Calcium-activated K⁺ channels increase cell proliferation independent of K⁺ conductance

Abstract

The intermediate-conductance calcium-activated potassium channel (IK1) promotes cell proliferation of numerous cell types including endothelial cells, T lymphocytes, and several cancer cell lines. The mechanism underlying IK1-mediated cell proliferation was examined in human embryonic kidney 293 (HEK293) cells expressing recombinant human IK1 (hIK1) channels. Inhibition of hIK1 with TRAM-34 reduced cell proliferation, while expression of hIK1 in HEK293 cells increased proliferation. When HEK293 cells were transfected with a mutant (GYG/AAA) hIK1 channel, which neither conducts K⁺ ions nor promotes Ca²⁺ entry, proliferation was increased relative to mock-transfected cells. Furthermore, when HEK293 cells were transfected with a trafficking mutant (L18A/L25A) hIK1 channel, proliferation was also increased relative to control cells. The lack of functional activity of hIK1 mutants at the cell membrane was confirmed by a combination of whole cell patch-clamp electrophysiology and fura-2 imaging to assess store-operated Ca²⁺ entry and cell surface immunoprecipitation assays. Moreover, in cells expressing hIK1, inhibition of ERK1/2 and JNK kinases, but not of p38 MAP kinase, reduced cell proliferation. We conclude that functional K⁺ efflux at the plasma membrane and the consequent hyperpolarization and enhanced Ca²⁺ entry are not necessary for hIK1-induced HEK293 cell proliferation. Rather, our data suggest that hIK1-induced proliferation occurs by a direct interaction with ERK1/2 and JNK signaling pathways.

cell proliferation is an important physiological process for embryonic development, wound healing, and the routine replacement of old cells with new cells as required. Cell proliferation in eukaryotes is a complex process that is controlled by extracellular signals or mitogens, such as growth factors, as well as the state of differentiation of the cell and involves the interaction of numerous regulatory proteins and signaling cascades. Proliferation is tightly regulated by the cell cycle. However, under certain conditions, cells can abnormally proliferate and this underlies the pathophysiology of atherosclerosis, angiogenesis, and tumor growth.

Potassium channels have been implicated in proliferation since 1984, when Decoursey and Cahalan demonstrated that voltage-gated K⁺ channels mediate proliferation in T lymphocytes (5). Subsequently, a plethora of K⁺ channels, including the intermediate-conductance calcium-activated potassium (IK1) channel, have been linked to proliferation in various cell types. IK1 channels are widely expressed and have been implicated in proliferation in a variety of nonexcitable tissues, endothelial cells, T lymphocytes, mesenchymal stem cells, vascular smooth muscle cells, and various types of cancer cells (12, 13, 15, 21, 23, 29, 30, 34–36).

The current hypothesis for K⁺ channel-mediated proliferation is that activation of K⁺ channels causes cell hyperpolarization which increases the driving force for Ca²⁺ entry, usually through store-operated Ca²⁺ channels. This leads to enhanced activation of Ca²⁺-dependent signaling pathways and thus ultimately increases cell proliferation. However, such experiments do not link K⁺ channel activity, Ca²⁺ influx, and proliferation per se. This proposed connection has not been rigorously tested in a system where changes in membrane potential, Ca²⁺ influx, and proliferation can be controlled and examined independently. In fact, recent evidence suggests that, in addition to their ability to pass ions across the membrane, ion channels can also have nonconducting functions that enable them to interact with cell signaling pathways to directly regulate biochemical events (18). One such example is the ability of the ether-à-go-go (EAG) K⁺ channel to regulate cell proliferation in fibroblasts via activation of the p38 mitogen-activated protein kinase (MAPK) pathway (14).

This study investigated the role of human IK channels (hIK1) in cell proliferation. Many cells that require expression of IK for proliferation, such as endothelial cells or T lymphocytes, express multiple types of ion channels, including a variety of K⁺ channels. Therefore, it would be easier to first dissect the role of hIK1 by means of overexpression of recombinant channels in a heterologous expression system before assessing the role in acutely isolated or primary cultured cells. We have devised a powerful molecular manipulation strategy, using mutant hIK1 channels that either cannot conduct K⁺ ions or cannot traffic to the plasma membrane. This strategy was used to examine the link between K⁺ channel function, Ca²⁺ entry, and cell proliferation.

MATERIALS AND METHODS

Cell culture and transfection of human embryonic kidney 293 cells.

Untransfected human embryonic kidney 293 (HEK293) cells and HEK293 cells stably expressing hIK1 channels (HEK293 hIK1 cells) were cultured in minimum essential medium containing Earle's salts and l-glutamine (Gibco), supplemented with 10% fetal bovine serum (Gibco), 1% nonessential amino acids (Gibco), and 1% antibiotic/antimycotic (Gibco). Selection for HEK293 hIK1 was maintained with 600 μg/ml G418 disulfate (Sigma).

Transient transfections of HEK293 cells with hemagglutinin (HA)-tagged HA-hIK1 (28), HA-hIK1^GYG/AAA (a hIK1 pore mutant), HA-hIK1^L18A/L25A [a hIK1 trafficking mutant (17)], or untagged voltage-gated sodium channel Na_v1.5 were performed using ExGen 500 in vitro transfection reagent (Fermentas) according to the manufacturer's instructions. Cells were cotransfected with monomeric red fluorescent protein (mRFP) or green fluorescent protein (GFP) to aid detection of transfected cells for electrophysiology and Ca²⁺ imaging experiments. Mock-transfected cells underwent the same transfection procedure except no plasmid DNA was added to the transfection mixture or cells were transfected with an empty vector.

Plasmids and construction.

All hIK1 constructs contained a HA tag YPYDVPDYA inserted into the second extracellular loop between Gly¹³² and Ala¹³³, as previously demonstrated (28). This was used to aid detection of the channel using anti-HA antibodies. This HA-tagged hIK1 channel was found to be functionally indistinguishable from native hIK1 with respect to regulation by Ca²⁺, pharmacology, and trafficking (28). All mutations in HA-hIK1 were produced using the QuickChange site-directed mutagenesis kit (Stratagene). HA-hIK1 constructs were fully sequenced and aligned with GenBank accession number AF022150. The Na_V1.5 construct was a gift from Jon Makielski (University of Wisconsin-Madison).

Immunofluorescence of transfected HEK293 cells.

