- Email: [email protected]
KCa2.3 channel-dependent hyperpolarization increases melanoma cell motility
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
available at www.sciencedirect.com
www.elsevier.com/locate/yexcr
Research Article
KCa2.3 channel-dependent hyperpolarization increases melanoma cell motility Aurelie Chantome a,b , Alban Girault a,b , Marie Potier a,b , Christine Collin b,e,f , Pascal Vaudin b,e,f,1 , Jean-Christophe Pagès b,e,f , Christophe Vandier a,b,⁎,2 , Virginie Joulin c,d,2 a
Inserm, U921, F-37032 Tours, France Université François Rabelais, F-37032 Tours, France c CNRS, FRE 2939, F-94805 Villejuif, France d Institut Gustave Roussy, F-94805 Villejuif, France e Inserm, U966, F-37032 Tours, France f CHRU, F-37032 Tours, France b
AR T IC L E I NF O R M AT I O N
AB S TR AC T
Article Chronology:
Cell migration and invasion are required for tumour cells to spread from the primary tumour bed
Received 27 May 2009
so as to form secondary tumours at distant sites. We report evidence of an unusual expression of
Revised version received 21 July 2009
KCa2.3 (SK3) protein in melanoma cell lines but not in normal melanocytes. Knockdown of the
Accepted 22 July 2009
KCa2.3 channel led to plasma membrane depolarization, decreased 2D and 3D cell motility.
Available online 30 July 2009
Conversely, enforced production of KCa2.3 protein in KCa2.3 non-expressing cells led to the plasma membrane becoming hyperpolarized, and enhanced cell motility. In contrast, KCa3.1
Keywords:
channels had no effect on cell motility despite an active role in regulating membrane potential. Our
KCa2.3
data also suggest that membrane hyperpolarization increases melanoma cell motility and that this
SK3
occurs through the KCa2.3 channel. Our findings reveal a previously unknown function of the
Potassium channel
KCa2.3 channel, and suggest that the KCa2.3 channel might be the only member of the Ca2+-
Motility
activated K+ channel family involved in melanoma cell motility pathways.
Membrane potential
© 2009 Elsevier Inc. All rights reserved.
Melanoma cells
Introduction The KCa2.3 channel (SK3) is a small-conductance potassium channel (SKCa) belonging to calcium-activated potassium channel (KCa) family that comprises three isoforms: KCa2.1–3 [1]. The KCa2.3 channel controls the excitability of a nerve cell of the central nervous system by mediating after-hyperpolarization [2]. We recently suggested that the KCa2.3 channel participates in cancer cell migration, but the mechanisms were not fully characterized [3]. In view of the other physiological functions of
KCa2.3, its involvement in cancer cell migration was unexpected and therefore needs to be confirmed. The KCa2.3 channel has several functional features that may contribute to cancerous cell migration: i) cyclical activation and inhibition of the KCa2.3 channel, following intracellular Ca2+ oscillations and/or cell volume changes, allow changes of cell shape ii) the KCa2.3 channel, an activated-membrane molecule, routinely adopts a polarized cell location, especially at protrusions (filopodia or lamellipodia) [4,5] and iii) the activity of the KCa2.3 channel is sensitive to variations in the free Ca2+ concentration often
⁎ Corresponding author. Inserm, U921, F-37032 Tours, France. Fax: +33 247366226. E-mail address: [email protected] (C. Vandier). 1 Present addresses: INRA, UMR PRC 6175, Nouzilly, F 37380 France; Université François Rabelais, Tours, F-37032 France. 2 Christophe Vandier and Virginie Joulin contributed equally to this work. 0014-4827/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2009.07.021
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
associated with cell motility/migration [1]. Thus, although the KCa2.3 channel had not previously been physiologically related to any cell migration process, its physical and functional features are compatible with an involvement in cancer cell migration. One main cause for the malignancy of a cancer is the ability of tumour cells to spread from the primary tumour to form secondary tumours at distant sites. Before the formation of metastasis, cancer cells have to go through a series of sequential steps. First they invade adjacent tissues. Next, they pass through blood and lymph vessels and spread to distant permissive sites. Several mechanical processes (e.g. cell adhesion and motility) and various molecular events (e.g. acto-myosin contraction) govern tumour cell dissemination. Despite growing research efforts, the molecular mechanisms underlying metastasis formation are still fragmentary. Discovering and understanding the molecular mechanisms underlying metastasis formation will help in the design of innovative therapeutic strategies. This is of particular importance for melanoma because the metastatic forms of this cancer are largely refractory to existing therapies. The ability of melanoma to spread was simply thought to be a consequence of melanocytes originating from highly motile cells with enhanced survival properties [6]. Here, we report that KCa2.3 protein, which is produced in melanoma cells but not in untransformed melanocytes, forms functional KCa2.3 channels which hyperpolarize the plasma membrane and promote cell motility. In contrast, KCa3.1 channels had no effect on cell motility despite actively regulating membrane potential. Our data show that KCa2.3-dependent hyperpolarization increases melanoma cell motility.
Materials and methods Cell culture Four human metastatic melanoma cell lines were used in this study 518A2 (a gift from Dr. van der Minne L., Leiden, The Netherlands), SKmel28, HBL, and Bris (a gift from Dr. S. Priest, IGR, Paris, France). All cell lines were maintained in Dulbecco's modified Eagle's medium except for HBL that was maintained in Ham's F10 medium. Culture media were supplemented with 10% (v/v) foetal bovine serum. Normal human epithelial melanocyte (a gift from Dr. J.E. Branka, Effiscience, Rennes, France) were maintained in serum-free Melanocyte growth medium M2 according to the manufacturer's protocol (PromoCell, Heidelberg, Germany). Cells were grown in a humidified atmosphere at 37 °C (95% air, 5% CO2). The absence of mycoplasma contamination was verified regularly using the Mycotech kit (Invitrogen).
Constructs, transfection, and transduction To inhibit KCa2.3 expression, we constructed a lentiviral vector encoding a short hairpin RNA (shRNA) specifically targeting the human KCa2.3. The sequence encoding shKCa2.3 was obtained by PCR elongation of two partially complementary primers; this also allowed us to introduce two restriction enzyme sites to facilitate manipulations. Forward primer: shKCa2.3-BamH I 5′ ggATCCCCCCATTCCTggCgAgTACAATTCAAgAgATT; shKCa2.3-Hind III 5′ AAgCTTAAAAACCATTCCTggCgAgTACAATCTCTTgAATT, as a reverse primer. The underscored sequences are complementary. The
3621
shKCa2.3 fragment was inserted into pGEMT easy vector (Promega) and transferred into pH1, a plasmid expressing the shRNA from RNA pol III promoter (a gift from Anne Galy, Généthon-Evry), at Bgl II/Hind III sites. The H1 promoter linked to the shKCa2.3 sequence was then inserted between the Spe I and Cla I sites in a pHR'-derived vector [7], expressing the GFP from a CMV promoter and an IRES. For control experiments we constructed, following the same protocol as above, a lentiviral vector expressing an untargeted shRNA (pLenti-shRD). For this, we used the following primers, shRD1_S: 5′ggATCCCCgCCgACCAATTCACggCCgTTCAAgAg ACg 3′ and shRD1_AS: 5′AAgCTTAAAAAgCCgACCAATTCACggCCgT CTCTTgAACg 3′. For KCa2.3 over-expression, the full length rat KCa2.3 cDNA was cloned by recombination with the system developed by Invitrogen, a shuttle vector was generated containing the PCR product of the full length cDNA: forward primer: ggCCCCAAgATggACACTTCTgggCACTTC and reverse primer: TTAgCAACTgCTTgAACTTgTg. 293T cells were plated at 3 × 106 cells per 100 ml culture and grown overnight before transfection. Transfection was performed with 10 μg of pCMV-8.74 encoding HIV1 Gag-Pol and all accessory proteins, 5 μg of pH-CMV-G encoding the VSV-G and 5 μg of the lentiviral vector (pLenti-shKCA2.3 or pLenti-shRD), using a calcium phosphate transfection kit (Invitrogen) according to the manufacturer's instructions. After 24 h, the medium was removed and the cells were washed with PBS, then fresh medium was added. Supernatants were harvested 48 and 72 h later and filtered through a 0.45 μm pore-size Millex HA filter (Millipore). The lentiviral vectors were then concentrated on 20% sucrose cushion by centrifugation at 26 krpm for 90 min (Sorvall, Discovery 90SE). Viral titres were determined by transducing 293T cells with serial dilutions of the viral stock, and then counting GFP cells using FACS analysis. Melanoma cells were transduced with lentiviral vector at multiplicities of infection (MOI) of 1 to 3 in the presence of polybrene (4 μg/ml, Sigma). The transduction rate determined by counting GFP cells was close to 90%.
Cell proliferation assays Cell proliferation was determined using the tetrazolium salt reduction method (MTT), as described elsewhere [8]. Cells were seeded on 24-well plates at a density of 20,000 cells per well and measurements were performed in triplicate daily for 5 days. Note that drugs and high external potassium concentration used in trans-well migration assays had no effect on cell proliferation/ viability (48 h).
Two-dimensional (2D) and 3D motility assays Cell motility was analyzed in 24-well plates receiving 8-μm poresize polyethylene terephthalate membrane cell culture inserts (Becton Dickinson, France), as previously described [8]. Briefly, 4 × 104 cells were seeded in the upper compartment with medium culture supplemented with 10% of FBS (±drugs/high external potassium concentration). The lower compartment was filled with medium culture supplemented with 20% FBS (± drugs/high external potassium concentration) as a chemoattractant. Twodimensional motility assays were performed without coating except for HBL cell lines for which bovine fibronectin and gelatin (0.05%; 0.02%) were used (Sigma-Aldrich). Three-dimensional
3622
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
motility assays were assessed as above but membrane was covered with a Matrigel® matrix. After 24 h, stationary cells were removed from the top side of the membrane, whereas migrated cells in the bottom side of the inserts were fixed, stained, and counted in five different fields (magnification, ×200). At least three independent experiments were each performed in triplicate.
Electrophysiology Experiments were performed with cells seeded into 35-mm Petri dishes at 2000 cells per cm2. All current-clamp (I = 0) and voltageclamp experiments were performed using the conventional whole-cell recording configuration of the patch-clamp technique as previously described [8]. PCa solution was 6 and 6.4 respectively for SKmel28 and 518A2 cells. Briefly, experiments were conducted using Axopatch 200B patch-clamp amplifier (Axon Instrument) and data, digitized with 1322-A Digidata converter (Axon Instrument), were stored on a computer using Clampex of pClamp 9.2 software (Axon Instrument). The patch-clamp data was analyzed using both Clampfit 9.2 and Origin 7.0 software (Microcal Inc., Northampton, MA, USA).
RT-PCR Total RNAs were extracted using the Nucleospin RNA II kit (Macherey-Nagel). Standard protocols were used for RT-PCR experiments as described previously [3]. Briefly, 1 μg of total RNA was used for cDNA synthesis using hexanucleotide primers and a Ready-to-Go You-Prime First-Strand Beads kit (GE Healthcare, UK). Polymerase chain reaction (PCR) was performed using puRE Taq Ready to go PCR Beads kit (Amersham Biosciences, UK). The primers used were as follows: KCa2.1 5′ primer AGGGAGACGTGGCTCATCTA and 3′ primer TTAGCCTGGTCGTTCAGCTT, KCa2.2 5′ primer GACTTGGCAAAGACCCAGAA and 3′ primer CCGCTCAGCATTGTAAGTGA, KCa2.3 5′ primer TGGACACTCAGCTCACCAAG and 3′ primer GTTCCATCTTGACGCTCCTC, KCa3.1 5' primer CATTCCTGACCATCGGCTAT and 3' primer ACG TGC TTC TCT GCC TTG TT, 18S as house keeping gene 5′ primer CGCGGTTCTATTTTGTTGGTTTT and 3′ primer TTCGCTCTGGTCCGTCTTG. The PCR reaction for KCa2.1-3 was as follows: 1 cycle at 94 °C for 5 min, 40 cycles of 94 °C for 30 s, 57° for 30 s, then 1 cycle at 72 °C for 30 s and a final incubation at 72 °C for 5 min. Protocol for 18s amplification was: 1 cycle at 94 °C for 5 min, 25 cycles of 94 °C for 30 s, 60° for 30 s, 72 °C for 30 s then 1 cycle at 72 °C for 5 min A negative control without RT product was included. Experiments were repeated three times.
