Oncogene (2013) 32, 3339–3349 & 2013 Macmillan Publishers Limited All rights reserved 0950-9232/13 www.nature.com/onc

ORIGINAL ARTICLE Transcriptional regulation of MDR-1 by HOXC6 in multidrug-resistant cells

K-J Kim1, S-M Moon1, S-A Kim2, K-W Kang3, J-H Yoon4 and S-G Ahn1

Resistance to chemotherapeutic drugs is a significant clinical problem in the treatment of cancer and this resistance has been linked to the cellular expression of multidrug-efflux transporters. The aim of this study was to explore the role of HOXC6 in the regulation of multidrug resistance (MDR) to chemotherapeutic drugs. The HOXC6 gene was identified as being overexpressed in drug-resistant cells compared with parental cell lines. Transfection assays demonstrated that HOXC6 activated MDR-1 promoter activity. A series of MDR-1 promoter deletion mutants was examined and the minimal HOXC6-responsive region was identified to be in the TAAT motif ( À 2243 bp) of the MDR-1 promoter. Interestingly, overexpression of HOXC6 in the parental cell lines resulted in the upregulation of MDR-1 expression. The inhibition of HOXC6 using small interfering RNA led to the repression of MDR-1. We determined that knockdown of HOXC6 expression in MDR cells increased their sensitivity to paclitaxel. Flow cytometry analysis suggested that siHOXC6 could induce paclitaxel-induced apoptosis and that this was accompanied by an increased accumulation and a decreased release of paclitaxel. Taken together, our findings suggest that HOXC6 expression is an important mechanism of chemotherapeutic drug resistance via its regulation of MDR-1.

Oncogene (2013) 32, 3339–3349; doi:10.1038/onc.2012.354; published online 20 August 2012 Keywords: HOXC6; MDR-1; multidrug resistance (MDR); apoptosis

INTRODUCTION PI3K/Akt, Notch and Wnt signaling pathways.13–17 Although The homeobox (HOX) genes belong to the homeoprotein family of HOXC6 is critical for various regulated cellular processes and transcription factors, which have an important role in morphogen- correlates with cancer progression, the function of HOXC6 is esis and cellular differentiation during the embryonic develop- largely unknown. ment of all multicellular organisms.1–4 There are 39 different Cancer chemotherapy is limited by acquired MDR, which is human HOX genes that are clustered into four different groups resistance to multiple structurally, mechanistically and functionally (HOXA, HOXB, HOXC, and HOXD).1,2 Recent studies have unrelated anticancer drugs.18,19 The overexpression of drug efflux demonstrated that HOX gene expression is deregulated in many transporters, including the ATP-binding cassette (ABC) membrane tumor types, including lung, breast, ovarian, sarcoma and transporters such as P-glycoprotein (ABCB1/MDR-1; Pgp), MDR leukemia.1,5 Although HOX genes are overexpressed in various protein 1 (ABCC1/MRP-1), and the breast-cancer resistance protein tumors and organisms, their functional differences are unknown. (ABCG2/BCRP) is recognized as a biological mechanism that The HOX family member HOXC6 is associated with several contributes to multidrug resistance (MDR).18,20,21 These membrane human carcinomas, including gastrointestinal, breast, lung and transporters pump anticancer agents out of the cell through the prostate, and HOXC6 is overexpressed in osteosarcomas and cell membrane to reduce intracellular drug accumulation.22 medulloblastomas.6–10 It has been demonstrated that HOXC6 is We hypothesize that HOXC6 may regulate MDR. Here, our data expressed in response to hormonal signals and that the level of provided novel evidence that HOXC6 induces the expression of HOXC6 expression varies based on developmental stage.3,11 MDR-1 in MDR cells. MDR-1 was identified as a direct regulatory In LNCaP cells, small interfering RNA (siRNA) knockdown of target of HOXC6 in MDR cell lines. We have also identified some of HOXC6 expression induces apoptosis, and the overexpression HOXC6’s biological function, including the apoptosis pathways of HOXC6 results in increased proliferation and decreased that are directly affected by MDR-1 repression. These data suggest apoptosis.12 In a recent study, HOXC6 was shown to have an a potential role for HOXC6 in the regulation of chemotherapeutic important role in several cellular events including prostate drug resistance. branching morphogenesis and bone metastasis through the regulation of its functional biological targets. Several HOXC6 functional targets include CD44, T-cell receptor alternate reading RESULTS frame protein, insulin-like growth factor binding protein 3, and Induction of HOXC6 and MDR-1 in MDR cancer cells neutral endopeptidase/membrane metallo-endopeptidase.13,15 In In our previous study, to confirm the expression of the HOX family addition, HOXC6 has been reported to modulate the expression of in the paclitaxel-resistant FaDu cell line (FaDu/PTX), reverse bone morphogenic protein 7, fibroblast growth factor receptor 2, transcription PCR (RT–PCR) was performed using specific oligonu- and platelet-derived growth factor receptor a and to regulate the cleotides against each of the HOX family genes (HOXA, HOXB,