The localization of each HA-tagged hIK1 channel construct was determined by immunofluorescence using an anti-HA antibody. For detection of HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A, HEK293 cells seeded on 25-mm coverslips were cotransfected with GFP and the appropriate plasmid DNA. Immunofluorescence experiments were carried out 2 days after transfection. For detection of HA-hIK1 and HA-hIK1^GYG/AAA, transfected cells did not need to be permeabilized, as it was expected that these constructs would be trafficked to the cell membrane and could therefore be detected extracellularly. Cells were rinsed three times with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde for 10 min. Following fixation, cells were washed three times for 10 min in PBS and incubated for 10 min with 50 mM glycine before being blocked for 1 h with blocking solution containing PBS with 5% normal goat serum (Jackson ImmunoResearch), 1% bovine serum albumin (Sigma), and 0.1% gelatin (Sigma). Cells transfected with HA-hIK1^L18A/L25A required an additional permeabilization step with 0.1% Triton X-100 (Sigma) in PBS and were permeabilized for 2 min after incubation with glycine. Cells were incubated with 5 μg/ml (1:200 dilution) anti-HA mouse monoclonal Alexa Fluor 594 conjugate (Invitrogen) in blocking solution overnight at 4°C in a humidity chamber. The use of a secondary antibody was not required since the primary antibody possessed a fluorescent conjugate. After overnight incubation the cells were washed three times for 10 min in PBS and incubated in 50 mg/ml 4′,6-diamidino-2-phenylindole (DAPI; Sigma) for 5 min. After a final PBS wash, coverslips were mounted on glass slides with AF1 glycerol/PBS (Citifluor) to prevent fading of the fluorescent signal. Images were collected on an Olympus BX5 upright microscope using a ×60/0.30 Plan Fln (Ph 3) oil immersion objective and captured using a Coolsnap ES camera (Photometrics) using MetaVue software (Molecular Devices).

Electrophysiology.

Cells were grown in 35-mm dishes and observed on an inverted microscope (Olympus IX71) while being continuously perfused with extracellular saline solution. Membrane currents were recorded using a Multiclamp 700B amplifier, with a Digidata 1322A interface and pCLAMP 9.2 software (Axon Instruments). For experiments measuring K⁺ currents, cells were perfused with an extracellular solution containing (in mM) 140 NaCl, 4 KCl, 1 CaCl₂, 2 MgCl₂, 1 KH₂PO₄, 10 HEPES, 10 glucose, pH 7.4 with NaOH. Pipettes were pulled from borosilicate glass (Kwik-Fil, World Precision Instruments) using a pipette puller (Narishige PC-10) and were fire polished with an MF 830 microforge (Narishige). Pipette resistance was between 3 and 7 MΩ when filled with intracellular solution. The intracellular solution contained (in mM) 135 KCl, 5.53 MgCl₂, 0.207 CaCl₂, 10 HEPES, 5 HEDTA, pH 7.2 with KOH, to give 3 μM free Ca²⁺. Cells were voltage clamped at −60 mV, and whole cell currents were activated by a voltage-ramp protocol (from −100 mV to 100 mV, 200 ms duration). Current-voltage relationships were determined by a voltage-step protocol (from −100 mV to 110 mV, in 10-mV increments, 100 ms per step). Application of 10 μM TRAM-34 (Sigma) to inhibit the hIK1 current was via addition to the extracellular solution.

For experiments measuring Na_V1.5 currents, cells were perfused with an extracellular solution containing (in mM) 140 NaCl, 4 KCl, 1.8 CaCl₂, 0.75 MgCl₂, 5 HEPES, 10 glucose, pH 7.4 with NaOH. The intracellular solution contained (in mM) 120 CsF, 20 CsCl, 2 EGTA, 5 HEPES, pH 7.2 with CsOH. Inclusion of CsF in the intracellular solution has been shown to decrease rundown of Na_V currents (10). Cells were voltage clamped at −140 mV, and current-voltage relationships were determined by a voltage-step protocol (from −160 mV to 100 mV, in 10-mV increments, 50 ms per step). Data were analyzed offline using Clampfit 9.2 (Axon Instruments) and are represented graphically using Origin 6.1 (OriginLab). Current size was expressed as current density (pA/pF). Data are expressed as means ± SE. Statistical significance was determined using either a paired t-test or a Mann-Whitney rank-sum test as appropriate (P < 0.05 was considered significant).

Surface membrane expression assay.

To determine the expression of IK1 and related mutant channels (HA-IK1^GYG/AAA and HA-IK1^L18A/L25A) at the plasma membrane, cell surface immunoprecipitation (CS-IP) experiments were carried out as previously described (17). Briefly, HEK293 cells were transiently transfected in 10-cm dishes using LipofectAMINE 2000 and 5 μg of either plasmid (HA-IK1 and HA-IK1^GYG/AAA in pBud, and HA-IK1^L18A/L25A in pcDNA3.1). Untransfected HEK cells were used as a control. Following 24 h posttransfection, cells were washed in ice-cold PBS and blocked in 1% BSA-PBS, and cell surface channels were labeled with polyclonal HA antibody (1:500; HA.11, Covance, Richmond, CA) for 90 min at 4°C. Unbound antibody was removed by extensive washing in 1% BSA followed by washes (3×) in PBS. All steps were performed at 4°C to prevent endocytosis of the channel and/or antibody. The cells were then lysed [50 mM HEPES, 150 mM NaCl, 1 mM EDTA, 1% vol/vol Triton X-100, complete EDTA-free protease inhibitor (Roche), pH 7.4], and protein concentrations were normalized. The immune-complexes were then precipitated for 2 h at 4°C with protein G-Sepharose beads (Invitrogen) followed by three washes in lysis buffer. After the final wash, the pellet was resuspended in Laemmli sample buffer, and the resulting proteins were separated by SDS-PAGE (8% gel) and transferred to nitrocellulose for immunoblot using monoclonal HA antibody (1:1,000). In addition to the IP, 20 μg of total protein were set aside following cell lysis for immunoblot analysis. In this way, we were able to evaluate the level of protein expression for different HA-IK1 constructs. This blot was then stripped and reprobed for tubulin (monoclonal α-tubulin was obtained from Sigma-Aldrich, St. Louis, MO) to confirm equivalent protein loading.

Proliferation assays.