Western blot Whole-cell lysates were prepared with 5% sodium dodecyl sulphate and protease inhibitor cocktail (Sigma, France). Proteins were separated by denaturing SDS-PAGE and transferred onto polyvinylidene fluoride membranes. Primary antibodies used were: rabbit anti-KCa2.3 N-term (1:500; Sigma-Aldrich), mouse anti-GAPDH (1:2000 dilution; Sigma-Aldrich) and rabbit antiactin (1:1000; Santa Cruz Biotechnology). Antibody binding was revealed with anti-rabbit (1:10,000; Jackson Immuno-Research Laboratories) or anti-mouse (1:3000, SouthernBiotech) IgG coupled to horseradish peroxidase, using an enhanced chemiluminescence kit (Pierce). Actin or GAPDH antibodies were used for loading control experiments.
Solutions and drugs The physiological saline solution (PSS) had the following composition (in mM): NaCl 140, MgCl2 1, KCl 4, CaCl2 2, D-glucose 11.1 and HEPES 10, adjusted to pH 7.4 with NaOH. The high external K+ PSS solution (K50) and high external Na+ PSS solution used were prepared by adding 46 mM KCl or 46 mM NaCl (control ionic solution) to PSS. The pipette solution for the whole-cell recordings contained (in mM): K-glutamate 125, KCl 20, MgCl2 1, Mg-ATP 1, HEPES 10, and pH was adjusted to 7.2 with KOH and various concentrations of CaCl2 and EGTA were added to obtain calculated pCa = 6 (8.7 mM CaCl2 and 10 mM EGTA) or pCa = 6.4 (0.7 mM CaCl2 and 1 mM EGTA). High external K+ DMEM solution (K50) and a high external + Na (Na50) DMEM solution used for motility assays were prepared by adding 45 mM KCl or 45 mM NaCl (control ionic solution) to the DMEM solution. All the drugs were obtained from Sigma-Aldrich.
Statistics Data are expressed as means ± SEM (n = number of cells). Comparisons between two means were made using Mann– Whitney or paired t-tests as appropriate. For comparison between more than two means we used Kruskal–Wallis one way analysis of variance followed by Dunn's test. Differences were considered significant when p < 0.05. SigmaStat (version 3.0.1a, Systat Software, Inc.) was used for statistical analysis.
Results SKCa channel expression in melanoma cell lines and primary human melanocytes To investigate the production of SKCa channels in melanoma, we used RT-PCR assays with various primer pairs specific for human KCa2.1, KCa2.2 and KCa2.3 mRNAs to test four human metastatic melanoma cells lines (SKmel28, Bris, 518A2 and HBL) and a primary culture of normal human epidermal melanocytes (NHEM). No KCa2.1 transcript was detected using a primer pair allowing amplification of the SK1 mRNA variants [9] in any of the samples (Fig. 1A). This is in agreement with the reported expression pattern of KCa2.1, which has mainly been found in neuronal tissues [10]. Three KCa2.2 variants have been described: short and long N-terminus variants KCa2.2-S and -L, respectively [11], and a novel shorter isoform called KCa2.2-sh [12]. RT-PCR with primers amplifying all three KCa2.2 variants revealed their mRNAs in all cells tested. Unfortunately, we were unable to detect KCa2.2 proteins because commercially available antibodies directed towards a rat KCa2.2 were poorly specific for the human proteins [13]. Interestingly, KCa2.3 mRNAs were detected in Bris, 518A2 and HBL, but not NHEM or SKmel28. Western-blot analyses confirmed these findings (Fig. 1B).
The KCa2.3 channel is involved in melanoma cell motility One of the most prominent features of malignant melanoma is the fast generation of metastasizing cells, resulting in a poor
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
3623
Fig. 1 – SKCa expression in NHEM and in melanoma cell lines, SKmel28, Bris, 518A2, and HBL. (A) RT-PCR products from total RNA extracts using primers specific for the human KCa2.1-3 mRNAs. Plasmid containing human KCa2.1 cDNA was used as a positive control for KCa2.1 detection. (B) Western-blot detection of KCa2.3 protein in NHEM and melanoma cells. Equal amounts of total lysates were immunoblotted with anti-KCa2.3. MDA-MB-435s total lysates were used as positive control. NS: non-specific band.
prognostic. In order to metastasize, cancer cells must gain the ability to move. Following our observations on MDA-MB-435s [3], we evaluated whether KCa2.3 expression by melanoma also promotes cell motility. No specific KCa2.3 channel blocker was available, so we tested apamin, a blocker of all SKCa channels. In 2D trans-well cell motility assays (Fig. 2A), apamin treatment decreased the migration of KCa2.3-expressing cells by approximately 40%. Therefore, 2D cell motility of Bris, 518A2 and HBL melanoma cell lines appeared partially dependent on the activity of SKCa channels. Next we used Matrigel matrix assay to assess cell motility in a 3D environment. Apamin treatment significantly reduced 3D motility of MDA-MB-435s and 518A2 cell lines (Fig. 2B). These results indicate that both 2D and 3D motility of melanoma cells are in part dependent on SKCa activity. To study the role of KCa2.3, its expression was specifically silenced in 518A2 cells. Melanoma cells were transduced with lentivectors containing either an interfering shRNA specific to the KCa2.3 transcript (shKCa2.3) or a non-targeting shRNA (shRD) as control. Patch-clamp experiments revealed that the currentdensity at +30 mV was substantially lower in 518A2shKCa2.3 cells than in 518A2shRD cells (Fig. 2C). Importantly, the residual current was insensitive to apamin. KCa2.3 protein was barely detectable in 518A2shKCa2.3 cells by Western blot, but was abundant in control wild-type 518A2 (518A2WT) and 518A2shRD cells (Fig. 2A). Also, 518A2shKCa2.3 cells exhibited a lower migrating capacity than 518A2WT and 518A2shRD cells, and the remaining cell motility was insensitive to apamin. To confirm the specificity of the effects of KCa2.3-shRNA, we performed a second infection of 518A2shKCa2.3 cells with a lentivector expressing rKCa2.3 mRNA under a strong promoter. The over-expression of rKCA2.3 in 518A2shKCa2.3 cells restored the expression of the KCa2.3 channel as assessed from the presence of the protein (Fig. 2A). Furthermore, the KCa2.3-shRNA phenotype was completely reversed in transduced cells: the current-density reduction was prevented (Fig. 2C) and apamin-sensitive cell migration was restored (Fig. 2A). Thus, this reversion was due to the action of the KCa2.3 provided by the channel expressed following lentiviral transduction. NHEM and SKmel28 cells do not contain KCa2.3 protein (Fig. 1B). As expected, the motility of these KCa2.3-negative cells was not sensitive to apamin (Fig. 3A). To confirm that aberrant KCa2.3
expression promotes melanoma cell motility, we expressed KCa2.3 in SKmel28 cells and then evaluated the phenotypic consequences. Expression of rKCa2.3 in SKmel28 cells (Fig. 3A, top) led to an increase of the current-density that became sensitive to apamin (Fig. 3B). Two-dimensional motility assays also revealed an increase, of up to 40%, of the number of migrating cells (Fig. 3A). After apamin treatment, the number of migrating cells was similar to that of parental SKmel28 cells. These results are consistent with KCa2.3 channel activity making a substantial contribution to melanoma cell motility.
KCa2.3 channels but not KCa3.1 channels promote melanoma cell migration by hyperpolarizing the plasma membrane We used the current-clamp mode of the patch-clamp technique to measure the membrane potential in melanoma cells expressing or not KCa2.3 channel. Silencing KCa2.3 in 518A2 cells induced a significant depolarization of the plasma membrane from −66.5 ± 0.8 mV (n = 12) to −34.1 ± 3.0 mV (n = 12; p < 0.05, Mann– Whitney test), whereas enforcing KCa2.3 expression in SKmel28 cells induced a significant hyperpolarization from −47.8 ± 3.6 mV (n = 12) to −74.5 ± 0.7 mV (n = 12; p < 0.05, Mann–Whitney test). Moreover, an acute application of apamin induced a depolarization of 518A2shRD plasma membrane from −65.8 ± 1.62 mV to − 36.4 ± 2.73 mV (n = 5; p < 0.001, Paired t-test) and of SKmel28KCa2.3 plasma membrane from − 75.1 ± 0.82 mV to −36.4 ± 2.73 mV (n = 8; p < 0.001, Paired t-test). Thus, expression of KCa2.3 channel confers to melanoma cells a more negative membrane potential than cells that do not express this potassium channel (Fig. 4A). Therefore, if the KCa2.3 channel is involved in melanoma cell motility by hyperpolarizing plasma membrane cells, a sufficient depolarisation may inhibit migration of KCa2.3-expressing cells. An easy way to depolarize membrane of KCa2.3-expressing cells is to increase the extracellular K+ concentration. As expected, a 10fold (K50) increase in the extracellular K+ concentration resulted in a larger depolarization of 518A2shRD than of 518A2shKCa2.3 (Fig. 4B, top) due to a more negative “resting” membrane potential in cells expressing KCa2.3 channels. Concomitantly, the migratory capacity of 518A2shRD cells was reduced whereas there was no
3624
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
Fig. 2 – KCa2.3 channel promotes the motility of KCa2.3-expressing melanoma cells. (A) Effect of apamin on 2D cell motility by Bris, HBL, 518A2WT, shRD, and shKCa2.3 melanoma cells. Left, normalized cell motility was calculated as the ratio of the total number of migrating cells in the presence of apamin to the total number of migrating cells in control experiments. The normalized cell motility of 518A2shRD, shKCa2.3 and shKCa2.3 + rKCa2.3 was calculated as the ratio of the total number of migrating cells to the total number of 518A2 WT migrating cells in control conditions. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann–Whitney test. #p < 0.05, significantly different from 518A2 WT in control conditions using Kruskal–Wallis one way analysis of variance followed by Dunn's test. Top of the left panel, Western-blot analysis of KCa2.3 protein in 518A2 cells. Right, representative pictures of migrated melanoma cells observed in a part of a field. The number in brackets indicates the mean number of migrating cells counted in one field (magnification, × 200). (B) Effect of apamin on 3D cell motility of MDA-MB-435s and 518A2 cells. The normalized cell motility was calculated as the ratio of the total number of migrating cells in the presence of apamin to the total number of migrating cells in control experiments. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann–Whitney test. (C) KCa2.3 channel activity in 518A2 cells measured using the patch-clamp technique at pCa 6.4. The numbers in brackets indicate the number of cells. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann–Whitney test. #p < 0.05, significantly different from 518shRD in control conditions using Kruskal– Wallis one way analysis of variance followed by Dunn's test. WT; wild-type, shRD; non-targeting short hairpin RNA named random, shKCa2.3; short hairpin RNA specific for the KCa2.3 transcript, rKCa2.3; rat KCa2.3 cDNA.
