Properties of WT and mutant hERG K+ channels expressed in neonatal mouse cardiomyocytes

Eric C. Lin, Katherine M. Holzem, Blake D. Anson, Brooke M. Moungey, Sadguna Y. Balijepalli, David J. Tester, Michael J. Ackerman, Brian P. Delisle, Ravi C. Balijepalli, Craig T. January

Abstract

Mutations in human ether-a-go-go-related gene 1 (hERG) are linked to long QT syndrome type 2 (LQT2). hERG encodes the pore-forming α-subunits that coassemble to form rapidly activating delayed rectifier K+ current in the heart. LQT2-linked missense mutations have been extensively studied in noncardiac heterologous expression systems, where biogenic (protein trafficking) and biophysical (gating and permeation) abnormalities have been postulated to underlie the loss-of-function phenotype associated with LQT2 channels. Little is known about the properties of LQT2-linked hERG channel proteins in native cardiomyocyte systems. In this study, we expressed wild-type (WT) hERG and three LQT2-linked mutations in neonatal mouse cardiomyocytes and studied their electrophysiological and biochemical properties. Compared with WT hERG channels, the LQT2 missense mutations G601S and N470D hERG exhibited altered protein trafficking and underwent pharmacological correction, and N470D hERG channels gated at more negative voltages. The ΔY475 hERG deletion mutation trafficked similar to WT hERG channels, gated at more negative voltages, and had rapid deactivation kinetics, and these properties were confirmed in both neonatal mouse cardiomyocyte and human embryonic kidney (HEK)-293 cell expression systems. Differences between the cardiomyocytes and HEK-293 cell expression systems were that hERG current densities were reduced 10-fold and deactivation kinetics were accelerated 1.5- to 2-fold in neonatal mouse cardiomyocytes. An important finding of this work is that pharmacological correction of trafficking-deficient LQT2 mutations, as a potential innovative approach to therapy, is possible in native cardiac tissue.

  • human ether-a-go-go-related gene 1
  • long QT syndrome 2
  • protein trafficking
  • E4031

human ether a-go-go-related gene 1 (hERG or Kv11.1) K+ channels are expressed in multiple tissue and cell types, including cardiac muscle (12). In the heart, hERG channels generate their maximal repolarizing current during phases 2 and 3 of the cardiac action potential to restore the membrane to the resting potential and to govern the duration of the surface ECG QT interval.

Curran and collaborators (12) first identified mutations in hERG channels linked to congenital long QT syndrome (LQT2). The hERG channel is also the target for most drugs causing acquired long QT syndrome (35). Mutations and drug block cause a loss-of-function phenotype in hERG current (IhERG). Several molecular mechanisms have been proposed for the loss-of-function phenotype in LQT2, including abnormalities in ion channel protein trafficking, abnormalities in ion channel gating or permeation, and nonsense-mediated decay of mRNA (3, 14, 20). The most common type of mutation in LQT2 is a single nucleotide change leading to a single amino acid substitution (missense mutation). Most missense mutant channels fail to traffic normally to the surface membrane; rather, they are retained in the endoplasmic reticulum as mutant proteins (3). Furthermore, for many LQT2 missense mutations, the trafficking-deficient phenotype can be improved (“rescue”) by several approaches to increase IhERG (3).

Previous investigations of LQT2 mutations have almost entirely used noncardiac, heterologous overexpression systems [Xenopus oocytes, human embryonic kidney (HEK)-293 cells, Chinese hamster ovary cells, etc.]. A few transgenic animal models have also been created for specific mutations (5, 8, 23, 37). In the present work, we studied the biochemical, biophysical, and pharmacological properties of wild-type (WT) hERG channels and three LQT2-linked channels expressed in native neonatal mouse cardiomyocytes. One mutation, ΔY475 hERG, is a single tyrosine deletion in the S2–S3 linker of the channel protein that has not been previously functionally characterized (38). The other two, N470D and G601S hERG, are missense mutations that have previously been reported to be trafficking deficient when expressed in noncardiac mammalian systems, whereas in Xenopus oocytes they traffic to the membrane to produce large IhERG. This is the first report to study the overexpression of LQT2 mutations in transfected native cardiomyocytes.