Proliferation assays using HEK293 hIK1 and untransfected HEK293 cells were set up by splitting the cells when they reached 70% confluency and seeding equal volumes of a single cell suspension in 35-mm dishes. Stock solutions of drug treatments were dissolved in DMSO or methanol and were applied to the culture media at the time of seeding to achieve the final concentration. Cell number was determined 3 days after seeding using a hemocytometer and is expressed as a percentage of untreated control cell number. Proliferation assays on transiently transfected HEK293 cells were carried out in the same way, except that 1 day after seeding, cells were transfected with a recombinant channel or mock transfected as described above. Cell number was determined 3 days after transfection and is expressed as a percentage of mock-transfected cell number. Transient transfection of HA-hIK1 resulted in an approximate transfection efficiency of 70%. All experiments were performed in duplicate, triplicate, or quadruplicate for each assay and were performed at least five times. Data were normalized and are expressed as a percentage of mock-transfected cell number (means ± SE). Statistical significance of the raw data (cell number) was determined using two-way analysis of variance with a Holm-Sidak post hoc test. P < 0.05 was considered significant.

Proliferation reagents.

PD98059 (ERK1/2 inhibitor which acts at MEK1/2), SB203580 (p38 MAPK inhibitor), and SP600125 (JNK inhibitor) were obtained from Calbiochem and dissolved in DMSO, and TRAM-34 (Sigma) was dissolved in methanol.

Fura-2 fluorescence in HEK293 cells.

HEK293 cells were loaded with 3 μM fura-2-acetoxymethyl ester (fura-2 AM, TefLabs) for 30 min at room temperature in HEPES-buffered physiological saline solution (HEPES-PSS) containing (in mM) 137 NaCl, 0.56 MgCl₂, 4.7 KCl, 1.28 CaCl₂, 1 Na₂HPO₄, 10 HEPES, 5.5 glucose, pH 7.4 with NaOH. Cells were then washed and incubated for 30 min in dye-free HEPES-PSS at 37°C and 5% CO₂, to allow deesterification of the dye to occur.

Cells were observed on an inverted microscope (Nikon Eclipse TE2000-S), using a Nikon ×40 oil immersion objective. Cells were excited with a xenon arc lamp (Cairn) at 340 and 380 nm (± 10 nm bandwidth for each wavelength) alternately via a monochromator (Cairn). The emitted fluorescence (at 510 nm) was separated from excitation light using a 400 nm dichroic mirror (DM400, Chroma) and 510 nm broadband emission filter. The intensity of the 340 and 380 nm signals were recorded and the F₃₄₀/F₃₈₀ ratio was determined. Background-subtracted 340 and 380 nm fluorescence images were acquired every second by a CoolSNAP HQ CCD camera (Photometrics) controlled and processed by Metafluor image acquisition and analysis software (Molecular Devices). All experiments were performed at room temperature.

Measurement of store-operated Ca²⁺ entry in HEK293 cells.

To examine the Ca²⁺-dependency of hIK1-mediated cell proliferation, store-operated Ca²⁺ entry (SOCE) was monitored in HEK293 cells transfected with HA-hIK1 or HA-hIK1^GYG/AAA and compared with that of untransfected control cells. HEK293 cells were seeded on glass coverslips and cotransfected with mRFP and either HA-hIK1 or HA-hIK1^GYG/AAA DNA. Cells were loaded with 3 μM fura-2 AM to monitor intracellular Ca²⁺ levels as described above. Experiments were performed on single cells rather than cells within clusters. Transfected cells were selected on the basis of appearance of red fluorescence (due to coexpression of mRFP), while cells that did not exhibit red fluorescence were assumed to be untransfected. All experiments were internally controlled by comparing SOCE in transfected cells to SOCE in untransfected cells in the same field of view.

To activate SOCE, the following protocol was followed. Cells were perfused with Ca²⁺-free HEPES-PSS (with 1 mM EGTA), and intracellular stores were depleted by perfusion with Ca²⁺-free HEPES-PSS (with 1 mM EGTA) containing 30 μM cyclopiazonic acid (CPA). Once the stores were depleted, extracellular Ca²⁺ was replaced by perfusion with a 20 mM Ca²⁺-containing HEPES-PSS (containing 30 μM CPA), which provides a large driving force allowing extracellular Ca²⁺ to enter the cell by SOCE. All HEPES-PSS solutions were phosphate free to prevent precipitation of Ca²⁺ salts.

The change in F₃₄₀/F₃₈₀ ratio induced by SOCE was calculated for each cell and compared with the change in F₃₄₀/F₃₈₀ ratio induced by SOCE in the corresponding untransfected control cell. Data are expressed as a percentage (means ± SE) of the SOCE in untransfected control cells. Statistical significance was determined using a one sample t-test and Mann-Whitney rank-sum test. P < 0.05 was considered significant.

RESULTS

Overexpression of hIK1 increases HEK293 cell proliferation.

Treatment of HEK293 cells stably expressing hIK1 with the selective IK1 blocker TRAM-34 caused a significant reduction in cell number compared with untreated HEK293 hIK1 cells. Cells were treated with 10 μM TRAM-34, a concentration that inhibited peak IK1 current amplitude (Fig. 1A) by 78%. Currents were reduced from 35 ± 4 pA/pF to 6 ± 2 pA/pF (mean current at +70 mV; P < 0.01; n = 5). This however, had no significant effect on the small endogenous K⁺ current observed in untransfected HEK293 cells (P > 0.05; n = 5). After 3 days of treatment, 10 μM TRAM-34 significantly reduced HEK293 hIK1 cell number, to 76 ± 3% of untreated cells (P < 0.001; n = 36 from 9 experiments) but had no significant effect on cell number in untransfected HEK293 cells (94 ± 3% of untreated controls, P > 0.05; n = 20 from 5 experiments; Fig. 1B). This result suggests that the reduction in cell number by TRAM-34 in HEK293 hIK1 cells is due to an inhibition of the IK1 channel and not due to any nonspecific effects of TRAM-34.

Fig. 1.

Inhibition of human intermediate-conductance calcium-activated potassium channel 1 (hIK1) by 10 μM TRAM-34 decreased cell proliferation in human embryonic kidney 293 (HEK293) cells stably expressing hIK1. A: whole cell patch-clamp recordings of HEK293 hIK1 demonstrated a typical IK current (○), which was inhibited by 10 μM TRAM-34 (●; n = 5). Treatment of untransfected HEK293 cells with 10 μM TRAM-34 (■) had no significant effect on whole cell current (□; n = 5). V, voltage; I, current. B: proliferation assays revealed that treatment of HEK293 hIK1 cells with 10 μM TRAM-34 decreased cell numbers compared with untreated control cells (*P < 0.001; n = 36 from 9 experiments), while treatment of untransfected HEK293 cells with 10 μM TRAM-34 had no significant effect on cell numbers (P > 0.05; n = 20 from 5 experiments).

