CHRNA5 as negative regulator of nicotine ... - Oxford Academic

Carcinogenesis vol.32 no.9 pp.1388–1395, 2011 doi:10.1093/carcin/bgr090 Advance Access publication May 17, 2011

CHRNA5 as negative regulator of nicotine signaling in normal and cancer bronchial cells: effects on motility, migration and p63 expression Annette M.Krais1, Agne`s H.Hautefeuille1, Marie-Pierre Cros2, Vladimir Krutovskikh2, Jean-Marie Tournier3, Philippe Birembaut3, Ame´lie The´pot1, Anupam Paliwal2, Zdenko Herceg2, Paolo Boffetta4, Paul Brennan5 and Pierre L.Hainaut1, 1 Group of Molecular Carcinogenesis, 2Group of Epigenetics, International Agency for Research on Cancer (IARC), 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France and 3INSERM UMRS-903, IFR 53, Universite´ de Reims, Reims, France, 4Group of Lifestyle, Environment and Cancer; and 5 Group of Genetic Epidemiology, International Agency for Research on Cancer (IARC), 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France To whom correspondence should be addressed. Tel: þ33 4 72 73 85 32; Fax: þ33 4 72 73 85 75; Email: [email protected]

Genome-wide association studies have linked lung cancer risk with a region of chromosome 15q25.1 containing CHRNA3, CHRNA5 and CHRNB4 encoding a3, a5 and b4 subunits of nicotinic acetylcholine receptors (nAChR), respectively. One of the strongest associations was observed for a non-silent singlenucleotide polymorphism at codon 398 in CHRNA5. Here, we have used pharmacological (antagonists) or genetic (RNA interference) interventions to modulate the activity of CHRNA5 in nontransformed bronchial cells and in lung cancer cell lines. In both cell types, silencing CHRNA5 or inhibiting receptors containing nAChR a5 with a-conotoxin MII exerted a nicotine-like effect, with increased motility and invasiveness in vitro and increasing calcium influx. The effects on motility were enhanced by addition of nicotine but blocked by inhibiting CHRNA7, which encodes the homopentameric receptor a7 subunit. Silencing CHRNA5 also decreased the expression of cell adhesion molecules P120 and ZO-1 in lung cancer cells as well as the expression of DeltaNp63a in squamous cell carcinoma cell lines. These results demonstrate a role for CHRNA5 in modulating adhesion and motility in bronchial cells, as well as in regulating p63, a potential oncogene in squamous cell carcinoma.

Introduction Lung cancer is the most frequent cancer among men in many countries and represents the primary cause of cancer death worldwide. In Europe, it accounts for an estimated 268 600 annual new cases among men and 82 900 cases among women (1). Smoking is the main risk factor worldwide. The cumulative risk by age 75 of continuous smokers is in the order of 15–20%, versus 1% in never smokers (2). In 2008, three independent genome-wide association studies identified a lung cancer susceptibility region on chromosome 15q25.1, with a consistent measure of effect between the studies (OR 5 1.30, P 5 5 1020; OR 5 1.31, P 5 1.5 108; OR 5 1.32, P 5 1 1017; OR, odds ratio) (3–5). This region encompasses three cholinergic nicotinic receptor genes (CHRNA3, CHRNA5 and CHRNB4) encoding nicotinic acetylcholine receptor (nAChRs) subunits a3, a5 and b4 in neuronal and other tissues. Given the frequency of at risk genotypes in the population, this association is responsible for up to 14% of lung cancer cases irrespective of histological subtype (4). In 2009, the same CHRNA locus was identified as associated with chronic obstructive pulmonary disease (6). Abbreviations: BrdU, 5-bromo-2#-deoxyuridine; DMEM, Dulbecco’s modified Eagle’s medium; nAChR, nicotinic acetylcholine receptor; NHBE, normal human bronchial epithelial cells; OR, odds ratio; PBS, phosphate-buffered saline; SNP, single-nucleotide polymorphism; siRNA, small interfering RNA; TDC, three-dimensional cultures.

The strong association with CHRNA 3/5 locus has led to suggestions that nAChR variants may increase lung cancer risk indirectly of tobacco addiction. In one of the genome-wide association studies, an association was identified between the same genetic region and smoking quantity (5). However, recent studies in Asian populations in which never smokers are better represented have reported a similar association among never and ever smokers, as well as among men and women (7). These results suggest that the 15q25.1 region has an independent effect on lung cancer risk above that of its effect on smoking propensity. One of the strongest associations within the 15q25.1 region is observed for the non-silent single-nucleotide polymorphism (SNP) rs16969968 in CHRNA5, a G-to-A substitution encoding either aspartic acid (D) or asparagine (N) at codon 398. This residue is located in the second intracellular loop of the four-transmembrane structure of the nAChR a5 subunit. SNP rs16969968 is in linkage disequilibrium with a non-coding SNP in CHRNA3 (4). The functional impact of these SNPs is unknown. nAChRs are ionotropic ligand-activated neurotransmitter receptors expressed by neuronal and non-neuronal cells throughout the body [reviewed in ref. (8)]. They form pentameric protein complexes activated in response to the endogenous neurotransmitter acetylcholine or to exogenous agonists, such as nicotineor tobacco-specific nitrosamines NNN or NNK (9–11). Nicotine mimics the effect of acetylcholine by inducing a conformational change at the internal side of the membrane that allows exchange of cations (Ca2þ, Naþ, Kþ) between extracellular space and cytoplasm. In turn, Ca2þ influx opens the gates of voltage-activated calcium channels, further increasing intracellular Ca2þ with pleiotropic effects on pathways controlling cell motility, proliferation, differentiation and apoptosis (8,12,13). NAChRs are expressed in most epithelial cells including the bronchial epithelium and lung cancer cells (14– 16). Six different receptor subunits have been detected in lung cancers, a3, a4, a5, a7, b2 and b4. The main receptor forms in bronchial cells are homopentameric a7 receptors and heteropentameric receptors, such as the (a3b2)2a5 receptor (16,17). Signaling through nAChRs modulates motility and migration in several types of cultured epithelial cells including bronchial cells. While the proliferative effects of nicotine in these cells appear to be mediated through homopentameric a7 receptors, heteropentameric (a3b2)2a5 receptors contribute to cell motility and wound healing of the respiratory epithelium by modulating intracellular calcium influx in migrating cells (16,17). To better understand the potential impact of CHRNA5 in susceptibility to lung cancer, we investigated the role of this subunit on migration, adhesion and calcium influx in non-transformed bronchial cells and in lung cancer cell lines. Materials and methods Cell lines, drug treatment and transfections Human tracheal epithelial SV40-transformed cells (16HBE) were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% of fetal bovine serum, 1% of glutamine and 1% of antibiotics. Normal human bronchial epithelial cells (NHBE) were purchased from Lonza (Walkersville, MD) and used at passages 2–3. They were grown in bronchial epithelial cell basal medium (Lonza) supplemented with 0.4% bovine pituitary extract, 0.1% hydrocortisone, 0.1% human epidermal growth factor, 0.1% epinephrine, 0.1% transferrin, 0.1% insulin, 0.1% retinoic acid, 0.1% triiodothyronine, 0.1% gentamicin/amphotericin B (GA-1000) (Lonza). The human cancer cell lines A549 (lung carcinoma), H1299 (carcinoma, non-small-cell lung cancer), TE1 and TE13 (esophageal squamous cell carcinoma) were used. H1299, TE1 and TE13 were cultured at 37°C and 10% CO2 in RPMI 1640 (PAA Laboratories GmbH, Pasching, Austria) medium and A549 at 37°C and 5% CO2 in DMEM medium (PAA Laboratories GmbH ). Both media were supplemented with 10% of fetal bovine serum, 1% of glutamine (Sigma, Saint Quentin Fallavier, France) and 1% of antibiotics (100 U/ml penicillin and 100 lg/ml streptomycin, Sigma).

