The a3 Isoform Vacuolar Type H⁺-ATPase Promotes Distant Metastasis in the Mouse B16 Melanoma Cells Free

Acidic microenvironment is one of the characteristic features of tumors. More than 50 years ago, Warburg first reported that tumors are acidic due to increased lactic acid production (1). Recent imaging studies also showed the presence of a peritumoral acid gradient (2, 3). It has been suggested that some tumor cells that can survive in this unfavorable condition are conferred malignancy and that the acidic microenvironment may contribute to the selection of further malignant subpopulation of tumor cells. Indeed, acidic microenvironments are reported to increase tumor malignancy by promoting chemoresistance, invasiveness, and proliferation (4–7). In addition, low extracellular pH can trigger the expression and the activity of several classes of proteases such as MMP-2 and MMP-9 and cause acid-induced upregulation of the proangiogenic factors VEGF-A and IL-8 (8–10). Therefore, targeting the acidic conditions around tumors may provide a new therapeutic intervention for cancer treatment. Earlier studies suggest that the production of lactate due to high rates of glycolysis accounts for tumor acidity. However, recent studies have shown that lactate is not necessarily a key player in the development of the acidic microenvironment around tumors (11, 12).

Vacuolar proton-ATPase (V-ATPase) is a multi-subunit enzyme that regulates proton transport and creates the acidic microenvironment. V-ATPase localizes in a variety of cellular membranes of eukaryotic cells, including endosomes, lysosomes, Golgi-derived vesicles, secretory vesicles. It is responsible for the regulation of the proton concentrations and various physiological functions associated with organelle pH regulation (13–16). In addition to the maintenance of intracellular pH, V-ATPase is expressed in plasma membrane and transports protons across the plasma membrane, inducing extracellular acidosis in specialized tissues. It is noted that extracellular acidification is involved in various biological functions including bone resorption and renal base homeostasis (17–19). Importantly, several recent studies showed that V-ATPase is involved in the creation of the acidic microenvironment of tumors. V-ATPase is highly expressed in the plasma membrane of highly metastatic cancer cells (20), whereas it is expressed in limited amounts in poorly metastatic cells (21). Furthermore, inhibiting V-ATPase function using siRNA or proton pump inhibitors suppresses cancer growth and metastasis in a number of tumor cell types, including hepatocellular carcinoma and breast cancer (21–23). These reports suggest that V-ATPase plays a critical role in tumor progression and invasiveness through the creation of the acidic microenvironments. It is therefore expected that identification of the particular V-ATPase subunit(s) involved in tumor metastasis would lead us to design new anticancer therapies.

V-ATPase is composed of a membrane component (V0) and a cytosolic catalytic component (V1). The integral V0 component functions in proton translocation, whereas the peripheral V1 component hydrolyzes ATP. The V1 domain consists of 3 A subunits, 3 B subunits, 2 G subunits, and 1 each of C, D, E, F, and H subunits. The V0 domain consists of 6 different subunits (a, c, c′, c″, d, and e; refs. 13, 24, 25). The V0 a subunit, the largest of the V0 subunits and major component of the V0 domain, is a 116-kDa protein and contains 4 isoforms, namely a1, a2, a3, and a4 (25). Although V-ATPase is a ubiquitous enzyme, several isoforms of subunit a have been described as being encoded by different genes with tissue-specific expression. Subunit a1 is expressed ubiquitously, whereas subunit a2 is found in kidney, lung, and spleen (23, 26, 27). In the mature osteoclast, the a3 isoform is strongly expressed in the plasma membrane (19). Moreover, V-ATPase a3 knockout mice show severe osteopetrosis due to the loss of extracellular acidity caused by the failure of bone resorption (17). The subunit a4 isoform is localized in the apical and basolateral plasma membranes of cortical intercalated cells and is important for renal acid/base homeostasis (18). Although accumulating data indicate that specific V-ATPase subunits have specialized physiological functions, the involvement of the V-ATPase subunits in tumor metastasis remains largely unclear.