1Department of Pathology, School of Dentistry, Chosun University, Gwangju, Republic of Korea; 2Department of Biochemistry, Oriental Medicine, Dongguk University, Gyeongju, Republic of Korea; 3College of Pharmacy, Seoul National University, Seoul, Republic of Korea and 4Department of Oral Pathology, Daejeon Dental Hospital, Wonkwang University, Daejeon, Republic of Korea. Correspondence: Dr S-G Ahn, Department of Pathology, School of Dentistry, Chosun University, #375, Seosuk-Dong, Dong-gu, Gwangju 501-759, Republic of Korea. E-mail: [email protected] Received 15 February 2012; revised 26 June 2012; accepted 30 June 2012; published online 20 August 2012 HOXC6 regulates MDR-1 expression K-J Kim et al 3340 HOXC and HOXD). We showed that 16 of 33 HOX genes were control cells) when HOXC6 expression was induced (Figures 2f, g expressed in FaDu/PTX and parental cells. Of these 16 expressed and h). The proliferation inhibition rate of HOXC6-transfected genes, two HOX genes (HOXA5 and HOXC6) were importantly FaDu cells treated with or without paclitaxel was detected using expressed in FaDu/PTX compared with parental cell. The other 14 an MTT assay. These results showed that HOXC6 overexpression HOX genes (HOXA4, HOXA6, HOXA10, HOXA11, HOXB6, HOXB13, increased the resistance of FaDu cells to paclitaxel (Figure 2i). HOXC9, HOXC10, HOXC11, HOXC13, HOXD8, HOXD9, HOXD10 and HOXD11) were downregulated in FaDu/PTX cells. The remaining siRNA targeting HOXC6 induces intracellular Rhodamine 123 17 of 33 HOX genes were not detectable in both the cell lines. One accumulation of these two expressed HOX genes (HOXA5) were slightly To determine the effect of HOXC6 depletion in MDR cells, we expressed in FaDu/PTX cells. Of particular interest, the expression downregulated HOXC6 using siRNA transfection and determined of HOXC6 was found to be significantly higher in FaDu/PTX cells the in vitro anticancer drug sensitivities of each cell line. We than in the parental cells (Supplementary data 1). Therefore, prepared total RNA and whole-cell protein lysates at 48 h post human MDR cancer cell lines were used in an initial screen to transfection. Parallel transfections using scramble siRNA were used identify potential associations between HOXC6 and MDR genes as negative controls. Western blotting indicated that HOXC6 related to drug resistance. expression was decreased in siHOXC6-transfected MDR cells. In this study, RT–PCR constantly showed that HOXC6 was HOXC6 siRNA inhibited MDR-1 mRNA expression in MDR cells upregulated in FaDu/PTX cells and other drug-resistant cancer cell compared with the scramble siRNA -transfected cells (Figure 3a). lines (MCF-7/ADR and SNU601/CIS) compared with the parental quantitative RT–PCR confirmed that expression of the MDR-1 cells (Figure 1a). The expression of MDR genes, which are involved transcript was inhibited in siHOXC6-transfected MDR cells in MDR, was evaluated by RT–PCR using specific primers. The (Figure 3b). In addition, we observed suppression (B4-fold mRNA and protein expression levels of MDR-1 were relatively relative to the control) of MDR-1 promoter activity when HOXC6 higher in FaDu/PTX, SNU601/CIS and MCF-1/ADR cells than in the expression was inhibited in FaDu/PTX cells (Figure 3c). These parental cell lines (Figures 1b and c). In addition, quantitative results were similar in both MCF-7/ADR and SNU601/CIS cell lines RT–PCR analysis indicated that HOXC6 and MDR-1 were significantly (Figures 3d and e). upregulated in three MDR cell lines (Figure 1d). The ratio of HOXC6 To determine whether the inhibition of HOXC6 expression was mRNA expression was46-fold in all MDR cell lines, and the highest sufficient to sensitize the drug-resistant FaDu/PTX cells to ratio was a ninefold increase, as determined using real-time RT–PCR. chemotherapy, an additional experiment was performed. FaDu/ Similarly, MDR-1 mRNA was increased in these cell lines (Figure 1d). PTX, MCF-1/ADR and SNU601/CIS cells were transfected with However, unlike MDR-1, no characteristic patterns in MRP-1 and HOXC6 siRNA and administered Rho123. The cells were counted at BCRP expression were observed in these cell lines. 