To further investigate the role of hIK1 in cell proliferation, HEK293 cells were transiently transfected with HA-hIK1 and cell number was compared with mock-transfected HEK293 cells. Figure 2A shows anti-HA staining (red) of a HEK293 cell transfected with HA-hIK1, revealing that HA-hIK1 was expressed at the cell membrane. Whole cell patch-clamp recordings (Fig. 2B) demonstrate a typical IK1 channel current, indicating that the channel was functional. Transient transfection of HEK293 cells with HA-hIK1 significantly increased cell number after 3 days to 161 ± 26% of mock-transfected cells (P < 0.001; n = 14 from 6 experiments; Fig. 2C). These results indicated a role for hIK1 in cell proliferation. However, it is worth highlighting that this is likely to be an underestimation of the true proliferation-inducing capacity of hIK1 expression, as the transfection efficiency of HA-hIK1 was only ∼70%.

Fig. 2.

hIK1-induced proliferation of HEK293 cells occurred independently of K⁺ flux and did not require hIK1 expression at the plasma membrane. HEK293 cells were transiently transfected with either hemagglutinin (HA)-hIK1, the pore mutant HA-hIK1^GYG/AAA, or the trafficking mutant HA-hIK1^L18A/L25A. A: anti-HA staining (red) revealed that HA-hIK1 and HA-hIK1^GYG/AAA were expressed at the cell membrane. However, coexpression of green fluorescent protein (GFP; green) with HA-hIK1^L18A/L25A confirmed that this mutant hIK1 channel was unable to reach the cell membrane. B: whole cell patch-clamp recordings from HEK293 cells transfected with HA-hIK1 demonstrated a typical IK current (○; n = 10). In contrast, transfection with HA-hIK1^GYG/AAA (●; n = 11) or HA-hIK1^L18A/L25A (□; n = 9) demonstrated that these channel mutants were unable to conduct K⁺ ions and were therefore nonfunctional. C: proliferation assays revealed that transient transfection with HA-hIK1 (*P < 0.001; n = 14 from 6 experiments), HA-hIK1^GYG/AAA (*P < 0.01; n = 14 from 6 experiments), or HA-hIK1^L18A/L25A (*P < 0.001; n = 18 from 6 experiments) increased cell numbers compared with mock-transfected cells.

The nonfunctional pore mutant of hIK1 increased cell proliferation.

To investigate the dependency of hIK1-induced cell proliferation on K⁺ efflux, HEK293 cells were transiently transfected with a nonconducting mutant hIK1 channel, hIK1^GYG/AAA. The GYG sequence within the selectivity filter was mutated to AAA rendering the channel nonconducting, similar to other studies on the HCN (40), K_IR (26), and GIRK families of K⁺ channels. Whole cell patch-clamp recordings (Fig. 2B) confirmed that the channel was unable to conduct K⁺ ions. Mean (± SE) currents at +60 mV were 7 ± 1 pA/pF (n = 11), which was significantly smaller than currents produced by transfection with HA-hIK1 (78 ± 20 pA/pF, n = 10; P < 0.001). It is worth noting that the current density of transiently transfected HA-hIK1 cells (78 ± 20 pA/pF, Fig. 2B) was much greater than stably transfected hIK1 cells (35 ± 4 pA/pF, Fig. 1A), most likely due to differences in expression. Anti-HA staining (red) of HEK293 cells transfected with HA-hIK1^GYG/AAA (Fig. 2A) revealed that the channel was correctly trafficked to the cell membrane. Therefore, the lack of current observed in Fig. 2B was due to the mutation of the selectivity filter. Transient transfection of HEK293 cells with HA-hIK1^GYG/AAA increased cell number after 3 days to 156 ± 20% of mock-transfected cells (P < 0.01; n = 14 from 6 experiments; Fig. 2C), a similar degree to that seen when transfecting with HA-hIK1. These results suggest that hIK1-induced cell proliferation is independent of K⁺ efflux but occurs via another mechanism that does not depend on its function as a conducting ion channel.

The trafficking mutant of hIK1 increased cell proliferation without plasma membrane expression.

An hIK1 trafficking mutant, HA-hIK1^L18A/L25A (17), was utilized to examine whether hIK1-induced proliferation required the channel to be expressed at the cell membrane. Anti-HA staining (red) of HEK293 cells transiently transfected with HA-hIK1^L18A/L25A (Fig. 2A) was observed in a ring surrounding the nucleus and in other intracellular compartments but not at the cell membrane. This confirmed that the channel was unable to reach the cell membrane consistent with previous studies (17). Whole cell patch-clamp recordings (Fig. 2B) demonstrated that HA-hIK1^L18A/L25A was unable to form a functional channel. Whole cell currents at +60 mV were 8 ± 2 pA/pF (n = 9), compared with 78 ± 20 pA/pF (n = 10) in HA-hIK1-transfected cells (P < 0.01). However, transient transfection of HEK293 cells with HA-hIK1^L18A/L25A increased cell number after 3 days, to 208 ± 13% of mock-transfected cells (P < 0.001; n = 18 from 6 experiments; Fig. 2C). This result indicated that hIK1-induced proliferation did not require hIK1 to be expressed at the cell membrane, but that hIK1 could induce proliferation even when located intracellularly.

Assessment of plasma membrane expression of HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A.

