HERG K⁺ Channel, a Regulator of Tumor Cell Apoptosis and Proliferation¹ Free

In the heart, the rapid delayed rectifier K⁺ current, the physiological counterpart of HERG,³ undergoes remarkable developmental changes, predominating in the fetal heart and dissipating in the adult (1, 2). Intriguingly, when the adult cardiac cells become dedifferentiated or cancerous, such as AT-1 and HL-1 (murine atrial tumor cell lines) cells, I_Kr regains its predominance among the K⁺ channels expressed (3, 4). Likewise, in neural crest neurons, HERG currents are transiently expressed at very early stages of their neuronal development, disappearing at later stages to be substituted by inward rectifier (IRK)-like currents (5, 6). Most strikingly, HERG expression has been found in a variety of tumor cell lines of different histogenesis but absent from the healthy cells from which the respective tumor cells are derived (3, 4, 7, 8, 9). These findings prompted us to hypothesize that HERG channels are involved in the regulation of cell growth and cell death. This study was designed to investigate the potential roles of HERG in regulating tumor cell apoptosis and proliferation.

HEK293 and SK-Mel-28 cells were grown in DMEM (American Type Culture Collection, Manassas, VA), SK-BR-3 cells in McCoy’s 5A Modified Medium, A549 cells in F-12K Medium, SH-SY5Y cells in DMEM/F-12, and HL-1 cells in Claycomb Medium (JRH Biosciences, Lenexa, KS). HEK293 cells with stable expression of HERG were a kind gift from Drs. Z. Zhou and C. T. January (University of Wisconsin Medical School, Madison, Wisconsin; Ref. 10), and HEK293 cells with stable expression of the dominant-negative HERG construct (S633A) were a kind gift from Prof. Jun Chen (Harbin Medical University, Harbin, People’s Republic of China).

The techniques have been described in detail elsewhere (11). Borosilicate glass electrodes (1-mm absorbance) had tip resistances of 1–3 MΩ when filled with pipette solution (0.1 mm GTP, 110 mm potassium aspartate, 20 mm KCl, 1 mm MgCl₂, 5 mm Mg-ATP, and 10 mm HEPES, pH 7.3). Junction potentials were zeroed before formation of the membrane-pipette seal in Tyrode’s solution (135 mm NaCl, 5 mm KCl, 1 mm MgCl₂, 1 mm CaCl₂, 5 mm HEPES, and 10 mm glucose, pH 7.4). For SK-Br-3, SH-SY5Y, A549, and SK-Mel-28 cells, the external KCl and NaCl were 40 and 100 mm, respectively. Experiments were conducted at 36°C.

The methods for ELISA and TUNEL quantification of DNA fragmentation and Annexin V detection of early apoptosis have been described in detail previously (12). To induce apoptosis, H₂O₂ was added to the culture medium to a final concentration of 400 μm and incubated with cells for 5 h. For experiments involving Dof, cells had been pretreated with the drug (1 μm) for 30 min before H₂O₂ was added.

Cell proliferation was assessed by characterizing the log phase growth with PDT calculated by the equation: 1/(3.32 × (logN_H − logN_I)/(t₂ − t₁), where N_H is the number of cells harvested at the end of the growth period (t₂) and N_I is the number of cells at 5 h (t₁) after seeded. Cells were counted by a flow cytometer (EPICS XL; Beckman Coulter Canada, Inc.), and the number obtained at 5 h (t₁) after seeding was taken as an initial cell number (N_I), and the number at 48 h (t₂) after seeding was taken as an endpoint number (N_H).

For immunostaining of wild-type and S633A (negative dominant mutant) HERG K⁺ channels, HEK293 cells transfected with wild-type or S633A mutant HERG channels were fixed with freshly prepared 3% paraformaldehyde in PBS for 20 min at 4°C. After washes with PBS, cells were permeabilized in 1% Triton for 5 min and blocked in 1% BSA. The cells were then incubated overnight at 4°C with primary antibody (anti-HERG residues 1106–1159; Alomone Labs, Jerusalem, Israel) diluted in 1% BSA, followed by incubation with donkey antirabbit FITC-conjugated secondary antibody (Jackson ImmunoResearch, Baltimore, MD) at room temperature for 2 h. The coverslip was mounted on a slide with 10 μl of anti-fading solution, and the cells were examined under a confocal microscope. For double staining of HERG and TNFR1, anti-HERG and anti-TNFR1 (Research Diagnostics, Flanders, NJ) antibodies were mixed and added to the cells in 1% BSA. Donkey antirabbit, FITC-conjugated antibodies and donkey antigoat, FITC-conjugated antibodies were used as secondary antibodies for HERG and TNFR1, respectively. Rabbit anti-cleaved caspase-3 antibody was purchased from Cell Signaling Technology (Beverly, MA) and used for analyzing caspase activity. The active form of NF-κB activity was analyzed with mouse anti-NF-κB p65 subunit monoclonal antibody (Chemicon International). Simultaneous PI staining was performed for identifying the nuclei caspase-3 and NF-κB experiments.