48 h post transfection. Rho123 accumulation increased after the To test the effect of HOXC6 expression on MDR-1 promoter administration of verapamil with Rho123 (positive control) activity, we generated a luciferase reporter construct driven by the compared with the negative control (without Rho123 treatment). 3 kb MDR-1 promoter. Parental and MDR cancer cells were The siHOXC6-transfected cells had significantly higher levels of transiently transfected with the MDR-1 promoter-luciferase Rho123 accumulation than the control cells (Figures 3f, g and h). reporter gene or an empty reporter vector. The higher endogen- ous expression of HOXC6 in MDR cancer cells (FaDu/PTX, SNU601/ CIS, and MCF-1/ADR) is associated with higher MDR-1 promoter Identification of the HOXC6-binding site in the MDR-1 promoter activity (Figures 1e, f, and g). These results are in agreement with We have demonstrated that HOXC6 can activate the MDR-1 our data from the real-time RT–PCR experiments, indicating that promoter–reporter activity; however, it is not known which region HOXC6 regulates MDR-1 expression in a positive manner. of the MDR-1 promoter is directly responsible for HOXC6 activation. We next attempted to define the binding sites of HOXC6 in the MDR-1 promoter by serially deleting regions within Overexpression of HOXC6 induces the upregulation of MDR-1 the promoter. AtB3 kb immediately upstream of the MDR-1 mRNA levels and increases drug resistance transcriptional start site, we identified several potential HOX It was previously demonstrated that HOXC6 indirectly regulates binding sites. We constructed a series of deletions aimed at the Notch signaling pathway,13 and that Notch-1 is involved in removing these sites in a stepwise manner. As depicted in MDR-1 expression.23 To determine whether Notch-1 is involved in Figure 4a, the deletion constructs pGL3-MDR-1-3001 and pGL3- MDR-1 expression, we measured Notch-1, Notch-2 and Hes-1 MDR-1-2859 had no drastic effects on their ability to bind HOXC6. (a major target of Notch-1). We observed higher levels of Notch-1 However, the deletion construct, pGL3-MDR-1-1966 demonstrated and Notch-2 proteins in the FaDu/PTX cells compared with the a dramatically decreased ability to be activated by HOXC6. These parental FaDu cells (Figure 2a). As expected, the higher Notch data suggest that there may be critical binding sites for HOXC6 in expression in HOXC6-expressing FaDu cells was associated with the MDR-1-2859 bp promoter region. higher Notch activity (fourfold) as measured by Hes-1 promoter It was previously demonstrated that HOXC6 is a member of the activity (Figure 2b). However, mRNA levels of MDR-1 in Notch HOX family of transcription factors, which bind to HOX core- intracellular domain (NICD) (the cleaved form of the Notch-1 binding motifs (TAAT) of target gene promoters to regulate gene protein)-overexpressing FaDu cells were not affected (Figure 2c). expression.6 According to the MDR-1 promoter sequence analysis These results indicate that HOXC6-induced Notch did not affect of the region from À 2859 to À 1966 bp, there are five tentative the expression of MDR-1. TAAT element binding sites for HOXC6 (Figure 4b). We then examined whether the overexpression of HOXC6 in the To determine which binding site is responsible for MDR-1 parental cells (FaDu, MCF-7 and SNU601) could affect the promoter activation by HOXC6, four MDR-1 promoter deletion expression of MDR-1 and drug resistance. We observed higher mutants were generated using MDR-1-2859 bp as a template. We levels of HOXC6 proteins in the HOXC6-transfected cells compared named the four mutants with deleted TAAT elements as pGL3-MDR- with the control cells. Remarkably, the overexpression of HOXC6 1-P1, pGL3-MDR-1-P2, pGL3-MDR-1-P3 and pGL3-MDR-1-P4 (Figures caused an increased expression of MDR-1 mRNA in FaDu cells 4b and c). As indicated in Figure 4c, the pGL3-MDR-1-P3 deletion (Figure 2d). A quantitative evaluation of HOXC6 overexpression mutant inhibited HOXC6 activation of the MDR-1 promoter. Thus, was determined in FaDu cells at 48 h after pcDNA3-HOXC6 using these results revealed that the À 2243 bp TAAT motif located in the flow cytometry (Figure 2e). In addition, we observed higher levels À 2859 bp to À 1966 bp region of the MDR-1 promoter was of MDR-1 promoter activity (an B6-fold increase compared with responsible for MDR-1 promoter activation by HOXC6.