significant change for the migration of 518A2shKCa2.3 cells (Fig. 4B, bottom). Also, when the plasma membrane was depolarized, the migration of 518A2shRD became insensitive to apamin (data not shown). Substitution of K+ by Na+ (Fig. 4B bottom) or the addition of mannitol (data not shown) did not affect 518A2 cell migration, indicating that the inhibition of migration of 518A2shKCa2.3 cells associated with the high extracellular K+ concentration was not due to the higher medium osmolarity. The membrane potential has been described to change K+ channel expression [14], so we tested the effect of K50-induced depolar-
ization on KCa2.3 expression. Using a western-blot analysis, we did not detect any difference in KCa2.3 expression (data not shown). Our findings indicate that the KCa2.3 channel promotes 518A2 cell motility associated with hyperpolarization of the plasma membrane. We next investigated the specificity of KCa2.3 action. If hyperpolarization is necessary, is it possible to trigger mobilization by inducing hyperpolarization through another K+ channel? RTPCR experiments have evidence of KCa3.1 products, a KCa channel, in NHEM and in the other melanoma cell lines tested (Fig. 5A). This
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
3625
Fig. 3 – Ectopic expression of KCa2.3 enhances the motility of melanoma cells lacking the KCa2.3 channel. (A) Effect of apamin on motility of NHEM and SKmel28 cells. Left, the normalized cell number was calculated as the ratio of the total number of migrating cells in the presence of apamin to the total number of migrating cells in control experiments. The normalized 2D motility of empty and rKCa2.3 cells was calculated as the ratio of the total number of migrating cells to the total number of SKmel28 WT migrating cells in control conditions. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann– Whitney test. #p < 0.05, significantly different from SKmel28 WT in control conditions using Kruskal–Wallis one way analysis of variance followed by Dunn's test. Top of the left panel, Western-blot analysis of KCa2.3 protein in WT, empty and rKCa2.3-SKmel28 cells. Equal amounts of total lysates were immunoblotted with anti-KCa2.3. Right, representative pictures of migrated melanoma cells observed in a part of a field. The number in brackets indicates the mean number of migrating cells counted in one field (magnification, ×200). (B) KCa2.3 channel activity in empty and rKCa2.3-SKmel28 cells measured by the patch-clamp technique at pCa6. Left, KCa2.3 channel activity in SKmel28 cells. Right, typical currents recorded in SKmel28 cells following voltage-clamp pulses from −100 mV to + 100 mV for 400 ms (holding potential = −70 mV). The numbers in brackets indicate the number of paired cells. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using the Wilcoxon test. #p < 0.05, significantly different from SKmel28 WT in control conditions using Mann–Whitney test. WT; wild-type, empty; empty lentivector, rKCa2.3; rat KCa2.3 cDNA.
channel is functional and regulates membrane potential like KCa2.3 channel. Indeed, TRAM-34, a specific KCa3.1 channel blocker [15], significantly depolarized the plasma membrane of 518A2 cells (1 μM TRAM-34; n = 3; from −67.6 ± 1.2 mV to −33.6 ± 3.3 mV; p < 0.001, paired t-test) but did not impair their cell motility (Fig. 5B). Since TRAM-34 exhibits a high avidity for proteins [16], the last experiments were performed in the presence of 10% SVF (same experimental conditions as for cell motility assays). Next, we inhibited KCa3.1 channels and other
voltage dependent K+ channels by treatment with 10 μM of clotrimazole [15]. This treatment also induced a large depolarization (data not shown) but still did not reduce 518A2 and SKmel28 cell motility (Fig. 5C). These experiments indicate that KCa3.1 channel activity or its inhibition had no effect on cell migration despite its active role in regulating membrane potential. Therefore, the melanoma cell motility induced by hyperpolarization appeared dependent on the KCa2.3 channel, and did not involve the KCa3.1 channel.
3626
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
Fig. 4 – KCa2.3 channels promote cell motility by hyperpolarizing the plasma membrane. (A) Membrane potential recorded in cells expressing KCa2.3 channel (KCa2.3+ cells; 518shRD, 518shKCa2.3+rKCa2.3 and SKmel28KCa2.3) and in cells that did not expressed KCa2.3 channel (KCa2.3− cells; 518shKCa2.3 and SKmel28empty). PCa solution was 6 and 6.4 respectively for SKmel28 and 518A2 cells. Boxes indicated the first quartile, the median and the third quartile, squares indicated the mean. #p < 0.001, significantly different between KCa3.2+ and KCa3.2− cells using Mann–Whitney test. (B) Top, examples of membrane potential recorded in one 518A2shRD and one 518shKCa2.3 cell using the patch-clamp technique in current-clamp mode. Membrane potentials were recorded in external solutions containing 5 mM K+ (K5) and 50 mM K+ (K50). Bottom, effect of K50 on 518A2shRD and shKCa2.3 cell motility. The normalized 2D cell motility was calculated as the ratio of the total number of migrating cells in the presence of K50 or Na50 to the total number of migrating cells in K5 experiments. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Kruskal–Wallis one way analysis of variance followed by Dunn's test.
KCa2.3 and KCa3.1 channels do not affect melanoma cell proliferation It has been reported that K+ channels may be involved in cell cycle progression of cancer cells [17]. We therefore explored KCa-dependent cancer cell proliferation. Cell proliferation was not affected in KCa2.3-negative 518A2 cells or in KCa2.3expressing SKmel28 cells (Fig. 6A). The same applied for the
KCa3.1 channel: two concentrations of TRAM-34 usually used for blocking KCa3.1 channels did not affect proliferation of either 518A2 or SKmel28 cells (Fig. 6B). In contrast, although 1 μM CLT had no effect on melanoma cell proliferation, increasing the concentration to 10 μM (a concentration 100 times higher than those used for specific inhibition of KCa3.1 channels) decreased the proliferation of both melanoma cell lines tested. High CLT concentrations do not have a specific effect on KCa3.1, and this
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
3627
Fig. 5 – KCa3.1-induced hyperpolarization did not promote cell motility. (A) RT-PCR products from total RNA extracts using primers specific for the human KCa3.1 mRNA. (B) Left, typical example of membrane potential recorded in one 518A2 cell in the absence or in the presence of 0.1 μM and 1 μM TRAM-34, using an external solution supplemented with 10% SVF. Right, effect of TRAM-34 on the motility of 518A2 cells. (C) Effect of 10 μM clotrimazole (CLT) on the motility of 518A2 and SKmel28 cells. The normalized motility was calculated as the ratio of the total number of migrating cells in the presence of TRAM-34 or CLT to the total number of migrating cells in control experiments. Columns, mean, bars, SEM. ns, p > 0.05, not significantly different from control conditions using Mann– Whitney test.
may explain why KCa3.1 channels have confusingly been suspected to be involved in cancer cell proliferation [17]. Here, we demonstrate that neither KCa2.3 nor KCa3.1 channels are involved in melanoma cell proliferation.
Discussion In the present study we report that KCa2.3 protein is produced in melanoma cells but not in untransformed melanocytes. Furthermore, we show that this channel promotes melanoma cell motility by hyperpolarizing plasma membrane. This new function of the KCa2.3 channel was revealed using a variety of approaches: (i) apamin treatment significantly reduced the migration of KCa2.3-expressing melanoma cells whereas it had no effect in KCa2.3-non-expressing melanoma cells; (ii) stable silencing of the KCa2.3 gene using a specific KCa2.3 shRNA inhibited 518A2 cell migration, with a remaining activity insensitive to apamin; conversely, the stable expression of KCa2.3 in
KCa2.3-defective melanoma cells, enhanced their migration, which became sensitive to apamin; iii) while patch-clamp experiments revealed that KCa2.3 and KCa3.1 channels regulate membrane potential, membrane potential per se had no effect on cell motility and iv) the increase of external potassium concentration only reduced SK3-dependant motility. Cancer cells, notably melanoma cells, may switch between mesenchymal movements (actin polymerisation with proteases activities) and amoeboid movements (actin–myosin contraction without protease activity) according to their microenvironment [18,19]. Interestingly, KCa2.3 channels appear to promote cell motility through several movement modes as suggested by 2D and 3D motility assays. Indeed, we previously showed that the KCa2.3 channel was a mediator of MDA-MB-435s breast cancer cell motility in a 2D environment and the present study shows a similar result in a 3D environment [3]. As MDA-MB-435s cells adopt an amoeboid form of movement in a 3D environment and a mesenchymal form of movement in a 2D environment [20], our results suggest that SKCa channels, and more exactly
3628
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
Fig. 6 – KCa2.3 and KCa3.1 channels are not involved in melanoma cell proliferation. (A) Proliferation rate of 518A2 and SKmel28 cells expressing or not KCa2.3 channels. (B) Proliferation rate of 518A2 and SKmel28 cells in the absence and in the presence of 0.1, 1 μM TRAM-34 and 1, 10 μM CLT. Cell proliferation was determined daily for 4 days using the tetrazolium salt reduction method (MTT). Columns, mean, bars, SEM, n = 3.
the KCa2.3 channel, may be involved in both types of cell movement. Expression of the KCa2.3 channel by melanoma cells appears explainable through their neural crest origin, as most other KCa2.3-expressing mature cells, i.e. neurons, glial cells and endocrine cells [12]. Re-expression of embryonic genes during cancer has been widely described [21]. It is therefore plausible that KCa2.3 channel expression during melanoma malignancy corresponds to the re-expression of an embryonic gene expressed by a neural progenitor common to neurons and melanocytes [22,23]. Of note, the tissue origin of MDA-MB-435s is in debate, since it has been shown that these cells express melanocytic markers and have the same genetic profile as the melanoma cell line M14 [24–26]. How do KCa2.3 channels regulate melanoma cell motility? Even if mesenchymal and amoeboid movements are driven by two different mechanisms, both are regulated by variations of the intracellular Ca2+ concentration. For example, [Ca2+] oscillations induce cell-rear detachment following calpain activation and are also necessary for actin–myosin contraction [27–29]. By changing the plasma membrane potential, activation of KCa2.3 channel may regulate these [Ca2+] oscillations (frequency, amplitude). Indeed, an increase of K+ efflux shifts the membrane potential towards negative values (hyperpolarization) and consequently increases the Ca2+ driving force
[3,30,31]. Inversely, a reduction of KCa2.3 channel activity may depolarize the plasma membrane and reduce Ca2+ entry. We have clearly demonstrated that KCa2.3-dependent K+ efflux regulate membrane potential of melanoma cells. KCa2.3 expression in SKmel28 cells induced membrane hyperpolarization and promoted cell motility whereas KCa2.3 inhibition led to membrane depolarization in 518A2 cells and inhibited cell motility. In addition, increasing extracellular K+ concentration depolarised plasma membrane cells and consequently reduced melanoma cell motility only in cells expressing KCa2.3 protein. It was found that K+ efflux could lead to water efflux and causes localized cell volume decreases that allow retraction of the rear cell body of migrating cells [32,33]. Nevertheless, if cell swelling was found to activate KCa2.3 channel in HEK 293 cells, endogenous KCa2.3 channel seems not to be activated by cell swelling in hepatoma cells [4,34]. More experiments are needed to determine if cell volume changes could regulate KCa2.3 channel motility. Does hyperpolarization induced by another KCa channels promote melanoma cell motility? KCa2.1 and KCa2.2 channels can be put aside since KCa2.1 is not expressed in melanoma cells and KCa2.2 was found to be not functional as we observed no apamin-sensitive currents in SKmel28 cells and 518A2shKCa2.3 cells. Concerning the large/big-conductance calcium-activated
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
potassium channels (BKCa channels), our previous study showed that despite their active role in regulating membrane potential those channels were not involved in MDA-MB-435s cell motility [8]. Consequently, we focused on the last KCa channel, the KCa3.1 channel, as it was previously involved in motility of non-cancerous cells (e.g. lung mast cells or human coronary smooth muscle cells) [16,35]. We find that the KCa3.1 channels had no effect on melanoma cell motility despite actively regulating membrane potential. A study with melanoma cells suggested that the KCa3.1 channel may have a role in SKmel28 cell motility, because treatment with charybdotoxin inhibited migration [36]. However, charybdotoxin is a non-specific K+ channel blocker, unlike TRAM34 a new specific KCa3.1 channel blocker developed by Wulff et al. [15]. With this compound, we show that the increase of melanoma cell mobility by hyperpolarization is KCa2.3 dependent. This suggests that KCa2.3 channel interacts in cooperation with a specific complex. Furthermore, a singular KCa2.3 channel spatial distribution was observed at cell protrusions, which may favour this interaction [4,5]. Such a coupling was already found between potassium and calcium channels [37]. The identity of the partner is not determined but it could be voltage-independent-calcium channel like transient-receptor-potential (TRP) that are known to be sufficient to activate Ca2+-regulated stimulatory pathways for cell motility [31,38–40]. As a general conclusion, we report that KCa2.3 protein is reexpressed in melanoma cells, leading to formation of a functional KCa2.3 channel; this channel hyperpolarizes the plasma membrane. Furthermore, we provide evidence from 2D and 3D motility assays that abnormal KCa2.3 channel expression in melanoma cells promotes a novel signal for cell motility. Therefore, these findings argue that it may be beneficial to assess KCa2.3 channel blockers as novel tools for prevention of melanoma metastasis.