Ó The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

1388

CHRNA5 regulates motility, invasiveness and p63 expression in lung cancer

Cells were seeded 24 h prior to treatment by nicotine (1 lM) (Sigma), a-bungarotoxin (1 lM) (Tocris Bioscience, Ellisville, MO) or a-conotoxin MII (10 nM) (Tocris Bioscience) for 72 h. Transfections of small interfering RNA (siRNA) CHRNA3 (L-006137-00; Thermo Scientific, Rockford, IL), siRNA CHRNA5 (ID 118985, Applied Biosystems, Foster City, CA) and siRNA CHRNA7 (ID 7044, Applied Biosystems) were performed with HiPerfect transfection reagent (Qiagen, Hilden, Germany) in optiMEM (Gibco, GibcoBRL, Invitrogen, Carlsbad, UK) and 2.5 nM siRNA (A549, TE1, TE13 cells) or 7.5 nM siRNA (H1299 cells). As a control, a generic scramble sequence was used (silencer negative control #2 siRNA—ID AM4613, Applied Biosystems). After 48 h, the medium was removed, a second transfection was performed to enhance gene silencing and cells were further incubated for 24 h prior to analysis. Migration assay Measurement of cell migration through microporous membranes coated with BD MatrigelTM (Matrigel invasion chambers, 354480; BD Bioscience, Bedford, MA) was performed using a Boyden Chamber assay. The membranes were rehydrated for 2 h in serum-free medium at 37°C. The lower chamber was filled with 750 ll serum-containing medium and the upper chamber with 350 ll of serum-free medium. Drugs were added to the upper chamber. For NHBE cells, we used a bronchial epithelial cell basal medium depleted of supplements in the upper chamber and supplemented bronchial epithelial cell basal medium in the lower chamber. Cell suspensions (150 ll, 1 106 cells/ml) were added to the upper chamber and the Boyden Chambers were kept at 37°C for 6 h (H1299), 12 h (A549), 24 h (NHBE) or 26 h (16HBE). The chambers were washed with Dulbecco’s phosphate-buffered saline (PBS) buffer (PAA Laboratories GmbH) and cells were fixed with methanol (Sigma) for 10 min at 20°C. The cells were stained using GIEMSA (Sigma) for 10 min at 4°C and rinsed with water. The cells on the upper side of the membrane were scraped off and the membranes were washed and dried overnight. After cutting out the membrane, cells were counted on four areas of the same size and the average number of cells per square millimeter was calculated. For each cell line, the migration time was determined experimentally in order to obtain a number of cells between 20 and 100 per membrane area. At least three independent experiments were performed for each condition. Three-dimensional tissue culture assay Substrates of collagen, chitosan and glycosaminoglycanes were prepared as described previously (18). Three-dimensional cultures (TDC) were generated using a protocol derived from the one developed for skin tissue equivalents and were prepared by adding a suspension of 2 105 human fibroblasts/cm2 on top of the substrate in Dulbecco’s modified Eagle’s medium (DMEM with Glutamax-1; Life Technologies, Cergy Pontoise, France) supplemented with 10% fetal calf serum (HyClone, Logan, UT), antibiotics and 1 mM ascorbic acid 2-phosphate (Sigma). The medium was replaced every day. After 21 days, lung adenocarcinoma cells (A549 and H1299) were plated on the substrate colonized by fibroblasts at a concentration of 2 105 cells/cm2. The TDCs were kept in immersion in DMEM medium as described above with 1 mM ascorbic acid 2-phosphate for 7 days. Treatment with siRNA was started on the second day after seeding the cell. One TDC of each cell line was kept as control, one was treated with scramble RNA and one with siRNA CHRNA5. Experiments were performed in triplicate. TDCs were then brought to the air– liquid interface for an additional 14 days in medium, supplemented with 0.4 lg/mL hydrocortisone (UpJohn, St. Quentin en Yvelines, France), 5 lg/ml insulin (Lilly, Saint-Cloud, France), 8 mg/ml Bovine Serum Albumine (Sigma), 1 mM ascorbic acid 2-phosphate and antibiotics. TDCs were fixed with formaldehyde (4%) and embedded in paraffin. Five micrometer-thick sections were cut and stained with hematoxylin (Sigma). Immunofluorescence A549, TE1 and TE13 cells were cultured on glass cover slides. After treatment with siRNA as indicated, cells were rapidly washed with PBS and fixed with formaldehyde (4% in PBS) for 10 min at room temperature. After washing with PBS, cells were permeabilized with Triton X-100 (0.5% in PBS), blocked in Image-iTTM FX Signal enhancer solution (Invitrogen) and immunostained with primary antibodies anti-P120 (12H6) (monoclonal, 1:500; BD Transduction Laboratories, San Jose, CA) or Anti-ZO-1 (monoclonal, 1:100; Zymed, South San Francisco, CA). After washing, cells were incubated to secondary goat anti-mouse antibodies conjugated to Alexa FluorÒ 488 (1:200 in PBS; Invitrogen) before DNA staining with ToPro3 (1:300 in PBS; Invitrogen). Stained slides were mounted using antifading medium (Invitrogen). For p63 staining, signal amplification was obtained using a horseradish peroxidase system. After permeabilization with Triton (0.5% in PBS), cells were treated with a H2O2 solution (5% in PBS; Sigma) prior to incubation on blocking

