (17, 46, 74). However, KChIP2 does not modulate I_to,s (17, 74), and, although it reduces Kv1.5-based current in a heterologous expression system (47), there is no indication that KChIP2 interferes with the 4-aminopyridine-resistant component of I_K,slow, which was found to be declined in NICD myocytes. Thus, the molecular basis for the observed reduction of Kv currents in cardiomyocytes in the presence of Notch activation remains to be defined. Interestingly, silencing of KChIP2 reduces the expression of the Na⁺ channel protein Nav1.5 (15, 56), a factor that may have contributed to the decline in I_Na density, attenuation of AP amplitude, and slower rate of depolarization in NICD-expressing cells.

Four Notch proteins (Notch1–Notch4) have been identified in mammals (42), raising the possibility that effects observed with γ-secretase inhibition in neonatal myocytes may have resulted from perturbation of multiple receptors. Therefore, caution has to be exercised when interpreting the consequences of the pharmacological modulation of Notch signaling in neonatal cells and overexpression of the Notch1 intracellular domain in adult transgenic myocytes. Specifically, whether the reduction of Kv currents is mediated exclusively by Notch1 activation or is achieved by the redundant action of Notch2–Notch4 signaling remains to be established. In this regard, a mutation of Notch3, which has been linked to a hereditary form of cerebral small vessel disease (CADSIL syndrome) (36), led to upregulation of Kv1 channel current density in smooth muscle cells from rodent cerebral arteries (13). This condition closely mimics the effects of γ-secretase inhibition in neonatal myocytes and supports the possibility of redundancy in the role of Notch1–Notch4 signaling.

Study limitations.

Kv current and I_Na were initially evaluated in NICD-expressing myocytes based on consistent differences in the early repolarization phase of the AP and reduced upstroke amplitude and velocity with respect to WT cells. Although our results indicate that conductance of Kv and Na⁺ channels is critically modulated by Notch signaling, we cannot exclude with certainty that other ionic currents are also affected. In this regard, KChIP2 expression alters L-type Ca²⁺ current by binding to the NH₂-terminal inhibitory module of Cav1.2 (75) and influencing Cav1.2 protein expression (56), a cascade of events that may have occurred in NICD-expressing myocytes.

Conclusions.

Our results support the notion that Notch signaling represents an important regulator of the physiological properties of neonatal and adult cardiomyocytes. Studies aimed at clarifying the sequence of events associated with Notch receptor activation and defining the role of this signaling axis on the electrical remodeling of myocytes in pathological conditions are warranted.

GRANTS

This work was supported by National Institute of Health Grants R01-HL-091021, R01- HL-114346, P01-AG-043353, P01-HL-092868, R01-AG-037495, R01-HL-111183, R01-AG-037490, R01-HL-105532, and R37-HL-081737 and intramural resources at New York Medical College (NYMC) including funds from the NYMC Translation Science Institute.

DISCLOSURES

P. Anversa is a member of Autologous, LLP, and P. Anversa and A. Leri are members of AAL Scientifics Corporation. All other authors have nothing to disclose.

AUTHOR CONTRIBUTIONS

G.B., C.A.E., S.S., A.S., K.K., A.A.-V., J.G.E., M.N., K.Q., L.M.E., and M.R. performed experiments; G.B., S.S., A.S., K.K., and M.R. analyzed data; G.B., S.S., A.S., and M.R. interpreted results of experiments; G.B., S.S., L.M.E., and M.R. prepared figures; G.B., S.S., and M.R. drafted manuscript; G.B., C.A.E., S.S., A.S., K.K., A.A.-V., J.G.E., M.N., K.Q., D.S., P.G., A.L., P.A., L.M.E., J.T.J., T.H.H., and M.R. edited and revised manuscript; G.B., C.A.E., S.S., A.S., K.K., A.A.-V., J.G.E., M.N., K.Q., D.S., P.G., A.L., P.A., L.M.E., J.T.J., T.H.H., and M.R. approved final version of manuscript; M.R. conceived and designed research.