[4]
[5]
[6] [7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Acknowledgments This work was funded by “Ligue contre le Cancer - Région Centre”, “Fondation Carrefour”, “INSERM” and “Cancéropole Grand Ouest”. A. C. and A. G. hold fellowships from the “INCa” and the “Region Centre”, respectively. We thank Dr. L. van der Minne and Dr. S. Priest for the kind gifts of melanoma cells and Dr. J.E. Branka for NHEM cells. We also thank Dr. S. Lidofsky for the kind gift of KCA2.3 plasmid and Dr. T. Peitersen for the SK1 plasmid. We thank Dr. M.L. Jourdan and E. Marais for their technical assistance, Prof. P. Bougnoux for helpful discussion and C. Leroy for secretarial support.
[15]
[16]
[17] REFERENCES [18] [1] A.D. Wei, G.A. Gutman, R. Aldrich, K.G. Chandy, S. Grissmer, H. Wulff, International Union of Pharmacology. LII. Nomenclature and molecular relationships of calcium-activated potassium channels, Pharmacol. Rev. 57 (2005) 463–472. [2] R. Hosseini, D.C. Benton, P.M. Dunn, D.H. Jenkinson, G.W. Moss, SK3 is an important component of K(+) channels mediating the afterhyperpolarization in cultured rat SCG neurones, J. Physiol. 535 (2001) 323–334. [3] M. Potier, V. Joulin, S. Roger, P. Besson, M.L. Jourdan, J.Y. Leguennec, P. Bougnoux, C. Vandier, Identification of SK3 channel
[19]
[20]
3629
as a new mediator of breast cancer cell migration, Mol. Cancer Ther. 5 (2006) 2946–2953. E.T. Barfod, A.L. Moore, M.W. Roe, S.D. Lidofsky, Ca2+-activated IK1 channels associate with lipid rafts upon cell swelling and mediate volume recovery, J. Biol. Chem. 282 (2007) 8984–8993. S. Liebau, B. Vaida, C. Proepper, S. Grissmer, A. Storch, T.M. Boeckers, P. Dietl, O.H. Wittekindt, Formation of cellular projections in neural progenitor cells depends on SK3 channel activity, J. Neurochem. 101 (2007) 1338–1350. V. Gray-Schopfer, C. Wellbrock, R. Marais, Melanoma biology and new targeted therapy, Nature 445 (2007) 851–857. R. Zufferey, T. Dull, R.J. Mandel, A. Bukovsky, D. Quiroz, L. Naldini, D. Trono, Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery, J. Virol. 72 (1998) 9873–9880. S. Roger, M. Potier, C. Vandier, J.Y. Le Guennec, P. Besson, Description and role in proliferation of iberiotoxin-sensitive currents in different human mammary epithelial normal and cancerous cells, Biochim. Biophys. Acta 1667 (2004) 190–199. B.M. Zhang, V. Kohli, R. Adachi, J.A. Lopez, M.M. Udden, R. Sullivan, Calmodulin binding to the C-terminus of the small-conductance Ca2+-activated K+ channel hSK1 is affected by alternative splicing, Biochemistry 40 (2001) 3189–3195. M.X. Chen, S.A. Gorman, B. Benson, K. Singh, J.P. Hieble, M.C. Michel, S.N. Tate, D.J. Trezise, Small and intermediate conductance Ca(2+)-activated K+ channels confer distinctive patterns of distribution in human tissues and differential cellular localisation in the colon and corpus cavernosum, Naunyn Schmiedebergs Arch. Pharmacol. 369 (2004) 602–615. T. Strassmaier, C.T. Bond, C.A. Sailer, H.G. Knaus, J. Maylie, J.P. Adelman, A novel isoform of SK2 assembles with other SK subunits in mouse brain, J. Biol. Chem. 280 (2005) 21231–21236. S.R. Murthy, G. Teodorescu, I.M. Nijholt, A.M. Dolga, S. Grissmer, J. Spiess, T. Blank, Identification and characterization of a novel, shorter isoform of the small conductance Ca2+-activated K+ channel SK2, J. Neurochem. 106 (2008) 2312–2321. R. Khanna, L. Roy, X. Zhu, L.C. Schlichter, K+ channels and the microglial respiratory burst, Am. J. Physiol. Cell Physiol. 280 (2001) C796–C806. M. Zanzouri, I. Lauritzen, F. Duprat, M. Mazzuca, F. Lesage, M. Lazdunski, A. Patel, Membrane potential-regulated transcription of the resting K+ conductance TASK-3 via the calcineurin pathway, J. Biol. Chem. 281 (2006) 28910–28918. H. Wulff, M.J. Miller, W. Hansel, S. Grissmer, M.D. Cahalan, K.G. Chandy, Design of a potent and selective inhibitor of the intermediate-conductance Ca2+-activated K+ channel, IKCa1: a potential immunosuppressant, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 8151–8156. K. Toyama, H. Wulff, K.G. Chandy, P. Azam, G. Raman, T. Saito, Y. Fujiwara, D.L. Mattson, S. Das, J.E. Melvin, P.F. Pratt, O.A. Hatoum, D.D. Gutterman, D.R. Harder, H. Miura, The intermediateconductance calcium-activated potassium channel KCa3.1 contributes to atherogenesis in mice and humans, J. Clin. Invest. 118 (2008) 3025–3037. N. Villalonga, J.C. Ferreres, J.M. Argiles, E. Condom, A. Felipe, Potassium channels are a new target field in anticancer drug design, Recent Patents Anticancer Drug Discov. 2 (2007) 212–223. V. Sanz-Moreno, G. Gadea, J. Ahn, H. Paterson, P. Marra, S. Pinner, E. Sahai, C.J. Marshall, Rac activation and inactivation control plasticity of tumor cell movement, Cell 135 (2008) 510–523. D.R. Croft, M.F. Olson, Regulating the conversion between rounded and elongated modes of cancer cell movement, Cancer Cell 14 (2008) 349–351. T.M. Kitzing, A.S. Sahadevan, D.T. Brandt, H. Knieling, S. Hannemann, O.T. Fackler, J. Grosshans, R. Grosse, Positive feedback between Dia1, LARG, and RhoA regulates cell morphology and invasion, Genes Dev. 21 (2007) 1478–1483.
3630
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
[21] M. Monk, C. Holding, Human embryonic genes re-expressed in cancer cells, Oncogene 20 (2001) 8085–8091. [22] S. Rasheed, Z. Mao, J.M. Chan, L.S. Chan, Is melanoma a stem cell tumor? Identification of neurogenic proteins in trans-differentiated cells, J. Transl. Med. 3 (2005) 14. [23] W.M. Klein, B.P. Wu, S. Zhao, H. Wu, A.J. Klein-Szanto, S.R. Tahan, Increased expression of stem cell markers in malignant melanoma, Mod. Pathol. 20 (2007) 102–107. [24] S. Sellappan, R. Grijalva, X. Zhou, W. Yang, M.B. Eli, G.B. Mills, D. Yu, Lineage infidelity of MDA-MB-435 cells: expression of melanocyte proteins in a breast cancer cell line, Cancer Res. 64 (2004) 3479–3485. [25] J.M. Rae, C.J. Creighton, J.M. Meck, B.R. Haddad, M.D. Johnson, MDA-MB-435 cells are derived from M14 melanoma cells—a loss for breast cancer, but a boon for melanoma research, Breast Cancer Res. Treat. 104 (2007) 13–19. [26] A.F. Chambers, MDA-MB-435 and M14 cell lines: identical but not M14 melanoma? Cancer Res. 69 (2009) 5292–5293. [27] R.S. Adelstein, Calmodulin and the regulation of the actin–myosin interaction in smooth muscle and nonmuscle cells, Cell 30 (1982) 349–350. [28] K. Clark, M. Langeslag, B. van Leeuwen, L. Ran, A.G. Ryazanov, C.G. Figdor, W.H. Moolenaar, K. Jalink, F.N. van Leeuwen, TRPM7, a novel regulator of actomyosin contractility and cell adhesion, EMBO J. 25 (2006) 290–301. [29] S.J. Franco, A. Huttenlocher, Regulating cell migration: calpains make the cut, J. Cell Sci. 118 (2005) 3829–3838. [30] J.N. Rao, O. Platoshyn, L. Li, X. Guo, V.A. Golovina, J.X. Yuan, J.Y. Wang, Activation of K(+) channels and increased migration of differentiated intestinal epithelial cells after wounding, Am. J. Physiol. Cell Physiol. 282 (2002) C885–C898. [31] J.N. Rao, O. Platoshyn, V.A. Golovina, L. Liu, T. Zou, B.S. Marasa, D.J. Turner, J.X. Yuan, J.Y. Wang, , TRPC1 functions as a store-operated Ca2+ channel in intestinal epithelial cells and regulates early
[32]
[33] [34]
[35]
[36]
[37]
[38]
[39]
[40]
mucosal restitution after wounding, Am. J. Physiol. Gastrointest. Liver Physiol. 290 (2006) G782–G792. A. Schwab, B. Schuricht, P. Seeger, J. Reinhardt, P.C. Dartsch, , Migration of transformed renal epithelial cells is regulated by K+ channel modulation of actin cytoskeleton and cell volume, Pflugers Arch. 438 (1999) 330–337. A. Schwab, V. Nechyporuk-Zloy, A. Fabian, C. Stock, , Cells move when ions and water flow, Pflugers Arch. 453 (2007) 421–432. N.K. Jorgensen, S.F. Pedersen, H.B. Rasmussen, M. Grunnet, D.A. Klaerke, S.P. Olesen, Cell swelling activates cloned Ca(2+)activated K(+) channels: a role for the F-actin cytoskeleton, Biochim. Biophys. Acta 1615 (2003) 115–125. G. Cruse, S.M. Duffy, C.E. Brightling, P. Bradding, Functional KCa3.1 K+ channels are required for human lung mast cell migration, Thorax 61 (2006) 880–885. A. Schwab, J. Reinhardt, S.W. Schneider, B. Gassner, B. Schuricht, K (+) channel-dependent migration of fibroblasts and human melanoma cells, Cell. Physiol. Biochem. 9 (1999) 126–132. E.Y. Kim, C.P. Alvarez-Baron, S.E. Dryer, Canonical transient receptor potential channel (TRPC)3 and TRPC6 associate with large-conductance Ca2+-activated K+ (BKCa) channels: role in BKCa trafficking to the surface of cultured podocytes, Mol. Pharmacol. 75 (2009) 466–477. M. Louis, N. Zanou, M. Van Schoor, P. Gailly, TRPC1 regulates skeletal myoblast migration and differentiation, J. Cell Sci. 121 (2008) 3951–3959. J. Waning, J. Vriens, G. Owsianik, L. Stuwe, S. Mally, A. Fabian, C. Frippiat, B. Nilius, A. Schwab, A novel function of capsaicinsensitive TRPV1 channels: involvement in cell migration, Cell Calcium 42 (2007) 17–25. R. Wondergem, T.W. Ecay, F. Mahieu, G. Owsianik, B. Nilius, HGF/SF and menthol increase human glioblastoma cell calcium and migration, Biochem. Biophys. Res. Commun. 372 (2008) 210–215.
available at www.sciencedirect.com
www.elsevier.com/locate/yexcr
Research Article
KCa2.3 channel-dependent hyperpolarization increases melanoma cell motility Aurelie Chantome a,b , Alban Girault a,b , Marie Potier a,b , Christine Collin b,e,f , Pascal Vaudin b,e,f,1 , Jean-Christophe Pagès b,e,f , Christophe Vandier a,b,⁎,2 , Virginie Joulin c,d,2 a
Inserm, U921, F-37032 Tours, France Université François Rabelais, F-37032 Tours, France c CNRS, FRE 2939, F-94805 Villejuif, France d Institut Gustave Roussy, F-94805 Villejuif, France e Inserm, U966, F-37032 Tours, France f CHRU, F-37032 Tours, France b
AR T IC L E I NF O R M AT I O N
AB S TR AC T
Article Chronology:
Cell migration and invasion are required for tumour cells to spread from the primary tumour bed
Received 27 May 2009
so as to form secondary tumours at distant sites. We report evidence of an unusual expression of
Revised version received 21 July 2009
KCa2.3 (SK3) protein in melanoma cell lines but not in normal melanocytes. Knockdown of the
Accepted 22 July 2009
KCa2.3 channel led to plasma membrane depolarization, decreased 2D and 3D cell motility.