buffer. The cells were then immunostained with the primary antibody antip63 (4A4, monoclonal, 1:500; GenTex, San Antonia, TX), washed again and incubated with peroxidases-conjugated secondary antibody, goat anti-mouse immunoglobulin, (GAM-HRP, 1:200 in PBS; Dako, Glostrup, Denmark) as secondary antibody. Cells were then exposed to tyramide-coupled fluorescent dye cyanine 3 (1:100 in amplification buffer; Perkin Elmer Life Sciences, Boston, MA) before DNA staining with ToPro3 (1:300 in PBS). The fluorescent samples were examined with a Zeiss laser scanning confocal microscope LSM 510 on Axiovert 200M. The LSM 4.0 sp1 (Zeiss, Iena, Germany) software was used for image acquisition and analysis. Real-time quantitative polymerase chain reaction Total RNA was extracted using a Nucleospin RNA II kit (Macherey-Nagel, Du¨ren, Germany) and reverse transcription was performed using random primers (Promega) and SuperScript II (Invitrogen). RNA expression was analyzed by real-time quantitative polymerase chain reaction using Brilliant SYBR Green Master Mix (Qiagen) and 0.4 mM of primers as described previously (19). Expression levels were normalized to GAPDH. All samples were analyzed in duplicate and three independent real-time quantitative polymerase chain reactions were performed for each experiment. Protein extraction and western blotting Cells were lyzed by adding lysis buffer [50 mM Tris–HCl, pH 7.4, 250 mM NaCl, 0.1% sodium dodecyl sulfate, 0.5% Nonidet P-40 (NP-40)] containing protease inhibitors [2 lg/ml aprotinin, 500 lM phenylmethylsulfonyl fluoride, 0.5 lg/ml leupeptin, 1 pg/ml pepstatin and 2 mM dithiothreitol (Sigma)] and placed on ice for 30 min. Extracts were subsequently centrifuged at 13 000 r.p.m. for 15 min at 4°C. Supernatants were collected. Protein concentration was determined using Bradford reagent (Bio-Rad; Bio-Rad Laboratories, Temeculah, CA). Thirty micrograms of protein extracts were separated on 7.5% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidinedifluoride (Amersham Biosciences, Piscataway, NJ) membranes by electroblotting. Membranes were blocked with 5% dry milk in PBS–NP-40 (0.5%) for 1 h at room temperature and incubated overnight with primary antibody diluted in 0.1% PBS–NP-40 containing 1% dry milk. The following antibodies were used: mouse monoclonal against p63 (4A4, 1:500 in PBS; Santa Cruz Biotechnology, Santa Cruz, CA) and Ku80 (1:40 000 in PBS; Serotec, Raleigh, NC). Membranes were washed and incubated with peroxidases-conjugated goat anti-mouse [1:10 000 in PBS (Dako)]. Proteins were detected using the ECLTM detection kit (Amersham Biosciences). Calcium influx The cell line H1299 (1 105 cells/ml) was seeded into glass chamber slides (Lab Tek, 154526; Thermo Scientific) and the cells were transfected with siRNA as described above. The medium was removed and the cells were incubated with a 1 lM solution of Fluo-4 (Invitrogen) in Ringer solution (pH 7.2 with NaOH, 320 mOsm). After 1 h at 37°C, cells were washed once with Ringer solution and then left in Ringer solution for another hour before analysis. NAChRs were stimulated by adding nicotine (final concentration 1 mM) directly into the Lab Tek chamber. Prior to nicotine treatment, cells were incubated for 30 min at 37°C with a-bungarotoxin (final concentration 1 lM) or a-conotoxin (final concentration 10 nM). The Fluo-4 was excited with an Argon laser at minimum intensity using a ZEISS LSM 5 exciter microscope. Images were analyzed using the program Image J (http://rsbweb.nih.gov/ij/, Jan 2010). Flow cytometry DNA synthesis was analyzed by 5-bromo-2#-deoxyuridine (BrdU) incorporation and measured by flow cytometry. Briefly, 48 h after siRNA or scramble RNA treatment, cells were incubated with BrdU (final concentration: 1 lg/ml; BD Biosciences, San Jose, CA) for 90 min. Cells were detached, fixed with 70% ethanol and incubated with 20 ll anti-BrdU-FITC (BD Biosciences) for 20 min. BrdU incorporation was measured using FACS Canto II as described by the suppliers (Becton Dickinson). Data were analyzed with BD FACSDIVA Software. Sequencing SNP rs16969968 was analyzed by PCR and direct sequencing. DNA was isolated from cell lines using Qiagen DNA Extraction Mini kit according to the manufacturer instructions (QIAamp DNA Mini Kit; Qiagen). PCR was performed with Taq Platinum (Invitrogen) and 0.4 lM (final concentration) primers [5#-GCCTTTGGTCCGCAAGATA-3# (forward) and 5#-CATGCTGT TTACTCTGCCAAA-3# (reverse); Eurogentec, Seraing, Belgium]. PCR included 2 min activation at 94°C, followed by 40 cycles of the program at 94°C for 30 s, 55°C for 30 s, 72°C for 30 s and 10 min extension at 72°C. Two independent PCRs were performed and negative controls were run in

1389

A.M.Krais et al.