MATERIALS AND METHODS

DNA constructs and site-directed mutagenesis.

hERG cDNA (WT or mutant hERG1a) was subcloned into the pcDNA3 vector (Invitrogen, Carlsbad, CA) (47). WT hERG with appropriate nucleotide changes (ΔTAC2342–2344) resulting in ΔY475 hERG was generated using site-directed mutagenesis and verified by DNA sequencing. G601S and N470D hERG were generated as previously described (16, 45). hERG cDNAs for neonatal mouse cardiomyocyte experiments were generated as endotoxin free using the EndoFree Plasmid Maxi Kit (Qiagen, Valencia, CA).

Isolation of neonatal mouse cardiomyocytes.

129 SVJ mouse cardiomyocytes were obtained from 1- to 3-day-old neonatal pups with ∼20 pup hearts furnishing sufficient tissue for myocyte isolation. The animal care and use protocol was approved by the University of Wisconsin-Madison Research Animal Resources Center and met National Institutes of Health guidelines for the health and well being of the animals. Briefly, hearts from neonatal mice were rapidly excised and washed to remove blood and debris. Whole hearts were then minced and dissociated into single cardiomyocytes by digestion five times with gentle stirring in HBSS solution (Mediatech, Manassas, VA) containing collagenase type II (100 μg/ml, Invitrogen) and pancreatin (1 mg/ml, Sigma) enzymes. Cardiomyocytes were centrifuged, washed, and then preplated (3 times, 1 h each to remove nonmuscle cells and cell debris) in growth medium (see below). The isolation procedure has been previously described (11). Untransfected neonatal mouse cardiomyocytes showed spontaneous beating within 2 days after the isolation procedure, and this could be observed for up to 10 days of cell culture.

Neonatal mouse cardiomyocyte and HEK-293 cell transient transfection.

cDNA (3–5 μg; WT, ΔY475, G601S, or N470D hERG or empty pCDNA3 vector) and 1 μg cDNA of green fluorescent protein (GFP) were transfected into neonatal mouse cardiomyocytes by an electroporation technique (7, 13, 30) using a Rat Cardiomyocyte-Neonatal Nucleofection Kit (Amaxa Biosystems, Cologne, Germany). HEK-293 cells were transfected using SuperFect (Qiagen) as previously described (3).

Cell culture and drug exposure.

Transfected neonatal mouse cardiomyocytes were seeded on laminin-coated coverslips (for patch-clamp and immunofluorescence analysis) or directly on laminin-coated plates (for Western blot analysis) and cultured at 37°C in 5% CO2 in complete growth medium (DMEM, Invitrogen) supplemented with medium 199 (1.6%, Invitrogen), l-glutamine (2 mM, Invitrogen), 100× penicillin (1.0%, Invitrogen), horse serum (10%, Invitrogen), and newborn calf serum (5%, Invitrogen). After 24 h, cardiomyocytes were recultured in maintenance medium (DMEM) supplemented with medium 199 (1%), 100× l-glutamine (1%), 100× penicillin (1%), horse serum (0.5%), and newborn calf serum (0.5%). Cardiomyocytes were cultured for up to 72 h in maintenance medium. For HEK-293 cells, culture conditions were the same as previous described (47). In some experiments, E4031 (10 μM, Sigma, 10 mM stock dissolved in distilled water) was added to the maintenance media for 24 h before study.

Immunofluorescence microscopy.