Available online 30 July 2009
Conversely, enforced production of KCa2.3 protein in KCa2.3 non-expressing cells led to the plasma membrane becoming hyperpolarized, and enhanced cell motility. In contrast, KCa3.1
Keywords:
channels had no effect on cell motility despite an active role in regulating membrane potential. Our
KCa2.3
data also suggest that membrane hyperpolarization increases melanoma cell motility and that this
SK3
occurs through the KCa2.3 channel. Our findings reveal a previously unknown function of the
Potassium channel
KCa2.3 channel, and suggest that the KCa2.3 channel might be the only member of the Ca2+-
Motility
activated K+ channel family involved in melanoma cell motility pathways.
Membrane potential
© 2009 Elsevier Inc. All rights reserved.
Melanoma cells
Introduction The KCa2.3 channel (SK3) is a small-conductance potassium channel (SKCa) belonging to calcium-activated potassium channel (KCa) family that comprises three isoforms: KCa2.1–3 [1]. The KCa2.3 channel controls the excitability of a nerve cell of the central nervous system by mediating after-hyperpolarization [2]. We recently suggested that the KCa2.3 channel participates in cancer cell migration, but the mechanisms were not fully characterized [3]. In view of the other physiological functions of
KCa2.3, its involvement in cancer cell migration was unexpected and therefore needs to be confirmed. The KCa2.3 channel has several functional features that may contribute to cancerous cell migration: i) cyclical activation and inhibition of the KCa2.3 channel, following intracellular Ca2+ oscillations and/or cell volume changes, allow changes of cell shape ii) the KCa2.3 channel, an activated-membrane molecule, routinely adopts a polarized cell location, especially at protrusions (filopodia or lamellipodia) [4,5] and iii) the activity of the KCa2.3 channel is sensitive to variations in the free Ca2+ concentration often
⁎ Corresponding author. Inserm, U921, F-37032 Tours, France. Fax: +33 247366226. E-mail address: [email protected] (C. Vandier). 1 Present addresses: INRA, UMR PRC 6175, Nouzilly, F 37380 France; Université François Rabelais, Tours, F-37032 France. 2 Christophe Vandier and Virginie Joulin contributed equally to this work. 0014-4827/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2009.07.021
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
associated with cell motility/migration [1]. Thus, although the KCa2.3 channel had not previously been physiologically related to any cell migration process, its physical and functional features are compatible with an involvement in cancer cell migration. One main cause for the malignancy of a cancer is the ability of tumour cells to spread from the primary tumour to form secondary tumours at distant sites. Before the formation of metastasis, cancer cells have to go through a series of sequential steps. First they invade adjacent tissues. Next, they pass through blood and lymph vessels and spread to distant permissive sites. Several mechanical processes (e.g. cell adhesion and motility) and various molecular events (e.g. acto-myosin contraction) govern tumour cell dissemination. Despite growing research efforts, the molecular mechanisms underlying metastasis formation are still fragmentary. Discovering and understanding the molecular mechanisms underlying metastasis formation will help in the design of innovative therapeutic strategies. This is of particular importance for melanoma because the metastatic forms of this cancer are largely refractory to existing therapies. The ability of melanoma to spread was simply thought to be a consequence of melanocytes originating from highly motile cells with enhanced survival properties [6]. Here, we report that KCa2.3 protein, which is produced in melanoma cells but not in untransformed melanocytes, forms functional KCa2.3 channels which hyperpolarize the plasma membrane and promote cell motility. In contrast, KCa3.1 channels had no effect on cell motility despite actively regulating membrane potential. Our data show that KCa2.3-dependent hyperpolarization increases melanoma cell motility.
Materials and methods Cell culture Four human metastatic melanoma cell lines were used in this study 518A2 (a gift from Dr. van der Minne L., Leiden, The Netherlands), SKmel28, HBL, and Bris (a gift from Dr. S. Priest, IGR, Paris, France). All cell lines were maintained in Dulbecco's modified Eagle's medium except for HBL that was maintained in Ham's F10 medium. Culture media were supplemented with 10% (v/v) foetal bovine serum. Normal human epithelial melanocyte (a gift from Dr. J.E. Branka, Effiscience, Rennes, France) were maintained in serum-free Melanocyte growth medium M2 according to the manufacturer's protocol (PromoCell, Heidelberg, Germany). Cells were grown in a humidified atmosphere at 37 °C (95% air, 5% CO2). The absence of mycoplasma contamination was verified regularly using the Mycotech kit (Invitrogen).
Constructs, transfection, and transduction To inhibit KCa2.3 expression, we constructed a lentiviral vector encoding a short hairpin RNA (shRNA) specifically targeting the human KCa2.3. The sequence encoding shKCa2.3 was obtained by PCR elongation of two partially complementary primers; this also allowed us to introduce two restriction enzyme sites to facilitate manipulations. Forward primer: shKCa2.3-BamH I 5′ ggATCCCCCCATTCCTggCgAgTACAATTCAAgAgATT; shKCa2.3-Hind III 5′ AAgCTTAAAAACCATTCCTggCgAgTACAATCTCTTgAATT, as a reverse primer. The underscored sequences are complementary. The
3621
shKCa2.3 fragment was inserted into pGEMT easy vector (Promega) and transferred into pH1, a plasmid expressing the shRNA from RNA pol III promoter (a gift from Anne Galy, Généthon-Evry), at Bgl II/Hind III sites. The H1 promoter linked to the shKCa2.3 sequence was then inserted between the Spe I and Cla I sites in a pHR'-derived vector [7], expressing the GFP from a CMV promoter and an IRES. For control experiments we constructed, following the same protocol as above, a lentiviral vector expressing an untargeted shRNA (pLenti-shRD). For this, we used the following primers, shRD1_S: 5′ggATCCCCgCCgACCAATTCACggCCgTTCAAgAg ACg 3′ and shRD1_AS: 5′AAgCTTAAAAAgCCgACCAATTCACggCCgT CTCTTgAACg 3′. For KCa2.3 over-expression, the full length rat KCa2.3 cDNA was cloned by recombination with the system developed by Invitrogen, a shuttle vector was generated containing the PCR product of the full length cDNA: forward primer: ggCCCCAAgATggACACTTCTgggCACTTC and reverse primer: TTAgCAACTgCTTgAACTTgTg. 293T cells were plated at 3 × 106 cells per 100 ml culture and grown overnight before transfection. Transfection was performed with 10 μg of pCMV-8.74 encoding HIV1 Gag-Pol and all accessory proteins, 5 μg of pH-CMV-G encoding the VSV-G and 5 μg of the lentiviral vector (pLenti-shKCA2.3 or pLenti-shRD), using a calcium phosphate transfection kit (Invitrogen) according to the manufacturer's instructions. After 24 h, the medium was removed and the cells were washed with PBS, then fresh medium was added. Supernatants were harvested 48 and 72 h later and filtered through a 0.45 μm pore-size Millex HA filter (Millipore). The lentiviral vectors were then concentrated on 20% sucrose cushion by centrifugation at 26 krpm for 90 min (Sorvall, Discovery 90SE). Viral titres were determined by transducing 293T cells with serial dilutions of the viral stock, and then counting GFP cells using FACS analysis. Melanoma cells were transduced with lentiviral vector at multiplicities of infection (MOI) of 1 to 3 in the presence of polybrene (4 μg/ml, Sigma). The transduction rate determined by counting GFP cells was close to 90%.
Cell proliferation assays Cell proliferation was determined using the tetrazolium salt reduction method (MTT), as described elsewhere [8]. Cells were seeded on 24-well plates at a density of 20,000 cells per well and measurements were performed in triplicate daily for 5 days. Note that drugs and high external potassium concentration used in trans-well migration assays had no effect on cell proliferation/ viability (48 h).
Two-dimensional (2D) and 3D motility assays Cell motility was analyzed in 24-well plates receiving 8-μm poresize polyethylene terephthalate membrane cell culture inserts (Becton Dickinson, France), as previously described [8]. Briefly, 4 × 104 cells were seeded in the upper compartment with medium culture supplemented with 10% of FBS (±drugs/high external potassium concentration). The lower compartment was filled with medium culture supplemented with 20% FBS (± drugs/high external potassium concentration) as a chemoattractant. Twodimensional motility assays were performed without coating except for HBL cell lines for which bovine fibronectin and gelatin (0.05%; 0.02%) were used (Sigma-Aldrich). Three-dimensional
3622
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
motility assays were assessed as above but membrane was covered with a Matrigel® matrix. After 24 h, stationary cells were removed from the top side of the membrane, whereas migrated cells in the bottom side of the inserts were fixed, stained, and counted in five different fields (magnification, ×200). At least three independent experiments were each performed in triplicate.
Electrophysiology Experiments were performed with cells seeded into 35-mm Petri dishes at 2000 cells per cm2. All current-clamp (I = 0) and voltageclamp experiments were performed using the conventional whole-cell recording configuration of the patch-clamp technique as previously described [8]. PCa solution was 6 and 6.4 respectively for SKmel28 and 518A2 cells. Briefly, experiments were conducted using Axopatch 200B patch-clamp amplifier (Axon Instrument) and data, digitized with 1322-A Digidata converter (Axon Instrument), were stored on a computer using Clampex of pClamp 9.2 software (Axon Instrument). The patch-clamp data was analyzed using both Clampfit 9.2 and Origin 7.0 software (Microcal Inc., Northampton, MA, USA).
RT-PCR Total RNAs were extracted using the Nucleospin RNA II kit (Macherey-Nagel). Standard protocols were used for RT-PCR experiments as described previously [3]. Briefly, 1 μg of total RNA was used for cDNA synthesis using hexanucleotide primers and a Ready-to-Go You-Prime First-Strand Beads kit (GE Healthcare, UK). Polymerase chain reaction (PCR) was performed using puRE Taq Ready to go PCR Beads kit (Amersham Biosciences, UK). The primers used were as follows: KCa2.1 5′ primer AGGGAGACGTGGCTCATCTA and 3′ primer TTAGCCTGGTCGTTCAGCTT, KCa2.2 5′ primer GACTTGGCAAAGACCCAGAA and 3′ primer CCGCTCAGCATTGTAAGTGA, KCa2.3 5′ primer TGGACACTCAGCTCACCAAG and 3′ primer GTTCCATCTTGACGCTCCTC, KCa3.1 5' primer CATTCCTGACCATCGGCTAT and 3' primer ACG TGC TTC TCT GCC TTG TT, 18S as house keeping gene 5′ primer CGCGGTTCTATTTTGTTGGTTTT and 3′ primer TTCGCTCTGGTCCGTCTTG. The PCR reaction for KCa2.1-3 was as follows: 1 cycle at 94 °C for 5 min, 40 cycles of 94 °C for 30 s, 57° for 30 s, then 1 cycle at 72 °C for 30 s and a final incubation at 72 °C for 5 min. Protocol for 18s amplification was: 1 cycle at 94 °C for 5 min, 25 cycles of 94 °C for 30 s, 60° for 30 s, 72 °C for 30 s then 1 cycle at 72 °C for 5 min A negative control without RT product was included. Experiments were repeated three times.