parallel to monitor possible contamination. PCR products were visualized on 2% agarose gel stained with ethidium bromide prior to sequencing. Five microliters of PCR products were purified with the enzyme ExoSap-IT (USB, Cleveland, OH) for 15 min at 37°C and 15 min at 80°C. Sequencing was performed on ABI Prism 3100 Genetic Analyser (Applied Biosystems) with the primers used for the amplification. Statistical analyses For migration assays using Boyden Chambers, each experiment was repeated at least three times. Cell counts were averaged for these triplicate experiments and standard errors were calculated. Statistical analysis was conducted using a paired t-test assuming that experimental conditions correspond to population of equal variance. P , 0.05 ( ) was considered statistically significant. These statistical analyses were repeated using the ADAM software version 2.54 (German Cancer Research Center 2008). Pairwise non-parametric Wilcoxon test without Bonferroni adjustment for multiple testing was performed to assess differences between treatment and control. The significance level alpha was set to 0.005 (Figure 1A) or 0.05 (Figure 1B and C).

Results Inhibition of CHRNA5 increases migration in non-transformed bronchial cells and in lung cancer cell lines Previous studies have shown that (a3b2)2a5 nAChRs are involved in cell migration and in the modulation of calcium influx during wound healing of the respiratory epithelium (16). To determine the nature of this effect in non-transformed human bronchial epithelial cells, we treated 16HBE cells (SV40-immortalized bronchial cells) and NHBE cells (normal human primary bronchial epithelial cells) with nicotine, a-conotoxin MII [Ctx, a specific inhibitor of nAChR complexes con-

taining a3b2 (20)] or a-bungarotoxin [Bgtx, a specific antagonist of a7 nAChR (21)]. Cell migration was quantified using Boyden Chamber assays. Migration times were experimentally determined to allow the counting of a number of cells between 20 and 100 per area at the lower face of the membrane in the control conditions. Figure 1A shows that in both cell types, migration significantly increased after treatment with either nicotine or Ctx (P , 0.005, Student’s t-Test; P , 0.005, pairwise non-parametric Wilcoxon test without Bonferroni adjustment). The effect of Ctx was blocked when cells were simultaneously treated with Bgtx. These observations suggest that blocking nAChR complexes containing a3b2 has a nicotine-like effect. To assess the role of the a5 nAChR subunit in cell migration, we silenced CHRNA5 expression using siRNA in two lung cancer cell lines with different CHRNA5 polymorphic status, A549 (rs16969968 GG, D398) and H1299 (rs16969968 AA, N398). Both cell lines were shown to express a wide range of nAChR subunits (Table I). The efficiency of gene silencing was assessed by reverse transcription polymerase chain reaction (see supplementary Figure S1, available at Carcinogenesis Online). In A549 and H1299 cells, expression of CHRNA5 was decreased by 80% compared with the siRNA control. Migration assays were performed in Boyden chambers with adapted migration times as detailed in Materials and Methods. Figure 1B shows the effects of nicotine, Ctx, Bgtx, siRNA to a5 or a7 nAChR, and combinations thereof, on the migration properties of A549 and H1299 cells. Treatment of either cell line with nicotine significantly increased invasiveness in both cell lines although the amplitude of this effect was greater in A549 (P , 0.05, Student’s t-test and P , 0.05, Wilcoxon test) than in H1299 cells (P , 0.05, Student’s t-test, P 5

Fig. 1. Effects of nicotine, antagonists and CHRNA silencing on the migration capacity of cancer and non-transformed bronchial cells. (A) Non-transformed bronchial cell migration using Boyden Chamber assay. NHBE or 16HBE cells were treated with nicotine (Nic, 1 lM) or a-conotoxin MII (Ctx, 10 nM) or both a-conotoxin MII and a-bungarotoxin (Ctx, 10 nM þ Bgtx, 1 lM). Results are expressed as percent of cells detected at the lower side of the Boyden Chamber membrane with respect to untreated cells, counted after 24 h (NHBE) or 26 h (16HBE), as explained in Materials and Methods. Each experiment was repeated three times. P , 0.005 (Student’s t-test); a , 0.005 (pairwise non-parametric Wilcoxon test without Bonferroni adjustment). (B) Effects of nicotine, antagonists and of CHRNA silencing on cell migration assay using Boyden Chamber assay, A549 cells (upper panel) and H1299 (lower panel). Cells were transfected with either control siRNA (white bars), CHRNA5 (gray), CHRNA7 (black) or both (hatched) siRNAs. They were then exposed to nicotine (Nic, 1 lM), a-bungarotoxin (Bgtx, 1 lM) or a-conotoxin MII (Ctx, 10 nM). Migration was expressed in percentage of cells at the lower side of the Boyden Chamber membrane with respect to control cells (transfected with control siRNA). At least three independent experiments were performed for each condition. P , 0.05 (Student’s t-test); P , 0.05 (pairwise non-parametric Wilcoxon test without Bonferroni adjustment). (C) DNA synthesis as measured by incorporation of BrdU in A549 and H1299 cells after silencing CHRNA5 or/and CHRNA7. Cells were transfected with siRNA 72 h prior to addition of BrdU. BrdU incorporation is expressed in percent with respect to cells transfected with control siRNA. Each experiment was repeated at least three times. P , 0.05 (Student’s t-test); a , 0.05 (pairwise non-parametric Wilcoxon test without Bonferroni adjustment).

