A Novel Function of CXCL13 to Stimulate RANK Ligand Expression in Oral Squamous Cell Carcinoma Cells

Published OnlineFirst August 20, 2009; DOI: 10.1158/1541-7786.MCR-08-0589

Sambandam Yuvaraj,1 Alfred C. Griffin,2 Kumaran Sundaram,1 Keith L. Kirkwood,2 James S. Norris,3 and Sakamuri V. Reddy1

1Charles P. Darby Children's Research Institute, 2College of Dental Medicine, and 3Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, South Carolina

Abstract (IL)-6 and IL-8 have been implicated as potential biomarkers Oral squamous cell carcinomas (OSCC) are malignant for OSCC (3). Chemokines are implicated in tumor progression tumors with a potent activity of local bone invasion/ and metastasis (4). More recently, gene expression profiling osteolysis. The chemokine ligand, CXCL13, has been studies implicated chemokine ligand-13 (CXCL13) in OSCC identified as a prognostic marker for OSCC development and tumor development/progression (5). CXCL13 (BCA-1) that progression. Here in, we show that recombinant hCXCL13 binds monogamously to the CXCR5 receptor is involved in treatment of OSCC cells stimulates (5-fold) RANK ligand B-cell chemotaxis (6). Osteoclast is the bone-resorbing cell (RANKL), a critical bone resorbing osteoclastogenic factor and RANK ligand (RANKL), a member of the tumor necrosis expression. Anti-CXCR5 chemokine receptor antibody factor family, which is expressed in bone marrow stromal/pre- abrogates CXCL13-induced RANKL expression in these osteoblast cells, is critical for osteoclastogenesis (7). Most re- cells. Also, CXCL13 stimulated (3.0-fold) hRANKL gene sorption stimuli in normal and pathologic conditions induce promoter activity in SCC14a cells. SuperArray screening for osteoclast formation by modulating RANKL gene expression transcription factors by real-time RT-PCR identified in marrow stromal/osteoblast cells. Osteoclast activation plays significant increase in the levels of c-Jun and NFATc3 mRNA an important role in several malignancies, including oral can- expression in CXCL13-stimulated OSCC cells. CXCL13 cers invasion of bone/osteolysis (8). However, a role for treatment significantly increased (3.5-fold) phospho-c-Jun CXCL13/CXCR5 in RANKL expression in OSCC cells is un- levels in these cells and a c-Jun-NH2-kinase inhibitor known. We previously cloned a 2-kbpromoter region of the abolished CXCL13-stimulated RANKL expression. hRANKL gene and characterized the fibroblast growth factor- Furthermore, we show that CXCL13 stimulation induced 2 transcriptional control of RANKL gene expression in human nuclear translocation of NFATc3 in OSCC cells. Chromatin- bone marrow stromal/osteoblast cells (9). immune precipitation assay confirmed NFATc3 binding to The nuclear factor of activated T cells (NFAT) family of pro- the RANKL promoter region. We also show that teins exists as phosphorylated proteins in the cytosol in their overexpression of NFATc3 stimulates RANKL expression/ transcriptionally inactive state. The NFAT transcription factor promoter activity and that siRNA suppression of NFATc3 family includes five members, NFATc1 (NFAT2), NFATc2 abolished CXCL13-stimulated RANKL expression. Thus, our (NFAT1), NFATc3 (NFAT4), NFATc4 (NFAT3), and NFAT5. results suggest that NFATc3 is a downstream target of the NFAT activation is mediated by calcineurin, a calcium/calmod- CXCL13/CXCR5 axis to stimulate RANKL expression in ulin-dependent phosphatase (10). NFAT proteins autoregulate OSCC cells and implicates CXCL13 as a potential and cooperate with the activator protein (c-Jun/Fos) transcrip- therapeutic target to prevent OSCC bone invasion/ tion factors at the cellular level (11). Several genes associated osteolysis. (Mol Cancer Res 2009;7(8):1399–407) with osteoclast development/bone resorbing activity such as tartrate-resistant acid phosphatase, matrix metalloproteinase 9, OSCAR, calcitonin receptor, and cathepsin K have been shown Introduction to be modulated by NFATc1 (10, 12). Recently, NFATc1 and Oral squamous cell carcinoma (OSCC) is the most common NFATc3 were reported to increase osterix transcription activity oral or pharyngeal malignant tumors estimated ∼30,000 cases upon coexpression with PIASxβ in osteoblast cells (13). In this each year in the United States. Etiologic factors for OSCC in- study, we showed that NFATc3 is a downstream target of cludes both a genetic predisposition (1) and exposure to carci- CXCL13 to induce RANKL expression in OSCC cells. Thus, nogens such as tobacco, alcohol, chronic inflammation, and the our results implicate CXCL13 as a potential therapeutic target human papilloma virus infection (2). Previously, interleukin to prevent osteolysis associated with OSCC.

Results Received 1/6/09; revised 4/15/09; accepted 5/6/09; published OnlineFirst 8/11/09. CXCL13 Stimulates RANKL Expression in OSCC Cells The costs of publication of this article were defrayed in part by the payment of Tumor cells activate bone resorbing osteoclasts, thereby advertisement page charges. This article must therefore be hereby marked in facilitating the osteolytic process and bone invasion (14). accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Sakamuri V. Reddy, Charles P. Darby Children's Research RANKL is a critical osteoclastogenic factor in the bone micro- Institute, 173 Ashley Avenue, Charleston, SC 29425. Phone: 843-792-8373; Fax: environment (7). Microarray analysis for gene expression 843-792-7927. E-mail: [email protected] Copyright © 2009 American Association for Cancer Research. profiling in OSCC identified gene signatures that include doi:10.1158/1541-7786.MCR-08-0589 chemokine (CXC motif) ligand-13 (CXCL13) that is implicated