In this study, we investigated whether the V-ATPase expressed in cancer cells plays a role in the pathophysiology of tumor metastasis by producing acidic microenvironments. We found that the a3 isoform of V-ATPase is strongly expressed in high-metastatic B16-F10 melanoma cells compared to low-metastatic B16 parental cells. Moreover, B16-F10 cells created evident acidic conditions when they spread to lung and bone. Knockdown of a3 V-ATPase in B16-F10 cells inhibited invasiveness and metastasis to lung and bone. Our results suggest that inhibition of a3 V-ATPase expression in tumor cells inhibits distant metastasis and thus could be a potential novel therapeutic approach for the treatment of cancer metastasis.

Mouse malignant melanoma B16 (Riken) and B16-F10 (Riken) cells were cultured in DMEM (Sigma) or DMEM-Ham's F12 media with 10% FBS and 100 μg/ml kanamycin sulfate (Meiji Seika, Ltd.). All cells were maintained in a humidified atmosphere of 5% CO₂ in air.

B16-F10 cells were i.v. injected into C57BL/6 mice via the tail vein. Two weeks after cell injection, acridine orange (AO; 1.0 mg/kg) was injected according to the procedure of Satonaka and colleagues (28). Two hours later, animals were sacrificed and lungs were excised and examined by fluorescence stereoscopic microscopy (Leica Microsytems).

Total RNA was isolated using NucleoSpin RNA II kit (Macherey Nagel) and single-stranded cDNA was synthesized using a PrimeScript 1st strand cDNA Synthesis kit (Takara Bio, Inc.). The primers used for the amplification were as follows: V-ATPase a1 (sense = 5′-TCAGTACCTGAGGAAGC-3′, anti-sense = 5′-CTGGTGGACCATGGTGTCGC-3′); a2 (sense = 5′-CGAGAAGTGACGTGTGAGGA-3′, anti-sense = 5′-ACTGAACTTGGAGGAGAGCA-3′); a3 (sense = 5′-CGAACCACCTGAGCTTTCTC-3′, anti-sense = 5′-CCCATGGAAGAGCAGATGAT-3′); a4 (sense = 5′-GCAGTGCATTGCCGAGATC-3′, anti-sense = 5′-GAACATAGGCTGGACACTCCAAG-3′); c (sense = 5′-CTGCTTGCAGACATGGCTGACATC-3′, anti-sense = 5′-GTCAGGCTGTTCGTTCTGGAATGAGGAG-3′); c″ (sense = 5′-GCTGCCATGACGGGGCTGGAGTTGCTCTAC-3′, anti-sense = 5′-GCTGAGGGACACAGCTCCAGCTGTCCCAGG-3′); d1 (sense = 5′-AGCAGATGGAGGCTGTGAACATC-3′, anti-sense = 5′-ACACCAAAATGGAACTGGTTCAGG-3′); d2 (sense = 5′-CAGAGATGGAAGCTGTCAACATTG-3′, anti-sense = 5′-ACACCATAATGGAATTGCCTGTTG-3′); and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; sense = 5′-TTGAAGGGTGGAGCCAAACG-3′, antisense = 5′-ACACATTGGGGGTAGGAACACG-3′). The PCR products were separated on 0.8% agarose gels containing ethidium bromide and visualized under ultraviolet light. Real-time PCR was performed using the Taqman PCR protocol and an ABI 7300 real-time PCR system (Applied Biosystems). Taqman primers and probes used for the amplification were as follows: mouse V-ATPase a3 (sense = 5′-TGGTGTCTTCCTTGGACCATATC-3′, anti-sense = 5′-GCTGAGGAACACCCCAAAGG-3′, probe = 5′-GCATTGACCCGATCTGGAGCCTGGCC-3′), mouse MMP2 (sense = 5′-CTCATCGCAGACTCCTGGAATG-3′, antisense = 5′-GTAATAAGCACCCTTGAAGAAGTAGC-3′, probe = 5′-TAACCTGGATGCCGTCGTGGACCTGC-3′), mouse MMP9 (sense = 5′-ACGTCAGCGGGCTTCTCC-3′, anti-sense = 5′-ATCCACCTTCTGAGACTTCAAGTC-3′, probe = 5′-TCTCCAGACACGCCCCTTGCTGAACA-3′), and mouse β-actin (sense = 5′-TTAATTTCTGAATGGCCCAGGTCT-3′, anti-sense = 5′-ATTGGTCTCAAGTCAGTGTACAGG-3′, probe = 5′-CCTGGCTGCCTCAACACCTCAACCC-3′). All mRNA expression levels were normalized to that of β-actin.

