Our protocol on human tissues has been approved by the Ethics Committee of our Institution (Comité d’éthique à la recherche du CHUM). All participants provided their written informed consent to participate in this study and all clinical investigation has been conducted according to the principles expressed in the Declaration of Helsinki. Tumor and matched, adjacent, non-neoplastic lung tissues were sourced from 26 patients with lung adenocarcinoma (AD), the most common subtype of NSCLC. Tissues were collected immediately after surgical resection at the CHUM Hôpital Notre-Dame (Montréal, Québec, Canada) or obtained from the tissue bank of the Respiratory Health Network of the FRQS. Tissues were dissected from surgical samples by a pathologist, who confirmed the diagnosis of primary lung cancer and established clinical and histological parameters. The different growth patterns were quantified using a 5% increment by a thoracic pathologist, who classified the AD tumor samples based on the latest IASLC classification (37), into the following sub-categories: i) AD in situ (1 patient), ii) lepidic-predominant AD [4 patients, including formely non-mucinous bronchioloalveolar carcinoma (BAC) class], iii) acinar-predominant AD (8 patients), iv) papillary-predominant AD (3 patients), v) micropapillary-predominant AD (2 patients) and vi) solid-predominant AD (4 patients) as well as vii) invasive mucinous AD (4 patients, formely mucinous BAC). As reported in Table I, among the 26 patients included in our study, 46% were men and 54% were women, with a median age of 66 years (range 39–77 years); 76% had a smoking history. No patients in our cohort had metastasis at the time of lung resection.

A549 and H460 lung cancer cells were purchased from the ATCC and maintained in culture medium (A549: DMEM, Gibco, Invitrogen, Burlington, ON, Canada; H460: RPMI, Gibco, Invitrogen) supplemented with 10% of fetal bovine serum (FBS, Gibco, Invitrogen), 2 mM L-glutamine (Gibco, Invitrogen) and 50 U/ml penicillin and 50 μg/ml streptomycin (Gibco, Invitrogen).

The commonly-employed ‘wound-healing assay’, which gives an estimation of the capacity of cells to migrate and proliferate (Fig. 1A) after mechanical injury, was used. Briefly, A549 and H460 cell monolayers were injured mechanically with a pipette tip (6 wounds/Petri dish), then washed to remove detached/injured cells, before photography with a Nikon camera under light microscopy (×4 enlargement). Marks on the Petri dishes allowed us to photograph the wounds exactly at the same place at various times, with measurement of the wound areas initially and after repair, by ImageJ software (National Institutes of Health), as described previously (38,39). The results, reported as the rates of wound closure in μm²/h, are compared to the controls and in the presence of K⁺ channel inhibitors, i.e., clofilium (Sigma-Aldrich, Oakville, ON, Canada), chromanol (Tocris, Bristol, UK), ergtoxin (Alomone Labs Ltd., Jerusalem, Israel) and astemizole (Tocris). Wound-healing rates were also evaluated in the presence of negative controls or KvLQT1 siRNAs (Invitrogen). For Ki67 staining during wound healing (Fig. 1A), confluent A549 cell monolayers, cultured on glass slides in Flexiperm (Dako Denmark, Glostrup, Denmark), were fixed in cold methanol at time 0 after injury and after 18 h of repair before staining with anti-Ki67 antibody (Dako), then Alexa Fluor 488 goat anti-mouse secondary antibody (proliferating cells stained in green) and finally counterstaining with DAPI (staining of nuclei in blue, Sigma).

The 2D cell migration rates (μm/h) of non-confluent A549 cells were evaluated by single-cell tracking with Axiovision software (Carl Zeiss, Toronto, ON, Canada) on sequence images captured at 5-min intervals over an 18-h period by digital camera connected to a Carl Zeiss microscope (Carl Zeiss, ×10 enlargement).

The 3D migratory capacities of A549 and H460 cells were evaluated by Boyden-type chamber migration assay (18,19). Briefly, cells obtained after trypsinization were counted, and cell viability was verified by trypan blue assay. The cell suspensions were then placed in the upper compartment of 8-μm pore filters (0.33 cm², ThinCerts-TC inserts, Greiner Bio-one; MJS Biolynx, Brockville, ON, Canada) coated on the lower side with a collagen matrix. After cell migration, the filters were washed with PBS, and the cells were fixed with paraformaldehyde-acetone solution, then stained with hematoxylin. Non-migrating cells of the upper compartment were scrapped-off with cotton-tipped applicators (Fisher, Napean, ON, Canada), whereas migrating cells that passed through the lower face of the filters were counted in 5 different randomly-chosen fields by a light-inverted microscope at ×20 magnification (18,19).