Cardiomyocytes were fixed and permeabilized with 4% paraformaldehyde for 20 min at room temperature and blocked with a buffer containing 5% goat serum, 0.2% Triton X-100, and 0.05% azide in PBS. To detect the contractile protein actin, cardiomyocytes were incubated with a monoclonal mouse anti-sarcomeric actin antibody (1:1,000, Sigma-Aldrich) that was counterstained with Alexa 405 (Molecular Probes, Eugene, OR). To detect hERG protein, cardiomyocytes were incubated with anti-hERG antibody as previously described (6). The immunofluorescence was viewed with a Nikon fluorescence microscope (7).

Western blot analysis.

Neonatal mouse cardiomyocytes from similarly confluent cultures were used to isolate cellular proteins. Cardiomyocytes were rinsed with PBS, treated with 50 μL Nonidet P-40 lysis buffer, scraped off into 1.5-ml eppendorf tubes, and centrifuged at 13,200 rpm at 4°C for 10 min to obtain a pellet. The protein was loaded on SDS-polyacrylamide gels for electrophoresis and transferred onto nitrocellulose membranes. Anti-hERG antibody against a COOH-terminal epitope was applied to the nitrocellulose membrane to detect hERG protein. The method was similar to that previously described for HEK-293 cells (45).

Image density analysis of hERG protein detected by Western blot analysis was performed using Bio-Rad GS 700 Imaging Densitometer and Molecular Analyst software (Bio-Rad Laboratories, Hercules, CA) (31). Using this software, a box of constant size was placed around each control and experimental 155- and 135-kDa band as well as an empty lane for background normalization. Each 135- and 155-kDa density reading was background subtracted to give normalized 135- and 155-kDa band densities. For each Western blot, a density ratio value (normalized 155 kDa/normalized 135 kDa) was generated and averaged.

Patch clamping.

Functional analysis was done using the whole cell patch-clamp technique at room temperature (22–23°C) within 3 h after cells were removed from culture conditions. Pipettes had resistance between 1.0 and 2.5 mΩ when filled with intracellular solution, which contained (in mM) 130 KCl, 1 MgCl2, 5 EGTA, 5 MgATP, and 10 HEPES (pH 7.2 with KOH). The extracellular solution was HEPES-buffered Tyrode solution, which contained (in mM) 137 NaCl, 4 KCl, 1.8 CaCl2, 1 MgCl2, 10 glucose, and 10 HEPES (7.4 with NaOH). Axopatch-1D and Axopatch 200 amplifiers (Molecular Devices, Sunnyvale, CA) were used to record membrane currents. The average cell capacitance for all neonatal mouse cardiomyocytes was 14.7 ± 1.9 pF (n = 52) and for all HEK-293 cells was 11.2 ± 0.5 pF (n = 19).

Statistics.

Origin 7.5 software (OriginLab, Northampton, MA) was used to fit peak tail current (Imax) amplitudes to a Boltzmann distribution in the following form: tail current = ImaxImax/{1 + exp[(VtV½)/k]}, where Vt is the test potential, V½ is the voltage at which Imax is half maximal, and k is the slope factor. Deactivation kinetic measurements were performed as previously described (4). Data are given as means ± SE. Student's t-test was used for statistical analysis with P < 0.05 considered significant.

RESULTS

We first characterized the electroporation technique for our isolated neonatal mouse cardiomyocytes. Figure 1 shows images of isolated cultured neonatal mouse cardiomyocytes. Figure 1, A and B, shows low-power phase-contrast images of the same field of the cardiomyocytes 72 h after electroporation of GFP cDNA. The light cells in Fig. 1A represent spontaneously beating cardiomyocytes and the dark spots represent cell debris. The results shown in Fig. 1B demonstrate that many of the cardiomyocytes in Fig. 1A express GFP with varying intensity; with our approach, ∼30–40% of the cardiomyocytes expressed GFP. We next compared GFP and WT hERG cDNA transfection in neonatal mouse cardiomyocytes. Figure 1, C–F, shows high-power confocal images of three cardiomyocytes stained for the muscle protein actin (C), expressing GFP (D), stained for hERG protein (E), and the merged image (F). The images show staining for sarcomeric actin throughout the three cardiomyocytes, that the GFP-positive cardiomyocyte also expressed hERG protein, and that WT hERG protein is present throughout the cardiomyocyte. We selected GFP-positive cardiomyocytes for subsequent patch-clamp study, and in these experiments, >90% of the GFP-positive cardiomyocytes expressed large-amplitude IhERG.