Oncogene (2013) 3339 – 3349 & 2013 Macmillan Publishers Limited HOXC6 regulates MDR-1 expression K-J Kim et al 3341

Figure 1. Expression of HOXC6 and MDR-1 was elevated in MDR cell lines. (a) RT–PCR of the HOX gene family in FaDu/PTX, MCF-7/ADR and SNU601/CIS cells. After mRNA was purified from these cells, RT–PCR was performed using primers as described in Table1. C represents the parental cells; and R represents the MDR cells (b) RT–PCR was performed to determine the expression levels of HOXC6 and MDR-related genes (MDR-1, MRP-1 and BCRP) in FaDu/PTX, MCF-7/ADR and SNU601/CIS cell lines. Glyceraldehyde 3-phosphate dehydrogenase was used as an internal control. C, parental cells; and R, MDR cells. (c) The cells were extracted as described in the Materials and Methods section. The HOXC6 and MDR-1 expression levels were analyzed using western blot. C, parental cells; and R, MDR cells. (d) The HOXC6 and MDR-1 mRNA levels in FaDu/PTX, MCF-7/ADR and SNU601/CIS cells were assessed using real-time RT–PCR analysis. The expression levels were normalized to glyceraldehyde 3-phosphate dehydrogenase. The results were presented as the percentage relative to control. (e–g) A luciferase reporter assay was used to determine the MDR-1 promoter activity in parental cells and MDR cells. The results are presented as mean±s.e.m. of triplicate experiments (*Po0.05).

To confirm the binding of HOXC6 to the MDR-1 promoter, we mutated reporters were transfected into MDR cells to analyze performed a chromatin immunoprecipitation assay. The cross- promoter activity. Interestingly, the mutated TAAT motif reduced linked extracts were immunoprecipitated with antibodies against MDR-1 promoter activity relative to the wild-type promoter, HOXC6 or control anti-IgG antibody. The cross-linked DNA was indicating HOXC6 specificity to the sequence containing TAAT analyzed using PCR with primers designed to amplify the HOXC6- (Figures 4e, f and g). responsive region, which covered the TAAT motif ( À 2243 bp) of the MDR-1-2859 bp promoter. Compared with the IgG control group, HOXC6 was determined to be associated with the MDR-1 HOXC6 siRNA reverses drug resistance and induces apoptosis in promoter region containing the TAAT motif (Figure 4d). The MDR-1 FaDu/PTX cells promoter–reporter plasmid containing a mutation of the TAAT motif To investigate whether drug resistance can be reversed by ( À 2243 bp) was constructed. The wild-type (pGL3-MDR-1-3001) and reducing MDR-1 activity/levels in siHOXC6-transfeced cells, we

& 2013 Macmillan Publishers Limited Oncogene (2013) 3339 – 3349 HOXC6 regulates MDR-1 expression K-J Kim et al 3342

Figure 2. HOXC6 regulates the expression of MDR-1 at the transcriptional level. (a) Expression of Notch-1 and Notch-2 in FaDu and FaDu/PTX cells. (b) Hes-1 promoter-luciferase assay by HOXC6-induced Notch-1. (c) RT–PCR of MDR-1 and Hes-1 in intracellular Notch-1-expressing FaDu cells. **Po0.01, compared with control cells. (d) HOXC6 overexpression in the pcDNA3-HOXC6-transfected cells. For this experiment, the cells were transiently transfected with the pcDNA3-HOXC6 plasmid for 48 h. HOXC6 expression was identified using western blot. MDR-1 mRNA levels were analyzed by RT–PCR analysis. Glyceraldehyde 3-phosphate dehydrogenase was used as an internal control. (e) A quantitative evaluation of HOXC6 overexpression was determined in FaDu cells at 48 h after pcDNA3-HOXC6, by flow cytometry after staining with specific antibodies against HOXC6. The bars indicate the populations counted as positive for expressed HOXC6. (f–h) Luciferase reporter assay to determine the regulatory effect of HOXC6 on MDR-1 promoter activity. The luciferase reporter assay was used to test the activity of the MDR-1 promoter in cells transiently transfected with HOXC6. After HOXC6 and MDR-1 promoter co-transfection, luciferase activity was measured in FaDu, MCF-7 and SNU601 cells. **Po0.01. (i) An MTT assay was used to test HOXC6 effects in cells treated with paclitaxel. Forty-eight hours after HOXC6 transfection, the FaDu cells were treated with the indicated concentrations of paclitaxel (20 nM) for 24 h. Cell growth was measured using an MTT assay. *Po0.05.