Mol Cancer Res 2009;7(8). August 2009 1399 Downloaded from mcr.aacrjournals.org on September 28, 2021. © 2009 American Association for Cancer Research. Published OnlineFirst August 20, 2009; DOI: 10.1158/1541-7786.MCR-08-0589

1400 Yuvaraj et al.

in OSCC development and progression (5). Recent evidence to modulate RANKL expression in OSCC cells. Total RNA indicated that down-regulation of CXCL5 expression inhibits isolated from SCC14a cells treated with recombinant head and neck squamous cell carcinoma development and in- hCXCL13 (0-25 ng/mL) for a 48-hour period were subjected vasion (15). Therefore, we examined the potential of CXCL13 to real-time RT-PCR analysis to quantify the levels of RANKL

Mol Cancer Res 2009;7(8). August 2009 Downloaded from mcr.aacrjournals.org on September 28, 2021. © 2009 American Association for Cancer Research. Published OnlineFirst August 20, 2009; DOI: 10.1158/1541-7786.MCR-08-0589

RANKL Expression in Oral Squamous Cell Carcinoma 1401

mRNA expression. SCC14a cells stimulated with CXCL13 by measuring β-galactosidase activity coexpressed in these showed a dose-dependent increase in the levels of RANKL cells. These data indicate that CXCL13 signaling through mRNA expression (Fig. 1A). The mRNA expression was CXCR5 up-regulated RANKL gene transcription in OSCC normalized with respect to glyceraldehyde-3-phosphate dehy- cells, which implicates CXCL13 play an important role in drogenase (GAPDH) amplification in these cells. The condi- OSCC tumor osteolysis. tioned media obtained from OSCC cells (SCC1, SCC12, and SCC14a) stimulated with and without CXCL13 (25 ng/mL) SuperArray Screening for CXCL13 Enhanced for 48 hours were measured for soluble RANKL (sRANKL) Transcription Factors by ELISA. As shown in Fig. 1B, CXCL13 stimulation signif- To identify the transcription factors that may be involved in icantly enhanced the levels of sRANKL secretion by OSCC CXCL13 modulation of RANKL expression, we screened a cells. SCC14a cells were stimulated with different concentra- SuperArray of 84 transcription factors using a set of house- tions of recombinant human CXCL13 (0-25 ng/mL) for 48 keeping genes as controls by real-time RT-PCR. SCC14a cells were hours and total cell lysates obtained were subjected to Western stimulated with CXCL13 (25 ng/mL) for 48 hours and total blot analysis of RANKL expression. As shown in Fig. 1C, RNA isolated was used as template for SuperArray screening CXCL13 stimulation resulted in a significant increase in by real-time RT-PCR and data obtained was analyzed by Web RANKL expression in a dose-dependent manner in these cells. portal as described in Materials and Methods. We thus identi- We further examined CXCL13 stimulation of RANKL expres- fied a 3.2- and 2.8-fold increase in c-Jun and NFATc3 mRNA sion in SCC14a cells at different time points. Total cell lysates expression in CXCL13-stimulated SCC14a cells, respectively obtained from cells stimulated with CXCL13 (25 ng/mL) for dif- (data not shown). Transfac database search identified the pres- ferent time points (0-48 hours) were analyzed by Western blot ence of a putative NFAT transcription factor–binding motif at for RANKL expression. SCC14a cells stimulated with −1434 bp to −1427 bp position in the hRANKL promoter region CXCL13 for 48 hours had a 5-fold increase in the levels (9). NFATc1 plays a critical role in RANKL signaling in osteo- of RANKL expression compared with unstimulated cells clast differentiation (11). We further quantified the mRNA ex- (Fig. 1D). We further determined the levels of CXCR5 receptor pression levels of NFAT transcription factor family members in expression in OSCC cells compared with human B lympho- CXCL13-stimulated SCC14a cells by real-time RT-PCR analy- cytes (FREV) by real-time reverse transcription-PCR (RT- sis. As shown in Fig. 2A, SCC14a cells stimulated with CXCL13 PCR) analysis. We detected significant levels of CXCR5 showed high levels of NFATc3 mRNA expression and a modest mRNA expression in OSCC cells, however, at low level com- increase in the levels of NFATc1, NFATc2, NFATc4, and NFAT5 pared with B lymphocytes (Fig. 1E). Furthermore, Western blot mRNA expression compared with unstimulated cells. We then analysis of total cell lysates obtained from SCC14a cells stim- examined CXCL13 stimulation of NFATc3 protein expression ulated with CXCL13 in the presence of anti-CXCR5 receptor in SCC14a cells at different time points. Total cell lysates ob- antibody abrogated CXCL13-induced RANKL expression tained from cells stimulated with CXCL13 (25 ng/mL) for differ- compared with control IgG–treated cells (Fig. 1F). We then ex- ent time points (0-48 hours) were analyzed by Western blot for amined CXCL13 stimulation of hRANKL gene promoter activ- NFATc3 expression. Cells stimulated with CXCL13 for 48 hours ity in OSCC cells. SCC14a cells were transiently transfected had a 3.5-fold increase in the levels of NFATc3 expression com- with hRANKL gene promoter-luciferase reporter plasmid pared with unstimulated cells (Fig. 2B). These results suggested (hRANKL P#3) by lipofectamine and treated with CXCL13 CXCL13 modulates RANKL expression through specific tran- for 48 hours. SCC14a cells mock-transfected with the empty scriptional regulatory mechanisms in OSCC cells. vector (EV) served as the control. Total cell lysates prepared after a 48-hour period were assayed for luciferase activity. CXCL13 Enhances Phospho-c-Jun Expression in OSCC CXCL13 stimulation of SCC14a cells transfected with hRANKL Cells P#3 had a significant increase (3.0-fold) in hRANKL gene pro- Previously, activator protein was reported to be constitutively moter activity (Fig. 1G). Transfection efficiency was normalized activated in OSCC cells (16, 17). We examined whether