Western blot Whole-cell lysates were prepared with 5% sodium dodecyl sulphate and protease inhibitor cocktail (Sigma, France). Proteins were separated by denaturing SDS-PAGE and transferred onto polyvinylidene fluoride membranes. Primary antibodies used were: rabbit anti-KCa2.3 N-term (1:500; Sigma-Aldrich), mouse anti-GAPDH (1:2000 dilution; Sigma-Aldrich) and rabbit antiactin (1:1000; Santa Cruz Biotechnology). Antibody binding was revealed with anti-rabbit (1:10,000; Jackson Immuno-Research Laboratories) or anti-mouse (1:3000, SouthernBiotech) IgG coupled to horseradish peroxidase, using an enhanced chemiluminescence kit (Pierce). Actin or GAPDH antibodies were used for loading control experiments.
Solutions and drugs The physiological saline solution (PSS) had the following composition (in mM): NaCl 140, MgCl2 1, KCl 4, CaCl2 2, D-glucose 11.1 and HEPES 10, adjusted to pH 7.4 with NaOH. The high external K+ PSS solution (K50) and high external Na+ PSS solution used were prepared by adding 46 mM KCl or 46 mM NaCl (control ionic solution) to PSS. The pipette solution for the whole-cell recordings contained (in mM): K-glutamate 125, KCl 20, MgCl2 1, Mg-ATP 1, HEPES 10, and pH was adjusted to 7.2 with KOH and various concentrations of CaCl2 and EGTA were added to obtain calculated pCa = 6 (8.7 mM CaCl2 and 10 mM EGTA) or pCa = 6.4 (0.7 mM CaCl2 and 1 mM EGTA). High external K+ DMEM solution (K50) and a high external + Na (Na50) DMEM solution used for motility assays were prepared by adding 45 mM KCl or 45 mM NaCl (control ionic solution) to the DMEM solution. All the drugs were obtained from Sigma-Aldrich.
Statistics Data are expressed as means ± SEM (n = number of cells). Comparisons between two means were made using Mann– Whitney or paired t-tests as appropriate. For comparison between more than two means we used Kruskal–Wallis one way analysis of variance followed by Dunn's test. Differences were considered significant when p < 0.05. SigmaStat (version 3.0.1a, Systat Software, Inc.) was used for statistical analysis.
Results SKCa channel expression in melanoma cell lines and primary human melanocytes To investigate the production of SKCa channels in melanoma, we used RT-PCR assays with various primer pairs specific for human KCa2.1, KCa2.2 and KCa2.3 mRNAs to test four human metastatic melanoma cells lines (SKmel28, Bris, 518A2 and HBL) and a primary culture of normal human epidermal melanocytes (NHEM). No KCa2.1 transcript was detected using a primer pair allowing amplification of the SK1 mRNA variants [9] in any of the samples (Fig. 1A). This is in agreement with the reported expression pattern of KCa2.1, which has mainly been found in neuronal tissues [10]. Three KCa2.2 variants have been described: short and long N-terminus variants KCa2.2-S and -L, respectively [11], and a novel shorter isoform called KCa2.2-sh [12]. RT-PCR with primers amplifying all three KCa2.2 variants revealed their mRNAs in all cells tested. Unfortunately, we were unable to detect KCa2.2 proteins because commercially available antibodies directed towards a rat KCa2.2 were poorly specific for the human proteins [13]. Interestingly, KCa2.3 mRNAs were detected in Bris, 518A2 and HBL, but not NHEM or SKmel28. Western-blot analyses confirmed these findings (Fig. 1B).
The KCa2.3 channel is involved in melanoma cell motility One of the most prominent features of malignant melanoma is the fast generation of metastasizing cells, resulting in a poor
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
3623
Fig. 1 – SKCa expression in NHEM and in melanoma cell lines, SKmel28, Bris, 518A2, and HBL. (A) RT-PCR products from total RNA extracts using primers specific for the human KCa2.1-3 mRNAs. Plasmid containing human KCa2.1 cDNA was used as a positive control for KCa2.1 detection. (B) Western-blot detection of KCa2.3 protein in NHEM and melanoma cells. Equal amounts of total lysates were immunoblotted with anti-KCa2.3. MDA-MB-435s total lysates were used as positive control. NS: non-specific band.
prognostic. In order to metastasize, cancer cells must gain the ability to move. Following our observations on MDA-MB-435s [3], we evaluated whether KCa2.3 expression by melanoma also promotes cell motility. No specific KCa2.3 channel blocker was available, so we tested apamin, a blocker of all SKCa channels. In 2D trans-well cell motility assays (Fig. 2A), apamin treatment decreased the migration of KCa2.3-expressing cells by approximately 40%. Therefore, 2D cell motility of Bris, 518A2 and HBL melanoma cell lines appeared partially dependent on the activity of SKCa channels. Next we used Matrigel matrix assay to assess cell motility in a 3D environment. Apamin treatment significantly reduced 3D motility of MDA-MB-435s and 518A2 cell lines (Fig. 2B). These results indicate that both 2D and 3D motility of melanoma cells are in part dependent on SKCa activity. To study the role of KCa2.3, its expression was specifically silenced in 518A2 cells. Melanoma cells were transduced with lentivectors containing either an interfering shRNA specific to the KCa2.3 transcript (shKCa2.3) or a non-targeting shRNA (shRD) as control. Patch-clamp experiments revealed that the currentdensity at +30 mV was substantially lower in 518A2shKCa2.3 cells than in 518A2shRD cells (Fig. 2C). Importantly, the residual current was insensitive to apamin. KCa2.3 protein was barely detectable in 518A2shKCa2.3 cells by Western blot, but was abundant in control wild-type 518A2 (518A2WT) and 518A2shRD cells (Fig. 2A). Also, 518A2shKCa2.3 cells exhibited a lower migrating capacity than 518A2WT and 518A2shRD cells, and the remaining cell motility was insensitive to apamin. To confirm the specificity of the effects of KCa2.3-shRNA, we performed a second infection of 518A2shKCa2.3 cells with a lentivector expressing rKCa2.3 mRNA under a strong promoter. The over-expression of rKCA2.3 in 518A2shKCa2.3 cells restored the expression of the KCa2.3 channel as assessed from the presence of the protein (Fig. 2A). Furthermore, the KCa2.3-shRNA phenotype was completely reversed in transduced cells: the current-density reduction was prevented (Fig. 2C) and apamin-sensitive cell migration was restored (Fig. 2A). Thus, this reversion was due to the action of the KCa2.3 provided by the channel expressed following lentiviral transduction. NHEM and SKmel28 cells do not contain KCa2.3 protein (Fig. 1B). As expected, the motility of these KCa2.3-negative cells was not sensitive to apamin (Fig. 3A). To confirm that aberrant KCa2.3
expression promotes melanoma cell motility, we expressed KCa2.3 in SKmel28 cells and then evaluated the phenotypic consequences. Expression of rKCa2.3 in SKmel28 cells (Fig. 3A, top) led to an increase of the current-density that became sensitive to apamin (Fig. 3B). Two-dimensional motility assays also revealed an increase, of up to 40%, of the number of migrating cells (Fig. 3A). After apamin treatment, the number of migrating cells was similar to that of parental SKmel28 cells. These results are consistent with KCa2.3 channel activity making a substantial contribution to melanoma cell motility.
KCa2.3 channels but not KCa3.1 channels promote melanoma cell migration by hyperpolarizing the plasma membrane We used the current-clamp mode of the patch-clamp technique to measure the membrane potential in melanoma cells expressing or not KCa2.3 channel. Silencing KCa2.3 in 518A2 cells induced a significant depolarization of the plasma membrane from −66.5 ± 0.8 mV (n = 12) to −34.1 ± 3.0 mV (n = 12; p < 0.05, Mann– Whitney test), whereas enforcing KCa2.3 expression in SKmel28 cells induced a significant hyperpolarization from −47.8 ± 3.6 mV (n = 12) to −74.5 ± 0.7 mV (n = 12; p < 0.05, Mann–Whitney test). Moreover, an acute application of apamin induced a depolarization of 518A2shRD plasma membrane from −65.8 ± 1.62 mV to − 36.4 ± 2.73 mV (n = 5; p < 0.001, Paired t-test) and of SKmel28KCa2.3 plasma membrane from − 75.1 ± 0.82 mV to −36.4 ± 2.73 mV (n = 8; p < 0.001, Paired t-test). Thus, expression of KCa2.3 channel confers to melanoma cells a more negative membrane potential than cells that do not express this potassium channel (Fig. 4A). Therefore, if the KCa2.3 channel is involved in melanoma cell motility by hyperpolarizing plasma membrane cells, a sufficient depolarisation may inhibit migration of KCa2.3-expressing cells. An easy way to depolarize membrane of KCa2.3-expressing cells is to increase the extracellular K+ concentration. As expected, a 10fold (K50) increase in the extracellular K+ concentration resulted in a larger depolarization of 518A2shRD than of 518A2shKCa2.3 (Fig. 4B, top) due to a more negative “resting” membrane potential in cells expressing KCa2.3 channels. Concomitantly, the migratory capacity of 518A2shRD cells was reduced whereas there was no
3624
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
Fig. 2 – KCa2.3 channel promotes the motility of KCa2.3-expressing melanoma cells. (A) Effect of apamin on 2D cell motility by Bris, HBL, 518A2WT, shRD, and shKCa2.3 melanoma cells. Left, normalized cell motility was calculated as the ratio of the total number of migrating cells in the presence of apamin to the total number of migrating cells in control experiments. The normalized cell motility of 518A2shRD, shKCa2.3 and shKCa2.3 + rKCa2.3 was calculated as the ratio of the total number of migrating cells to the total number of 518A2 WT migrating cells in control conditions. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann–Whitney test. #p < 0.05, significantly different from 518A2 WT in control conditions using Kruskal–Wallis one way analysis of variance followed by Dunn's test. Top of the left panel, Western-blot analysis of KCa2.3 protein in 518A2 cells. Right, representative pictures of migrated melanoma cells observed in a part of a field. The number in brackets indicates the mean number of migrating cells counted in one field (magnification, × 200). (B) Effect of apamin on 3D cell motility of MDA-MB-435s and 518A2 cells. The normalized cell motility was calculated as the ratio of the total number of migrating cells in the presence of apamin to the total number of migrating cells in control experiments. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann–Whitney test. (C) KCa2.3 channel activity in 518A2 cells measured using the patch-clamp technique at pCa 6.4. The numbers in brackets indicate the number of cells. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann–Whitney test. #p < 0.05, significantly different from 518shRD in control conditions using Kruskal– Wallis one way analysis of variance followed by Dunn's test. WT; wild-type, shRD; non-targeting short hairpin RNA named random, shKCa2.3; short hairpin RNA specific for the KCa2.3 transcript, rKCa2.3; rat KCa2.3 cDNA.