A549 and H460 cell growth was evaluated by counting cell number over a 72-h period. Briefly, the cells were cultured at low density (15,000/cm²) in 35-mm Petri dishes for 24 h in the absence of inhibitor. At this time (T0), they were exposed or not to K⁺ channel inhibitors for 24–72 h, then separated with trypsin-EDTA (0.05%, Gibco, Invitrogen) at times 0, 24, 48 and 72 h, before counting by hemacytometer. The absence of drug cytotoxicity was verified by trypan-blue exclusion assay. At the end of growth assays, A549 cells were fixed in ethanol (70%) for cell cycle analysis. After centrifugation, they were re-suspended in PBS supplemented with pancreatic RNase A (Roche, Mannheim, Germany). After 30 min, propidium iodide solution (1 mg/ml, Sigma-Aldrich) was added to each sample, and flow cytometry performed with a flow Coulter EpicsXL cytometer (Beckman Coulter, Inc., Mississauga, ON, Canada). Data were recorded for at least 10,000 events, and signals were analyzed by Expo32 software (Applied Cytometry Systems, Sheffield, UK).

As described previously (40), A549 cells (40–50% confluence) were exposed to control siRNAs (Stealth siRNA-negative control, 12935-400, 100 pmol, Invitrogen) or siRNA directed against KvLQT1 (Stealth™ siRNA duplex oligonucleotides; Kcnq1-HSS142716, Invitrogen, 100 pmol), mixed with Lipofectamine™ RNAiMax transfection reagent in OptiMEM (Invitrogen). Fresh, complete DMEM was added onto cells after 5-h exposure to siRNAs, and then replaced every 24 h for 72–96 h. Transfection efficiency >90% was observed under these experimental conditions (40).

Total RNA purified from A549 cells with TRIzol reagent (Invitrogen) was reverse-transcribed with MMLV reverse transcriptase (Invitrogen) in the presence of oligodT primers. cDNAs were amplified with Taq polymerase (Invitrogen), and specific primers designed from sequences of human KvLQT1 [Kv7.1 (41), NM_000218, 5′-taaggaagagccc aacactg-3′, 5′-cgatccttgctcttttctga-3′, 1 μM, 355-bp product], human ether-a-go-go 1 (hEAG1, Kv10.1, NM_002238.3, 5′-gag aacgtggatgagggcat-3′, 5′-gtagtggtccagcttacggg-3′, 1 μM, 96-bp product), human ether-a-go-go-related gene (hERG1, Kv11.1, NM_000238.3, NM_172056.2, 5′-ggccagagccgtaagttcat-3′, 5′-aagccgtcgttgcagtagat-3′, 1 μM, 74-bp product) and β-actin (hβ-actin, 5′-agagctacgagctgcctgac-3′, 5′-aaagccatgccaatctcatc-3′, 0.25 μM, 499-bp product), for normalization. Primer pairs were designed in distinct exons to avoid genomic DNA amplification. KvLQT1, EAG and ERG products were amplified for 30 cycles, while β-actin amplification was stopped after 18 cycles. RT-PCR products were finally separated on 1% agarose gel and stained with SYBR^® Safe (Invitrogen). Signals were detected by Typhoon Gel Imager and analyzed by ImageQuant software (Molecular Dynamics, Baie d’Urfe, QC, Canada). KvLQT1 silencing with siRNA was also verified by RT-PCR (40).

Immunoblotting was undertaken according to a well-established laboratory protocol (18,40). Total A549 proteins were solubilized in lysis buffer [150 mM NaCl, 50 mM Tris-HCl, pH 7.6, 1% Triton X-100, 0.1% SDS, protease inhibitor cocktail (Complete Mini EDTA-free protease inhibitor cocktail, Roche)] on ice for 30 min. Proteins from lung tissues were extracted by homogenization, with a glass putter for 2 min in lysis buffer (150 mM NaCl, 25 mM Tris-HCl, 1 mM EDTA, leupeptin, aprotinin, PMSF protease and orthovana-date phosphatase inhibitors). Cell lysates were centrifuged, supernatants collected, and protein concentration measured by Bradford assay. Proteins were separated by SDS-PAGE and transferred onto nitrocellulose membranes. After blocking, the membranes were incubated with anti-KvLQT1 (sc-10645, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) and anti-β-actin antibodies (CLT9001, Cedarlane Laboratory Ltd., Burlington, ON, Canada) on the same blot, to ensure equivalent loading. KvLQT1 antibody specificity was verified with its blocking peptide [sc10645P, Santa Cruz, (18)]. After washing, the membranes were incubated with donkey anti-goat (for KvLQT1, Santa Cruz Biotechnology, Inc.) and horse anti-mouse (for β-actin; Cell Signaling Technology, Boston, MA, USA) IgG linked to horseradish peroxidase for 1 h. The intensity of each specific band was quantified with ChemiDoc XRS+ Molecular Imager and Image Lab software (Bio-Rad, Mississauga, ON, Canada), then normalized to the β-actin signal.

The data are presented as mean ± standard error of the mean (SEM) and were compared with Statview software (SAS Institute, Cary, NC, USA). Statistical analyses were made using paired Student’s t-test, Wilcoxon signed rank test and one sample sign test. For comparison between more than two means, we used Friedman analysis followed by Dunn’s test. Differences were considered significant when p<0.05.