FIGURE 1. CXCL13 stimulates RANKL expression and promoter activity in OSCC cells. A. CXCL13 stimulates RANKL mRNA expression in SCC14a cells. Cells were stimulated with different concentration of recombinant (E. coli expressed) human CXCL13 (0-25 ng/mL) for 48 h. Total RNA was isolated from these cells and RANKL mRNA expression was quantified by real-time RT-PCR analysis. The mRNA expression was normalized with respect to GAPDH amplification. Columns, mean for three independent experiments (*, P < 0.05); bars, SD. B. CXCL13 (25 ng/mL) stimulates the secretion of RANKL in OSCC cells. sRANKL levels in conditioned media obtained from OSCC cells stimulated with CXCL13 for 48 h were measured by ELISA. Columns, mean of three independent experiments (*, P < 0.05); bars, SD. C. CXCL13 stimulates RANKL protein expression in SCC14a cells. Cells were stimulated with different concentrations of CXCL13 (0-25 ng/mL) for 48 h. Total cell lysates obtained were analyzed by Western blot for RANKL expression. Data represent three independent experiments (D). The SCC14a cells were stimulated with CXCL13 (25 ng/mL) at different time points (0-48 h). Total cell lysates were analyzed by Western blot for RANKL expression. β-Actin expression served as loading control. Data represent three independent experiments. The band intensity was quantified by densitometric analysis using the NIH ImageJ Program. E. CXCR5 expression in OSCC cells. Total RNA isolated from FREV (human B lymphocytes) and OSCC cells were subjected to real-time RT-PCR analysis for CXCR5 mRNA expression. Columns, mean of three independent experiments (*, P < 0.05); bars, SD. F. Blockade of the chemokine receptor, CXCR5 inhibits CXCL13 stimulated RANKL expression in SCC14a cells. Cells were stimulated with CXCL13 (25 ng/mL) for 48 h with or without CXCR5 antibody (10 μg/mL) and control IgG. Total cell lysates were analyzed by Western blot for RANKL expression. β-Actin expression serves as a control. Data represent three independent experiments. G. CXCL13 stimulates hRANKL gene promoter activity. SCC14a cells were transiently transfected with hRANKL gene promoter-luc reporter plasmid (hRANKL P#3) using lipofectamine. Cells were cultured with 25 ng/mL of recombinant hCXCL13 for 48 h. SCC14a cells mock-transfected with the EV served as the control. Total cell lysates prepared after a 48-h period were assayed for luciferase activity. Transfection efficiency was normalized by measuring β-galactosidase activity coexpressed in these cells. Columns, mean of three independent experiments (*, P < 0.05); bars, SD.

1402 Yuvaraj et al.

CXCL13 stimulation elevates phospho-c-Jun (p-c-Jun) expres- of NFATc3 (Fig. 4A). We further analyzed NFATc3 binding sion levels in these cells. Consistent with previous reports, to the hRANKL gene promoter element by chromatin immuno- SCC1, SCC12, and SCC14a cells had significant levels of p- precipitation (ChIP) assay. SCC14a cells stimulated with c-Jun expression, indicating constitutive activation of c-Jun in CXCL13 were assayed by ChIP with anti-NFATc3 antibodies these cells (Fig. 3A). Western blot analysis of total cell lysates as described in methods. PCR analysis of chromatin immune obtained from SCC14a cells stimulated with CXCL13 (25 ng/ complexes using hRANKL gene promoter–specific primers for mL) for different time points (0-30 minutes) revealed a signifi- the NFATc3 binding region identified NFATc3 binding to the cant increase (3.5-fold) in the level of p-c-Jun expression at 15- RANKL promoter region in these cells (Fig. 4B). These results minute period (Fig. 3B). We confirmed CXCL13 stimulation of suggest that NFATc3 is a downstream target for CXCL13 sig- c-Jun NH2-terminal kinase (JNK) activity in SCC14a cells as naling to stimulate RANKL expression in OSCC cells. We then described in Materials and Methods. As shown in Fig. 3C, examined the capacity of NFATc3 overexpression to stimulate CXCL13 treatment significantly increased (2.5-fold) JNK activ- RANKL expression in OSCC cells. SCC14a cells were tran- ity over basal levels of expression in these cells in a time-depen- siently transfected with human NFATc3 expression vector, dent manner. We further tested if a JNK inhibitor abrogate the and total cell lysates were analyzed by Western blot for NFATc3 CXCL13-stimulated RANKL expression in OSCC cells. West- and RANKL expression. As shown in Fig. 5A, overexpres- ern blot analysis of total cell lysates obtained from SCC14a sion of NFATc3 stimulated RANKL expression (3.5-fold) cells treated with JNK inhibitor (10 μmol/L) for a 6-hour period compared with mock-transfected cells. We further confirmed before the CXCL13 treatment (48 hours) abolished CXCL13- the specificity of NFATc3 regulation of RANKL expression induced RANKL expression in these cells (Fig. 3D). These re- through siRNA suppression of NFATc3 expression in these sults suggest that CXCL13 stimulates RANKL gene expression cells. Western blot analysis of total cell lysates obtained from through up-regulation of JNK activity in OSCC cells. cells transiently transfected with double stranded NFATc3 siRNA revealed inhibition of RANKL expression in CXCL13- NFATc3 Modulates RANKL Expression stimulated cells (Fig. 5B). siRNA suppression of NFATc3 Previously, it has been shown that JNK activation of c-Jun expression in unstimulated cells resulted in inhibition of endog- plays a critical role in expression and function of NFAT family enous levels of RANKL expression in OSCC cells. These results (18). Because CXCL13 enhanced p-c-Jun levels and NFATc3 indicate NFATc3 transcription factor plays a critical role in expression in OSCC cells, we next examined the potential of basal and CXCL13-stimulated expression of RANKL in OSCC NFATc3 to modulate RANKL expression in these cells. We cells. We next examined whether NFATc3 overexpression stim- tested CXCL13-activated NFATc3 translocation to the nucleus ulates RANKL gene promoter activity. A stable OSCC cell in OSCC cells. SCC14a cells were stimulated with CXCL13 line (SCC11B) with hRANKL-luc construct was transiently (25 ng/mL) for 24 hours and immunostained with anti-NFATc3 transfected with NFATc3 expression plasmid or EV using antibody. Confocal microscopy revealed that CXCL13 treat- lipofectamine. Total cell lysates prepared after 48 hours were ment induced nuclear translocation of NFATc3 in these cells. assayed for luciferase activity. As shown in Fig. 5C, NFATc3 In contrast, unstimulated cells showed cytosolic localization overexpression significantly (4.7-fold) stimulated hRANKL