significant change for the migration of 518A2shKCa2.3 cells (Fig. 4B, bottom). Also, when the plasma membrane was depolarized, the migration of 518A2shRD became insensitive to apamin (data not shown). Substitution of K+ by Na+ (Fig. 4B bottom) or the addition of mannitol (data not shown) did not affect 518A2 cell migration, indicating that the inhibition of migration of 518A2shKCa2.3 cells associated with the high extracellular K+ concentration was not due to the higher medium osmolarity. The membrane potential has been described to change K+ channel expression [14], so we tested the effect of K50-induced depolar-
ization on KCa2.3 expression. Using a western-blot analysis, we did not detect any difference in KCa2.3 expression (data not shown). Our findings indicate that the KCa2.3 channel promotes 518A2 cell motility associated with hyperpolarization of the plasma membrane. We next investigated the specificity of KCa2.3 action. If hyperpolarization is necessary, is it possible to trigger mobilization by inducing hyperpolarization through another K+ channel? RTPCR experiments have evidence of KCa3.1 products, a KCa channel, in NHEM and in the other melanoma cell lines tested (Fig. 5A). This
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
3625
Fig. 3 – Ectopic expression of KCa2.3 enhances the motility of melanoma cells lacking the KCa2.3 channel. (A) Effect of apamin on motility of NHEM and SKmel28 cells. Left, the normalized cell number was calculated as the ratio of the total number of migrating cells in the presence of apamin to the total number of migrating cells in control experiments. The normalized 2D motility of empty and rKCa2.3 cells was calculated as the ratio of the total number of migrating cells to the total number of SKmel28 WT migrating cells in control conditions. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Mann– Whitney test. #p < 0.05, significantly different from SKmel28 WT in control conditions using Kruskal–Wallis one way analysis of variance followed by Dunn's test. Top of the left panel, Western-blot analysis of KCa2.3 protein in WT, empty and rKCa2.3-SKmel28 cells. Equal amounts of total lysates were immunoblotted with anti-KCa2.3. Right, representative pictures of migrated melanoma cells observed in a part of a field. The number in brackets indicates the mean number of migrating cells counted in one field (magnification, ×200). (B) KCa2.3 channel activity in empty and rKCa2.3-SKmel28 cells measured by the patch-clamp technique at pCa6. Left, KCa2.3 channel activity in SKmel28 cells. Right, typical currents recorded in SKmel28 cells following voltage-clamp pulses from −100 mV to + 100 mV for 400 ms (holding potential = −70 mV). The numbers in brackets indicate the number of paired cells. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using the Wilcoxon test. #p < 0.05, significantly different from SKmel28 WT in control conditions using Mann–Whitney test. WT; wild-type, empty; empty lentivector, rKCa2.3; rat KCa2.3 cDNA.
channel is functional and regulates membrane potential like KCa2.3 channel. Indeed, TRAM-34, a specific KCa3.1 channel blocker [15], significantly depolarized the plasma membrane of 518A2 cells (1 μM TRAM-34; n = 3; from −67.6 ± 1.2 mV to −33.6 ± 3.3 mV; p < 0.001, paired t-test) but did not impair their cell motility (Fig. 5B). Since TRAM-34 exhibits a high avidity for proteins [16], the last experiments were performed in the presence of 10% SVF (same experimental conditions as for cell motility assays). Next, we inhibited KCa3.1 channels and other
voltage dependent K+ channels by treatment with 10 μM of clotrimazole [15]. This treatment also induced a large depolarization (data not shown) but still did not reduce 518A2 and SKmel28 cell motility (Fig. 5C). These experiments indicate that KCa3.1 channel activity or its inhibition had no effect on cell migration despite its active role in regulating membrane potential. Therefore, the melanoma cell motility induced by hyperpolarization appeared dependent on the KCa2.3 channel, and did not involve the KCa3.1 channel.
3626
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
Fig. 4 – KCa2.3 channels promote cell motility by hyperpolarizing the plasma membrane. (A) Membrane potential recorded in cells expressing KCa2.3 channel (KCa2.3+ cells; 518shRD, 518shKCa2.3+rKCa2.3 and SKmel28KCa2.3) and in cells that did not expressed KCa2.3 channel (KCa2.3− cells; 518shKCa2.3 and SKmel28empty). PCa solution was 6 and 6.4 respectively for SKmel28 and 518A2 cells. Boxes indicated the first quartile, the median and the third quartile, squares indicated the mean. #p < 0.001, significantly different between KCa3.2+ and KCa3.2− cells using Mann–Whitney test. (B) Top, examples of membrane potential recorded in one 518A2shRD and one 518shKCa2.3 cell using the patch-clamp technique in current-clamp mode. Membrane potentials were recorded in external solutions containing 5 mM K+ (K5) and 50 mM K+ (K50). Bottom, effect of K50 on 518A2shRD and shKCa2.3 cell motility. The normalized 2D cell motility was calculated as the ratio of the total number of migrating cells in the presence of K50 or Na50 to the total number of migrating cells in K5 experiments. Columns, mean, bars, SEM. ⁎p < 0.05, significantly different from control conditions using Kruskal–Wallis one way analysis of variance followed by Dunn's test.
KCa2.3 and KCa3.1 channels do not affect melanoma cell proliferation It has been reported that K+ channels may be involved in cell cycle progression of cancer cells [17]. We therefore explored KCa-dependent cancer cell proliferation. Cell proliferation was not affected in KCa2.3-negative 518A2 cells or in KCa2.3expressing SKmel28 cells (Fig. 6A). The same applied for the
KCa3.1 channel: two concentrations of TRAM-34 usually used for blocking KCa3.1 channels did not affect proliferation of either 518A2 or SKmel28 cells (Fig. 6B). In contrast, although 1 μM CLT had no effect on melanoma cell proliferation, increasing the concentration to 10 μM (a concentration 100 times higher than those used for specific inhibition of KCa3.1 channels) decreased the proliferation of both melanoma cell lines tested. High CLT concentrations do not have a specific effect on KCa3.1, and this
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
3627
Fig. 5 – KCa3.1-induced hyperpolarization did not promote cell motility. (A) RT-PCR products from total RNA extracts using primers specific for the human KCa3.1 mRNA. (B) Left, typical example of membrane potential recorded in one 518A2 cell in the absence or in the presence of 0.1 μM and 1 μM TRAM-34, using an external solution supplemented with 10% SVF. Right, effect of TRAM-34 on the motility of 518A2 cells. (C) Effect of 10 μM clotrimazole (CLT) on the motility of 518A2 and SKmel28 cells. The normalized motility was calculated as the ratio of the total number of migrating cells in the presence of TRAM-34 or CLT to the total number of migrating cells in control experiments. Columns, mean, bars, SEM. ns, p > 0.05, not significantly different from control conditions using Mann– Whitney test.
may explain why KCa3.1 channels have confusingly been suspected to be involved in cancer cell proliferation [17]. Here, we demonstrate that neither KCa2.3 nor KCa3.1 channels are involved in melanoma cell proliferation.
Discussion In the present study we report that KCa2.3 protein is produced in melanoma cells but not in untransformed melanocytes. Furthermore, we show that this channel promotes melanoma cell motility by hyperpolarizing plasma membrane. This new function of the KCa2.3 channel was revealed using a variety of approaches: (i) apamin treatment significantly reduced the migration of KCa2.3-expressing melanoma cells whereas it had no effect in KCa2.3-non-expressing melanoma cells; (ii) stable silencing of the KCa2.3 gene using a specific KCa2.3 shRNA inhibited 518A2 cell migration, with a remaining activity insensitive to apamin; conversely, the stable expression of KCa2.3 in
KCa2.3-defective melanoma cells, enhanced their migration, which became sensitive to apamin; iii) while patch-clamp experiments revealed that KCa2.3 and KCa3.1 channels regulate membrane potential, membrane potential per se had no effect on cell motility and iv) the increase of external potassium concentration only reduced SK3-dependant motility. Cancer cells, notably melanoma cells, may switch between mesenchymal movements (actin polymerisation with proteases activities) and amoeboid movements (actin–myosin contraction without protease activity) according to their microenvironment [18,19]. Interestingly, KCa2.3 channels appear to promote cell motility through several movement modes as suggested by 2D and 3D motility assays. Indeed, we previously showed that the KCa2.3 channel was a mediator of MDA-MB-435s breast cancer cell motility in a 2D environment and the present study shows a similar result in a 3D environment [3]. As MDA-MB-435s cells adopt an amoeboid form of movement in a 3D environment and a mesenchymal form of movement in a 2D environment [20], our results suggest that SKCa channels, and more exactly
3628
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
Fig. 6 – KCa2.3 and KCa3.1 channels are not involved in melanoma cell proliferation. (A) Proliferation rate of 518A2 and SKmel28 cells expressing or not KCa2.3 channels. (B) Proliferation rate of 518A2 and SKmel28 cells in the absence and in the presence of 0.1, 1 μM TRAM-34 and 1, 10 μM CLT. Cell proliferation was determined daily for 4 days using the tetrazolium salt reduction method (MTT). Columns, mean, bars, SEM, n = 3.
the KCa2.3 channel, may be involved in both types of cell movement. Expression of the KCa2.3 channel by melanoma cells appears explainable through their neural crest origin, as most other KCa2.3-expressing mature cells, i.e. neurons, glial cells and endocrine cells [12]. Re-expression of embryonic genes during cancer has been widely described [21]. It is therefore plausible that KCa2.3 channel expression during melanoma malignancy corresponds to the re-expression of an embryonic gene expressed by a neural progenitor common to neurons and melanocytes [22,23]. Of note, the tissue origin of MDA-MB-435s is in debate, since it has been shown that these cells express melanocytic markers and have the same genetic profile as the melanoma cell line M14 [24–26]. How do KCa2.3 channels regulate melanoma cell motility? Even if mesenchymal and amoeboid movements are driven by two different mechanisms, both are regulated by variations of the intracellular Ca2+ concentration. For example, [Ca2+] oscillations induce cell-rear detachment following calpain activation and are also necessary for actin–myosin contraction [27–29]. By changing the plasma membrane potential, activation of KCa2.3 channel may regulate these [Ca2+] oscillations (frequency, amplitude). Indeed, an increase of K+ efflux shifts the membrane potential towards negative values (hyperpolarization) and consequently increases the Ca2+ driving force
[3,30,31]. Inversely, a reduction of KCa2.3 channel activity may depolarize the plasma membrane and reduce Ca2+ entry. We have clearly demonstrated that KCa2.3-dependent K+ efflux regulate membrane potential of melanoma cells. KCa2.3 expression in SKmel28 cells induced membrane hyperpolarization and promoted cell motility whereas KCa2.3 inhibition led to membrane depolarization in 518A2 cells and inhibited cell motility. In addition, increasing extracellular K+ concentration depolarised plasma membrane cells and consequently reduced melanoma cell motility only in cells expressing KCa2.3 protein. It was found that K+ efflux could lead to water efflux and causes localized cell volume decreases that allow retraction of the rear cell body of migrating cells [32,33]. Nevertheless, if cell swelling was found to activate KCa2.3 channel in HEK 293 cells, endogenous KCa2.3 channel seems not to be activated by cell swelling in hepatoma cells [4,34]. More experiments are needed to determine if cell volume changes could regulate KCa2.3 channel motility. Does hyperpolarization induced by another KCa channels promote melanoma cell motility? KCa2.1 and KCa2.2 channels can be put aside since KCa2.1 is not expressed in melanoma cells and KCa2.2 was found to be not functional as we observed no apamin-sensitive currents in SKmel28 cells and 518A2shKCa2.3 cells. Concerning the large/big-conductance calcium-activated
E XP E RI ME N T AL C E L L R E S EA RC H 31 5 ( 20 0 9) 3 6 2 0– 3 63 0
potassium channels (BKCa channels), our previous study showed that despite their active role in regulating membrane potential those channels were not involved in MDA-MB-435s cell motility [8]. Consequently, we focused on the last KCa channel, the KCa3.1 channel, as it was previously involved in motility of non-cancerous cells (e.g. lung mast cells or human coronary smooth muscle cells) [16,35]. We find that the KCa3.1 channels had no effect on melanoma cell motility despite actively regulating membrane potential. A study with melanoma cells suggested that the KCa3.1 channel may have a role in SKmel28 cell motility, because treatment with charybdotoxin inhibited migration [36]. However, charybdotoxin is a non-specific K+ channel blocker, unlike TRAM34 a new specific KCa3.1 channel blocker developed by Wulff et al. [15]. With this compound, we show that the increase of melanoma cell mobility by hyperpolarization is KCa2.3 dependent. This suggests that KCa2.3 channel interacts in cooperation with a specific complex. Furthermore, a singular KCa2.3 channel spatial distribution was observed at cell protrusions, which may favour this interaction [4,5]. Such a coupling was already found between potassium and calcium channels [37]. The identity of the partner is not determined but it could be voltage-independent-calcium channel like transient-receptor-potential (TRP) that are known to be sufficient to activate Ca2+-regulated stimulatory pathways for cell motility [31,38–40]. As a general conclusion, we report that KCa2.3 protein is reexpressed in melanoma cells, leading to formation of a functional KCa2.3 channel; this channel hyperpolarizes the plasma membrane. Furthermore, we provide evidence from 2D and 3D motility assays that abnormal KCa2.3 channel expression in melanoma cells promotes a novel signal for cell motility. Therefore, these findings argue that it may be beneficial to assess KCa2.3 channel blockers as novel tools for prevention of melanoma metastasis.