treated the MDR cells with either a non-toxic dose of chemo- To support these results, we investigated the effect of siHOXC6 therapeutic drugs or HOXC6 siRNA to knock down HOXC6 and PTX on apoptosis induction in FaDu/PTX cells using flow expression levels. We then tested cell viability for changes in cytometry analysis. We determined that apoptosis was slightly drug sensitivity. Western blotting indicated that HOXC6 expres- induced by siHOXC6 treatment in FaDu/PTX cells. Exposure of the sion decreased in siHOXC6-transfected FaDu/PTX cells. The cells to siHOXC6 with paclitaxel (400 nM) caused increased viability of the FaDu/PTX cells was slightly decreased by paclitaxel apoptosis (34%) compared with siHOXC6-treated cells (11%; compared with untreated cells. Similarly, HOXC6 siRNA slightly Figure 5b), supporting the results of the MTT assay. In addition, inhibited cell viability in the FaDu/PTX cells compared with the similar apoptosis was observed in both the MCF-7/ADR and scramble siRNA -transfected cells. These results were similar in SUN601/CIS cell lines (data not shown). both MCF-7/ADR and SNU601/CIS cell lines (Figure 5a). Interest- To determine caspase activation during apoptosis in FaDu/PTX- ingly, treatment with siHOXC6 combined with paclitaxel siHOXC6 cells, we investigated caspase-3 activity after the cells significantly inhibited cell proliferation compared with siHOXC6 were treated with paclitaxel and/or siHOXC6. The colorimetric or paclitaxel alone. These results strongly support the hypothesis assay revealed that caspase-3 activity was remarkably increased in that HOXC6 exerts its effect on MDR-1-mediated drug the FaDu/PTX-siHOXC6 cells. Indeed, the caspase-3 activity in responsiveness. FaDu/PTX-siHOXC6 cells treated with paclitaxel was significantly

Oncogene (2013) 3339 – 3349 & 2013 Macmillan Publishers Limited HOXC6 regulates MDR-1 expression K-J Kim et al 3343

Figure 3. Effect of HOXC6 siRNA on MDR-1 mRNA expression and Rho123 accumulation. (a) The cells were transfected with either HOXC6 siRNA (20 nM) or control siRNA. After 48 h, the cells were collected, and total cellular proteins were used for western blotting analysis with anti-HOXC6 antibody as described. Actin served as an internal control. The MDR-1 mRNA levels were analyzed by RT–PCR. (b) FaDu/PTX and MCF-7/ADR cells were transfected with scramble siRNA or siHOXC6 for 48 h. The MDR-1 mRNA levels in these cells were assessed using real- time RT–PCR analysis. The expression levels were normalized to glyceraldehyde 3-phosphate dehydrogenase. (c–e) FaDu/PTX, MCF-7/ADR and SNU601/CIS cells were transfected for 48 h with either HOXC6 siRNA or control scramble siRNA. After 48 h, the transcriptional activity of MDR-1 was analyzed using the luciferase reporter assay. The error bars in c, d and e denote s.e.m. (n ¼ 3). **Po0.01. (f–h) A representative drug uptake is shown depicting the levels of Rho123 accumulation in FaDu/PTX, MCF-7/ADR and SNU601/CIS cells treated with HOXC6 siRNA or with scramble siRNA for 48 h. The results of three accumulation experiments (n ¼ 3) are summarized. The verapamil was used as the positive control. Values are presented as the means±s.e.m. (n ¼ 3). *Po0.05. higher than in the FaDu/PTX-siHOXC6 cells (Figure 5c). Poly (ADP- DISCUSSION ribose) polymerase cleavage was examined by western blotting. HOX transcription factors are essential for embryonic develop- siHOXC6-transfected MDR cells (FaDu/PTX, MCF-7/ADR, and ment and have critical roles in apoptosis, differentiation and SUN601/CIS) treated with paclitaxel, doxorubicin or cisplatin had proliferation.1,2,12 HOXC6 is crucial to the development and strongly increased levels of cleaved poly (ADP-ribose) polymerase, proliferation of epithelial cells in response to hormonal signals. suggesting that these cells undergo apoptosis (Figures 5d, e and f). Deregulation of HOXC6 has been detected in many cancer types Our results suggest that the downregulation of HOXC6 in MDR such as prostate, breast and lung cancers.3,7,10,22,24–26 However, cells improves their sensitivity to chemotherapeutic drugs by the mechanism by which HOXC6 regulates target genes is not well initiating the apoptosis pathway. understood.

& 2013 Macmillan Publishers Limited Oncogene (2013) 3339 – 3349 HOXC6 regulates MDR-1 expression K-J Kim et al 3344