FIGURE 2. CXCL13 stimulates NFATc3 transcription factor expression in SCC14a cells. A. SCC14a cells were stimulated with CXCL13 (25 ng/mL) for 48 h. Total RNA was isolated from these cells and subjected to real-time RT-PCR analysis for NFAT family transcription factors mRNA expression. The mRNA expression was normalized with respect to GAPDH amplification. Columns, mean for three independent experiments (*, P < 0.05); bars, SD. B. Time-dependent expression of NFATc3 in CXCL13-stimulated SCC14a cells. Cells were treated with CXCL13 (25 ng/mL) for different time points (0-48 h) and the cell lysates obtained were analyzed by West- ern blot for NFATc3 expression. Data represent three independent experiments. The band intensity was quantified by densitometric analysis using the NIH ImageJ Program.

RANKL Expression in Oral Squamous Cell Carcinoma 1403

as IL-1β have been shown to induce CXCL13 production in dif- ferentiated osteoblasts (20) and human osteoblasts have been reported to express functional CXCR5 receptors (21). Although CXCR5 expression by OSCC in situ is yet to be elucidated, we identified the CXCL13-specific CXCR5 chemokine receptor expression in OSCC-derived cells. However, evidence also indi- cates that CXCL13 has low affinity binding to CCX-CKR expressed in dendritic cells and T cells (22). Thus, CXCL13 may have a paracrine role in regulating RANKL expression in osteoblast and other cell types in the tumor-bone microenvironment. Several osteotropic factors/resorption stimuli including 1,25-(OH)2D3, parathyroid hormone (PTH), IL-1β, IL-11, and prostaglandin E2 have been shown to up-regulate RANKL expression on marrow stromal/osteoblast cells (23-25). OSCC showed mRNA expression of osteotropic cytokines such as IL-6, tumor necrosis factor α, and PTHrP (26). Therefore, RANKL expression in OSCC cells could be regulated through complex autocrine transcriptional regulatory mechanisms oper- ative in tumor cells. Recently, it has been shown that CXCL13 is overexpressed in tumor tissues and in peripheral blood of breast cancer patients (27). Thus, our findings that CXCL13 stimulates RANKL expression implicate a novel function of CXCL13 in the tumor osteolytic process and in metastasis. Nuclear factor-κB has been reported to be constitutively activated in OSCC, and inactivation of nuclear factor-κB has been shown to suppress a malignant phenotype (28). NFATc2 and nuclear factor-κB cooperatively activate the NFATc1 promoter in response to RANKL stimulation (29). Also, c-Jun ac- FIGURE 3. c-Jun activation in CXCL13-stimulated OSCC cells. A. Western blot analysis for p-c-Jun expression in OSCC cell lines. B. tivation is associated with transformation and malignancy CXCL13 stimulates p-c-Jun expression in SCC14a cells. The SCC14a progression in OSCC cells (30, 31). Our findings that CXCL13 cells were stimulated with CXCL13 (25 ng/mL) for different time points enhances p-c-Jun levels and that JNK activity is up-regulated in (0-30 min). Total cell lysates obtained were analyzed by Western blot for p-c-Jun expression. C. CXCL13 stimulates JNK activity in SCC14a cells. SCC14a cells implicate CXCL13 in constitutive activation of The cells were stimulated with CXCL13 (25 ng/mL) for different time points activator protein complex transcription factors and in the malig- (0-30 min). The total cell lysates prepared were assayed for JNK activity nant phenotype in OSCC cells. Studies indicate that STAT-1 using an SAPK/JNK assay kit (Cell Signaling) by Western blot analysis. β- Actin levels shown represent equal amount of total cell lysates used for the transcription factor is up-regulated in oral cancer and implicat- study. D. JNK inhibitor blocked expression of CXCL13-induced RANKL ed in OSCC development and growth (2). Previously it has expression. SCC14a cells were treated with JNK inhibitor (SP600125) at 10 μmol/L concentration for 6 h before the CXCL13 (25 ng/mL) treatment been shown that signal transducers and activators of transcrip- for 48 h. Total cell lysates were analyzed by Western blot for RANKL ex- tion 3 activation in stromal/osteoblastic cells is required for in- pression. β-Actin expression serves as loading control. Data shown in duced expression of RANKL by gp130-utilizing cytokines or Western blot analysis represent three independent experiments. The band intensity on Western blots was quantified by densitometric analysis using IL-1 but not 1,25-dihydroxyvitamin D3 or PTH (32). There- the NIH ImageJ Program. fore, CXCL13/CXCR5 axis–independent transcriptional regulatory mechanisms in response to other cytokines such as PTHrP could also play a role in RANKL expression in OSCC promoter activity in these cells compared with control EV plas- cells. We found no significant change in the levels of RANKL mid–transfected cells. Collectively, these results suggest that mRNA expression in CXCL13-treated cells in the presence of CXCL13 activation of NFATc3 transcription factor modulates actinomycin D (data not shown), which suggests transcriptional RANKL expression in OSCC cells. regulation of RANKL expression. Enhanced levels of RANKL expression and nuclear localization of NFATc3 in CXCL13- stimulated OSCC cells indicate CXCL13-specific NFAT activa- Discussion tion in these cells. Although the levels of NFATc3 expression is RANKL is a critical osteoclastogenic factor in the bone mi- high in CXCL13-stimulated cells compared with unstimulated croenvironment (7). Activation of bone-resorbing osteoclasts at cells, we cannot exclude the possibility that other NFAT mem- the tumor-bone interface facilitates the osteolytic process and bers that are modestly increased may regulate RANKL expres- bone invasion of cancer cells (14). Recent studies indicated sion as they have been shown to cooperate with c-Jun signaling CXCR5 chemokine receptor expression by colon carcinoma to modulate gene expression (18). Our findings that NFATc3 cells and potential role in tumor growth (19). Therefore, our find- overexpression stimulated RANKL expression and that ings that CXCL13 stimulates RANKL expression in OSCC cells CXCL13 enhanced NFATc3 binding to the hRANKL promoter implicates a potential role in osteolysis associated with OSCC region in OSCC cells suggests that NFATc3 is a downstream and tumor cell invasion of bone. Inflammatory cytokines such target to induce RANKL expression in these cells. Thus, our

