Pore mutants of HERG and KvLQT1 downregulate the reciprocal currents in stable cell lines

Abstract

We previously reported a transgenic rabbit model of long QT syndrome based on overexpression of pore mutants of repolarizing K⁺ channels KvLQT1 (LQT1) and HERG (LQT2).The transgenes in these rabbits eliminated the slow and fast components of the delayed rectifier K⁺ current (I_Ks and I_Kr, respectively), as expected. Interestingly, the expressed pore mutants of HERG and KvLQT1 downregulated the remaining reciprocal repolarizing currents, I_Ks and I_Kr, without affecting the steady-state levels of the native polypeptides. Here, we sought to further explore the functional interactions between HERG and KvLQT1 in heterologous expression systems. Stable Chinese hamster ovary (CHO) cell lines expressing KvLQT1-minK or HERG were transiently transfected with expression vectors coding for mutant or wild-type HERG or KvLQT1. Transiently expressed pore mutant or wild-type KvLQT1 downregulated I_Kr in HERG stable CHO cell lines by 70% and 44%, respectively. Immunostaining revealed a severalfold lower surface expression of HERG, which could account for the reduction in I_Kr upon KvLQT1 expression. Deletion of the KvLQT1 NH₂-terminus did not abolish the downregulation, suggesting that the interactions between the two channels are mediated through their COOH-termini. Similarly, transiently expressed HERG reduced I_Ks in KvLQT1-minK stable cells. Coimmunoprecipitations indicated a direct interaction between HERG and KvLQT1, and surface plasmon resonance analysis demonstrated a specific, physical association between the COOH-termini of KvLQT1 and HERG. Here, we present an in vitro model system consistent with the in vivo reciprocal downregulation of repolarizing currents seen in transgenic rabbit models, illustrating the importance of the transfection method when studying heterologous ion channel expression and trafficking. Moreover, our data suggest that interactions between KvLQT1 and HERG are mediated through COOH-termini.

delayed rectifier K⁺ current (I_K) plays an important role in repolarizing the cardiac action potential. I_K is composed of two main distinct components, namely, rapidly activating I_Kr and slowly activating I_Ks (24). HERG was originally cloned from the hippocampus and has been identified as the pore-forming α-subunit for cardiac I_Kr (8, 31). I_Ks can be reproduced by coexpression of KvLQT1, the gene product of KCNQ1 and its β-regulatory subunit minK (3, 27). Long QT (LQT) syndrome is a heritable disorder associated with prolonged cardiac repolarization, ventricular tachycardia, syncope, and sudden cardiac death (25). Type 1 LQT syndrome (LQT1) is caused by loss-of-function mutations in the KCNQ1 (KvLQT1) gene (7, 23). Similarly, type 2 LQT syndrome (LQT2) results from loss-of-function mutations in the KCNH2 (HERG) gene (29). Since normal cardiac repolarization is the result of the intricate interplay of a number of K⁺ channels, it is conceivable that if one fails (e.g., due to mutation or drug exposure), the remaining K⁺ currents may still provide sufficient repolarization, as the concept of “repolarization reserve” proposes (26). Remodeling of K⁺ currents in the heart could be one mechanism that contributes to the repolarization reserve (32). For example, studies (5, 30, 32) using dog ventricular preparations or primary canine cardiomyocyte cultures demonstrated a functional interaction between I_Kr and I_Ks. In these cases, inhibition of I_Kr led to increased I_Ks, minimizing the elongation of the action potential. Additionally, Ehrlich et al. (9) showed that cotransfection of KvLQT1 with HERG upregulated the surface expression of HERG and I_Kr in a Chinese hamster ovary (CHO) cell-based transient expression system. However, coexpressed HERG had no effect on KvLQT1 cell surface expression or current properties (9). Finally, we (6) have recently reported two transgenic rabbit models for LQT1 and LQT2 using overexpressed dominant negative human pore mutants of KvLQT1 (KvLQT1-Y315S) and HERG (HERG-G628S), respectively, in the heart. As expected, both transgenes eliminated the corresponding repolarizing currents. Additionally, we found a significant downregulation of the reciprocal repolarizing current by the transgene, lowering of I_Kr by KvLQT1-Y315S, and lowering of I_Ks by HERG-G628S. Therefore, we sought to establish an in vitro system that would mimic the phenomenon of I_Kr and I_Ks reduction seen in vivo. Compared with the rabbit model, such a system would facilitate the study of the underlying molecular mechanisms responsible for the downregulation of complementary repolarizing K⁺ currents.

MATERIALS AND METHODS

K⁺ channel DNA constructs.