Serum-free CM of confluent B16-F10 cells cultured in pH 7.4 or pH 5.5 medium were concentrated 50-fold on Centricon-10 (10 kDa cut-off; Amicon, Inc.), determined for protein amounts using Bio-Rad DC protein assay. MMP activity in the media was assessed by gelatin zymography following the methods described previously (29). In brief, the concentrated media (20 mg/lane) were run on SDS-PAGE (10% containing 1 mg/mL gelatin). The gels were incubated in Triton X-100 (2.5%, v/v) for 1 hour, then in 50 mM Tris, 200 mM NaCl, 10 mM CaCl₂, 0.02% Brij 35 pH 7.6 overnight at 37°C. The gels were then stained with 0.5% (w/v) Coomassie brilliant blue and destained in methanol/acetic acid/water (50:10:40). White bands indicate gelatinase activity.

Western blotting was performed as described previously (30). Briefly, cells were rinsed twice with PBS and solubilized in lysis buffer (20 mM Hepes, pH 7.4, 150 mM NaCl, 1 mM EGTA, 1.5 mM MgCl₂, 10% glycerol, 1% Triton X-100, 10 μg/mL aprotinin, 10 μg/mL leupeptin, 1 mM AEBSF, 0.2 mM sodium orthovanadate). The lysates were centrifuged for 10 minutes at 4°C at 15,000 × g and boiled in SDS sample buffer containing 0.5M β-mercaptoethanol for 5 minutes. The supernatants were separated by SDS-PAGE, transferred to nitrocellulose membranes, immunoblotted with primary antibodies against V-ATPase a3, and visualized with horseradish peroxidase-coupled anti-mouse or anti-rabbit IgG antibody using an ECL detection kit.

Cultured cells were washed 3 times with ice-cold PBS and fixed with 3.7% paraformaldehyde-PBS for 20 minutes. After a 20-minute incubation with 0.1% Triton X-100-PBS, the cells were blocked for 2 hours with PBS containing 1% bovine serum albumin, incubated with anti-a3 rabbit polyclonal antibodies in 1% bovine serum albumin-PBS, washed 6 times with PBS, and visualized with Alexa 488-labeled antibodies. Antibodies against V-ATPase a1 (H-140), V-ATPase a2 (Y-20), and V-ATPase a4 (D-15) were purchased from Santa Cruz Biotechnology (Santa Cruz). The cytoskeleton was labeled with TRIC-phalloidin (Sigma). Fluorescence images were obtained using a confocal microscope (Leica Microsystems).

Mice were anesthetized with pentobarbital (0.05 mg/g body weight) and fixed by perfusion with 4% paraformaldehyde in 0.1M phosphate buffer through the left cardiac ventricle. The lungs and femurs were removed and postfixed for 24 hours in the same fixative, and the femurs were decalcified in 15% EDTA. Frozen sections were made following a conventional method. Immunohistochemical staining of V-ATPase a3 was carried out using a chick monoclonal anti-V-ATPase a3 antibody (31) and visualized using the Vectastain peroxidase ABC kit and DAB/nickel (Vector Laboratories, Inc.) according to the manufacturer's protocol. The slides were counterstained with hematoxylin.

V-ATPase a3 expression was knocked down by transfection with SureSilencing small-hairpin RNA plasmids (SuperArray Bioscience Corporation, http://www.superarray.com). A shRNA plasmid containing the V-ATPase shRNA sequence (5′-GGTGTCTTCCTTGGACCATAT-3′) or the nontargeting control shRNA plasmid (5′-GGAAtCTCATTCGATGCATAC-3′) was transfected into B16-F10 cells using FuGENE6 (Roche). Colonies resistant to 1 mg/mL puromycin (Sigma) were isolated and cloned. The knockdown effects of the shRNA were confirmed by real-time RT-PCR and Western blotting, and the stable transfectants were designated as sh-NT or sh-a3, respectively.