1404 Yuvaraj et al.

FIGURE 4. CXCL13 activation of NFATc3. A. CXCL13 induces NFATc3 nuclear translocation in OSCC cells. SCC14a cells were stimulated with CXCL13 (25 ng/mL) for 24 h. Immunostaining for NFATc3 is shown as detected by Alexa 488–conjugated anti- goat antibody. Nuclear staining was done with DRAQ5. B. ChIP assay for NFATc3 binding to the hRANKL promoter region. SCC14a cells were stimulated with and without CXCL13 (25 ng/mL) for 24 h and ChIP assay was done using anti-NFATc3–specific antibody as described in methods. Data represent three independent experiments.

results implicate CXCL13 as a potential therapeutic target for ing RNAzol reagent (Biotecx Laboratories). The reverse tran- the prevention of OSCC bone invasion/osteolysis. scription reaction was done using poly-dT primer and Moloney murine leukemia virus reverse transcriptase (Applied Biosys- tems) in a 25-μL reaction volume containing total RNA Materials and Methods (2 μg), 1× PCR buffer, and 2 mmol/L MgCl2, at 42°C for Reagents and Antibodies 15 min followed by 95°C for 5 min. The quantitative real-time The cell culture and DNA transfection reagents were pur- RT-PCR was done using IQ SYBR Green Supermix in an iCy- chased from Invitrogen. Recombinant Escherichia coli–derived cler (iCycler iQ Single-color Real-time -PCR detection system; human CXCL13 (purity >95%) with an effective dose at 0.005 Bio-Rad). The primer sequences used to amplify GAPDH to 0.02 μg/mL to chemoattract mouse BaF/3 cells, anti- mRNA were 5′-CCTACCCCCAATGTATCCGTTGTG-3 CXCR5, and anti-human RANKL antibodies were purchased (sense) and 5′-GGAGGAATGGGAGTTGCTGTTGAA-3′ from R&D systems, Inc. Anti-NFATc3, anti–c-Jun, peroxi- (anti-sense); hRANKL mRNA 5′-ACCAGCATCAAAATCC- dase-conjugated secondary antibodies, and siRNA against CAAG-3′ (sense) and 5′-TAAGGAGTTGGAGACCT-3′ (anti- NFATc3 were purchased from Santa Cruz Biotechnology. Su- sense); CXCR5 mRNA 5′-CTTCGCGAAAGTCAGCCAAG-3′ per signal-enhanced chemiluminescence reagent was obtained (sense) and 5′-TGGTAGAGGAATCGGGAGGT-3′ (anti-sense). from Amersham Biosciences and nitrocellulose membranes Thermal cycling parameters were 94°C for 3 min, followed by were purchased from Millipore. A luciferase reporter assay sys- 40 cycles of amplifications at 94°C for 30 s, 60°C for 1 min, tem was obtained from Promega. Protease inhibitor cocktail 72°C for 1 min, and 72°C for 5 min as the final elongation was purchased from Sigma. step. Relative levels of mRNA expression were normalized in all the samples analyzed with respect to the levels of GAPDH Quantitative Real-time RT-PCR amplification. RANKL and CXCR5 mRNA expression levels were measured by real-time RT-PCR as described earlier (12). Briefly, Quantification of sRANKL total RNA was isolated from FREV (human B lymphocyte cell sRANKL levels in OSCC cell–conditioned media were line), SCC1, SCC12, and SCC14a cells treated with or without measured using an ELISA kit (R&D Systems) following the different concentrations of CXCL13 (0-25 ng/mL) for 48 h us- manufacturer's protocol.