Fig. 1.

Cultured neonatal mouse cardiomyocytes. A: phase-contrast image of isolated, cultured cardiomyocytes. B: phase-contrast image of the same field showing green fluorescent protein (GFP) expression. Higher-power confocal images (C–F) show three isolated neonatal mouse cardiomyocytes expressing sacromeric actin (C), GFP (D), wild-type (WT) hERG (E), and a merged image (F). Calibration bars = 50 μm in A and B or 5 μm in C–F.

We measured native rapidly activating delayed rectifier K+ current (IKr) in untransfected spontaneously beating cardiomyocytes. The voltage protocol and representative current traces are shown in Fig. 2A. From the holding potential of −60 mV, 2.5-s-long steps were applied to −50 to 30 mV in 10-mV increments and then repolarized to −40 mV for 2.5 s to generate a tail current, with each voltage step sequence repeated every 10 s. Depolarizing steps elicited a brief inward current consistent with Na+ and Ca2+ current followed by sustained outward current, and with the repolarizing step to −40 mV, a rapidly decaying outward tail current was present. Figure 2A, inset a, shows a family of capacitance-corrected inward current traces with a peak amplitude of nearly −800 pA; this inward current was observed in all cardiomyocytes. After the control currents had been recorded, E4031 (10 μM, n = 6) was added to block IKr. Figure 2A, inset b, shows two current traces for repolarizing voltage steps from 30 to −40 mV before and after the addition of E4031; the tail current was nearly completely abolished consistent with IKr. Figure 2A also shows the activation current-voltage (I-V) relation for peak tail IKr fit with a Boltzmann distribution. V½ was −11.5 ± 3.4 mV with a slope factor of 4.9 ± 0.7 mV/e-fold change, and the cell capacitance was 16.6 ± 3.1 pF (n = 5).

Fig. 2.

Currents in untranfected and transfected native cardiomyocytes. A: representative current traces from an untransfected cardiomyocyte. Scale = 25 pA by 0.5 s. The voltage-clamp protocol is shown above the currents. Inset a shows inward currents elicited with depolarization (scale = 200 pA by 0.05 s), and inset b shows block of tail current by E4031 (scale = 10 pA by 2 s). The activation relation for peak tail rapidly activating delayed rectifier K+ current is shown at the bottom. B and C: WT hERG-transfected cardiomyocytes. B: block of tail current by E4031. Scale = 50 pA by 2 s. C: a family of current traces and the activation relation for peak tail hERG current (IhERG). Scale = 200 pA by 2 s. Dotted line = 0 pA.

Figure 2, B and C, shows currents from neonatal mouse cardiomyocytes transfected with WT hERG. Figure 2B shows two current traces for a depolarizing step from −60 to 40 mV followed by a repolarizing step to −50 mV to elicit tail current. After the control trace had been recorded, E4031 was added to block native IKr and transfected IhERG (n = 4), confirming that the outward current is E4031 sensitive and that the tail current was nearly completely abolished (arrow). Figure 2C shows a family of current traces from another WT hERG-transfected neonatal mouse cardiomyocyte. From a holding potential of −60 mV, 5-s-long steps were applied to −80 to 40 mV in 10-mV increments and then repolarized to −50 mV for 5 s to generate a tail current, with each voltage step sequence repeated every 10 s (4). During depolarizing steps, outward current activated to reach a maximum amplitude that then decreased with further depolarization, consistent with hERG channel inactivation. With repolarization to −50 mV, a large-amplitude rapidly decaying outward tail current was present. Figure 2C also shows the activation I-V relation for Imax fit with a Boltzmann distribution. Tail current began to activate at −40 mV and was fully activated by 0–10 mV. When fit with a Boltzmann distribution, V½ was −14.6 ± 3.0 mV with a slope factor of 5.9 ± 0.3 mV/e-fold change, and the cell capacitance was 13.1 ± 1.5 pF (n = 4). Our experiments did not test for coassembly of hERG with native IKr channels. Compared with native IKr, the Imax density was ∼10-fold greater, it was blocked by E4031, and the V½ and slope factor values were similar. We conclude that the tail current in WT hERG-transfected neonatal mouse cardiomyocytes is composed almost entirely of IhERG.