For pcDNA3-Flag-HERG-G628S and pcDNA3-Flag-KvLQT1-Y315S constructs, the coding sequences of Flag-HERG-G628S and Flag-KvLQT1-Y315S were cut from the corresponding transgenic rabbit constructs (6) and inserted in the EcoRI site of the pcDNA3 vector. pcDNA3-Flag-HERG and pcDNA3-Flag-KvLQT1 wild-type constructs were generated by exchanging the fragments containing the mutation with corresponding wild-type fragments. Plasmid pcDNA3-KCNE1-KCNQ1 encodes a fusion of human KvLQT1 and minK (21), whereas pcDNA3-HERG-S1HAS2 is an expression vector for human HERG carrying a hemagglutinin (HA) tag inserted in its extracellular S1-S2 loop (10). DNA encoding an NH₂-terminal Flag epitope was added in frame to Kir2.1 by PCR using rabbit Kir2.1 cDNA and the appropriate primers (forward primer: 5′-GGGGAATTCACCATGGACTACAAGGACGACGATGACAAGGGCAGTGTGCGAACCAAC-3′ and reverse primer: 5′-GGGGAATTCTCATATCTCCGATTCTCGCCG-3′; the underlined sequence codes for the Flag epitope, and the bold sequence indicates an EcoRI site). The PCR product was digested with EcoRI and cloned into EcoRI-digested pcDNA3. All constructs were sequence verified.

Cloning, expression, and purification of K⁺ channel COOH-terminal domains.

DNA fragments encoding the full-length KvLQT1 COOH-terminus (KCNQ1-CT, amino acids 349–676) and a 100-amino acid fragment of the HERG COOH-terminus (HERG-14, amino acids 771–870) were obtained by PCR amplification of human KvLQT1 and HERG cDNAs using Pfu DNA polymerase and primers containing EcoRI and HindIII restriction sites, respectively. PCR products were cloned into an EcoRI- and HindIII-digested pMAL-2C vector. Plasmids were expressed in Esherichia coli strain BL21 (DE3) pLysS, which was grown at 37°C to an optical denisty (600-nm absorbance) of 0.5 in LB medium containing 50 μg/ml carbenicillin and 34 μg/ml chloramphenicol. Cultures were induced with 0.5 mM IPTG, and growth was continued for an additional 6–8 h at 25°C. Maltose-binding protein (MBP)-KCNQ1-CT and MBP-HERG-14 fragments were purified by an amylose resin column. Briefly, cells (∼10 g) were suspended in MBP buffer containing 20 mM Tris·HCl (pH 7.4), 200 mM NaCl, 1 mM EDTA, and 1 mM β-mercaptoethenol, protease inhibitors (Boehringer Mannheim), and 100 μg/ml DNase I. Cells were lysed by sonication, and debris was removed by centrifugation. The supernatant was loaded onto 3-ml amylose columns (New England BioLabs), and proteins were eluted with three column volumes of MBP buffer containing 10 mM maltose. After reaching a concentration of 5 ml, proteins were applied to a 1.6 × 70-cm Superdex 200 gel filtration column. Proteins were eluted with 20 mM HEPES containing 150 mM NaCl, 5 mM KPO₄, and 1 mM β-mercaptoethanol (pH 7.8) at 0.5 ml/min, and the correct size fractions were pooled and concentrated.

Cell culture and stable cell line generation.

CHO cells were obtained from the American Type Culture Collection (Manassas, VA) and cultured at 37°C with 5% CO₂ in F-12 medium (Invitrogen, Carlsbad, CA) supplemented with 10% heat-inactivated FBS (Sigma, St. Louis, MO). Transient or stable transfections into CHO cells were performed using Fugene 6 (Roche Applied Science, Nutley, NJ) following the manufacturer's instructions. Cells were studied by patch clamp or immunostaining 24–48 h after transient transfections. A plasmid carrying cDNA for green fluorescent protein (GFP; 0.2–0.3 μg) served as the control. To isolate stable cell lines expressing Flag-HERG, HA-HERG, or KvLQT1-minK, CHO cells were transfected with the corresponding linearized expression plasmid. Forty-eight hours after transfection, cells were split 1:10, 1:50, and 1:100 into 96 wells containing F-12 medium supplemented with 1 mg/ml neomycin (Invitrogen). After 7–10 days, single clones were isolated and expanded, and the expression of protein and surface currents was determined. The clones in which >90% of the cells exhibited a high surface current were selected for future experiments. A human embryonic kidney (HEK)-293 cell line stably expressing wild-type HERG channels (37) was cultured at 37°C with 5% CO₂ in DMEM (Invitrogen) supplemented with 10% FBS, 0.1 mM nonessential amino acids solution (Invitrogen), 2 mM GlutaMAX (Invitrogen), and 400 μg/ml geneticin (Invitrogen). Transient transfections into HEK-293 cells were performed using Lipofectamine 2000 (Invitrogen) or Fugene 6 following the manufacturer's instructions.

Electrophysiological recording and data analysis.

Patch-clamp recordings in CHO and HEK-293 cells were performed with an Axopatch-200B amplifier (Axon Instruments, Foster City, CA) using a standard whole cell configuration of the patch-clamp technique as previously described (11). Generally, patch-clamp recordings were performed with cells of early cell passages (not more than 20 passages, 100% of cells with relevant current) since transgene silencing was observed with increasing passage numbers. Briefly, the pipette resistances were 2–4 MΩ when filled with 50 mM KCl, 65 mM K-glutamate, 5 mM MgCl₂, 5 mM EGTA, 10 mM HEPES, 5 mM K₂-ATP, and 0.2 mM Tris-GTP (pH 7.2). The extracellular bath solution contained 140 mM NaCl, 5.4 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, 0.33 mM NaH₂PO₄, 7.5 mM glucose, and 5 mM HEPES (pH 7.4). To record currents in the presence of Kir2.1 current, thus minimizing the overlap of HERG and inward rectifier K⁺ currents, 3.6 mM KCl and 0.2 mM CaCl₂ rather than 5.4 mM KCl and 1 mM CaCl₂ were used in the bath solution. Currents were recorded at room temperature (21–23°C). The recording protocol is described in Figs. 2–4.