1390


Table I. Expression of CHRNA subunits in bronchial or esophageal cell lines Cells

Characteristics

Origin

CHRNA subunits (DCt)a a3 a4 a5 a6 a7 b2 b4

A549 Adenocarcinoma Lung 31 H1299 Adenocarcinoma Lung 41 TE1 Squamous cell cancer Esophagus 31 TE13 Squamous cell cancer Esophagus 31 16HBE Human bronchial epithelium Lung ND NHBE Human bronchial epithelium Lung 32

33 32 30 32 ND ND

33 31 31 29 23 25

33 36 ND 36 ND ND

26 25 27 29 30 33

28 29 31 36 32 31

ND ND ND ND 28 28

a

Expression of messenger RNA of different nAChR subunits by RT-PCR and DCt (cycle threshold) were calculated as described in (22). ND (not detectable).

non-significant, Wilcoxon test). Enhanced migration was also observed with Ctx but not with Bgtx, the latter even slightly decreasing migration in H1299 cells. Upon silencing CHRNA5, migration significantly increased in both cell lines. This effect was enhanced by addition of nicotine but was blocked by Btgx, whereas Ctx had no significant effect. Finally, silencing CHRNA7 or of both CHRNA5 and 7 strongly reduced migration. Overall, these results suggest that A549 or H1299 cells responded to nicotine or nAChR antagonists in a manner similar to non-transformed bronchial cells. Blocking nAChR containing a5 with a-conotoxin MII or silencing CHRNA5 had a nicotine-like effect. Moreover, silencing CHRNA5 in A549 and H1299 cells increased the effects of nicotine. Taking into account that Ctx is a specific antagonist of receptor complexes containing a3 and b2 subunits which in bronchial cells are essentially represented by (a3b2)2a5 receptors, the lack of effect of Ctx in cells silenced for CHRNA5 suggests that inhibition of CHRNA5 expression suppresses signaling through these heteropentameric receptors. Thus, CHRNA5 appears to operate as a negative regulator of nicotine-induced cell migration. In contrast, inhibition of these effects in the presence of Btgx or after silencing CHRNA7 indicates that the homopentameric a7 receptor operates as a positive regulator of cell migration. To rule out that the above effects may be affected by changes in cell proliferation patterns induced by various treatments, we silenced CHRNA5 and/or CHRNA7 in A549 or H1299 cells and measured the incorporation of BrdU as a marker of S-phase. Figure 1C shows that BrdU incorporation significantly decreased in cells with silenced CHRNA5 or CHRNA7 and that this effect was more marked in H1299 (60% decrease) than in A549 (20% decrease) (P , 0.005, Student’s t-test and Wilcoxon test). To further evaluate the role of the nAChR a5 subunit on motility at the single cell level, we used videomicroscopy to quantify the twodimensional movements of A549 and H1299 cells. Silencing of CHRNA5 in A549 cells increased the distance covered by the cells (P 5 0.01, Kruskal–Wallis test) (supplementary Figure S2 is available at Carcinogenesis Online). Interestingly, in H1299, silencing of CHRNA5 did not influence their two-dimensional movements (data not shown), suggesting that migration (as measured in Boyden Chamber assays) and two-dimensional cell motility properties are distinctly affected by nicotine signaling. Effects on CHRNA5 silencing on cell adhesion properties To evaluate cell proliferation and migration in a TDC system, we seeded A549 or H1299 cells over a synthetic dermis equivalent made of polymerized glycosaminoglycans, chitosans and collagens, colonized with primary fibroblasts prior to cancer cell seeding (23). Comparison between cells transfected with either control or a5 siRNA showed that in both cell lines, the silencing of CHRNA5 increased the thickness of the reconstituted epithelium, blurred the delineation between the epithelium and the underlying dermis equivalent and increased the invasion of epithelial cells within the synthetic dermis layer (Figure 2A). These observations are consistent with an increased

Fig. 2. Increased invasion in vitro and decreased expression of adhesion markers after silencing CHRNA5. (A) Three-dimensional cell cultures of A549 or H1299 cells. Upper panels: control cells with control siRNA. Lower panel: cells with siRNA to CHRNA5 (bars: 50 lM). Arrows indicate cells infiltrating the culture substrate. (B) Immunofluorescence for adhesion molecules P120 (left) and ZO-1 (right) in A549 cells after silencing CHRNA5. Upper panels (green fluorescence): P120 and ZO-1. Lower panels: merged pictures of adhesion molecules (green) and nucleus (blue).

capacity of CHRNA5-silenced cells to invade the extracellular matrix. This increase in invasive capacity was observed with both A549 and H1299 cells but was particularly marked in the latter cell line (arrows on Figure 2A). Increased cell motility and migration suggest that nicotinic receptor signaling may alter the expression of components of cell–cell and/or cell–matrix adhesion complexes. To assess the effects of the a5 receptor subunit on cell adhesion complexes, we analyzed the distribution of P120-catenin and ZO-1, which are components of adherens junctions and tight junctions, respectively, in A549 cells. Figure 2B shows that silencing of CHRNA5 reduced the expression of both adhesion molecules, with an almost complete disappearance of P120-catenin from cell–cell contacts and reduction and disorganization of ZO-1 staining. Overall, these results show that silencing CHRNA5 led to changes in the expression of cell adhesion patterns, compatible with loss of adhesion properties that may account for increased motility and invasiveness. Effect of CHRNA5 silencing on calcium influx The nicotine-like effects of CHRNA5 silencing on cell motility, invasiveness and expression of P120-catenin and ZO-1 are consistent with previous reports showing that nicotine induced proliferation, invasion and epithelial-to-mesenchyme transition in a variety of human cancer cell lines (24). Nicotinic receptors belong to the family of ligand-gated calcium channels, the activation of which mediates Ca2þ influx. Our results suggest that, similar to nicotine, silencing CHRNA5 may increase calcium influx and may also further enhance the influx triggered by nicotine. To test this hypothesis, we measured Ca2þ influx in H1299 cells using fluorescent probe Fluo-4. Silencing CHRNA5 considerably increased the effect of nicotine (1 mM) on Ca2þ influx, consistent with a negative effect of CHRNA5 on Ca2þ influx (Figure 3A and B). To further understand the different roles

1391

A.M.Krais et al.

Fig. 3a. Effects of nicotine and antagonists on calcium influx in cells with silenced CHRNA5. Calcium influx in H1299 (A–D), 16HBE (E–F) and NHBE (G–E) was detected using the fluorescent marker Fluo-4. Panels A, C, E and G: microphographs of cells with artificial colors expressing Ca2þ levels. Color ranges from blue (low amount of Ca2þ) to green, yellow and red (high amount of Ca2þ). Panels B, D, F and G: quantitative analysis of Ca2þ influx in H1299 was performed using the Image J analysis program (http://rsbweb.nih.gov/ij/, Jan 2010). Nicotine was added at t 5 0 sec (final concentration of nicotine: 1 mM) to cells transfected with siRNA CHRNA5 (A and B) or treated with a-conotoxin MII (Ctx, 10 nM) or a-bungarotoxin (Bgtx, 1 lM) (C–G).