We used Western blot analysis to study hERG channel protein processing in neonatal mouse cardiomyocytes. As shown in Fig. 3, lane 1, cardiomyocytes were transfected with empty pcDNA3 vector and showed protein bands at ∼165 and 205 kDa and additional bands at ∼95 and 114 kDa. These bands were consistently observed in all Western blot data. Cardiomyocytes transfected with WT hERG showed new protein bands at 135 and 155 kDa. Western blot analyses of WT hERG overexpression in several systems, including HEK-293 cells, have shown two protein bands at 135 kDa (immature) and 155 kDa (mature) that represent posttranslational core and complex glycosylation of hERG protein, respectively (2, 15, 19, 45). We next transfected the missense mutations, G601S and N470D hERG (12, 16, 36), that form trafficking-deficient channel proteins when studied in noncardiac mammalian systems (3, 16, 45). Western blot analysis in the cardiomyocytes showed that G601S and N470D hERG generate 135-kDa protein bands but only weak or no 155-kDa protein bands, indicating that for control conditions these two mutations generate immature protein that then does not undergo normal Golgi processing to the mature protein. E4031 has been suggested to act as a stabilizing ligand that can correct the protein folding and processing of many LQT2 trafficking-deficient hERG missense mutations, including G601S and N470D hERG, to increase surface membrane expression and IhERG (3, 46). After 24 h of incubation with 10 μM E4031, the 155-kDa band was increased for both G601S and N470D hERG. ΔY475 hERG has not been previously functionally characterized. Western blot analysis of ΔY475 hERG expressed in cardiomyocytes showed both the 135- and 155-kDa protein bands, indicating that this mutant protein undergoes posttranslational core and complex glycosylation, similar to WT hERG. Incubation for 24 h in 10 μM E4031 had little effect on the 135- and 155-kDa bands. Western blot analysis also showed that untransfected neonatal mouse cardiomyocytes lack the 135- and 155-kDa bands. Figure 3 also shows quantitative analyses (density ratio of the 155- to 135-kDa bands; see materials and methods) of the Western blots. The quantitative data show that the 155- and 135-kDa bands for WT hERG have similar densities (mean density ratio: 1.04). In contrast, for G601S and N470D hERG, the control density ratios were smaller, and with culture in E4031, these ratios increased (P < 0.05), suggesting that G601S and N470D hERG protein trafficking can undergo pharmacological correction by E4031 in cardiomyocytes. Quantitative analysis of the Western blot band density ratios of ΔY475 hERG for control conditions and after culture in E4031 confirmed that the ratio did not increase significantly (P > 0.05).

Fig. 3.

Western blot analysis for hERG channels in neonatal mouse cardiomyocytes. Cardiomyocytes transfected with empty pcDNA3 vector (lane 1) have protein bands of 165 and 205 kDa, consistent with mouse ERG1a (mERG1a), and 95 and 114 kDa, consistent with mERG1b (small arrows). Cardiomyocytes transfected with WT hERG (hERG1a; lane 2) have additional protein bands at 135 and 155 kDa (large arrows). Lanes 3 and 4 show G601S hERG and lanes 5 and 6 show N470D hERG without (−) or with (+) E4031 incubation. E4031 incubation resulted in increased density of the 155-kDa bands. Lanes 7 and 8 show ΔY475 hERG without or with E4031 incubation and show both 135- and 155-kDa bands. Lane 9 shows that untransfected cardiomyocytes lack the 135- and 155-kDa bands. Quantitative Western blot density ratio analysis of hERG protein is shown at the bottom. See text for detail.