Cell invasion assay was performed using a Cell Invasion Assay kit (Chemicon, Inc.) according to the manufacturer's instructions. Briefly, B16-F10/parental cells, B16-F10/sh-control cells, and B16-F10/sh-a3 cells were suspended in serum-free media without phosphate and sodium bicarbonate. The cells (1 × 10⁶) were placed in the upper compartment of the 24-well cell culture chamber using inserts with a polycarbonate membrane (8-μm pore size) over a thin layer of extracellular matrix. Lower compartments were filled with DMEM containing 10% FBS. After incubation for 6 hours at 37°C in a CO₂ (5%) incubator, cells that invaded and migrated through the matrix-containing membrane and reached the lower surface of the invasion chamber were stained with Chemicon cell stain. After washing, the stain was extracted with a solution of 10% acetic acid, and 100 μL of the dye mixture was transferred to a 96-well plate and the absorbance was measured using a microplate reader (Model 550; Nippon Bio-Rad Laboratories) at a wavelength of 560 nm.

For the wound-healing assay, 1 × 10⁵ B16-F10 cells per well were plated in DMEM-Ham's F12 containing 10% FBS without phosphate and sodium bicarbonate in 24-well plates and incubated for 24 hours. After confirming that a complete monolayer had formed, the monolayers were wounded by scratching lines in them with a standard 200-μL plastic tip. Migration and cell movement throughout the wound area was observed with a phase-contrast microscope after 22 hours. The distance that the cells had migrated was measured on the photograph. The percent of the wounded area filled was calculated as follows: [(mean wounded breadth − mean remained breadth)/mean wounded breadth] × 100 (%).

The cell proliferation assay was performed using Promega's CellTiter-Glo Luminescent Cell Viability Assay (Promega) according to the manufacturer's protocol. Briefly, B16-F10 cells were plated into 96-well plates (500 cells/well) and incubated at 37°C in a 5% CO₂ atmosphere. On days 1, 2, 3, and 4, cell proliferation reagent was added to each well and incubated for 1 hour. The cell number was measured by GloMax 96 Microplate Luminometer (Promega).

V-ATPase a3 specific inhibitor, FR167356, was kindly provided by Astellas Pharma, Inc. (32). FR167356 suspended in 0.5% methylcellulose was orally administered for 10 days at a volume of 200 mg/kg/d.

B16-F10 parental, sh-NT, and sh-a3 cells (5 × 10⁵ in 0.05 mL PBS) were inoculated s.c. into the right hind footpads of 7-week-old male C57BL/6 mice (SLC). The tumor volume was determined at 10, 12, 15, and 18 days after inoculation according to the following formula: tumor volume (mm³) = S² × L × 0.5, where L and S represent the size in millimeters of the longest and shortest axis of the tumor, respectively.

Mice were anesthetized with pentobarbital (0.05 mg/g body weight; Dainippon Pharmaceutical Co., Ltd.), and B16-F10 parental, sh-NT, and sh-a3 cells (2 × 10⁵ in 0.1 mL PBS) were injected through the tail vein to evaluate lung metastasis. The number of metastatic tumors was counted under a binocular microscope. For the bone metastasis assay, parental B16-F10, B16-F10/sh-NT, and B16-F10/sha3 cells (2 × 10⁵ in 0.1 mL PBS) were injected into the left heart ventricle in 7-week-old male C57BL/6 mice (SLC). Twelve days after injection, animals were sacrificed and the femurs were harvested for histomorphometric examination. The area of tumor burden/total area in the metastases found in the distal femurs was determined using longitudinal H&E sections. The number of mice used in each experiment is described the figures. All experiments were conducted according to the ethical guidelines of the institutional review boards and approved by the Institutional Animal Use Committee of the Osaka University Graduate School of Dentistry.