Cells were rinsed with ice-cold PBS and scraped off into the lysis buffer (200 mm NaCl, 33 mm NaF, 10 mm EDTA, and 50 mm HEPES, pH 7.4) plus protease inhibitor cocktail (100 μm phenylmethylsulfonyl fluoride, 1 μg/ml pepstatin A, 1 μg/ml leupeptin, and 4 μl/ml aprotinin; Sigma). The cells were sonicated and spun at 500 × g for 10 min. The membrane fractions were pelleted from the low-speed supernatants by centrifugation at 60,000 rpm for 1 h at 4°C and resuspended in 50 mm Tris-HCl, 15 mm β-mercaptoethanol, and 1% SDS. For immunoprecipitation, a protein sample (60 μg) was incubated with anti-TNFR1 or anti-HERG antibody for 1 h at 4°C, followed by the addition of 1:1 slurry of protein G-Sepharose beads (Sigma) and incubated overnight at 4°C. The beads were washed with extraction buffer, and bound proteins were eluted with SDS-PAGE sample buffer and boiled for 5 min. Samples were subjected to SDS-PAGE and immunoblotting (11, 12) with anti-HERG or anti-TNFR1.

We first compared the cell death, in response to H₂O₂ stimulation, between cells expressing endogenous HERG current (I_HERG) and cells without HERG expression. The cells expressing endogenous I_HERG used in this study include SK-BR-3 (human mammary gland adenocarcinoma cells), SH-SY5Y (neuroblastoma cells), and HL-1 (rat atrial tumor cells; Refs. 4, 7). The cells used in this study, which do not express I_HERG, include tumor cell line A549 and SK-Mel-28. The presence and absence of HERG channels in these cells were verified by whole-cell patch-clamp (Fig. 2) and immunostaining using antibody raised against the COOH-terminal sequence of HERG (Fig. 4). The cells grown in the normal culture medium to 80% confluence were incubated with H₂O₂ (400 μm) for 5 h and then examined under a light microscope for morphological alterations to identify the apoptotic cells (shrinking, blebbing, detached, and round up; Ref. 12). As shown in Fig. 1, H₂O₂ caused considerable cell death in cells expressing endogenous I_HERG relative to cells devoid of I_HERG (Fig. 1,A). To confirm that the observed cell death was caused by apoptosis, ELISA (12) and TUNEL (12) methods were used to detect DNA/chromosomal fragmentation, a biochemical hallmark of apoptosis, and the results are presented in Fig. 1, B and C. Consistent with the morphological data, DNA fragmentation is substantially higher in HERG-expressing cells than in cells lacking HERG channel conductance. In addition, Annexin V and PI double staining was performed to identify the cells in the early stage of apoptosis (12). Our data showed clearly that after exposure to H₂O₂ for 2 h, a majority of cells expressing HERG, but not the cells lacking HERG, displayed positive Annexin V staining (Fig. 1 D, green).

We then evaluated the effects of Dof, a specific blocker of HERG channels, on apoptosis. As illustrated in Fig. 1, Dof prevented cell death and DNA fragmentation in cells with endogenous I_HERG but did not affect the apoptosis in cells lacking I_HERG.

Also noteworthy is that the concentration of H₂O₂ required to reach the same degree of apoptosis was much higher in I_HERG-lacking cells relative to I_HERG-expressing cells. For example, 400 μm H₂O₂ induced ∼50% cell death in all I_HERG-expressing tumor cells tested, but only ∼5% in the I_HERG-lacking cells, and to obtain 50% cell death in these I_HERG-lacking cells, H₂O₂ concentration had to be increased to >1.2 mm.