1392


of CHRNA5 and CHRNA7, we treated H1299 cells with either Ctx or Bgtx and measured the Ca2þ influx induced by nicotine. As shown in Figure 3C and D, Ctx increased calcium influx by .10-fold, whereas Bgtx almost completely inhibited calcium influx. Similar effects, although of lower amplitude, were observed in both non-transformed primary NHBE cells and immortalized 16HBE cells (Figure 3E–H). Overall, these results support that the expression CHRNA5 contributes to the formation of nAChR subunits that exert an inhibitory effect on calcium influx mediated by nicotine through the homopentameric a7 nAChR, as well as on the subsequent signaling pathways controlling cell motility and migration. Downregulation of CHRNA5 decreases the expression of p63 Recent studies have shown that p63, a protein with structural similarities with the tumor suppressor protein p53, is a critical regulator of epithelial cell differentiation, adhesion and survival, including bronchial cells (25–27). In bronchial cells as in most epithelial cells, p63 is expressed as multiple isoforms including in particular TA and DeltaN isoforms, the latter being devoid of the N-terminal, transactivation domain of the TA isoforms. In squamous cell lung carcinoma, TP63, the gene encoding p63, is often amplified and systematically overexpressed, leading to the accumulation of the DeltaNp63 isoform, which is constitutively expressed only in the basal cells of the normal epithelium (28). To determine if silencing of specific CHRNA subunits may affect p63 expression, we used two squamous cell carcinoma cell lines derived from esophageal cancers, TE1 and TE13. We selected these cells for this experiment on the basis of their previous characterization as expressing either low (TE1) or high (TE13) levels of DeltaNp63a (29,30). These two cell lines express multiple nAChR subunits (Table I). In both cell lines, CHRNA3, 5 or 7 were silenced and the efficiency of gene silencing was assessed by reverse transcription

polymerase chain reaction (supplementary Figure 1C is available at Carcinogenesis Online). Figure 4A shows that silencing of either CHRNA3 or CHRNA5 induced a decrease in DeltaNp63 messenger RNA levels as measured by reverse transcription polymerase chain reaction, this effect being significant for CHRNA3 silencing in TE13 and for CHRNA5 silencing in both cell lines. In contrast, no significant reduction was observed with siRNA to CHRNA7. Figure 4B shows the reduction in nuclear levels of DeltaNp63 protein after silencing either CHRNA3 or CHRNA5, whereas silencing CHRNA7 had no effect, and even increased DeltaNp63 expression in TE13 cells. Western blot analysis of DeltaNp63a protein levels (Figure 4C) showed a strong reduction of protein levels in both cells silenced for CHRNA5. With siRNA to CHRNA3, DeltaNp63a became undetectable in TE1 and was reduced in TE13. In contrast, after CHRNA7 silencing, p63 protein levels were essentially unaffected and even increased in comparison with Ku80 levels in both cell lines. These results indicate that a7 subunits, on the one hand, and a3 or a5 subunits, on the other, have opposite effects on the regulation of DeltaNp63a. Discussion The strong association of loci encoding nAChR on 15q25.1 with risk for lung cancer or chronic obstructive pulmonary disease has been interpreted as the result of an increased propensity to smoking due to nicotine addiction in subjects carrying distinct nAChR variants [reviewed in ref. (31)]. However, the association is detectable in never smokers although there is controversy between different studies in this respect. A detailed analysis of CHRNA5 SNP rs16969968 in a total of 17 300 subjects from five lung cancer studies and four upper aerodigestive tract cancer studies has identified that subjects with the homozygous minor genotype (AA) smoked on average 1.2 cigarettes per

Fig. 4. Effects of CHRNA silencing on p63 expression. (A) RT-PCR analysis of DeltaNp63 expression in TE1 and TE13 cells after transfection with control siRNA or siRNA CHRNA3, CHRNA5 or CHRNA7. Changes in messenger RNA levels (DCt) were calculated with respect to cells treated with control siRNA. P , 0.05 and P , 0.005 (Student’s t-test). (B) Immunofluorescence staining of p63 (Pan-p63 4A4) in TE1 cells (upper panel) and TE13 cells (lower panel) after transfection with control siRNA or siRNA CHRNA3, CHRNA5 or CHRNA7. (C) Western blot analysis of p63 (Pan-p63 4A4) at 48 h after siRNA or scramble RNA transfection. The band at 73 kDa corresponding to DeltaNp63a was identified on the basis of its immunological reactivity and electrophoretic motility as detailed in ref. (30). Ku80 was used as loading control. TE1 cells (upper panel) and TE13 cells (lower panel).

1393

A.M.Krais et al.

day more than subjects with a GG genotype (P , 0.001). Moreover, the association between the variant and lung cancer risk was virtually unchanged after adjusting for smoking (32). There is biological evidence for a role for nicotine and nAChRs in regulating the proliferation, differentiation and migration of bronchial epithelial cells. Studies in normal bronchial cells have demonstrated that (a3b2)2a5 receptors were expressed during wound healing in the normal mucosa and that this expression was particularly high at the edges of the wounds (16). In contrast, the homopentameric a7 receptors, which are not encoded by genes located in a lung cancer susceptibility region, were involved in the control of the proliferation of human airway epithelial basal cells (17). In this study, we show that CHRNA5 plays a significant role in modulating adhesion and motility in both transformed and non-transformed bronchial cells, as well as in regulating the expression of p63, a potential oncogene in squamous cell carcinoma (25,28). Using a-conotoxin MII as a specific antagonist of heteropentameric receptor complexes (containing a5) or silencing of CHRNA5 expression by RNA interference, we show that inhibition of receptors containing the a5 subunit induces a nicotinetype effect and increases the effects of nicotine on cell migration in a three-dimensional invasion assay (Boyden Chamber). In contrast, silencing CHRNA7 or inhibition of a7 with the antagonist a-bungarotoxin decreased cell migration and counteracted the effects of inhibition of a5. Similar effects (although with different amplitudes) were observed in transformed and non-transformed cells, independently of the polymorphic status of rs16969968. The effects of nicotine, antagonists and RNA silencing on cell migration were compatible with the effects on calcium influx. In agreement with the nicotine-like effect of a5 inhibition, silencing of CHRNA5 or treatment with a-conotoxin MII increased the wave of Ca2þ induced by nicotine. Conversely, treatment with a-bungarotoxin inhibited Ca2þ influx almost completely. Again, similar effects were observed in transformed and non-transformed cells. Overall, these results are compatible with the hypothesis that the a5 subunit is a component of receptors operating as negative regulators of calcium influx through the homopentameric a7 receptor. The effects of a5 inhibition on cell motility and migration were correlated with changes in expression of cell–cell adhesion molecules,