[4]
[5]
[6] [7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Acknowledgments This work was funded by “Ligue contre le Cancer - Région Centre”, “Fondation Carrefour”, “INSERM” and “Cancéropole Grand Ouest”. A. C. and A. G. hold fellowships from the “INCa” and the “Region Centre”, respectively. We thank Dr. L. van der Minne and Dr. S. Priest for the kind gifts of melanoma cells and Dr. J.E. Branka for NHEM cells. We also thank Dr. S. Lidofsky for the kind gift of KCA2.3 plasmid and Dr. T. Peitersen for the SK1 plasmid. We thank Dr. M.L. Jourdan and E. Marais for their technical assistance, Prof. P. Bougnoux for helpful discussion and C. Leroy for secretarial support.
[15]
[16]
[17] REFERENCES [18] [1] A.D. Wei, G.A. Gutman, R. Aldrich, K.G. Chandy, S. Grissmer, H. Wulff, International Union of Pharmacology. LII. Nomenclature and molecular relationships of calcium-activated potassium channels, Pharmacol. Rev. 57 (2005) 463–472. [2] R. Hosseini, D.C. Benton, P.M. Dunn, D.H. Jenkinson, G.W. Moss, SK3 is an important component of K(+) channels mediating the afterhyperpolarization in cultured rat SCG neurones, J. Physiol. 535 (2001) 323–334. [3] M. Potier, V. Joulin, S. Roger, P. Besson, M.L. Jourdan, J.Y. Leguennec, P. Bougnoux, C. Vandier, Identification of SK3 channel
[19]
[20]
3629
as a new mediator of breast cancer cell migration, Mol. Cancer Ther. 5 (2006) 2946–2953. E.T. Barfod, A.L. Moore, M.W. Roe, S.D. Lidofsky, Ca2+-activated IK1 channels associate with lipid rafts upon cell swelling and mediate volume recovery, J. Biol. Chem. 282 (2007) 8984–8993. S. Liebau, B. Vaida, C. Proepper, S. Grissmer, A. Storch, T.M. Boeckers, P. Dietl, O.H. Wittekindt, Formation of cellular projections in neural progenitor cells depends on SK3 channel activity, J. Neurochem. 101 (2007) 1338–1350. V. Gray-Schopfer, C. Wellbrock, R. Marais, Melanoma biology and new targeted therapy, Nature 445 (2007) 851–857. R. Zufferey, T. Dull, R.J. Mandel, A. Bukovsky, D. Quiroz, L. Naldini, D. Trono, Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery, J. Virol. 72 (1998) 9873–9880. S. Roger, M. Potier, C. Vandier, J.Y. Le Guennec, P. Besson, Description and role in proliferation of iberiotoxin-sensitive currents in different human mammary epithelial normal and cancerous cells, Biochim. Biophys. Acta 1667 (2004) 190–199. B.M. Zhang, V. Kohli, R. Adachi, J.A. Lopez, M.M. Udden, R. Sullivan, Calmodulin binding to the C-terminus of the small-conductance Ca2+-activated K+ channel hSK1 is affected by alternative splicing, Biochemistry 40 (2001) 3189–3195. M.X. Chen, S.A. Gorman, B. Benson, K. Singh, J.P. Hieble, M.C. Michel, S.N. Tate, D.J. Trezise, Small and intermediate conductance Ca(2+)-activated K+ channels confer distinctive patterns of distribution in human tissues and differential cellular localisation in the colon and corpus cavernosum, Naunyn Schmiedebergs Arch. Pharmacol. 369 (2004) 602–615. T. Strassmaier, C.T. Bond, C.A. Sailer, H.G. Knaus, J. Maylie, J.P. Adelman, A novel isoform of SK2 assembles with other SK subunits in mouse brain, J. Biol. Chem. 280 (2005) 21231–21236. S.R. Murthy, G. Teodorescu, I.M. Nijholt, A.M. Dolga, S. Grissmer, J. Spiess, T. Blank, Identification and characterization of a novel, shorter isoform of the small conductance Ca2+-activated K+ channel SK2, J. Neurochem. 106 (2008) 2312–2321. R. Khanna, L. Roy, X. Zhu, L.C. Schlichter, K+ channels and the microglial respiratory burst, Am. J. Physiol. Cell Physiol. 280 (2001) C796–C806. M. Zanzouri, I. Lauritzen, F. Duprat, M. Mazzuca, F. Lesage, M. Lazdunski, A. Patel, Membrane potential-regulated transcription of the resting K+ conductance TASK-3 via the calcineurin pathway, J. Biol. Chem. 281 (2006) 28910–28918. H. Wulff, M.J. Miller, W. Hansel, S. Grissmer, M.D. Cahalan, K.G. Chandy, Design of a potent and selective inhibitor of the intermediate-conductance Ca2+-activated K+ channel, IKCa1: a potential immunosuppressant, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 8151–8156. K. Toyama, H. Wulff, K.G. Chandy, P. Azam, G. Raman, T. Saito, Y. Fujiwara, D.L. Mattson, S. Das, J.E. Melvin, P.F. Pratt, O.A. Hatoum, D.D. Gutterman, D.R. Harder, H. Miura, The intermediateconductance calcium-activated potassium channel KCa3.1 contributes to atherogenesis in mice and humans, J. Clin. Invest. 118 (2008) 3025–3037. N. Villalonga, J.C. Ferreres, J.M. Argiles, E. Condom, A. Felipe, Potassium channels are a new target field in anticancer drug design, Recent Patents Anticancer Drug Discov. 2 (2007) 212–223. V. Sanz-Moreno, G. Gadea, J. Ahn, H. Paterson, P. Marra, S. Pinner, E. Sahai, C.J. Marshall, Rac activation and inactivation control plasticity of tumor cell movement, Cell 135 (2008) 510–523. D.R. Croft, M.F. Olson, Regulating the conversion between rounded and elongated modes of cancer cell movement, Cancer Cell 14 (2008) 349–351. T.M. Kitzing, A.S. Sahadevan, D.T. Brandt, H. Knieling, S. Hannemann, O.T. Fackler, J. Grosshans, R. Grosse, Positive feedback between Dia1, LARG, and RhoA regulates cell morphology and invasion, Genes Dev. 21 (2007) 1478–1483.
3630
E XP E RI ME N T AL C E L L R E SE A RC H 31 5 ( 20 0 9) 3 6 20 – 3 63 0
[21] M. Monk, C. Holding, Human embryonic genes re-expressed in cancer cells, Oncogene 20 (2001) 8085–8091. [22] S. Rasheed, Z. Mao, J.M. Chan, L.S. Chan, Is melanoma a stem cell tumor? Identification of neurogenic proteins in trans-differentiated cells, J. Transl. Med. 3 (2005) 14. [23] W.M. Klein, B.P. Wu, S. Zhao, H. Wu, A.J. Klein-Szanto, S.R. Tahan, Increased expression of stem cell markers in malignant melanoma, Mod. Pathol. 20 (2007) 102–107. [24] S. Sellappan, R. Grijalva, X. Zhou, W. Yang, M.B. Eli, G.B. Mills, D. Yu, Lineage infidelity of MDA-MB-435 cells: expression of melanocyte proteins in a breast cancer cell line, Cancer Res. 64 (2004) 3479–3485. [25] J.M. Rae, C.J. Creighton, J.M. Meck, B.R. Haddad, M.D. Johnson, MDA-MB-435 cells are derived from M14 melanoma cells—a loss for breast cancer, but a boon for melanoma research, Breast Cancer Res. Treat. 104 (2007) 13–19. [26] A.F. Chambers, MDA-MB-435 and M14 cell lines: identical but not M14 melanoma? Cancer Res. 69 (2009) 5292–5293. [27] R.S. Adelstein, Calmodulin and the regulation of the actin–myosin interaction in smooth muscle and nonmuscle cells, Cell 30 (1982) 349–350. [28] K. Clark, M. Langeslag, B. van Leeuwen, L. Ran, A.G. Ryazanov, C.G. Figdor, W.H. Moolenaar, K. Jalink, F.N. van Leeuwen, TRPM7, a novel regulator of actomyosin contractility and cell adhesion, EMBO J. 25 (2006) 290–301. [29] S.J. Franco, A. Huttenlocher, Regulating cell migration: calpains make the cut, J. Cell Sci. 118 (2005) 3829–3838. [30] J.N. Rao, O. Platoshyn, L. Li, X. Guo, V.A. Golovina, J.X. Yuan, J.Y. Wang, Activation of K(+) channels and increased migration of differentiated intestinal epithelial cells after wounding, Am. J. Physiol. Cell Physiol. 282 (2002) C885–C898. [31] J.N. Rao, O. Platoshyn, V.A. Golovina, L. Liu, T. Zou, B.S. Marasa, D.J. Turner, J.X. Yuan, J.Y. Wang, , TRPC1 functions as a store-operated Ca2+ channel in intestinal epithelial cells and regulates early
[32]
[33] [34]
[35]
[36]
[37]
[38]
[39]
[40]
mucosal restitution after wounding, Am. J. Physiol. Gastrointest. Liver Physiol. 290 (2006) G782–G792. A. Schwab, B. Schuricht, P. Seeger, J. Reinhardt, P.C. Dartsch, , Migration of transformed renal epithelial cells is regulated by K+ channel modulation of actin cytoskeleton and cell volume, Pflugers Arch. 438 (1999) 330–337. A. Schwab, V. Nechyporuk-Zloy, A. Fabian, C. Stock, , Cells move when ions and water flow, Pflugers Arch. 453 (2007) 421–432. N.K. Jorgensen, S.F. Pedersen, H.B. Rasmussen, M. Grunnet, D.A. Klaerke, S.P. Olesen, Cell swelling activates cloned Ca(2+)activated K(+) channels: a role for the F-actin cytoskeleton, Biochim. Biophys. Acta 1615 (2003) 115–125. G. Cruse, S.M. Duffy, C.E. Brightling, P. Bradding, Functional KCa3.1 K+ channels are required for human lung mast cell migration, Thorax 61 (2006) 880–885. A. Schwab, J. Reinhardt, S.W. Schneider, B. Gassner, B. Schuricht, K (+) channel-dependent migration of fibroblasts and human melanoma cells, Cell. Physiol. Biochem. 9 (1999) 126–132. E.Y. Kim, C.P. Alvarez-Baron, S.E. Dryer, Canonical transient receptor potential channel (TRPC)3 and TRPC6 associate with large-conductance Ca2+-activated K+ (BKCa) channels: role in BKCa trafficking to the surface of cultured podocytes, Mol. Pharmacol. 75 (2009) 466–477. M. Louis, N. Zanou, M. Van Schoor, P. Gailly, TRPC1 regulates skeletal myoblast migration and differentiation, J. Cell Sci. 121 (2008) 3951–3959. J. Waning, J. Vriens, G. Owsianik, L. Stuwe, S. Mally, A. Fabian, C. Frippiat, B. Nilius, A. Schwab, A novel function of capsaicinsensitive TRPV1 channels: involvement in cell migration, Cell Calcium 42 (2007) 17–25. R. Wondergem, T.W. Ecay, F. Mahieu, G. Owsianik, B. Nilius, HGF/SF and menthol increase human glioblastoma cell calcium and migration, Biochem. Biophys. Res. Commun. 372 (2008) 210–215.