Fig. 2.

Expression of HERG- and KvLQT1-minK-encoded currents. A: original recordings of stable HERG CHO cells transfected with expression vectors for green fluorescent protein (GFP; left), KvLQT1-Y315S (middle), and KvLQT1 (right). No KvLQT1 inhibitor was added in this experiment. The holding potential was at −70 mV, and test potentials ranged from −60 to +50 mV. Tail currents were recorded when the test potential had returned to −50 mV. B: current-voltage (I-V) relationships at the end of the test pulse. CHO cells expressing HERG and GFP (n = 17), HERG and KvLQT1-Y315S (n = 10), and HERG and KvLQT1 (n = 20) were tested. All currents were normalized to cell capacitance. C: normalized peak tail current analysis of the aforementioned three groups. D: original recordings of stable KvLQT1-minK CHO cells transfected with expression vectors for GFP (left), HERG-G628S (middle), and HERG (right). No HERG inhibitor was added in this experiment. E: I-V relationships at the end of the test pulse. CHO cells expressing KvLQT1-minK and GFP (n = 14), KvLQT1-minK and HERG-G628S (n = 10), and KvLQT1-minK and HERG (n = 10) were tested. F: normalized peak tail current analysis of the aforementioned three groups. Data are expressed as means ± SE. ANOVA was applied to analyze the data. P values of <0.05 were considered to indicate statistical significance. 293B (50 μM) was added when HERG currents were measured in stable HERG cells transiently expressing KvLQT1 (B and C). Similarly, E-4031 (5 μM) was added when KvLQT1 currents were measured in stable KvLQT1 cells transiently expressing HERG (E and F).

Fig. 3.

Expression of HERG- and KvLQT1-minK-encoded currents in stable cell lines transiently expressing Kir2.1. A: original recording of stable HERG CHO cells transfected with expression vectors for Kir2.1. The holding potential was at −70, and test potentials ranged from −100 to +50 mV. Tail currents were recorded when the test potential had returned to −50 mV. B: I-V relationships at the end of the test pulse. CHO cells expressing HERG and GFP (n = 7) or HERG and Kir2.1 (n = 7) were tested. Inward rectifier K⁺ current was noticeable only at potentials below −40 mV and was not evaluated. C: normalized peak tail current analysis of the aforementioned two groups. D: original recording of stable KvLQT1-minK CHO cells transfected with expression vectors for Kir2.1. E: I-V relationships at the end of the test pulse. CHO cells expressing KvLQT1-minK and GFP (n = 7) or KvLQT1-minK and Kir2.1 (n = 7) were tested. F: normalized peak tail current analysis of the aforementioned two groups.

Fig. 4.

Expression of HERG-encoded currents in human embryonic kidney (HEK)-293 cells. A: original recordings of stable HERG cells transfected with expression vectors for GFP (left) and KvLQT1-Y315S (right). The holding potential was at −80 mV, and test potentials ranged from −70 to +40 mV. Tail currents were recorded when the test potential had returned to −60 mV. B: I-V relationships at the end of the test pulse. HERG cells expressing HERG and GFP (n = 12), HERG and KvLQT1-Y315S (n = 9), HERG and KvLQT1 (n = 12), and HERG and tKvLQT1 (n = 9) were tested. All currents were normalized to cell capacitance. C: normalized peak tail current analysis of the aforementioned three groups. 293B (50 μM ) was added when HERG currents were measured in stable HERG cells transiently expressing KvLQT1.

Antibodies.

A monoclonal antibody against the Flag epitope (M2, Sigma) and polyclonal antibodies against HERG and KvLQT1 (Alomone Labs, Jerusalem, Israel) were used for immunoblot analysis. Monoclonal HA antibody 16B12 was obtained from Covance (Berkeley, CA). Goat anti-mouse and goat anti-rabbit horseradish peroxidase (HRP)-conjugated secondary antibodies were from Invitrogen. Alexa fluor 488-conjugated goat anti-mouse and Alexa fluor 594-conjugated goat anti-rabbit secondary antibodies were obtained from Molecular Probes.

Immunostaining.