To further examine the total and relative expression of the “wild-type” and mutant IK1 channels (HA-hIK1^GYG/AAA and HA-hIK1^L18A/L25A) at the plasma membrane, we performed CS-IP assays. Briefly, this involved labeling the cell surface channels of intact cells by incubating with an anti-HA polyclonal antibody, and following extensive washes to remove unbound antibody, the cells were lysed and immune-complexes were immunoprecipitated using conventional methods. Each CS-IP was separated by SDS-PAGE and immunoblotted with a monoclonal HA antibody and compared with total cell lysates (Fig. 3A, representative blot from at least 3 representative experiments). Blots were stripped and reprobed with anti-tubulin antibody to ensure that equal amounts of protein were loaded into each lane (Fig. 3A, bottom; n = 3). Untransfected HEK293 cells were used as a negative control (first lane, Fig. 3A). CS-IP of wild-type (HA-hIK1) and mutant IK1 channels (HA-hIK1^GYG/AAA and HA-hIK1^L18A/L25A) was quantified using densitometry and normalized to HA-hIK1 (Fig. 3B, mean data from at least 3 separate experiments). On average, the cell surface expression of HA-hIK1^GYG/AAA was 63 ± 30% of HA-hIK1 surface expression (n = 3), whereas the cell surface expression of the trafficking mutant, HA-hIK1^L18A/L25A, was only 6 ± 4% of HA-hIK1 surface expression (Fig. 3B, n = 3). However, this small relative surface expression of HA-hIK1^L18A/L25A is likely to be much lower relative to total expression, which was 3.6 ± 0.6-fold greater than HA-hIK1 total expression (n = 3). These data suggest that cell surface expression and functional activity of hIK1 at the plasma membrane are not required per se for hIK1-induced cell proliferation.

Fig. 3.

Assessment of plasma membrane expression of HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A by cell surface immunoprecipitation. A: intact HEK293 cells expressing HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A were incubated with an anti-HA polyclonal antibody to label the cell surface channels. Following extensive washes to remove unbound antibody, cells were lysed and cell surface channels were immunoprecipitated (CS-IP), separated by SDS-PAGE, and immunoblotted (IB) with a monoclonal HA antibody (top). Equivalent total cell lysates from each corresponding cell type were run during parallel experiments and immunoblotted with a monoclonal HA antibody (middle). Blots were stripped and reprobed with anti-tubulin antibody as a loading control (bottom). Untransfected HEK293 cells were used as a negative control (first lane). B: mean data showing the densitometric quantification ( ± SE; n = 3) of cell surface expression (CS-IP) normalized to HA-hIK1 (n = 3).

TRAM-34 inhibits hIK1 but not hIK1^GYG/AAA or hIK1^L18A/L25A-induced cell proliferation.

Since TRAM-34 inhibited cell proliferation induced by stable expression of hIK1 (Fig. 1), the next logical step was to test the effect of TRAM-34 on cell proliferation induced by the transient expression of the wild-type and mutant hIK1 channels (HA-hIK1^GYG/AAA and HA-hIK1^L18A/L25A). In these experiments, 24 h posttransfection, cells were reseeded at a constant density (5 × 10⁵ cells) in the absence or presence of TRAM-34 (10 μM) and cells were counted 3 days later. Similar to Fig. 2, cell number was normalized to mock-transfected cells. Consistent with data in Fig. 2, transient transfection of HEK293 cells with HA-hIK1 or the mutant HA-hIK1^GYG/AAA and HA-hIK1^L18A/L25A significantly increased cell number after 3 days to 139 ± 2%, 131 ± 8%, and 132 ± 6% of mock-transfected cells, respectively (P < 0.05; n = 6–8 from 4 experiments; Fig. 4). Consistent with data in Fig. 1, TRAM-34 had no effect on mock-transfected cell number (102 ± 1%) but markedly inhibited the increase in cell number induced by transient transfection of HA-hIK1 (101 ± 4% compared with 139 ± 2% without TRAM-34; P < 0.05). However, to our surprise TRAM-34 had no significant effect on the increase in cell number induced by transient transfection of the mutant HA-hIK1^GYG/AAA (125 ± 9%) or HA-hIK1^L18A/L25A (127 ± 5%). These data suggest that TRAM-34 inhibits wild-type hIK1 channel-induced cell proliferation but has no effect on “mutant” hIK1 channel-induced cell proliferation.

Fig. 4.

TRAM-34 inhibits hIK1 but not hIK1^GYG/AAA or hIK1^L18A/L25A-induced cell proliferation. HEK293 cells were transiently transfected with “wild-type” HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A, and following 24 h, cells were reseeded at a constant density (5 × 10⁵ cells) in the absence or presence of TRAM-34 (10 μM). Cells were counted 3 days later, and cell number was normalized to mock-transfected cells in time-matched parallel experiments (*P < 0.05; n = 6–8 from 4 experiments).

hIK1-induced cell proliferation did not require an enhanced Ca^{2 +} entry.

Since K⁺ efflux was not necessary for hIK1-induced proliferation in HEK293 cells, we hypothesized that enhanced Ca²⁺ influx driven by cell hyperpolarization would also not be essential. The Ca²⁺ dependency of hIK1-induced cell proliferation was examined by comparing the SOCE evoked in HEK293 cells expressing either HA-hIK1 or HA-hIK1^GYG/AAA with that of untransfected control cells. This was achieved by depleting the endoplasmic reticulum (ER) Ca²⁺ stores with CPA in the absence of external Ca²⁺. This resulted in a small increase followed by a small decrease in fura-2 340/380 ratio, due to leak of Ca²⁺ from the ER and the subsequent clearance of intracellular Ca²⁺ concentration. Addition of high external Ca²⁺ back to the cells resulted in a much larger rapid increase in fura-2 340/380 ratio, due predominantly to SOCE. The magnitude of this increase in fura-2 340/380 ratio was therefore used as an indirect measure of SOCE, and the response in hIK1 or hIK1^GYG/AAA-expressing cells was normalized to nonexpressing cells in the same experiment. The expression of hIK1 or hIK1^GYG/AAA was defined as the presence of red fluorescence due to the coexpression of mRFP (left image of Fig. 5, A and B). The lack of red fluorescence was therefore an indication of nonexpressing cells (indicated by the black box in right image of Fig. 5, A and B). Example traces of SOCE evoked a HA-hIK1 (gray trace, Fig. 5A) or HA-hIK1^GYG/AAA transfected cell (gray trace, Fig. 5B) compared with that of an untransfected control cells in the same field of view shown in the black traces in Fig. 5, A and B. The magnitude of SOCE evoked in HA-hIK1-transfected HEK293 cells was 380 ± 84% of the SOCE in untransfected control cells (P < 0.05, n = 5), while transfection with hIK1^GYG/AAA had no significant effect on SOCE (88 ± 6% of control) compared with untransfected control cells (P > 0.05, n = 6; Fig. 3C). The magnitude of SOCE evoked was significantly greater in cells transfected with HA-hIK1 than in cells transfected with HA-hIK1^GYG/AAA (P < 0.01), presumably because cells transfected with HA-hIK1 were able reach a more hyperpolarized membrane potential than the cells transfected with the hIK1^GYG/AAA mutant, therefore creating a greater driving force for Ca²⁺ influx into the cell. However, both HA-hIK1 and HA-hIK1^GYG/AAA increased cell proliferation by a similar degree, which indicates that the enhanced SOCE elicited by HA-hIK1 is not necessary for the observed increase in cell proliferation.