To further assess the role of I_HERG in regulating H₂O₂-induced apoptosis, we performed similar experiments in HEK293 cells stably transfected with HERG (HERG-HEK; Ref. 10). Our data demonstrated that H₂O₂ (400 μm) caused minimal apoptosis in nontransfected or mock-transfected HEK cells. By comparison, substantial apoptosis was consistently seen in HERG-HEK cells, an effect effectively prevented by pretreatment with Dof (Fig. 1). To investigate whether the apoptosis promotion in HERG-expressing cells was ascribed to HERG channel expression per se or was attributable to HERG channel conductance, we used the HEK cells stably transfected with the negative dominant construct of HERG channel (S633A mutant). Expression of S633A HERG proteins in the cytoplasmic membrane was verified by immunostaining, and the absence of HERG conductance was confirmed by whole-cell patch-clamp recording (Fig. 2,B). As shown in Fig. 1 D, the HEK cells transfected with S633A lost the ability to enhance H₂O₂-induced apoptosis. Our data thus suggest that HERG conductance or I_HERG promotes apoptosis induced by H₂O₂ insult.

Proapoptotic effect of other K⁺ currents has been recognized recently. The elegant studies reported by several groups including Yu et al. (13, 14), Bortner and Cidlowski (15), Vu et al. (16), and Maeno et al. (17) open up a new era in the field of K⁺ channel research. Their data, together with some other studies, provide convincing evidence that decreased intracellular K⁺ content attributable to K⁺ efflux through K⁺ channels results in activation of apoptotic signaling pathways. Loss of intracellular K⁺ can lead to cell shrinkage because K⁺ is a major determinant of intracellular osmolarity (17) and can also mean a withdrawal of inhibitory factor for caspase activation (executioners of apoptosis; Refs. 15, 16). We performed experiments to assess the effects of H₂O₂ on I_HERG in the tumor cells. As illustrated in Fig. 2 A, H₂O₂ (400 μm) significantly increased outward I_HERG in various cells tested and shifted the HERG activation to more negative potentials. For example, in HERG-HEK cells, minimal I_HERG was seen at potentials negative to −40 mV under control conditions. However, 10 min after H₂O₂, ample activation of I_HERG was observed starting from −60 mV. Particularly noticeable is that the outward tail currents that were virtually absent under control conditions became prominent in the presence of H₂O₂ in HERG-expressing tumor cells. Dof (1 μm) effectively inhibited I_HERG in the various cells tested.

To study whether HERG conductance was correlated with caspase activation, immunostaining analysis of caspase-3 activity using antibody directed against the cleaved form of the enzyme was performed. Our data show clearly that H₂O₂ produced more pronounced activation of caspase-3 in HERG-expressing cells than in cells lacking HERG or in HEK cells transfected with S633A. Preincubation of cells with Dof significantly prevented caspase-3 from activating in HERG-expressing cells (Fig. 2 C).

Moreover, promotion of apoptosis by HERG is also observed when 10 ng/ml TNF-α was used to induce apoptosis in both HERG-expressing and HERG-lacking cells, an effect effectively abrogated by Dof only in the former but not in the latter. And similar to the results obtained with H₂O₂, TNF-α consistently caused more apoptotic cell death (Fig. 3 A) and more pronounced activation of caspase-3 (data not shown) in HERG-expressing cells than in HERG-lacking cells.

However, enhancement of proliferation in the tumor cells by TNF-α was also seen and became more obvious when the apoptosis was less prominent with reduced TNF-α concentrations to 1 and 0.1 ng/ml. This growth-facilitating effect of TNF-α was significantly more pronounced in the HERG-expressing cells than in the HERG-lacking tumor cells. PDT was determined to quantify the cell growth. PDT was markedly shortened by TNF-α (0.1 or 1 ng/ml) in HERG-expressing cells. For example, TNF-α shortened PDT by 60% in SK-Br-3 cells and by 70% in HL-1 cells (Fig. 3 B). A similar proliferation-facilitating effect of TNF-α was also seen in HEK293 cells expressing the dominant-negative HERG (S633A), although HERG conductance or function was lost in this construct. Noticeably, Dof failed to affect the TNF-α-induced cell proliferation.