RANKL Expression in Oral Squamous Cell Carcinoma 1405

hRANKL Promoter-Luciferase Reporter Gene Assay with 0.3 mL cell lysis reagent. The cells were scraped and spun Human OSCC–derived SCC14a cells were cultured in briefly in a microfuge to pellet the debris. A 20-μL aliquot of DMEM supplemented with 10% fetal bovine serum and 100 each sample was mixed with 100 μL of the luciferase assay units/mL penicillin/streptomycin in a humidified atmosphere reagent. The light emission was measured for 10 s of integrated with 5% CO2 at 37°C. We have previously developed time using a Sirius Luminometer following the manufacturer's hRANKL promoter-luciferase reporter plasmid construct instructions (Promega) (hRANKL P#3) as previously described (9). DNA transfections were done using lipofectamine-plus transfection reagent (Invi- Western Blot Analysis trogen). The transfection efficiency was normalized by cotrans- SCC14a cells were seeded (5 × 105 cells per well) in six- fection with 0.2 μg of pRSV β-gal plasmid and β-galactosidase well plates and supplemented with DMEM containing 10% activity was measured in the cells (Promega). LacZ cytochem- FBS. The cells were stimulated with or without CXCL13 as ical activity staining (Invitrogen, Inc.) indicated the DNA trans- indicated, and total cell lysates were prepared in a buffer con- fection efficiency of 80% in SCC14a cells. The cells were taining 20 mmol/L Tris-HCl at pH 7.4, 1% Triton X-100, stimulated with E. coli–expressed recombinant hCXCL13 (25 1 mmol/L EDTA, 1.5 mmol/L MgCl2 10% glycerol, 150 ng/mL) for 48 hours. The cell monolayer was washed twice mmol/L NaCl, 0.1 mmol/L Na3VO4, and 1× protease inhibitor with PBS and incubated at room temperature for 15 minutes cocktail. The protein content of the samples was measured

FIGURE 5. NFATc3 regulates RANKL expression. A. Overexpression of NFATc3 in SCC14a cells. Cells were transiently transfected with NFATc3 expression plasmid. After 48 h, total cell lysates obtained were analyzed with Western blot for NFATc3 and RANKL expression. Data represent three independent experiments. B. siRNA suppression of NFATc3 in SCC14a cells. The cells were transiently transfected with double-stranded NFATc3 siRNA by oligofectamine and stimulated with or without CXCL13 (25 ng/mL) for 48 h. Total cell lysates obtained were analyzed with Western blot for NFATc3 and RANKL expression. β-Actin expression served as loading control. Data represent three independent experiments. The band intensity was quantified by densitometric analysis using the NIH ImageJ Program. C. NFATc3 overexpression stimulates gene promoter activity in hRANKL-luc SCC11B stable cell line. SCC11B cells were transiently transfected with or without NFATc3 expression plasmid by lipofectamine. Total cell lysates prepared after a 48-h period were assayed for luciferase activity per milligram protein. Columns, mean for three independent experiments (*, P < 0.05); bars, SD.

1406 Yuvaraj et al.

using the BCA protein assay reagent (Pierce). Protein (100 μg) Confocal Microscopy samples were then subjected to SDS-PAGE using 12% Tris- SCC14a cells were cultured (1 × 103/well) in a Lab-Tek HCl gels and blot transferred on to a polyvinylidene difluoride four-well chamber slides (Nunc, Inc.). Cells were stimulated membrane, immunoblotted with anti-RANKL, anti-NFATc3, with and without CXCL13 (25 ng/mL) for 24 h and fixed with and anti–p-c-Jun antibodies. The bands were detected using 4% paraformaldehyde in PBS for 10 min at room temperature. the enhanced chemiluminescence detection system (Pierce). The cells were permeabilized with 0.1% Triton X-100 for 10 The band intensity was quantified by densitometric analysis us- min and blocked for 1 h with PBS containing 2% horse serum ing the NIH ImageJ Program. at room temperature. Cells were incubated with anti–rabbit- NFATc3 (10 μg/mL) antibody in PBS containing 2% horse serum for 3 h and treated with Alexa 488–conjugated anti-rabbit SuperArray Screening 6 IgG in PBS containing 2% horse serum for 1 h at room temper- SCC14a cells (5 × 10 cells per well) were cultured in a 60- ature. The nuclear staining was done with DRAQ5 and cellular mm tissue culture plate with or without CXCL13 (25 ng/mL) localization of NFATc3 was visualized by confocal microscopy for 48 h and total RNA was isolated using RNAzol reagent (LSM 510; Carl Zeiss, Inc.). (Biotecx Laboratories). Reverse transcription reaction was done using poly-dT primer and Moloney murine leukemia virus ChIP Assay μ reverse transcriptase (Applied