Figure 4. Identification of the HOXC6-binding site in the MDR-1 promoter. (a) Deletion analysis of the MDR-1 promoter–reporter constructs. pGL3-MDR-1-3001 bp, pGL3-MDR-1-2859 bp, pGL3-MDR-1-1966 bp, and pGL3-MDR-1-1391 bp promoter–reporter plasmids were transfected into FaDu/PTX cells. Forty-eight hours post transfection, the cell extracts were prepared and analyzed for luciferase activity. *Po0.05, compared with control cells (pGL3-MDR-1-3001 bp) (b) MDR-1-2859 bp promoter sequence containing five putative HOXC6-binding sites (TAAT): I ( À 2533), II ( À 2424), III ( À 2243), IV ( À 2108), and V ( À 2042). The þ 1 represents the transcriptional start site. The deleted MDR-1 promoter was amplified by PCR. The primers (P1BP5) are underlined. (c) MDR-1 promoter-luciferase assay with different sequence lengths of the MDR-1 promoter-luciferase constructs covering À 2859 to À 1966 bp (relative to the transcriptional start site) in FaDu/PTX cells. pGL3- MDR-1-P1, pGL3-MDR-1-P2, pGL3-MDR-1-P3 and pGL3-MDR-1-P4 are individual deletion mutants of the MDR-1-2859 bp construct. These mutants were transfected as described in panel A. The luciferase activity was normalized. *Po0.05, **Po0.01 compared with control cells (pGL3-MDR-1-3001 bp) (d) Chromatin immunoprecipitation-PCR analysis. Chromatin was prepared and immunoprecipitated with specific antibodies against HOXC6 or IgG. The input DNA and DNA isolated from the precipitated chromatin were amplified by PCR using Taq polymerase and separated on a 1.5% agarose gel. Lanes: M, marker; 1, input; 2, HOXC6 antibody; 3, IgG (negative control). (e–g) pGL3-MDR-1 promoter assay with TAAT motif ( À 2243 bp) point mutation constructs in FaDu/PTX, MCF-7/ADR and SNU601/CIS cells. FaDu/PTX, MCF-7/ADR and SNU601/CIS cells were transfected with wild-type pGL3-MDR-1 promoter or point mutated pGL3-MDR-1 (as indicated). pRL(Renilla luciferase) plasmid was cotransfected as an internal control. The cells were harvested 48 h after transfection. The promoter activity of each preparation was normalized to the Renilla value. The relative promoter activity is averaged from at least three independent experiments. The data are represented as the mean±s.e.m. *Po0.05, **Po0.01, compared with control cells (pGL3-MDR-1-WT). A full colour version of this figure is available at the Oncogene journal online.

The resistance of chemotherapeutic drugs in cancer seems to carcinogenesis. Our results suggested that the one or more of be much more complicated, as both losses and gains in many these HOX genes can be related to MDR in cancer cells. In HOX gene expression is related to tissue development or Supplementary data Figure 1, the fourteen HOX genes (HOXA4,

Oncogene (2013) 3339 – 3349 & 2013 Macmillan Publishers Limited HOXC6 regulates MDR-1 expression K-J Kim et al 3345

Figure 5. Downregulation of HOXC6 expression sensitizes the FaDu/PTX, mcf-7/ADR and SNU601/CIS cells to chemotherapeutic drugs. (a) Cell viability was assessed by an MTT assay in the presence of paclitaxel (PTX, 400 nM), doxorubicin (ADR, 5 mM), and cisplatin (CIS, 1 mg/ml) in MDR cells transfected with HOXC6 siRNA. The cells were transfected for 48 h with either HOXC6 siRNA (siHOXC6-1 and siHOXC6-2) or control scramble siRNA. After 48 h, cell proliferation was analyzed using an MTT assay. (b) Forty-eight hours after siHOXC6 transfection, the FaDu/PTX cells were treated with paclitaxel (400 nM) for 24 h and then stained using Annexin V-FITC/PI. The apoptotic cells were then analyzed by flow cytometry. (c) The caspase-3 activity of cells treated with the combination of siHOXC6 and paclitaxel was measured. FaDu/PTX cells were transfected with siHOXC6 before the addition of paclitaxel (400 nM). Untreated cells were used as the control. The caspase-3 activity was measured 24 h after treatment. (d–f) Western blots were probed for the caspase-cleaved form of poly (ADP-ribose) polymerase as a marker of apoptosis in scramble siRNA and HOXC6 siRNA transfected FaDu/PTX, MCF-7/ADR and SNU601/CIS cells. The cells were transfected for 48 h with either HOXC6 siRNA (siHOXC6-1 and siHOXC6-2) or control scramble siRNA. After 48 h, the poly (ADP-ribose) polymerase active form was analyzed using western blotting. The blot was reprobed using monoclonal antibodies specific to actin to monitor for equal protein loading.

HOXA6, HOXA10, HOXA11, HOXB6, HOXB13, HOXC9, HOXC10, precise, causal relationship between individual HOX genes and the HOXC11, HOXC13, HOXD8, HOXD9, HOXD10, and HOXD11) were specific cancer phenotypes or the chemotherapeutic drugs to downregulated in FaDu/PTX cells compared with parental cells. which they contribute, or to understand how transcriptional and One possibility is that their HOX genes are hypermethylated by biological specificity can be obtained. Further studies of the post-modification in MDR cancer cells. However, the current state specific expression patterns of HOX genes in MDR cancer cells are of our knowledge about HOXC6 is insufficient to establish a needed to fill these gaps in knowledge. We also showed that two