We studied the cell electrophysiology of the LQT2 mutations expressed in neonatal mouse cardiomyocytes. Figure 4A shows representative IhERG traces for G601S, N470D, and ΔY475 hERG for control conditions and after culture in 10 μM E4031 for 24 h followed by drug washout for 1 h with maintenance media (see materials and methods). From a holding potential of −60 mV, a 5-s-long step was applied to 20 mV and then repolarized to −50 mV for 5 s to generate a tail current. For G601S and N470D hERG, the control current amplitudes were small, whereas after pharmacological correction by E4031, the amplitudes were markedly increased. In addition, N470D hERG generated outward current at the holding potential of −60 mV (see below) and had altered gating properties (34, 46). In contrast, for ΔY475 hERG, the control IhERG amplitude was larger and culture in E4031 caused only a small increase in IhERG. Peak tail IhERG was measured for G601S, N470D, and ΔY475 hERG mutations for control conditions and after pharmacological correction by E4031 (summarized in Fig. 4B). Imax values of G601S and N470D hERG after pharmacological correction by E4031 were significantly larger than control values (P < 0.05, n ≥ 9 for each group). In contrast, Imax of ΔY475 hERG was not statistically different from its control value (P > 0.05, n = 7 for each group). Figure 4C shows activation I-V relations for peak tail IhERG (n ≥ 5) fit with a Boltzmann distribution for pharmacologically corrected G601S, N470D, and ΔY475 hERG-transfected cardiomyocytes using the voltage protocol described in Fig. 2C. V½ values and slope factors for WT, G601S, N470D, and ΔY475 hERG are shown in Table 1.

Fig. 4.

G601S, N470D, and ΔY475 hERG currents in cardiomyocytes. A: representative current traces for control conditions and after pharmacological correction with E4031. Scale = 50 pA by 2 s. Dotted line = 0 pA. B: mean peak tail IhERG amplitudes for G601S, N470D, and ΔY475 hERG for control conditions and after pharmacological correction with E4031 (*P < 0.05). C: activation current-voltage relations for G601S, N470D, and ΔY475 hERG peak tail IhERG after pharmacological correction with E4031.

View this table:
Table 1.

hERG channel activation (V½ and slope factor) values in HEK-293 cell and neonatal mouse cardiomyocyte expression systems

ΔY475 hERG has not been previously functionally characterized in a noncardiac expression system; thus, we also expressed it in HEK-293 cells to permit a comparison with cardiomyocytes and compare it with WT hERG. Figure 5A shows current traces for control conditions and after culture for 24 h in E4031 (10 μM) followed by drug washout for 1 h in cardiomyocytes (left traces) and HEK-293 cells (right traces) expressing ΔY475 hERG. The voltage protocol is shown above the respective control current traces. Figure 5B shows the mean peak tail IhERG amplitudes recorded for WT and ΔY475 hERG channels in neonatal mouse cardiomyocytes and HEK-293 cells as well as the effect on ΔY475 hERG of 24 h of culture in 10 μM E4031 followed by 1 h of drug washout. Mean peak tail IhERG amplitudes for channels expressed in HEK-293 cells were significantly larger than the respective mean peak tail IhERG amplitudes for channels expressed in cardiomyocytes for both WT and ΔY475 hERG (P < 0.05, n ≥ 4), and ΔY475 hERG did not undergo pharmacological correction in HEK-293 cells, similar to the findings in cardiomyocytes. Figure 5C shows normalized activation I-V relations for peak tail IhERG (n ≥ 5) fit with a Boltzmann distribution for WT hERG- and control ΔY475 hERG- transfected cardiomyocytes compared with HEK-293 cells. V½ values and slope factors are shown in Table 1.

Fig. 5.