Since the aforementioned polyclonal antiserum against HERG was raised against a peptide of the intracellular COOH-terminus of HERG, we used the HA antibody to detect the HA-tagged extracellular S1-S2 loop of HA-HERG found on the cell surface of HA-HERG stable cells. For immunostaining of cell surface HA-HERG channels, cells were plated on 12-mm circular glass coverslips (Fisher Scientific, Pittsburgh, PA), transfected, and cultured for an additional 24 h in 24-well plates. Cells were fixed with 3% paraformaldehyde in PBS for 5 min at room temperature. After four washes with PBS, cells were blocked with 3% BSA in PBS for 30 min. Cells were then incubated with a monoclonal antibody against HA (1:1,000 dilution) and, after three washes with PBS, incubated with Alexa fluor 488-conjugated goat anti-rat secondary antibody (1:1,000 dilution). For immunostaining of intracellular Flag-tagged protein, transfected CHO cells were fixed and then permeabilized with 0.1% Triton X-100 in PBS containing 3% BSA for 30 min at room temperature. Intracellular Flag staining was performed by incubation with a polyclonal antibody against Flag (1:1,000 dilution). Alexa fluor 594-conjugated goat anti-rabbit IgG (1:1,000 dilution) was used as a secondary antibody.

Fluorescence microscopy.

Cells were examined using an Eclipse TE2000 inverted microscope (Nikon, Melville, NY) equipped with an Intensilight C-HGFI illuminator and a ×40 plan apochromat objective lens. Images were captured using Elements software (Nikon), FITC, Cy3 and Cy5 emission filters, and a monochromatic camera. For image analysis, captured images were pseudocolored and processed using Elements and Adobe Photoshop software.

Immunoblot analysis.

Coimmunoprecipitations and immunoblot analyses were carried out as previously described (16). For coimmunoprecipitations, 2 μl of anti-Flag antibody or control IgG were added to the whole cell lysate and incubated with agitation at 4°C overnight. Proteins were subjected to SDS-PAGE and transferred to nitrocellulose membranes. Blots were incubated with antibodies diluted in 5% nonfat milk powder. Binding of the primary antibody to proteins was detected using HRP-conjugated secondary antibody followed by ECL detection (Pierce, Rockford, IL).

Surface plasmon resonance analysis.

Interaction experiments were carried out on a Biacore 3000 Instrument (Biacore, Piscataway, NJ) in running buffer (20 mM KPO₄, 130 mM KCl, 3.4 mM EDTA, and 0.005% Tween 20). The MBP-HERG-14 domain was immobilized on biosensor chip CM5, and MBP-KCNQ1-CT (analyte) or MBP (control analyte) was injected for flow over this interaction surface. An equivalent volume of each protein analyte was injected over a chip surface with no protein immobilized to serve as a blank phase for the background noise subtraction. For kinetic analyses, different concentrations of analytes (0.3125–20 μM) were tested. The binding capacity (R_max) and dissociation kinetic constant (K_d) were calculated by BIAevaluation software (version 3.1, GE Healthcare) using a two-state binding model (19, 28, 35).

RESULTS

Generation of HERG and KvLQT1-mink stable cell lines.

Our study (6) on transgenic rabbits with LQT syndrome showed that the expression of pore mutants of HERG (HERG-G228S) and KvLQT1 (KvLQT1-Y315S) not only eliminated I_Kr and I_Ks but also caused downregulation of the remaining complementary K⁺ current, implying an interaction between mutant channels and their reciprocal wild-type forms. To facilitate relevant studies, we wanted to create an in vitro model of the observed in vivo effect. We created stable CHO cell lines to obtain more consistent I_Kr and I_Ks compared with transient transfections. Since KvLQT1 produced I_Ks only when minK was coexpressed (34), an expression construct encoding a fusion between KvLQT1 and minK was used for the stable transfection. Individual clones stably expressing Flag-tagged HERG or a fusion of KvLQT1 and minK were isolated, expanded, and shown to have persistently high expression of the integrated gene. Figure 1 shows Flag-HERG and KvLQT1-minK levels of relevant clones compared with wild-type CHO cells.

Fig. 1.

Creation of Chinese hamster ovary (CHO) cell lines stably expressing HERG and KvLQT1-minK. Immunoblots of stable CHO cell lines expressing HERG or KvLQT1-minK1 are shown. Left: 25 μg of membrane proteins from nontransfected CHO cells and a CHO clone stably expressing Flag-HERG were subjected to Western blot analysis using anti-HERG antiserum. The core-glycosylated immature band of 135 kDa and the fully mature channel protein at 155 kDa are indicated by arrows. Right: 50 μg of membrane proteins from nontransfected CHO cells and a CHO clone stably expressing KvLQT1-minK were analyzed using an immunoblot based on polyclonal anti-KvLQT1 antiserum.

Downregulation of I_Kr and I_Ks upon overexpression of their complementary channel proteins KvLQT1 and HERG.