Biosystems) in 25 L reaction ChIP was done using the ChIP Assay kit (Upstate) following μ volume containing total RNA (2 g), 1× PCR buffer, and the manufacturer's instructions. Briefly, human SCC14a cells 2 mmol/L MgCl2, at 42°C for 15 min followed by 95°C for were stimulated with and without CXCL13 (25 ng/mL) for 5 min. Real-time PCR was done using 2× SuperArray RT 2 24 h. Cells were cross-linked with 1% formaldehyde for 10 qPCR Master Mix [RT ProfilerTM PCR Array System, min. Soluble chromatin was prepared following sonication with SuperArray (PAHS-075A-02)] in 96-well plate to quantify ex- a Branson-250 digital sonifier (Branson Ultrasonics) to an av- pression levels of 84 transcription factors. Thermal cycling erage DNA length of 200 to 1,000 bp. Approximately 5 × 105 parameters were 95°C for 10 min, followed by 40 cycles of cell equivalent (1/6th) of the sheared soluble chromatin was amplifications at 95°C for 15 s, 55°C for 30 s, 72°C for 30 s, precleared with blocked Protein G agarose, and 10% of the pre- and 72°C for 5 min as the final elongation step. Relative levels cleared chromatin was set aside as input control. Immunopre- of mRNA expression were normalized in all the samples with cipitation was carried out with anti-NFATc3 antibody (5 μg) or expression levels of housekeeping genes, and data analysis was 4 equivalent concentrations of rabbit IgG as negative control done using the Webportal. overnight at 4°C. Immune complexes were pulled down using Protein G agarose, washed and eluted twice with 250 μLof JNK Assay elution buffer (0.1 mol/L NaHCO3, 1% SDS), and cross-linking μ The in vitro nonradioactive stress-activated protein kinase reversed in 200 mmol/L NaCl at 65°C overnight with 20 g (SAPK)/JNK assay was done using the SAPK/JNK assay kit RNase A (Sigma). DNA was purified following proteinase K (#9810; Cell Signaling Technology) following the manufactur- treatment (Invitrogen Life Technologies) with the Qiagen er instructions. SCC14a cells were stimulated with CXCL13 PCR purification kit (Qiagen). To analyze the NFATc3 binding hRANKL (25 ng/mL) for 0- to 30-min period. Total cell lysates (250 μg to the gene promoter region, the immunoprecipitated protein) obtained were mixed with 20 μL of c-Jun 1-89 fusion chromatin DNA samples were subjected to PCR analysis using − − protein bound to glutathione agarose beads (GST-c-Jun1-89) in primer pairs for NFATc3 binding region ( 1524 to 1324 bp) in ′ 1× lysis buffer, and incubated overnight at 4°C with constant the hRANKL promoter (9): 5 -GAT ACA CAT ATA AAT GCT ′ ′ ′ rotation. The immunoprecipitated beads were washed twice with AA-3 (sense) and 5 -CGC TAA TGA GTA TTT CTC TA-3 μ lysis buffer and twice with kinase buffer, and then they were (anti-sense). Thermal cycling parameters of a 20 L reaction resuspended in 50 μL of kinase buffer containing 2.5 μLof mixture were 94°C for 30 s, followed by 35 cycles of amplifi- 2 mmol/L ATP. The kinase reaction was carried out at 30°C cations at 94°C for 30 s, 55°C for 30 s, 72°C for 30 s, and 72°C for 30 min, and terminated by addition of 25 μLof3×SDS for 5 min as the final elongation step. PCR products were elec- sample loading buffer. The reaction mix was boiled for 5 min trophoresed on 2% agarose gels and visualized by ethidium and the supernatant was collected and loaded on a precast 4% bromide. to 20% SDS-PAGE. The gels were subsequently blotted onto polyvinylidene difluoride membranes, which were processed NFATc3 Overexpression and siRNA Interference 5 as described for the Western blots using an enhanced chemilumi- SCC14a cells were seeded (5 × 10 cells per well) in six- nescence detection kit. The level of c-Jun phosphorylation by well plates and supplemented with DMEM containing 10% fe- SAPK/JNK was determined using an anti-phospho (Ser63) tal bovine serum. One day after seeding, cells were transfected μ c-Jun antibody. The intensity of the phosphorylated c-Jun band with EV or NFATc3 expression construct (2 g) using μ represents the relative SAPK/JNK activity precipitated by c-Jun lipofectamine or with double-stranded siRNA (10 mol/L) fusion protein agarose beads. against NFATc3 (Santa Cruz Biotechnology, Inc.) by oligofectamine (Invitrogen). Nonspecific siRNA-transfected cells served as control. SiRNA-transfected cells were cultured in the presence and absence of CXCL13 (25 ng/mL) for 48 h. Total cell lysates obtained were analyzed by Western blot for 4 http://www.superarray.com/pcrarraydataanalysis.php NFATc3 and RANKL expression using specific antibodies.