Data are expressed as the mean ± SD. The data were analyzed by 1-way ANOVA followed by the Tukey–Kramer test (StatView; SAS Institute Inc.) for determination of differences between groups. P < 0.05 was considered significant.

We first examined whether an acidic microenvironment was created in the metastatic tumor in vivo using the pH-sensitive dye AO, a well-known marker for acidity. As shown in Figure 1A, B16-F10 melanoma cells showed exclusive accumulation of AO in lung metastasis. Notably, AO was detected beyond the tumor boundary of B16-F10 cells (Fig. 1A3), suggesting that B16-F10 cells create acidic microenvironment in lung metastasis.

We next examined whether acidic microenvironment promotes protease expression and activity in B16 melanoma cells. High-metastatic B16-F10 cells (Fig. 1B) and low-metastatic B16 cells (Supplementary Fig. S1) treated with acidic medium (pH 5.5) exhibited marked elevation in MMP2 and MMP9 expression. Moreover, we observed that B16-F10 cells cultured in low pH medium showed increased secretion of gelatinase compared to normal medium (Fig. 1C). Acidic condition also increased the expression of V-ATPase a1, a2, a3, c, c” and d1 isoforms (Supplementary Fig. S2). These data suggest that lung metastatic B16-F10 melanoma cells can create acidic microenvironments and is a model system that could be used to investigate the role of V-ATPase in distant metastasis.

To elucidate the role of V-ATPase in tumor metastasis, a detailed analysis of V-ATPase subunit expression was performed using B16 (low-metastatic) and B16-F10 (high-metastatic) cells. A comprehensive analysis of V-ATPase subunit expression by RT-PCR revealed that expression of the V-ATPase a3 isoform was increased in B16-F10 cells compared to B16 cells (Fig. 2A). Real-time PCR analysis (Fig. 2B) and Western blotting (Fig. 2C) also showed that B16-F10 cells expressed increased amounts of the V-ATPase a3 compared to B16 cells.

Notably, V-ATPase a3 was expressed in both the plasma membrane and cytoplasm in high-metastatic B16-F10 cells, whereas low-metastatic B16 cells expressed the V-ATPase a3 only in the cytoplasm (Fig. 2D). Immunohistochemical analysis further showed that the V-ATPase a3 is expressed in lung and bone metastasis of B16-F10 cells (Fig. 3). Collectively, these data indicate that the expression of the V-ATPase a3 correlates with the metastatic potential of B16 melanoma cells.

It is also noted that the expression of the V-ATPase a4 was reduced in B16-F10 cells, raising a possibility that a4 plays a role as well. However, knockdown of the V-ATPase a4 showed no effects on V-ATPase subunit composition, extracellular acidification, and MMP expression in B16 cells (Supplementary Fig. S3). Moreover, immunocytochemical examination showed that the V-ATPase a4 was not localized in the plasma membrane (Supplementary Fig. S4C), suggesting that the V-ATPase a4 was not involved in the generation of tumor acidity.

To determine whether the increased expression of the V-ATPase a3 is responsible for the high metastatic ability of B16-F10 cells, we introduced a stable, short hairpin RNA targeting V-ATPase a3 (sh-a3) or a nontargeting short hairpin RNA (sh-NT) into B16-F10 cells. We confirmed the knockdown of the V-ATPase a3 expression by real-time PCR analysis (Fig. 4A) and Western blotting (Fig. 4B). Importantly, the knockdown of a3 inhibited the development of acidic conditions by B16-F10 in vivo, whereas sh-NT cells failed to impair acidic conditions (Fig. 4C). We confirmed that knockdown of the V-ATPase a3 had no effects on the V-ATPase subunit composition (Fig. 4D). These data indicate that sh-a3 cells allow us to examine the role of tumor acidity induced by the V-ATPase a3 in cancer metastasis.