Next, we addressed whether expression of a mutant channel would affect its reciprocal current in the respective stable cell line. Stable HERG and KvLQT1-minK clones were transiently transfected with expression vectors encoding KvLQT1-Y315S and HERG-G628S, respectively, and patch clamping was performed ∼24 h posttransfection. First, I_Kr was examined in stable HERG cells mock transfected with a GFP expression vector and stable HERG cells transfected with KvLQT1-Y315S plasmid (Fig. 2, A–C). The current density at +10 mV was 25.14 ± 3.78 pA/pF in control CHO cells, whereas it was reduced to 7.24 ± 2.28 pA/pF in cells expressing KvLQT1-Y315S (P ≤ 0.01; Fig. 2B). KvLQT1-Y315S expression caused a minor but significant shift to the left of the voltage dependence of steady-state activation (Fig. 2C), although this could not explain the drop in I_Kr density. Voltages at half-maximal activation (V_1/2) were −4.75 ± 0.54 mV (HERG) and −10.94 ± 0.39 mV (HERG/KvLQT1-Y315S, P < 0.05), respectively. Next, we sought to determine whether the opposite phenomenon, i.e., lowering of I_Ks by mutant HERG protein, was also reproducible in a heterologous expression system. Again, current densities of stable KvLQT1-minK CHO cells mock transfected or transiently expressing HERG-G628S were measured (Fig. 2, D–F). HERG-G628S reduced the KvLQT1-minK-encoded current at +10 mV from 490.67 ± 60.99 to 145.74 ± 31.37 pA/pF (P ≤ 0.01; Fig. 2E). In addition, there was a significant change in the activation voltage dependence (Fig. 2F). V_1/2 values were 8.16 ± 0.27 mV (KvLQT1-minK) and 15.33 ± 0.59 mV (KvLQT1-minK/HERG-G628S, P < 0.05), respectively. This altered voltage dependence may partially account for the approximately threefold drop in I_Ks. Thus, the observed downregulation of HERG- or KvLQT1-minK-encoded current densities by overexpressed reciprocal mutant channels reflected the aforementioned in vivo scenario in transgenic LQT rabbits (6). To determine whether wild-type channels have the same effect as pore mutant channels on reciprocal currents, stable HERG and KvLQT1-minK CHO cells were transfected with expression vectors for KvLQT1 and HERG, respectively. Similar to mutant channels, wild-type channels lowered reciprocal repolarizing currents. KvLQT1 expression decreased HERG current density at +10 mV from 25.14 ± 3.78 to 14.14 ± 2.14 pA/pF (Fig. 2B), whereas HERG reduced I_Ks density from 490.67 ± 60.99 to 328.91 ± 35.18 pA/pF (Fig. 2E). In addition, we noted differences in the voltage dependence of activation of cells expressing HERG or HERG and KvLQT1 (V_1/2 of −4.75 ± 0.54 and −12.45 ± 0.28 mV, P < 0.05) as well as cells expressing KvLQT1-minK or KvLQT1-minK and HERG (V_1/2 of 8.16 ± 0.27 and 10.72 ± 0.39 mV). To underline the specificity of repolarizing current downregulation, HERG and KvLQT1-minK stable CHO cells were transiently transfected with Kir2.1, which encodes a cardiac inward rectifier channel (18). However, we found no effect of Kir2.1 on I_Kr and I_Ks density (Fig. 3, A, B, D, and E) or tail currents of I_Kr and I_Ks (Fig. 3, C and F).

To further confirm our aforementioned results in stable CHO cells, we also used HEK-293 cells stably expressing wild-type HERG channels (37) to study the effect of KvLQT1 expression on HERG in another cell system and to map the critical domains involved in the interactions. This additional approach allowed us to rule out any artifacts caused by the transgene integration site or copy number in our stable CHO cell clone. HERG currents were examined in HEK-293 cells stably expressing HERG and transiently transfected with expression vectors encoding KvLQT1-Y315S, wild-type KvLQT1, or tKvLQT1, a truncated isoform in which the first 127 amino acids (i.e., the entire NH₂-terminus) are absent (15) (Fig. 4, A–C). The HERG current density at 0 mV was 74.43 ± 9.8 pA/pF in control (HERG only) cells, whereas it was reduced in HEK-293 cells coexpressing KvLQT1 constructs to 27.97 ± 6.3 pA/pF (KvLQT1-Y315S), 27.87 ± 5.69 pA/pF (wild-type KvLQT1), and 60.26 ± 13.3 pA/pF (tKvLQT1; Fig. 4B), suggesting that the COOH-terminus of KvLQT1 plays a critical role in mediating the functional interactions with HERG channels. Expression of all KvLQT1 constructs caused a significant shift of the voltage dependence of steady-state activation (Fig. 4C). V_1/2 values were −26.63 ± 1.42 mV (HERG), −15.6 ± 3.15 mV (HERG + KvLQT1-Y315S), −11.94 ± 1.86 mV (HERG + wild-type KvLQT1), and −13.78 ± 0.99 mV (HERG + tKvLQT1, P < 0.05). In contrast to the effect seen in CHO cells stably expressing HERG (Fig. 2C), coexpression of KvLQT1 constructs in stable HEK cells caused a major shift to the right of the voltage dependence of steady-state activation (Fig. 4C). Since both stable cell lines express the same human HERG gene, it is highly likely that the different shifts of the activation voltage dependence caused by KvLQT1 constructs are caused by the assembly of HERG with different accessory subunits or other channel-interacting proteins found in CHO and HEK-293 cells. Further study is warranted in the future to address this difference. Finally, as in stable CHO cells, transient transfection of an expression vector for Kir2.1 into stable HEK-293 cells had no effect on HERG current density and tail currents (data not shown).

Reduction of HERG surface expression by overexpressed KvLQT1 in a HERG stable cell line.