Fig. 5.

Enhanced store-operated Ca²⁺ entry (SOCE) elicited by hIK1 was not necessary for cell proliferation. HEK293 cells were cotransfected with monomeric red fluorescent protein (mRFP) and either hIK1 or hIK1^GYG/AAA and were loaded with fura-2 AM to monitor intracellular Ca²⁺. Transfected cells were identified by expression of red fluorescence. To activate SOCE, cells were perfused with Ca²⁺-free cyclopiazonic acid (CPA)-containing HEPES-buffered physiological saline solution (HEPES-PSS) until the intracellular Ca²⁺ stores were depleted. Once the stores were depleted, extracellular Ca²⁺ was replaced by perfusion with a 20 mM Ca²⁺-containing HEPES-PSS, which provides a large driving force allowing extracellular Ca²⁺ to enter the cell, causing a large increase in fura-2 340/380 ratio due predominantly to SOCE. Therefore, the magnitude of this increase in fura-2 340/380 ratio was used as an indirect measure of SOCE. This response in hIK1 or hIK1^GYG/AAA-expressing cells was normalized to nonexpressing cells in the same experiment, as defined by the coexpression of mRFP (SOCE,% of control in C). A: transient transfection of HEK293 cells with hIK1 (gray box/trace) increased SOCE compared with untransfected (black box/trace) control cells (P < 0.05; n = 6 cells from 5 experiments). B: in contrast, SOCE in cells transfected with hIK1^GYG/AAA (gray box/trace) was not significantly different from untransfected (black box/trace) control cells (P > 0.05; n = 10 cells from 6 experiments). C: mean magnitude of SOCE elicited by hIK1 or hIK1^GYG/AAA-transfected cells compared with untransfected control cells (expressed as a percentage of SOCE in untransfected cells). *P < 0.05.

Overexpression of Na_v1.5 in HEK293 cells has no effect on cell proliferation.

The effect of nonfunctional hIK1 channels on cell proliferation was unexpected and raised the question of whether the increases in cell number seen by transient transfection of HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A might be a response induced by overexpression of a membrane ion channel protein that is not implicated in cell proliferation. To address this question, HEK293 cells were transiently transfected with another large transmembrane protein, Na_V1.5, which is not known to be involved in cell proliferation. If the increases in cell number induced by transient transfection of HA-hIK1, HA-hIK1^GYG/AAA, and HA-hIK1^L18A/L25A were in fact due to an overexpression of large transmembrane proteins, then it was expected that transfection with Na_V1.5 would also increase cell proliferation compared with mock-transfected cells. Whole cell patch-clamp recordings (Fig. 6A) demonstrated a typical Na_V current representative of five such recordings, indicating that the channel was functional. Transient transfection of HEK293 cells with Na_V1.5 had no significant effect on cell number. Cell numbers were 97 ± 2% of mock-transfected cells (P > 0.05; n = 15 from 5 experiments; Fig. 6B). This result confirmed that overexpression of a large transmembrane protein itself was not sufficient to stimulate cell proliferation, but that the increase in cell number seen in hIK1-transfected cells was due to expression of the hIK1 channel.

Fig. 6.

Transient transfection of HEK293 cells with voltage-gated sodium channel Na_V1.5 confirmed that overexpression of large transmembrane proteins does not cause proliferation. A: whole cell patch-clamp recordings from cells expressing Na_V1.5 (n = 5). B: proliferation assays revealed that transfection with Na_V1.5 had no effect on cell numbers compared with mock-transfected cells (P > 0.05; n = 15 from 5 experiments).

hIK1-induced cell proliferation required the ERK1/2 and JNK MAP kinase pathways.

Since K⁺ efflux and enhanced SOCE were not necessary for hIK1-induced proliferation in transfected HEK293 cells, we hypothesized that hIK1-induced proliferation occurred via a mechanism that does not depend on its function as a conducting ion channel and instead involved regulation of intracellular signaling pathways. MAP kinase signaling pathways have been shown to be vital for cell proliferation in various cell types (1, 3, 7, 8, 14, 19, 20, 22, 24, 31). We investigated whether hIK1-induced proliferation requires MAP kinase signaling pathways by treating HEK293 hIK1 cells with specific MAP kinase inhibitors.

Treatment of HEK293 hIK1 cells with 20 μM PD98059, an ERK1/2 inhibitor, significantly reduced cell numbers after 3 days, to 60 ± 2% of untreated control cells (P < 0.001, n = 20 from 5 experiments; Fig. 7A). Inhibition of JNK with 3 μM SP600125 in HEK293 hIK1 cells also significantly reduced cell numbers after 3 days, to 73 ± 3% of untreated control cells (P < 0.001, n = 24 from 6 experiments; Fig. 7A). In contrast, inhibition of p38 MAP kinase with 30 μM SB203580 had no significant effect on cell number. Cell numbers were 98 ± 4% of untreated control cells (P > 0.05, n = 20 from 5 experiments; Fig. 7A). Treatment of untransfected HEK293 cells with each of the MAP kinase inhibitors had no significant effect on cell number compared with untreated control cells (Fig. 7B).

Fig. 7.

hIK1 may cause proliferation via direct regulation of MAP kinase signaling pathways. A: inhibition of ERK1/2 with 20 μM PD98059 (*P < 0.001; n = 20 from 5 experiments) and JNK with 3 μM SP600125 (*P < 0.001; n = 24 from 6 experiments) reduced cell numbers in HEK293 cells stably expressing hIK1. Inhibition of p38 MAP kinase with 30 μM SB203580 had no significant effect on cell numbers compared with untreated control cells (P > 0.05; n = 20 from 5 experiments). B: inhibition of ERK1/2, JNK, and p38 MAP kinase in untransfected HEK293 cells had no effect on cell numbers (P > 0.05; n = 20 from 5 experiments).