ΔY475 hERG currents in cardiomyocytes and human embryonic kidney (HEK)-293 cells. A: current traces from ΔY475 hERG expressed in cardiomyocytes (scale = 50 pA by 2 s) and HEK-293 cells (scale = 200 pA by 2 s). Dotted line = 0 pA. B: mean peak tail IhERG amplitudes for WT, control ΔY475, and ΔY475 hERG channels cultured in E4031 for 24 h followed by drug washout for 1 h and expressed in cardiomyocytes and HEK-293 cells (*P < 0.05, cardiomyocytes vs. HEK-293 cells). C: activation current-voltage relations for peak tail IhERG for WT and control ΔY475 hERG channels expressed in cardiomyocytes and HEK-293 cells.

Finally, we analyzed IhERG deactivation kinetics for the ΔY475 hERG mutation and WT hERG recorded after repolarizing steps from 20 to −50 mV in both neonatal mouse cardiomyocytes and HEK-293 cells to test whether these gating properties might be different between the cardiac and noncardiac expression systems (see Figs. 2C and 5). Tail currents were fit as a double-exponential process to give fast and slow deactivation rates (Table 2). We also tested whether we obtained similar kinetic measurements in the two expression systems. In HEK-293 cells compared with neonatal mouse cardiomyocytes (Table 2), the deactivation rates were slower for WT hERG (∼2-fold) and ΔY475 hERG (∼1.5-fold). Thus, small differences in deactivation kinetics were found between the two expression systems.

View this table:
Table 2.

Deactivation rates of WT and ΔY475 hERG in cardiomyocytes and HEK-293 cells at −50 mV

DISCUSSION

Central to understanding genetic mutations has been ion channel gene expression in several model systems, including mRNA injection into Xenopus oocytes, cDNA transfection (chemical, ionic, and electroporation) into noncardiac immortalized cell lines, and, in more limited studies, viral cDNA infection. In this report, we studied WT and LQT2 hERG channel α-subunit overexpression in a model of freshly isolated, cultured neonatal mouse cardiomyocytes. Neonatal mouse cardiomyocytes differ from adult mouse cardiomyocytes in several ways (1): neonatal cardiomyocytes are not fully terminally differentiated, they are small and have a rounded or triangular shape, they lack a well-developed sarcomeric contractile protein pattern, and they tolerate electroporation, which does not work well in older cardiomyocytes. Compared with adult mouse cardiomyocytes, neonatal mouse cardiomyocytes lack transverse tubules, which may have an advantage in voltage-clamp experiments (27). Neonatal mouse cardiomyocytes have been used as a model for ion channel expression, including for the Na+ channel Nav1.5 (39, 44), T- and L-type Ca2+ channels (17, 21), the K+ channel Kv4.2 (18), and the pacemaker HCN channel (9). The study of WT and mutant ion channels in native cardiac myocyte models is important since ion channel mutations have been reported to function differently in native myocyte and HEK-293 cell expression systems (26).

We focused on three LQT2 mutations along with WT hERG channels and native IKr. In neonatal mouse cardiomyocytes, native IKr was identified as an E4031-sensitve current that was small in amplitude (peak tail IKr ∼1.6 pA/pF), and our IKr recordings are similar to those in previous reports (24, 27, 40, 41). WT hERG channel overexpression generated an E4031-sensitive current that activated over a voltage range similar to IKr but with an amplitude ∼10-fold greater. ΔY475 hERG is located in the S2–S3 linker, and in both cardiomyocytes and HEK-293 cells, ΔY475 hERG traffics similar to WT hERG channels (135- and 155-kDa bands present) but gates at more negative voltages compared with WT hERG and native IKr. It also deactivates rapidly compared with WT hERG, and this increase in the rate of channel closing would be expected to decrease IhERG available for action potential repolarization [class 3 mechanism (14)]. In neonatal mouse cardiomyocytes transfected with G601S and N470D hERG, small-amplitude tail currents were recorded under control conditions, probably representing native mouse IKr. More importantly, in cardiomyocytes, both G601S and N470D hERG undergo pharmacological correction (“rescue”) by culture in E4031 to increase IhERG, and this is the first report demonstrating this property in a native cardiac myocyte expression system. With IhERG pharmacologically corrected, G601S hERG gates at voltages similar to WT hERG, whereas N470D hERG gates at more negative voltages, and the pharmacologically corrected tail current densities for G601S and N470D hERG are close to that of WT hERG. The voltage-dependent properties of the channel activation are shown in Table 1.