Generally, a decrease in current density can be caused by changes in voltage dependence, lowered cell surface expression of the channel, or altered single channel conductance. Since tail current analyses (Fig. 2C) ruled out voltage dependent changes as a possible factor for I_Kr reduction, we investigated whether cell surface expression of HERG could account for this phenomenon. We created a stable CHO cell line expressing HA-tagged HERG. Although the HA epitope is inserted in the extracellular linker between transmembrane segments 1 and 2 of HERG, HA-HERG functions like a wild-type HERG channel when expressed in mammalian cells (10). HA immunostaining of nonpermeabilized, stable HA-HERG cells therefore detected only HERG channels on the cell surface. To study the effect of KvLQT1 on HERG surface expression, HA-HERG stable CHO cells were transiently transfected with Flag-tagged KvLQT1 or Flag-tagged Kir2.1. Surface expression of HA-HERG was first determined by staining with a monoclonal antibody against HA, visualized as green fluorescence (Fig. 5A,a and d). After cell membrane permeabilization, Flag-tagged KvLQT1 and Kir2.1 were detected with anti-Flag antiserum, visualized as red fluorescence (Fig. 5A,b and e). Cells positive for Flag-KvLQT1 and Flag-Kir2.1 were selected, and green and red fluorescence intensities were measured to assess the expression of cell-surface HA-HERG and total Flag-KvLQT1 or Flag-Kir2.1 in a semiquantitative manner. Although expression levels of Flag-KvLQT1 and Flag-Kir2.1 were comparable, HERG surface expression was reduced to 17% in Flag-KvLQT1-positive cells compared with Flag-Kir2.1-positive cells (Fig. 5B). Thus, the reduced HERG surface expression caused by KvLQT1 could account for the observed drop in I_Kr density.

Fig. 5.

Cell surface expression of HERG. A: immunostaining analysis of transfected CHO cells. A CHO cell line stably expressing hemagluttinin (HA)-tagged HERG was transiently transfected with an expression vector for Flag-KvLQT1 (left) or Flag-Kir2.1 (right). Two days after transfection, cells were fixed with paraformaldehyde and probed with anti-HA monoclonal antibody for surface HERG channel staining. After unbound antibodies were washed out, cells were permeabilized and incubated with polyclonal antibody against the Flag epitope. CHO cells were immunolabeled with anti-HA (a and d) or anti-Flag (b and e) using Alexa fluor 488-conjugated goat anti-mouse and Alexa fluor 594-conjugated goat anti-rabbit secondary antibodies, respectively. Merged patterns of a and b are shown in c; merged patterns of d and e are shown in f. B: fluorescence intensity measurements. Elements software was used to semiquantitate green and red fluorescence signals of Flag-KvLQT1- and Flag-Kir2.1-positive cells. Fluorescence intensity signals (means ± SE) for KvLQT1-transfected (n = 12) and Kir2.1-transfected (n = 8) cells are shown.

Interaction of KvLQT1 and HERG proteins.

Next, we addressed whether a functional interaction between HERG and KvLQT1 based on the reciprocal downregulation of respective repolarizing currents might be linked to a physical interaction of the two channels. Coimmunoprecipitations were carried out to determine whether pore mutants and wild-type channels would interact with the reciprocal channel. For this purpose, CHO cells were transiently cotransfected with suitable expression vectors encoding the two channel proteins, their pore mutants, or Flag-tagged versions thereof (Fig. 6A). Immunoblots of respective cell lysates showed that expression levels of HERG, Flag-HERG, and Flag-HERG-G628S did not differ among the five groups (Fig. 6A, top left). Also, protein levels of KvLQT1, Flag-KvLQT1, and Flag-KvLQT1-Y315S were similar (Fig. 6A, bottom left). Coimmunoprecipitations were performed with anti-Flag antibody or IgG control. Precipitated immunocomplexes were resolved by SDS-PAGE and subsequently immunoblotted with anti-HERG and anti-KvLQT1 polyclonal antisera. Cells coexpressing HERG and KvLQT1 were used as negative controls. As expected, immunoblots did not detect any protein bands in control cells (Fig. 6A, right, lanes 1 and 2). For KvLQT1, coimmunoprecipitations demonstrated an interaction with HERG as well as with the pore mutant HERG-G628S (Fig. 6A, right, lanes 3–6). Similarly, we saw interactions between wild-type HERG and KvLQT1 or its pore mutant KvLQT1-Y315S (Fig. 6A, right, lanes 7–10). Another set of coimmunoprecipitations using anti-Flag antiserum shows the specificity of the interaction between the two channel proteins (Fig. 6B). CHO cells stably expressing HA-HERG were transiently transfected with expression vectors for Flag-tagged KvLQT1, Flag-tagged pore mutant KvLQT1-Y315S, Flag-Kir2, or empty expression vector. Although HA-HERG expression levels were similar in all four samples (Fig. 6B, left), only Flag-tagged KvLQT1 and KvLQT1-Y315S (but not Flag-tagged Kir2.1) were able to pull down HA-tagged HERG channel protein (Fig. 6B, top right, lanes 2–4).

Fig. 6.