These results suggested that the decreases in cell number observed in HEK293 hIK1 cells treated with PD98059 and SP600125 were due to inhibition of the ERK1/2 and JNK MAP kinase signaling pathways and not due to any nonspecific effects of the compounds used. Cell proliferation induced by hIK1 in HEK293 hIK1 cells requires ERK1/2 and JNK, but not p38 MAP kinase signaling pathways.

DISCUSSION

The results of this study indicate a role for hIK1 in cell proliferation. However, the ability of hIK1 to induce cell proliferation is via a mechanism that does not depend on either its ability to conduct K⁺ ions or enhancing Ca²⁺ influx. This is because the nonconducting pore mutant, hIK1^GYG/AAA, increased proliferation by similar amounts to cells expressing hIK1. We also showed that hIK1 need not be expressed at the cell membrane, because the trafficking mutant hIK1^L18A/L25A also increased cell proliferation by similar amounts to hIK-expressing cells, despite only a small fraction (6%) being expressed at the cell surface. It is unclear which intracellular compartment the hIK1^L18A/L25A trafficking mutant was localized to, as organelle markers were not used. Nevertheless, these findings support the idea that hIK1 can regulate cell proliferation without functional K⁺ efflux at the plasma membrane. In addition, ERK1/2 or JNK inhibitors, but not p38 MAP kinase inhibitors, decreased proliferation in hIK1-expressing HEK293 cells, suggesting that both ERK1/2 and JNK pathways could be involved in hIK1-mediated cell proliferation. However, the specific molecular mechanism for how these signaling pathways are regulated by hIK1 to induce increased cell proliferation remains unclear and is beyond the scope of this study. One possible explanation is that Ca²⁺-dependent activation of hIK1 induces a conformational change that not only gates the channel, but also leads to the activation of proliferative MAPK signaling pathways. This could occur by either direct interaction with either ERK1/2 or JNK or by an indirect association with an intermediate protein, protein complex, or signaling pathway.

The current model to explain K⁺ channel-induced cell proliferation involves membrane hyperpolarization, which increases the driving force for Ca²⁺ entry and the consequent enhanced activation of Ca²⁺-dependent cell proliferative signaling pathways. However, these data clearly demonstrate regulation of proliferation in cells expressing nonconducting mutant hIK1 channels and do not lead to enhanced Ca²⁺ entry. This suggests that hIK1 has additional functional properties that lead to activation of cell proliferation. Such a phenomenon is not unprecedented, and several lines of evidence suggest that numerous ion channels are multifunctional and can perform functions other than that of simply mediating ion flux. For example, K_V1.3 channels can regulate β₁-integrin function (2), TRP-PLIK, a member of the long transient receptor potential family, exhibits kinase activity (27), the slowpoke (K_Ca channel) channel binding protein SLOB from Drosophila exhibits protein kinase activity (41), and the β-subunits of K_V1 channels function as aldo reductases, converting aldehydes to alcohols while converting NADPH to NADP⁺ (37).

While the idea of hIK1 regulating cell proliferation independently of conducting K⁺ ions contradicts the widely accepted model, there are other examples of K⁺ channels that can induce proliferation independently of K⁺ flux. Transfection of NIH 3T3 fibroblasts with the ether-à-go-go voltage-dependent K⁺ channel (EAG) or EAG^F456A (a nonconducting mutant containing a point mutation in the selectivity filter) significantly increased cell proliferation compared with vector controls (14). Additionally, HERG (human ether-à-go-go-related gene) mediates tumor necrosis factor-α-induced proliferation in several tumor cells lines independent of K⁺ conductance (33). Collectively, these findings corroborate our observations that hIK1-induced proliferation does not require K⁺ flux per se.

However, some K⁺ channels do require the ability to conduct K⁺ ions to regulate proliferation, including TASK3 (25) and TREK-1 (32). Given these contrasting examples and the enormous diversity within the K⁺ channel superfamily, it is unlikely that one hypothesis fits all cases of K⁺ channel-mediated proliferation. Instead, each family or members within a K⁺ channel family may mediate proliferation via a different or multiple mechanisms dependent on the individual characteristics of the channel and the cell type in which they are expressed.

The results of this study suggest that hIK1-induced proliferation can occur without enhanced store-operated Ca²⁺ entry (SOCE). In support of this finding, enhanced Ca²⁺ influx was also not an essential downstream mechanism for EAG-induced proliferation, since incubation of cells in Ca²⁺-free (EGTA containing) media, thereby preventing Ca²⁺ influx, had no effect on EAG-induced proliferation (14). Furthermore, cancer cells continued to proliferate in cell culture medium containing very little Ca²⁺ (0.01 mM), which would also reduce the driving force for Ca²⁺ entry (4). It has also been observed that mitogens can induce T lymphocyte proliferation in the absence of an increase in intracellular Ca²⁺ (11).

Taken together, the current study and the results of Hegle et al. (14) provide evidence that K⁺ channels can regulate proliferation via a mechanism that is not dependent on promoting cell hyperpolarization to enhance Ca²⁺ entry. Our data do not dispute the importance of Ca²⁺ signaling in mediating cell proliferation. Moreover, our data do not rule out the possibility that hIK1 and hIK1^GYG/AAA may couple to Ca²⁺ signaling pathways by a Ca²⁺ entry-independent mechanism, which could provide an explanation as to how the hIK1^GYG/AAA mutant also increased cell proliferation.

The results from this study suggest that the ERK1/2 and JNK, but not the p38 MAP kinase families, may be involved in hIK1-induced proliferation in HEK293 cells. In contrast, EAG-induced proliferation in NIH 3T3 fibroblasts was dependent on the p38 MAP kinase pathway but not the ERK1/2 or JNK MAP kinase pathways (14). The involvement of MAP kinase in cell proliferation is extremely complex, and the specific role of each MAP kinase family can vary between cell types and also within cell types depending on the mitogenic stimulus [e.g., Swiss 3T3 fibroblasts (20)]. MAP kinases are widely implicated in cell proliferation, with evidence for involvement of each of the three main MAP kinase families (ERK1/2, p38, and JNK). ERK1/2 activity has been reported to be required for proliferation of fibroblasts (22), endothelial cells (3, 24), and cancer cell lines (1). The JNK and p38 MAP kinase families are generally associated with cellular responses to stress and are linked to apoptosis (39). However, there is growing evidence for a role that supports cell survival and cell proliferation. It has been demonstrated that JNK is required for proliferation of fibroblasts (7), endothelial cells (8, 24, 31), and various cancer cell lines (7, 8), as well as the growth of tumors (8).