RANKL Expression in Oral Squamous Cell Carcinoma 1407

Statistical Analysis Down-regulation of CXCL5 inhibits squamous carcinogenesis. Cancer Res 2006;66:4279–84. Results are presented as mean ± SD for three independent 16. Domann FE, Jr., Levy JP, Finch JS, Bowden GT. Constitutive AP-1 DNA experiments and compared by Student's t test. Values were con- binding and transactivating ability of malignant but not benign mouse epidermal sidered significant at a P value of <0.05. cells. Mol Carcinog 1994;9:61–6. 17. Turatti E, da Costa Neves A, de Magalhaes MH, de Sousa SO. Assessment of c-Jun, c-Fos and cyclin D1 in premalignant and malignant oral lesions. J Oral Sci Disclosure of Potential Conflicts of Interest 2005;47:71–6. No potential conflicts of interest were disclosed. 18. Ikeda F, Nishimura R, Matsubara T, et al. Critical roles of c-Jun signaling in regulation of NFAT family and RANKL-regulated osteoclast differentiation. J Clin Invest 2004;114:475–84. Acknowledgments 19. Meijer J, Zeelenberg IS, Sipos B, Roos E. The CXCR5 chemokine receptor We thank Dr. Thomas E. Carey, University of Michigan, Ann Arbor, MI for gen- is expressed by carcinoma cells and promotes growth of colon carcinoma in the erously providing OSCC cell lines (SCC1, SCC12, SCC14a, and SCC11B) used liver. Cancer Res 2006;66:9576–82. in this study. 20. Lisignoli G, Cristino S, Toneguzzi S, et al. IL1β and TNFα differently modulate CXCL13 chemokine in stromal cells and osteoblasts isolated from osteoar- thritis patients: evidence of changes associated to cell maturation. Exp Gerontol References 2004;39:659–65. 1. Califano J, van der Riet P, Westra W, et al. Genetic progression model for 21. Lisignoli G, Toneguzzi S, Piacentini A, et al. Human osteoblasts express head and neck cancer: implications for field cancerization. Cancer Res 1996; functional CXC chemokine receptors 3 and 5: activation by their ligands, 56:2488–92. CXCL10 and CXCL13, significantly induces alkaline phosphatase and β-N-acet- – 2. Choi S, Myers JN. Molecular pathogenesis of oral squamous cell carcinoma: ylhexosaminidase release. J Cell Physiol 2003;194:71 9. implications for therapy. J Dent Res 2008;87:14–32. 22. Gosling J, Dairaghi DJ, Wang Y, et al. Cutting edge: identification of a novel 3. St John MA, Li Y, Zhou X, et al. Interleukin 6 and interleukin 8 as potential chemokine receptor that binds dendritic cell- and T cell-active chemokines includ- – biomarkers for oral cavity and oropharyngeal squamous cell carcinoma. Arch ing ELC, SLC, and TECK. J Immunol 2000;164:2851 6. Otolaryngol Head Neck Surg 2004;130:929–35. 23. Hofbauer LC, Heufelder AE. Role of receptor activator of nuclear factor-κB – 4. Krieg C, Boyman O. The role of chemokines in cancer immune surveillance ligand and osteoprotegerin in bone cell biology. J Mol Med 2001;79:243 53. by the adaptive immune system. Semin Cancer Biol 2008;19:76–83. 24. Lee SK, Kalinowski J, Jastrzebski S, Lorenzo JA. 1,25(OH)2 vitamin D3- 5. Ziober AF, Patel KR, Alawi F, et al. Identification of a gene signature for rapid stimulated osteoclast formation in spleen-osteoblast cocultures is mediated in part α κ screening of oral squamous cell carcinoma. Clin Cancer Res 2006;12:5960–71. by enhanced IL-1 and receptor activator of NF- B ligand production in osteoblasts. J Immunol 2002;169:2374–80. 6. Legler DF, Loetscher M, Roos RS, Clark-Lewis I, Baggiolini M, Moser B. B cell-attracting chemokine 1, a human CXC chemokine expressed in lymphoid 25. Nakashima T, Kobayashi Y, Yamasaki S, et al. Protein expression and func- κ tissues, selectively attracts B lymphocytes via BLR1/CXCR5. J Exp Med tional difference of membrane-bound and soluble receptor activator of NF- B 1998;187:655–60. ligand: modulation of the expression by osteotropic factors and cytokines. Bio- chem Biophys Res Commun 2000;275:768–75. 7. Hsu H, Lacey DL, Dunstan CR, et al. Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by os- 26. Nomura T, Shibahara T, Katakura A, Matsubara S, Takano N. Establishment teoprotegerin ligand. Proc Natl Acad Sci U S A 1999;96:3540–5. of a murine model of bone invasion by oral squamous cell carcinoma. Oral Oncol 2007;43:257–62. 8. Deyama Y, Tei K, Yoshimura Y, et al. Oral squamous cell carcinomas stimulate osteoclast differentiation. Oncol Rep 2008;20:663–8. 27. Panse J, Friedrichs K, Marx A, et al. Chemokine CXCL13 is overexpressed in the tumour tissue and in the peripheral blood of breast cancer patients. Br J 9. Roccisana JL, Kawanabe N, Kajiya H, Koide M, Roodman GD, Reddy SV. Cancer 2008;99:930–8. Functional role for heat shock factors in the transcriptional regulation of human κ RANK ligand gene expression in stromal/osteoblast cells. J Biol Chem 2004;279: 28. Loercher A, Lee TL, Ricker JL, et al. Nuclear factor- B is an important mod- 10500–7. ulator of the altered gene expression profile and malignant phenotype in squamous cell carcinoma. Cancer Res 2004;64:6511–23. 10. Takayanagi H. The role of NFAT in osteoclast formation. Ann N Y Acad Sci 2007;1116:227–37. 29. Asagiri M, Sato K, Usami T, et al. Autoamplification of NFATc1 expression determines its essential role in bone homeostasis. J Exp Med 2005;202: 11. Takayanagi H, Kim S, Koga T, et al. Induction and activation of the tran- 1261–9. scription factor NFATc1 (NFAT2) integrate RANKL signaling in terminal differentiation of osteoclasts. Dev Cell 2002;3:889–901. 30. Robinson CM, Prime SS, Huntley S, et al. Overexpression of JunB in undifferentiated malignant rat oral keratinocytes enhances the malignant phe- 12. Sundaram K, Nishimura R, Senn J, Youssef RF, London SD, Reddy SV. notype in vitro without altering cellular differentiation. Int J Cancer 2001;91: RANK ligand signaling modulates the matrix metalloproteinase-9 gene expres- 625–30. sion during osteoclast differentiation. Exp Cell Res 2007;313:168–78. 31. Dong Z, Birrer MJ, Watts RG, Matrisian LM, Colburn NH. Blocking of tu- β 13. Ali MM, Yoshizawa T, Ishibashi O, et al. PIASx is a key regulator of os- mor promoter-induced AP-1 activity inhibits induced transformation in JB6 terix transcriptional activity and matrix mineralization in osteoblasts. J Cell Sci mouse epidermal cells. Proc Natl Acad Sci U S A 1994;91:609–13. 2007;120:2565–73. 32. O'Brien CA, Gubrij I, Lin SC, Saylors RL, Manolagas SC. STAT3 activation 14. Tada T, Jimi E, Okamoto M, Ozeki S, Okabe K. Oral squamous cell carci- in stromal/osteoblastic cells is required for induction of the receptor activator of noma cells induce osteoclast differentiation by suppression of osteoprotegerin NF-κB ligand and stimulation of osteoclastogenesis by gp130-utilizing cytokines – expression in osteoblasts. Int J cancer 2005;116:253 62. or interleukin-1 but not 1,25-dihydroxyvitamin D3 or parathyroid hormone. 15. Miyazaki H, Patel V, Wang H, Edmunds RK, Gutkind JS, Yeudall WA. J Biol Chem 1999;274:19301–8.

A Novel Function of CXCL13 to Stimulate RANK Ligand Expression in Oral Squamous Cell Carcinoma Cells

Sambandam Yuvaraj, Alfred C. Griffin, Kumaran Sundaram, et al.

Mol Cancer Res 2009;7:1399-1407. Published OnlineFirst August 20, 2009.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-08-0589

Cited articles This article cites 31 articles, 14 of which you can access for free at: http://mcr.aacrjournals.org/content/7/8/1399.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/7/8/1399. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.