HERG and KvLQT1 coimmunoprecipitations. A: immunoblots for HERG and KvLQT1 proteins. CHO cells were transiently transfected with expression vectors encoding wild-type, pore mutant, and respective Flag-tagged forms of HERG and KvLQT1, as shown. Left, cell lysates were separated by SDS-PAGE, blotted, and probed with anti-HERG (top) or anti-KvLQT1 (bottom) antisera. Right, cell lysates were immunoprecipitated with anti-Flag/protein A/G PLUS-agarose (*) or IgG/protein A/G PLUS-agarose (o). Immunoprecipitates were separated by SDS-PAGE, blotted, and probed with anti-HERG (top) or anti-KvLQT1 (bottom) antisera. B: immunoblots for HERG and Flag-tagged KvLQT1 and Kir2.1 proteins. CHO cells stably expressing HA-HERG were transiently transfected with the following expression plasmids: pcDNA3 (lane 1), pcDNA3-Flag-KvLQT1 (lane 2), pcDNA3-KvLQT1-Y315S (lane 3), or pcDNA3-Flag-Kir2.1 (lane 4). Left, cell lysates were separated by SDS-PAGE, blotted, and probed with anti-HERG antisera. Right, cell lysates were immunoprecipitated with anti-Flag/protein A/G PLUS-agarose. Immunoprecipitates were separated by SDS-PAGE, blotted, and probed with anti-HERG (top) or anti-Flag (bottom) antisera . All experiments were carried out in triplicate. Shown are representative immunoblots.

Surface plasmon resonance (SPR) analysis was also performed to examine the specific domains that might be involved in the interaction between KvLQT1 and HERG. The full-length COOH-terminus of KvLQT1 and a 100-amino acid fragment of the HERG COOH-terminus were purified as MBP-fusion proteins from E. coli cells using an amylose affinity column followed by Superdex 200 gel filtration column chromatography. Subunit molecular weights of MBP-fusion proteins were verified by SDS-PAGE (Fig. 7A). To assay for physical interactions between the two proteins, we subjected them to SPR, where the HERG-14 fragment was immobilized on the sensor surface and the KvLQT1 fragment was flowed across in varying concentrations. An unambiguous and strong response was detected (R_max = 4,460 ± 33 response units), indicating a physical association between the COOH-terminus of KvLQT1 and the HERG-14 fragment (Fig. 7B). By analyzing various concentrations of KvLQT1-CT in the analyte, we estimated a K_d of 2.7 ± 0.2 μM. To control for a possible contribution of MPB, we applied the same concentrations of MBP as the analyte to immobilized HERG-14 and observed no response (R_max <10 response units; Fig. 7C). Together, the coimmunoprecpitation and SPR results suggested a specific, physical interaction between the COOH-termini of HERG and KvLQT1 in vitro.

Fig. 7.

Surface plasmon resonance (SPR) analysis of KvLQT1-HERG associations. A: Commassie blue-stained gel showing the results of purification of maltose-binding protein (MBP)-KCNQ1-CT and MBP-HERG-14 cytoplasmic domains. B: representative SPR tracing of KCNQ1-CT interacting with immobilized HERG-14. Each trace represents the SPR response to a different concentration (between 0.31 and 20 μM) of soluble KCNQ1-CT flowing across immobilized HERG-14-MBP. C: representative SPR tracing of MBP applied to immobilized HERG-14. The same concentration range of soluble MBP as the analyte was used as in B. No resonance changes were observed for MBP and HERG-14. R_max, binding capacity.

DISCUSSION

Transgenic rabbit models of LQT syndrome have demonstrated, in vivo, that each dominant negative K⁺ channel α-subunit, KvLQT1-Y315S or HERG-G628S, eliminated not only the current encoded by its respective wild-type protein but also significantly downregulated the remaining complementary repolarizing current, I_Kr or I_Ks, without affecting the steady-state levels of the native polypeptides (6). For example, at +10 mV, I_Kr was reduced by ∼33% in KvLQT1-Y315S transgenic animals, whereas HERG-G628S expression lowered I_Ks by 25%. Minor changes in voltage dependence were noticed but were not sufficient to explain the observed downregulation of repolarizing currents. This reciprocal downregulation was in sharp contrast to the findings of several studies (12, 13, 36) using mice with knockouts of K⁺ channels or respective pore mutant dominant negative transgenes. In all these cases, downregulation of a specific K⁺ current gave rise to electrical remodeling characterized by the upregulation of other K⁺ currents. For instance, mice expressing a dominant negative form of Kv1.1 lost the rapidly activating, slowly inactivating K⁺ current (I_K,slow1) (20) but showed a selectively upregulated, Kv2-encoded current (I_K,slow2) due to higher Kv2.1 transcript levels (36).