Western blot analysis of neonatal mouse cardiomyocytes showed two protein bands at ∼165 and 205 kDa and two additional protein bands at ∼95 and 114 kDa. These bands have previously been attributed to endogenous mouse ether-a-go-go-related gene 1 (mERG1) protein, which undergoes extensive posttranslational modification with the 165- and 205-kDa bands, representing mERG1a, and the 95- and 114-kDa bands, representing mERG1b (28, 42, 43). Compared with mERG1a protein, hERG channel protein has 96% sequence homology at the amino acid level (25). In our Western Blot analyses, transfection of WT hERG or the LQT2 mutants into neonatal mouse cardiomyocytes resulted in the expression of new protein bands at 135 and 155 kDa, similar to previous reports (2, 15, 19, 45) of hERG expression in noncardiac mammalian heterologous systems. For control conditions, WT hERG- and ΔY475 hERG-transfected cardiomyocytes generated both 135- and 155-kDa bands, whereas G601S hERG- and N470D hERG-transfected cardiomyocytes generated 135-kDa bands with weak or absent 155-kDa bands. These findings are consistent with the previous identification in HEK-293 cells of G601S and N470D hERG as trafficking-deficient hERG mutations. Culture in E4031 resulted in the enhancement of the 155-kDa bands for G601S and N470D hERG to further confirm that pharmacological correction occurs in the cardiomyocytes. Since missense mutations are predominant in LQT2, and abnormal channel trafficking is the common mechanism proposed for the loss of IhERG in these mutations (3), an important finding of this work is that pharmacological correction of trafficking-deficient LQT2 mutations, as a potential innovative approach to therapy, is possible in native cardiac tissue.

Differences in WT and ΔY475 hERG current were found between the cardiomyocyte and HEK-293 cell expression systems. The largest difference was in IhERG density, which was reduced nearly 10-fold in cardiomyocytes (peak tail IhERG of ∼12.5–15 vs. ∼110 pA/pF, respectively). This could result simply from the different transfection techniques as well as from differences in cellular mechanisms controlling gene transcription, protein processing, and channel gating in the two expression models. Consistent differences were also found in the deactivation rates of both WT and ΔY475 hERG between cardiomyocytes and HEK-293 cells, with the decay rates being 1.5- to 2-fold faster in cardiomyocytes. Possible explanations include differences in cell signaling pathways (22), membrane structure and lipid content (6), cytoskeleton interactions (10), α-subunit coassembly with β-subunits (e.g., KCNE2) (29), and that the neonatal mouse heart expresses the ERG1b (mERG1b) isoform that accelerates the deactivation kinetics of mERG1a (25, 32, 33). One implication of the faster deactivation kinetics found in cardiomyocyte expression is that occupancy of the open state of the hERG channel will be reduced during a cardiac action potential, and this may need to be considered in computational models of cardiac action potentials that use data from heterologous expression systems.

GRANTS

This work was supported by American Heart Association Predoctoral Training Award 0815624G (to E. Lin), by American Heart Association Grants SDG 0535068N (to B. P. Delisle) and SDG 0730010N (to R. C. Balijepalli), and by National Heart, Lung, and Blood Institute R01-HL-60723 (to C. T. January).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

ACKNOWLEDGMENTS

The authors thank Jing Wang for assistance with the mouse cardiomyocyte isolation procedure.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
View Abstract