To determine the role of the V-ATPase a3 in tumor metastasis, the capacity of parental, sh-NT, and sh-a3 cells to develop distant metastases following tail or intracardiac inoculation was examined. There was no difference in tumor growth between these cells when inoculated subcutaneously into the plantar region of the unilateral hind paw (Fig. 5A and Supplementary Fig. S5). Notably, sh-a3 cells displayed significantly decreased lung metastases compared with parental or sh-NT cells (Fig. 5B and C). Moreover, histomorphometric examination revealed that bone metastasis significantly decreased in animals injected with sh-a3 cells compared with parental or sh-NT cells (Figs. 5D and E). These data suggest that suppression of the V-ATPase a3 expression inhibits pulmonary and bone metastases.

We next examined the mechanism underlying the decreased lung and bone metastases in sh-a3 cells. To approach this, we investigated the effects of V-ATPase a3 knockdown on in vitro metastatic activity. To reduce the buffering of the culture medium, the cells were cultured in media-lacking phosphate and sodium bicarbonate. We found that in vitro invasion activity was suppressed in sh-a3 cells compared to parental and sh-NT cells as determined by an invasion assay using a chamber coated with extracellular matrix (Fig. 6A). Consistent with this result, the expression of MMP2 and MMP9 (Fig. 6B) and in vitro migration activity (Fig. 6C) were also decreased in sh-a3 cells compared to parental cells and sh-NT cells. In contrast, we found that the invasion activity and MMP expression were not altered in sh-a3 cells compared to control when cultured in the normal media containing HEPES buffer (Supplementary Fig. S6), suggesting that decreased pH but not actin-remodeling is likely involved in the reduction of cancer invasiveness in sh-a3 cells. Consistent with the in vivo results, there were no significant differences in in vitro proliferation between these cells (Fig. 6D).

Finally, we examined the effect of the specific V-ATPase a3 inhibitor FR167356 on B16-F10 cancer metastasis. As shown in Figure 7A, FR167356 significantly reduced MMP9 expression in vivo, while the expression of V-ATPase a3 was not altered. More importantly, bone metastasis of B16-F10 cells was significantly reduced by FR167356 compared to untreated group (Fig. 7B).

Because acidic microenvironments are known to be associated with tumor aggressiveness, it is plausible that targeting the molecules responsible for the creation of tumor-associated acidic microenvironments is a novel mechanism-based anticancer treatment. To pursue this, identification of the specific V-ATPase subunit involved is a necessary first step. Here, we showed that the a3 isoform of V-ATPase is strongly expressed in high-metastatic B16-F10 melanoma cells, which these cells created acidic environments in the metastases to lung and bone and, most importantly, that suppression of a3 V-ATPase decreases the metastatic potential of these cells through an inhibition of invasiveness.

Our data suggest that a3 V-ATPase promotes distant metastasis of B16-F10 melanoma cells by stimulating invasiveness through increasing the expression and activity of MMP-2 and MMP-9. Acidic microenvironments are associated with the activation, secretion, and cellular distribution of many proteases involved in the degradation of ECM (33–35). Although we examined the expression and activation of MMP2 and MMP9 in this study, other proteases including cathepsin B, cathepsin L are also up-regulated by acidic microenvironments (8, 36). In addition to increased protease expression and activity, our data show that knockdown of V-ATPase a3 caused the decrease in cellular motility. Several reports indicated that V-ATPase in plasma membranes also colocalizes with actin at the leading edge of migrating cells, suggesting that V-ATPase is involved in dynamic actin remodeling (37, 38). These data collectively indicate that V-ATPase a3 in plasma membranes possesses dual roles and thus contributes to the increase in cancer invasiveness.

In this study, we focused on a3 V-ATPase as the key modulator of tumor acidity in B16 melanoma cells, because the expression of a3 V-ATPase was much greater in high-metastatic cells compared to low-metastatic cells. Several groups have also reported a correlation between V-ATPase expression and metastatic ability in the human breast cancer cell line, MDA-MB-231 (21, 39). Moreover, suppression of V-ATPase subunit c using siRNA inhibits the metastasis of human hepatocellular carcinoma and breast cancer cells (22, 40). We found that V-ATPase subunit a2 was expressed in plasma membrane of B16-F10 cells (Supplementary Fig. S4). These data indicate that identification of the specific components of V-ATPase that are solely involved in tumor malignancy is crucial for successful determination of the molecular target for anticancer treatment.