In the present study, we transiently transfected HERG- or KvLQT1-minK-expressing stable cell lines with expression vectors for mutant or wild-type KvLQT1 or HERG, respectively. This approach allowed us to recapitulate the downregulation of I_Kr or I_Ks by its reciprocal K⁺ channel seen in the rabbit LQT models (6). Yet, in the in vitro CHO cell system, downregulation of the complementary current by pore mutants was significantly greater than that seen in vivo (at +10 mV, CHO: I_Ks 70% and I_Kr 70% vs. LQT rabbits: I_Ks 25% and I_Kr 33%). It is possible that the higher expression of pore mutant channels in CHO cells due to the strong cytomegalovirus promoter-driven transgene expression could explain the more pronounced effect on complementary current densities. Interestingly, transiently expressed wild-type HERG or KvLQT1 channels were less effective than their pore mutants (despite comparable protein expression levels) and showed only a slightly higher degree of downregulation of the complementary current compared with in vivo data (at +10 mV, CHO: I_Ks 33% and I_Kr 44% vs. LQT rabbits: I_Ks 25% and I_Kr 33%).

A previous study by Nattel and colleagues (9), however, used a transient CHO cotransfection system and reported findings seemingly contradictory to ours. This group showed that coexpressed KvLQT1 increased HERG current density approximately twofold. This was corroborated by the fact that KvLQT1 expression also increased the cell surface expression of HERG. In contrast, no effect of HERG coexpression on I_Ks and KvLQT1 membrane localization was noted. Moreover, Hayashi et al. (14) recently showed that trafficking-competent KvLQT1 was capable of increasing the current density of several trafficking-deficient HERG mutants in CHO cells, but this effect was not seen for all mutants examined. Currently, we can only speculate as to the reason(s) for the disparity among the results obtained by us and those of Nattel and coworkers or Hayashi et al. The most obvious difference among the studies lies in the transfection method. Whereas we used a combination of stable and transient transfections, the other groups relied solely on transient cotransfections. The rationale behind our strategy was twofold. In our hands, results obtained from transient HERG and KvLQT1 cotransfection experiments were very inconsistent (data not shown), possibly due to variations in current density often encountered in transient transfection systems. Therefore, we sought to minimize such variations by creating stable cell lines. Furthermore, when examining the long-term effects of the transgene, i.e., the wild-type channel, HERG, or KvLQT1, stable cells will more closely model an in vivo situation than transiently transfected cells. Other possible but less likely explanation for the obvious discrepancy could be a difference in the physiology of CHO cells used (17) or different transfection efficiencies affecting expression levels of the channel protein. Since we also reproduced the downregulation of I_Kr by KvLQT1 in HEK-293 cells stably expressing HERG channels, corroborating our observations in stable CHO cells, we believe that the transfection method is the primary reason for the differences between our findings and the data published by others. To our knowledge, this is the first report highlighting the importance of the transfection method on heterologous channel expression.

Despite differences, both we and Nattel and coworkers (9) demonstrated that the modulation of current density was due to changes in membrane localization of the responsible channel and suggested that the two channel proteins somehow affected each other's synthesis in the endoplasmic reticulum, assembly, processing, trafficking, or membrane targeting. A tight link between HERG and KvLQT1 trafficking would not be surprising, as they encode the two components of I_K. Similarities in HERG and KvLQT1 trafficking due to the use of the same β-subunit, minK, was also implied by Bianchi and coworkers (4), who reported that the coexpression of minK increased both KvLQT1 and HERG currents. However, there remains controversy as to the contribution of a β-subunit, viz. Mirp1 or minK, to I_Kr in vivo (2). Both Yang and colleagues (33) as well as McDonald et al. (22) reported that minK affected I_Kr. However, we were able to detect an interaction between Flag-tagged pore mutant KvLQT1 and minK in rabbit hearts using anti-Flag immunoprecipitations and failed to detect an interaction between minK and the pore mutant of HERG (data not shown); therefore, we can probably rule out a competition between HERG and KvLQT1 for minK in the rabbit heart.

Additionally, Ehrlich et al. (9) demonstrated that KvLQT1 coimmunoprecipitated with a GST fusion to the COOH-terminal of HERG (amino acids 681–1159), suggesting that this region of HERG contains domains capable of a physical interaction with KvLQT1. In this work, we have shown via cellular electrophysiology that the transient expression of an NH₂-terminal truncation of KvLQT1 downregulated HERG currents, consistent with results from cells expressing either wild-type or pore mutant KvLQT1. This implies that the NH₂-terminus of KvLQT1 is not essential for the interaction with HERG and leads to the hypothesis that the COOH-terminal regions of both proteins are critically involved in the interaction. Using the highly sensitive SPR analysis, we showed a specific, physical interaction between the COOH-terminus of KvLQT1 and a region of the HERG COOH-terminus (amino acids 771–870). Interestingly, this region of HERG corresponds to a large portion of the cyclic nucleotide-binding domain (1).

In summary, we have created an in vitro system using stable CHO and HEK-293 cell lines that accurately recapitulates the reduction of the complementary repolarizing K⁺ current by HERG or KvLQT1 reported in our in vivo LQT rabbit models (6). Certainly, we are aware of the possibility of different mechanisms underlying the interaction between HERG and KvLQT1 in our stable cell line systems and transgenic rabbits. Nonetheless, in vitro, cellular model systems allowed us to demonstrate a physical interaction between KvLQT1 and HERG mediated by the COOH-termini and will enable us to perform future studies examining the location, duration, and regulation of the functional interactions between these K⁺ channels.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants R01-HL-046005-18 (to G. Koren) and R01-HL-093440 (to T. McDonald).

DISCLOSURES

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