An important caveat to the interpretation of data from the current study is that an increase in cell number could equally be attributed to inhibition of cell death rather than an activation of cell proliferation. This is because the nature of our cell proliferation assay did not take into account cell viability. However, previous optimization of growth rates of untransfected HEK293 cells, where cell viability was determined using Trypan blue exclusion, showed that cell death only occurred in ∼1–5% of cells during any particular growth phase (preliminary unpublished observations). Therefore, any maneuver that would inhibit cell death is unlikely to increase cell number by 150–200%. This suggests that the observed increase in cell number following hIK1 overexpression is due to activation of cell proliferation rather than inhibition of cell death. Moreover, the involvement of hIK1 in cell proliferation is well documented, whereas there is little evidence linking hIK1 to cell death.

We have shown that hIK1-induced proliferation of HEK293 cells requires ERK1/2 and JNK activity; however, it is unknown how the two MAP kinases mediate proliferation and whether both or just one of the MAP kinases translocates to the nucleus. It is possible that cross talk between ERK and JNK occurs and that one of them acts as the final effector for proliferation. Inhibition of either ERK1/2 or JNK separately decreased VEGF-stimulated proliferation, indicating that activation of both the ERK1/2 and JNK pathways were required for proliferation (24), similar to our own data. Moreover, a recent study has shown that IK1 channels are necessary for prolactin-induced proliferation of breast cancer cells that involves the JAK2 signaling pathway, a potential intermediary step in the activation of the MAPK pathways (9). This further highlights the signaling complexity of the IK1-mediated cell proliferation.

It remains unclear how hIK1 signals to the nucleus to activate transcription of key components required for cell proliferation. Clues may be sought when considering the mechanism by which L-type voltage-gated Ca²⁺ channels (VGCC) can signal to the nucleus via the ERK1/2 pathway. Influx of Ca²⁺ ions, specifically through the L-type VGCC, bind to calmodulin (CaM) that is constitutively bound to the channel. This stimulates ERK1/2 and causes activation of transcription factors that drive the expression of a specific set of genes. Specificity is thought to arise through scaffolding proteins bound to the carboxyl terminal of the channel, which recruit activators of Ras/MAPK which can then be activated by CaM either directly or indirectly through a conformational change in the channel (6). Like the L-type VGCC, hIK1 contains a CaM-binding domain in its carboxyl terminal, where CaM is constitutively bound to the channel. Upon binding of Ca²⁺ to the CaM-binding domain, a conformational change occurs that causes the hIK1 channel to open. It is possible that in addition to channel opening, such a conformational change may facilitate the recruitment of activators of MAP kinase and thus induce cell proliferation, similarly to the L-type VGCC. Such a theory implies that the conformational change induced by Ca²⁺ binding or channel opening per se, rather than ion conductance, is required to induce cell proliferation. Although the results of the current study do not directly demonstrate this, such a phenomenon would help to explain how the nonconducting hIK1^GYG/AAA mutant was capable of inducing cell proliferation. This is because the gating mechanism and thus the conformational change induced by an increase in Ca²⁺ would be expected to be unaffected in hIK1^GYG/AAA mutant channels. This is potentially important because specific inhibitors of IK1, such as TRAM-34, inhibit K⁺ conductance and cell proliferation, suggesting that this drug is an allosteric modulator rather than a pore blocker. However, the binding site of TRAM-34 and its parent compound clotrimazole has been identified to be within the pore of the IK1 channel, suggesting that these drugs may act as pore blockers (38). However, to our surprise, TRAM-34 had no effect on hIK1^GYG/AAA or hIK1^L18A/L25A-induced cell proliferation, which seems counterintuitive. This is because if TRAM-34 behaves as an allosteric modulator, rather than a pore blocker, one might expect TRAM-34 to inhibit cell proliferation regardless of whether the channel can conduct K⁺ or reach the cell membrane. One possible explanation for this observation is that the specific mutations within the mutant channels (hIK1^GYG/AAA or hIK1^L18A/L25A) prevent TRAM-34 from binding to the channels. In the case of the hIK1^GYG/AAA mutant, the GYG region within the pore is only two amino acids downstream of one of the two critical residues (Thr²⁵⁰) of the putative TRAM-34 binding site (38). In fact, mutation of Thr²⁵⁰ has been shown to reduce TRAM-34 binding by 1,000-fold and markedly reduce IK currents (38). In the case of the hIK1^L18A/L25A, the mutated amino acids are within the NH₂ terminus of the channel, which is much further upstream from the putative TRAM-34 binding site. However, we know that the L18/L25 mutations alter trafficking and channel assembly (17) and that mutations (¹⁵RKR¹⁷/AAA) very close to the L18/L25 alter the gating of IK1 by reducing the P_o by fourfold at saturating Ca²⁺ concentration. This therefore suggests that mutations in the NH₂ terminus can alter gating in such a way that changes in the three-dimensional structure of the pore may in turn alter TRAM binding (16). In any case, the future design of allosteric modulators that could specifically modify the interaction with downstream proliferative signaling pathways without affecting K⁺ conductance, and thus membrane excitability, would be highly desirable. Such drugs would be very useful for the treatment of a variety of human pathologies including cancer, atherosclerosis, and ischemic heart disease.

In conclusion, the results of this study support a role for hIK1 in cell proliferation. Using a model system of stable and transient transfection of hIK1 in HEK293 cells, we have shown that the ability of hIK1 to increase proliferation does not depend on its ability to conduct K⁺ ions, nor does it depend on an enhanced Ca²⁺ influx. We have also demonstrated that hIK1 does not need to be expressed at the cell membrane to increase proliferation and that hIK1 may directly regulate cell signaling pathways such as ERK1/2 and JNK, to promote proliferation. We have provided further evidence that ion channels can have nonconducting functions and should not simply be viewed as transmembrane proteins that facilitate ion flux.

GRANTS

This work was supported by a Biotechnology and Biological Sciences Research Council (BBSRC) grant awarded to J. I. E. Bruce and by a British Heart Foundation grant awarded to I. M. Fearon while he was at the University of Manchester. J. E. Millership was funded by a BBSRC studentship. D. C. Devor was funded by National Heart, Lung, and Blood Institute Grants HL-083060 and HL-092157, and K. L. Hamilton was funded by a Lottery Health Board grant.

DISCLOSURES

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