The presence of human KvLQT1, EAG1 and ERG1 mRNA was first detected by RT-PCR in lung AD A549 cells (Fig. 1B). To evaluate the contribution of these channels, A549 cells were exposed to various pharmacological Kv channel inhibitors in scratch-assays, which engages cell migration and proliferation processes (Fig. 1A). Neither astemizole (EAG and ERG channel inhibitor, 5 and 20 μM), nor ergtoxin (specifically blocking ERG channels, 100 nM) significantly affected A549 wound-healing rates (Fig. 1C and D). On the contrary, clofilium, which is frequently used to study KvLQT1 function in respiratory epithelia (42–44), dose-dependently decreased wound-healing rates of A549 cell monolayers (11, 21 and 31% inhibition in the presence of 5, 10 and 25 μM clofilium, respectively, Fig. 1E). Similarly, the KvLQT1 blocker chromanol reduced A549 wound healing (Fig. 1F).

To confirm the role of KvLQT1, we adopted a complementary approach with stealth KvLQT1 siRNAs to test if its specific silencing would subsequently affect wound-healing rates. First, the efficiency of KvLQT1 siRNA in downregulating KvLQT1 mRNA (Fig. 2A) and protein (Fig. 2B) expression was verified. We then found that partial KvLQT1 silencing elicited a significant decrease in wound-healing (28.9±6.9×10³ μm²/h, Fig. 2C), compared to A549 cells exposed to negative control siRNA (44.3±8.9×10³ μm²/h, i.e., 35% inhibition). Data from both pharmacological and molecular approaches thus point toward KvLQT1 involvement in the regulation of A549 wound healing. We then undertook complementary assays to dissect its implication in both cell migration and proliferation.

The effect of clofilium on 2D cell migration dynamics was assessed by single-cell tracking of sub-confluent A549 cells in time-lapse video-microscopy experiments. As illustrated in Fig. 3A, cell migration rates were reduced by 42% in the presence of 5 μM clofilium; no further inhibition (45%) was elicited by higher (10 μM) clofilium concentration. 3D cell motility, an important determinant of cancer propagation, was then estimated by Boyden-type chamber assay. It was observed that the number of cells migrating through the chambers was significantly reduced by clofilium (25, 31 and 42% inhibition at 5, 10 and 25 μM, respectively, Fig. 3B). Altogether, these data indicate that clofilium could depress 2D and 3D A549 cell motility.

Because cell proliferation is a crucial component of tumor growth, we then evaluated KvLQT1’s contribution to cell proliferation, by counting the number of A549 cells after a 24- to 72-h period in the presence of KvLQT1 inhibitors. We noted that 20 μM chromanol significantly reduced cell growth, compared to untreated cells (Fig. 4A). Similarly, increasing clofilium concentrations (from 5 to 25 μM) decreased A549 cell proliferation in a dose- and time-dependent manner (Fig. 4B). At the 72-h post-treatment time-point, we estimated that <5 μM of clofilium was required to inhibit 50% of cell growth.

It has been proposed that cell cycle control by K⁺ channels may be a possible mechanism whereby they could regulate cell proliferation (10,25–27,45). We thus decided to examine the effect of 24-, 48- and 72-h clofilium application on A549 cell cycle distribution, by flow cytometry. First, it has to be mentioned that clofilium, even at 25 μM concentration for 72 h, did not elicit cytotoxicity as indicated by the absence of significant accumulation of sub-G0 apoptotic cells (Fig. 4C, insert). Moreover, we discerned a dose-dependent effect of clofilium on the cell cycle with an increasing number of cells in the G0/G1 phase and subsequent reduction of cells in the S and G2/M phases (Fig. 4C). These results showed that KvLQT1 inhibition in A549 cells significantly curbed cell proliferation and impacted cell cycle progression.

The effect of clofilium was also assessed in a second lung cancer cell model, the H460 cell line. As depicted in Fig. 5, KvLQT1 inhibition severely reduced H460 cell growth (Fig. 5A) and migration in the Boyden-type chamber (Fig. 5B). Accordingly, healing rates, measured at 24 and 48 h after wounding, were affected by clofilium (Fig. 5C).

KvLQT1 expression levels were then examined by immunoblotting of human lung AD tumor tissues, compared to matched, adjacent, non-neoplastic lung parenchyma, from 26 patients (Fig. 6 and Table I). An elevation (1.5- to 7-fold) of KvLQT1 protein was apparent in AD tumor tissues from 17 among 26 patients studied (i.e., 65%). More precisely, KvLQT1 overexpression (>1.5-fold) was found in 1/1 patient with AD in situ, 3/4 patients with lepidic-predominant AD, 4/8 patients with acinar-predominant AD, 3/3 patients with papillary-predominant AD, 1/2 patients with micropapillary-predominant AD, 2/4 patient with solid-predominant AD and 3/4 patients with invasive mucinous AD. KvLQT1 overexpression was detected in both males and females, with or without smoking history (Table I).