with a reduction of components of tight junctions (ZO-1) and of adherens junctions (P120) in A549 cells silenced for CHRNA5. Diminished expression of adhesion molecules and increased invasiveness are two properties of epithelial cells undergoing epithelial-to-mesenchyme transition (33). Furthermore, silencing either CHRNA3 or CHRNA5 decreased the expression and nuclear accumulation of DeltaNp63, an isoform of the p63 protein that is constitutively expressed in basal and proliferating bronchial epithelial cells (27). Again, this effect was not observed upon silencing CHRNA7, which appeared to increase DeltaNp63 expression. These results support that nAChR a5 subunit is part of a receptor operating as repressor of adhesion and migration in bronchial cells. Of note, silencing CHRNA3 in bronchial cancer cell lines H1299 and H1650 had similar effects on Ca2þ influx as silencing CHRNA5 (34), suggesting that this repressor effect is a property of the (a3b2)2a5 receptor complex rather than a specific property of the a5 subunit. Recent studies have shown that the expression of several nAChR subunits is deregulated in primary lung cancer tissues (19,34,35). Moreover, CHRNA3, but not CHRNA5, is often hypermethylated and downregulated in cancer tissues, and a 30-fold upregulation of CHRNA5 expression has been observed in lung cancers as compared with the normal lung. Restoration of CHRNA3 by ectopic expression in the H1975 lung cancer cell line induced apoptotic cell death (34). These observations indicate that the expression and functions of (a3b2)2a5 receptor complexes are altered in primary lung cancers. Our observations do not explain how polymorphisms of CHRNA5 may affect lung cancer susceptibility. The non-silent polymorphism rs16969968 falls within a conserved region of the second intracytoplasmic loop of the protein. The frequent allele (G) encodes an aspartic acid (D) at position 398, whereas the rare A allele (carrying the increased risk of lung cancer) encodes an asparagine (N). It is possible that the high-risk 398N protein variant exerts a less potent negative effect on nicotine signaling than the 398D protein (Figure 5). Alternatively, polymorphisms in linkage disequilibrium with rs16969968 may modulate the expression of CHRNA5 (36,37). The expression and function of (a3b2)2a5 receptors may be critical for the integrity of the bronchial mucosa in the face of injury by physical, chemical, infectious or inflammatory factors that locally damage the mucosa. Depending upon

Fig. 5. Role (a3b2)2a5 nAChR complexes in regulating nicotine-induced calcium flux: a model. This diagram illustrates the negative role of (a3b2)25 nAChR complexes in the regulation of nicotine signaling through homopentameric a7 receptors and its impact on cell adhesion, migration/invasiveness and p63 expression. This model suggests that the susceptibility associated with CHRNA3 variants may be due to differences in their capacity to exert a negative regulation on signaling through a7 receptors. According to this hypothesis, the high-risk a3 protein variant (398N) may exert a less potent negative regulatory effect on nicotine signaling than the low-risk variant (398D) protein variant, thus making bronchial cells more susceptible to the proliferating or motility-promoting effects of nicotine mediated through a7 homopentameric receptor complexes.

1394


the activity of these receptors, the mucosa may undergo rapid repair and recovery or in contrast undergo modifications to form hyperplastic or metaplastic areas that are precursors of cancer lesions. Polymorphisms in CHRNA genes may thus influence cancer risk by modulating the dynamics of the normal bronchial epithelium under stress conditions (31). Another, non-exclusive mechanism may consist in enhanced invasiveness and metastatic capacity. Moreover, the effects of the CHRNA polymorphism at the mucosal level may be compounded by neuronal effects on addiction to nicotine, increasing tobacco consumption and exposure to tobacco carcinogens (38). Together, these effects may account for the role of CHRNA5 variations as risk factors for a broad range of lung cancer aetiologies and histologies, as well as in non-cancer lung diseases that involve extensive tissue repair and remodeling, such as chronic obstructive pulmonary disease (6). Supplementary material Supplementary Figures S1 and S2 can be found at http://carcin. oxfordjournals.org/ Funding French National Cancer Institute (INCA; Bio-SILC program). Acknowledgements The authors thank Be´atrice Vozar for technical assistance with reconstituted 3D-cell culture and Matthieu Boniol (IPRI, Lyon) and Simone Braun (DKFZ, Heidelberg) for help with statistical analyses. We thank Anne Beghin, Yves Tourneur, Sandrine Pouvreau and Vincent Jacquemont for their help and support regarding the experiments of microscopy and calcium influx. Conflict of Interest Statement: None declared.