The fact that TNF-α produced greater effects on growth and apoptosis in HERG-expressing cells than in HERG-lacking cells suggests a possibility that TNFR is expressed more abundantly in the former than in the latter cells. To test this notion, double staining with antibodies directed against HERG and TNFR1 was carried out, and the results are shown in Fig. 4,A. Clearly, cells that express HERG channels (green staining) demonstrated strong TNFR1 staining (red), with the antibody targeting the COOH-terminal sequence of the protein. By comparison, cells lacking HERG (SK-Mel-28 and A549) did not show positive HERG green staining, and only weak TNFR1 red staining was seen (Fig. 4,A). Similar results were obtained when another anti-TNFR1 antibody targeting the NH₂-terminal region of the protein was used (data not shown). These data indicated that HERG channels could somehow recruit TNFR1 to the cytoplasmic membrane, and there might exist physical interactions between HERG and TNFR1 proteins. To examine the notion, immunocoprecipitation was performed with membrane protein samples extracted from various cells. Anti-HERG antibody recognized a single discrete band of M_r 155,000 size in anti-TNFR1 pull-down samples, as well as in the samples without precipitation steps, from HERG-expressing cells. In contrast, anti-TNFR1 antibody identified a M_r 55,000 band in the protein samples immunoprecipitated with anti-HERG antibody. Pretreatment of the antibodies with their respective antigens abolished the bands. The data are displayed in Fig. 4 B, and only examples from A549, SK-Br-3, and HERG-HEK cells are shown in the figure. Similar results were consistently observed in other HERG-expressing and HERG-lacking cells, respectively. It is unclear what the structural components are for the interactions of two completely different categories of membrane proteins.

No significant differences of immunoreactivity to anti-TNFR2 or anti-Fas were found between cells with and without HERG expression (data not shown).

TNFR1 is known to be critical for regulating cell proliferation and apoptosis in many cells (18, 19). Overexpression of TNFR1 has indeed been found in many malignant tumor cells (19, 20). TNFR1 induction of apoptosis is fulfilled by sequential activation of caspase-2 and caspase-3, whereas TNFR1 induction of cell proliferation may be mediated by NF-κB. NF-κB is involved in the control of numerous cellular functions, particularly regulation of survival and proliferation (21, 22). Translocation from cytoplasm to the nucleus is an event essential for NF-κB activation. An increase in constitutive NF-κB activity has been observed in a variety of malignant tumors, and it may have an important role in tumorigenesis and chemotherapy resistance because it facilitates cell proliferation and antagonizes apoptosis (21, 22). Here, NF-κB activity was analyzed by immunostaining with the antibody raised against the active form of NF-κB in various cells. The data consistently showed higher immunoreactivity of basal active NF-κB in the nuclei of the cells expressing HERG than in those of the cells that do not express HERG. After treatment with TNF-α (0.1 ng/ml), the immunostaining was further enhanced (reflecting the inducible activation of NF-κB), and the increase was more in the HERG-expressing cells relative to the cells without HERG (Fig. 4,C). These results, together with the data on TNFR1 expression, might explain the more pronounced increase in proliferation of HERG-expressing cells in response to TNF-α stimulation (Fig. 3 B).

We show here that H₂O₂ induced substantial apoptosis in cells expressing endogenous or cloned HERG channels, and the concentration of H₂O₂ required to induce a similar degree of apoptosis was approximately three times higher in two lines of tumor cells that do not express HERG channels. H₂O₂-induced apoptosis was significantly prevented by HERG blockade. Similarly, higher concentrations of TNF-α (1–10 ng/ml) also induced apoptosis, which was greatly prevented by Dof. On the other hand, lower concentrations of TNF-α (0.1–1 ng/ml) promoted cell proliferation in tumor cells expressing HERG channels but not in those without HERG expression, an effect unaffected by HERG blockade. Immunostaining and immunocoprecipitation experiments revealed coexpression of HERG and TNFR1 on the surface membrane, and TNFR1 expression was markedly higher in cells expressing HERG channel proteins than in cells lacking HERG. It appears from our data that HERG channel conductance promotes H₂O₂ (or TNF-α)-induced apoptosis, and HERG protein expression seems to recruit TNFR1 to the membrane, which facilitates TNF-α-induced tumor cell growth. Therefore, HERG expression may represent an advantage for tumor cell growth and cancer development in the absence of apoptotic inducers such as H₂O₂ and higher concentrations of TNF-α, indicating that HERG K⁺ channels might contribute to tumorigenesis. Yet, promotion of apoptosis by I_HERG suggests that manipulating HERG conductance to enhance apoptotic tumor cell death may be a novel strategy for cancer therapy.

We are grateful to Drs. Z. Zhou and C. T. January (University of Wisconsin Medical School, Madison, WI) for providing HERG-transfected HEK293 cells (10). We also thank XiaoFan Yang for excellent technical support.