& 2013 Macmillan Publishers Limited Oncogene (2013) 3339 – 3349 HOXC6 regulates MDR-1 expression K-J Kim et al 3346 HOX genes (HOXA5 and HOXC6) were importantly expressed in approaches clearly demonstrated that endogenous HOXC6 was FaDu/PTX compared with parental cell; especially the expression associated with the TAAT motif in the native MDR-1 promoter in level of HOXC6 was higher in FaDu/PTX cells than parental cell. FaDu/PTX, MCF-7/ and SNU601/CIS cells. These studies support In the present study, we confirmed that the expression of the established HOXC6 mechanism that involves direct binding of HOXC6 was upregulated in our MDR cell lines, including FaDu/PTX, HOXC6 to specific promoter elements to activate MDR-1 expres- MCF-7/ADR and SNU601/CIS cells. Moreover, using gene transfec- sion in MDR cells. Moreover, siRNA directed against HOXC6 tion and RNA interference techniques, we demonstrated that the resulted in decreased MDR-1 mRNA levels and reduced luciferase in vitro drug sensitivity to paclitaxel in FaDu/PTX cells and reporter activity in MDR cells. Therefore, we believe that HOXC6 chemotherapeutic drug-induced apoptosis were increased in cells induces MDR-1 expression in MDR cancer cells by directly transfected with siHOXC6. This strongly indicates that HOXC6 is increasing its promoter activity. involved in the regulation of MDR in MDR cancer cells. HOXC6 could be acting both to prevent apoptosis and to Chemotherapeutic drug resistance is a major clinical problem promote cell cycle progression. The silencing of HOXC6 expression and a cause of cancer treatment failure in many human cancers. using siRNA resulted in decreased cell proliferation and the Recently, various mechanisms, including the upregulation of induction of apoptosis in prostate cancer cells.12 Our results drug efflux ABC transporters, such as P-glycoprotein (ABCB1), demonstrate that HOXC6 siRNA can effectively downregulate MDR protein 1 (ABCC1) and ABCG2 have been identified and HOXC6 overexpression in FaDu/PTX, MCF-7/ADR and SNU601/CIS characterized as transporting and conferring resistance to virtually cells. However, the siRNA suppression of HOXC6 was not strongly the entire spectrum of cancer drugs. Therefore, the upregulation correlated with the induction of apoptosis in these cells. This of these genes can cause MDR in cancers.18–20 Unfortunately, differential response may be due to differences in the many target there is currently no solution to MDR. molecules and signaling pathways in various cancer cell lines. In our previous study, we attempted to elucidate the mechan- Therefore, the roles of HOXC6 differ across a variety of MDR cell ism of MDR by creating a FaDu/PTX by exposing wild-type FaDu lines. Our findings suggest that the silencing of HOXC6 sensitizes cells to increasing concentrations of paclitaxel for more than MDR cells to chemotherapeutic drugs by downregulating 10 months. We identified that the IC50 values for paclitaxel in MDR-1 expression. In FaDu/PTX, MCF-7/ADR and SNU601/CIS FaDu/WT and FaDu/PTX cells were 20 nM±3.5 and 60 mM±7.2, cells, treating cells with siHOXC6 combined with paclitaxel, respectively. The FaDu/PTX cells also had cross-resistance to DOX doxorubicin and cisplatin significantly induced apoptosis (fivefold resistance; unpublished data). In addition, the mRNA compared with siHOXC6 or PTX alone. These results strongly expression of HOXC6 and MDR-1 was increased in FaDu/PTX cells. support the hypothesis that HOXC6 exerts its effect on MDR-1- It has been shown that MDR-1 (Pgp) is the most important drug mediated drug responsiveness. In future studies, we will test the transporter in mammalian cells.22 In this study, we demonstrated unique expression and function of HOX gene confirmed in the that the expression of MDR-1 was upregulated by HOXC6, present study in tissues from different chemotherapeutic drug- suggesting that the upregulation of MDR-1 is likely responsible resistant patients. We believe will provide considerable evidence for the decrease in both drug accumulation and retention. to our understanding of the relationship of deregulated HOX gene Therefore, HOXC6 may be one of the mechanisms that confer family and MDR in cancer and the functions of HOXC6 genes in resistance to many cancer chemotherapies. general and specific phenotypes of MDR cells. In addition, we observed increased expression of Notch-1 and In conclusion, this study provided multiple lines of evidence Notch-2 in FaDu/PTX cells. A recent study has suggested that indicating that HOXC6 increases resistance to chemotherapeutic Notch is involved in the transcriptional regulation of ABCC1/MRP-1 drugs by inducing MDR-1, suggesting that HOXC6 could serve as a in MCF-7/VP cells and likely contributes to the ABCC1-mediated therapeutic target for MDR cancer. Further identification of other resistance to etoposide. However, this group did not investigate HOXC6 regulatory targets will provide us with a better under- the connection between Notch-1 and either ABCB1/MDR-1 or standing of how HOXC6 signaling pathways contribute to altering ABCG2/BCRP.23 Although our results suggest that there is the MDR/sensitivity and survival of cancer cells. increased activation/expression of Notch-1 and Notch-2 in FaDu/ PTX cells overexpressing HOXC6, Notch signaling is not directly involved in MDR-1 expression. In our study, the exact mechanism MATERIALS AND METHODS of increased Notch expression in MDR cells is not known. Cell cultures However, MDR cells are continuously exposed to sublethal The cell culture media (Dulbecco’s modified Eagle’s medium, minimum concentrations of an anticancer drug, and consequently, MDR essential medium (MEM), and RPMI-1640) were supplemented with 10% cells experience various epigenetic and genetic changes. FBS, 100 units/ml penicillin, and 100 mg/ml streptomycin. The cells were Accordingly, the increased level of Notch expression in FaDu/ cultured in a humidified incubator in an atmosphere of 95% air and 5% PTX cells may be a consequence of one of these changes. CO2. The MCF-7 wild-type (MCF-7/WT) and adriamycin-resistant MCF-7 cell We show, for the first time, that HOXC6 directly targets MDR-1 lines (MCF-7/ADR) were cultured in Dulbecco’s modified Eagle’s medium expression. HOXC6 is a HOX family member that presumably (GIBCO, Grand Island, NY, USA). The FaDu human oral cancer cell line and functions by binding directly to DNA promoter elements via its paclitaxel-resistant FaDu cells (FaDu/PTX) were maintained in minimum essential medium (GIBCO). The human gastric cancer cells, SNU601, and homeodomain, which is a DNA-binding domain located within its 27 the cisplatin-resistant SNU601 cells (SNU601/CIS) were maintained in RPMI- N-terminus. The DNA core sequence of the homeodomain- 1640 medium (GIBCO). binding sites frequently contains the sequence TAAT. In this study, the results of the promoter deletion mutants confirm that HOXC6 directly binds to the specific TAAT-a sequence motif in the MDR-1 RNA purification and RT–PCR gene promoter. Previous studies also showed that HOXC6 bound Total RNA was isolated from both parental and drug-resistant cell lines to similar elements on the neural cell adhesion molecule using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). To avoid genomic promoter.28 This promoter also contains the TAAT-a motif. Our DNA contamination, the extracted RNA was purified using the RNeasy kit studies define HOXC6-binding elements as being critical for (Invitrogen). The quantity and quality of the RNA were determined by measuring the optical density at 260 and 280 nm. Two micrograms of RNA mediating the transcriptional activation of the MDR-1 promoter by were used for cDNA synthesis using oligo-(dt)15 primer and Moloney- HOXC6. The TAAT element deletion mutants (pGL3-MDR-1-P3 and Murine Leukemia Virus (M-MLV) reverse transcriptase. The RT reaction pGL3-MDR-1-P4) reduced HOXC6-directed activation of the MDR-1 began with 10 min incubation at room temperature followed by 42 1C for promoter by B60%, and mutating this element nearly abolished 60 min, and 10 min at 70 1C to terminate the reaction. Subsequently, 2 ml the response (Figure 4). Chromatin immunoprecipitation aliquot of cDNA was PCR-amplified in a total volume of 25 ml containing