References 1. Ferlay,J. et al. (2007) Estimates of the cancer incidence and mortality in Europe in 2006. Ann. Oncol., 18, 581–592. 2. Crispo,A. et al. (2004) The cumulative risk of lung cancer among current, ex- and never-smokers in European men. Br. J. Cancer, 91, 1280–1286. 3. Amos,C.I. et al. (2008) Genome-wide association scan of tag SNPs identifies a susceptibility locus for lung cancer at 15q25.1. Nat. Genet., 40, 616–622. 4. Hung,R.J. et al. (2008) A susceptibility locus for lung cancer maps to nicotinic acetylcholine receptor subunit genes on 15q25. Nature, 452, 633–637. 5. Thorgeirsson,T.E. et al. (2008) A variant associated with nicotine dependence, lung cancer and peripheral arterial disease. Nature, 452, 638–642. 6. Pillai,S.G. et al. (2009) A genome-wide association study in chronic obstructive pulmonary disease (COPD): identification of two major susceptibility loci. PLoS Genet., 5, e1000421. 7. Shiraishi,K. et al. (2009) Contribution of nicotine acetylcholine receptor polymorphisms to lung cancer risk in a smoking-independent manner in the Japanese. Carcinogenesis, 30, 65–70. 8. Albuquerque,E.X. et al. (2009) Mammalian nicotinic acetylcholine receptors: from structure to function. Physiol. Rev., 89, 73–120. 9. Arredondo,J. et al. (2006) Nicotinic receptors mediate tumorigenic action of tobacco-derived nitrosamines on immortalized oral epithelial cells. Cancer Biol. Ther., 5, 511–517. 10. Schuller,H.M. et al. (1998) Tobacco-specific carcinogenic nitrosamines. Ligands for nicotinic acetylcholine receptors in human lung cancer cells. Biochem. Pharmacol., 55, 1377–1384. 11. Schuller,H.M. (2007) Nitrosamines as nicotinic receptor ligands. Life Sci., 80, 2274–2280. 12. Grando,S.A. et al. (2007) Recent progress in understanding the non-neuronal cholinergic system in humans. Life Sci., 80, 2181–2185. 13. Schuller,H.M. (2009) Is cancer triggered by altered signalling of nicotinic acetylcholine receptors? Nat. Rev. Cancer, 9, 195–205.

14. Grando,S.A. (2006) Cholinergic control of epidermal cohesion. Exp. Dermatol., 15, 265–282. 15. Proskocil,B.J. et al. (2004) Acetylcholine is an autocrine or paracrine hormone synthesized and secreted by airway bronchial epithelial cells. Endocrinology, 145, 2498–2506. 16. Tournier,J.M. et al. (2006) alpha3alpha5beta2-Nicotinic acetylcholine receptor contributes to the wound repair of the respiratory epithelium by modulating intracellular calcium in migrating cells. Am. J. Pathol., 168, 55–68. 17. Maouche,K. et al. (2009) {alpha}7 Nicotinic acetylcholine receptor regulates airway epithelium differentiation by controlling basal cell proliferation. Am. J. Pathol., 168, 55–68. 18. Noblesse,E. et al. (2004) Lysyl oxidase-like and lysyl oxidase are present in the dermis and epidermis of a skin equivalent and in human skin and are associated to elastic fibers. J. Invest. Dermatol., 122, 621–630. 19. Lam,D.C. et al. (2007) Expression of nicotinic acetylcholine receptor subunit genes in non-small-cell lung cancer reveals differences between smokers and nonsmokers. Cancer Res., 67, 4638–4647. 20. Cartier,G.E. et al. (1996) A new alpha-conotoxin which targets alpha3beta2 nicotinic acetylcholine receptors. J. Biol. Chem., 271, 7522–7528. 21. Zhang,Z.W. et al. (1994) Neuronal acetylcholine receptors that bind alphabungarotoxin with high affinity function as ligand-gated ion channels. Neuron, 12, 167–177. 22. Schefe,J.H. et al. (2006) Quantitative real-time RT-PCR data analysis: current concepts and the novel ‘‘gene expression’s CT difference’’ formula. J. Mol. Med., 84, 901–910. 23. Auxenfans,C. et al. (2009) Characterisation of human fibroblasts as keratinocyte feeder layer using p63 isoforms status. Biomed. Mater. Eng., 19, 365–372. 24. Dasgupta,P. et al. (2009) Nicotine induces cell proliferation, invasion and epithelial-mesenchymal transition in a variety of human cancer cell lines. Int. J. Cancer, 124, 36–45. 25. Candi,E. et al. (2007) TAp63 and DeltaNp63 in cancer and epidermal development. Cell Cycle, 6, 274–285. 26. Carroll,D.K. et al. (2007) p63, cell adhesion and survival. Cell Cycle, 6, 255–261. 27. Daniely,Y. et al. (2004) Critical role of p63 in the development of a normal esophageal and tracheobronchial epithelium. Am. J. Physiol. Cell Physiol., 287, C171–C181. 28. Hibi,K. et al. (2000) AIS is an oncogene amplified in squamous cell carcinoma. Proc. Natl Acad. Sci. U S A, 97, 5462–5467. 29. Roman,S. et al. (2007) Downregulation of p63 upon exposure to bile salts and acid in normal and cancer esophageal cells in culture. Am. J. Physiol. Gastrointest. Liver Physiol., 293, G45–G53. 30. Thepot,A. et al. (2010) Intraepithelial p63-dependent expression of distinct components of cell adhesion complexes in normal esophageal mucosa and squamous cell carcinoma. Int. J. Cancer, 127, 2051–2062. 31. Brennan,P. et al. (2010) Genetics of lung-cancer susceptibility. Lancet Oncol., 12, 399–408. 32. Lips,E.H. et al. (2010) Association between a 15q25 gene variant, smoking quantity and tobacco-related cancers among 17 000 individuals. Int. J. Epidemiol., 39, 563–577. 33. Polette,M. et al. (2007) Beta-catenin and ZO-1: shuttle molecules involved in tumor invasion-associated epithelial-mesenchymal transition processes. Cells Tissues Organs, 185, 61–65. 34. Paliwal,A. et al. (2010) Aberrant DNA methylation links cancer susceptibility locus 15q25.1 to apoptotic regulation and lung cancer. Cancer Res., 70, 2779–2788. 35. Minna,J.D. (2003) Nicotine exposure and bronchial epithelial cell nicotinic acetylcholine receptor expression in the pathogenesis of lung cancer. J. Clin. Invest., 111, 31–33. 36. Falvella,F.S. et al. (2009) Transcription deregulation at the 15q25 locus in association with lung adenocarcinoma risk. Clin. Cancer Res., 15, 1837– 1842. 37. Falvella,F.S. et al. (2010) Promoter polymorphisms and transcript levels of nicotinic receptor CHRNA5. J. Natl Cancer Inst., 102, 1366–1370. 38. Changeux,J.P. (2010) Nicotine addiction and nicotinic receptors: lessons from genetically modified mice. Nat. Rev. Neurosci., 11, 389–401. Received November 25, 2010; revised March 18, 2011; accepted May 06, 2011

1395