Although several previous reports described potential inhibitory effects of V-ATPase on cancer metastasis, targeting V-ATPase for anticancer therapy is rather complicated and difficult. V-ATPase expressed on the plasma membrane has many biological functions in various organs including renal intercalated cells, macrophages, and osteoclasts. Moreover, V-ATPase expressed in intracellular compartments isinvolved in a number of critical physiological cellular steps such as endocytosis, protein processing and degradation, and coupled transport of small molecules (reviewed in ref. 13). Accordingly, inhibition of V-ATPase may cause unpredictable adverse effects in various organs and cellular events. In this regard, however, targeting selectively a3 V-ATPase for the treatment of bone metastases may have some advantages because the major abnormal phenotype in mice deficient of a3 V-ATPase was severe osteopetrosis due to dysfunction of osteoclasts (17), suggesting that osteoclasts in bone may be the only cells in which a3 V-ATPase plays critical physiological roles. Osteoclasts play a central role in the development and progression of bone metastases (41, 42). Inhibition of osteoclasts by agents such as bisphosphonates has now been established as an effective therapeutic intervention for bone metastases of breast, lung, and prostate cancers (43). Furthermore, acidic conditions are shown to activate osteoclastic bone resorption (44). Along this line, it is noted that a specific inhibitor of a3 V-ATPase, FR167356 significantly reduced bone metastasis of B16-F10 cells (Fig. 7). Moreover, Niikura and colleagues also found FR167356 significantly inhibited bone metastases of 4T1 mouse breast cancer cells (45). Taken together, we believe that inhibition of a3 V-ATPase is a promising mechanism-based specific approach to suppress distant metastases.

Although acidification is mediated by a V-type proton ATPase, it has been known that a parallel anion pathway conducted by chloride channels is also important to allow bulk proton transport (46). Thus, we examined the expression pattern of several chloride channels implicated in cancer metastasis (47–49). We found that Clca1, Clca5, and Clc-4 were highly expressed in B16-F10 cells compared to B16 cells (Supplementary Fig. S7). Moreover, the expression of Clc-7, which plays an essential role as a Cl⁻/H⁺ antiporter in osteoclasts (50, 51), was also increased (Supplementary Fig. S7). These data suggest that chloride channels could function as a Cl⁻/H⁺ antiporter in highly metastatic cancer cells.

It has been well recognized that the interactions between cancer cells and endothelial cells, pericytes, and fibroblasts are important for cancer pathophysiology. Therefore, influences of tumor-associated acidic environments on the activity and function of these associated cells may in turn affect the progression of cancer itself. Cancer-associated fibroblasts is one of the most crucial components of the tumor microenvironment, which promotes the growth and invasion of cancer cells (52). We observed that acidic condition increased the expression of MMP2 and MMP9 in bone marrow stromal cells (data not shown). Moreover, we have recently reported that acid activation of the capsaicin receptor Trpv1 leads to an upregulation of the algesic substance calcitonin gene-related peptide expression in dorsal root ganglion neurons via the CaMK-CREB cascade, suggesting that tumor-associated acidic environments play a role in causing cancer pain (53). These data collectively indicate that tumor-associated acidic microenvironments are involved in varieties of aspects of the pathogenesis of cancer.

In conclusion, our data suggest that the expression of the a3 isoform of V-ATPase is closely associated with metastatic potential in B16 melanoma cells. Moreover, the acidic condition generated by a3 V-ATPase promotes invasiveness and cell motility of cancer cells. Our results suggest that inhibiting a3 V-ATPase function in tumor cells could be a novel mechanism-based therapeutic approach for the treatment of tumor metastasis.

No potential conflicts of interest were disclosed.

We are grateful to Dr. Riko Nishimura (Osaka University Graduate School of Dentistry) for helpful discussions.

This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas (T. Yoneda) from the Ministry of Education, Culture, Sports, Science and Technology of Japan, the 21st Century COE Program (T. Yoneda) and The Naito Foundation (K. Hata). This study is partially supported by a Research Grant from the Princess Takamatsu Cancer Research Fund (No. 08-24020).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.