Oncogene (2013) 3339 – 3349 & 2013 Macmillan Publishers Limited HOXC6 regulates MDR-1 expression K-J Kim et al 3347 2.5 mlof10Â PCR buffer (0.2 M Tris–HCl (pH 8.4), and 0.5 M KCl), 0.2 mM and reverse 50-GCCAAGACCTCTTCAGCTACTGC-30. The glyceraldehyde deoxyribonucleotide triphosphate each primer, and 1.25 units of Platinum 3-phosphate dehydrogenase real-time PCR primers were forward Taq DNA polymerase (Invitrogen). The thermal cycler profile was 95 1C for 50-AGCCAAAAGGGTCATCATCTCTGC-30 and reverse 50-GCATTGCTGATGA 5 min, followed by 30 cycles of 94 1C for 30 s, 52 1C for 30 s, and 72 1C for TCTTGAGGCTG-30. The cycling conditions were as follows: denaturation at 1 min, with a final extension step at 72 1C for 10 min. The primers specific 95 1C for 10 min, followed by 40 cycles of 95 1C for 20 s, 58 1C for 20 s, and for MDR-1 and HOXC6 are provided in Table 1. The PCR products were 70 1C for 20 s. analyzed using gel electrophoresis with 1.2% agarose gel. MTT assay 5 Quantitative real-time PCR Briefly, the cells were seeded at 1 Â 10 cells per well into 12-well plates. Before testing, the MTT solution (5 mg/ml in PBS) was added, and the cells Quantitative real-time PCR was performed using SYBR Green (Applied were incubated at 37 1C for 3 h (h). The culture medium was aspirated and Biosystems, Foster City, CA, USA). PCR runs and fluorescence an acid-isopropanol mixture (0.04 mol/l HCl in isopropanol) was added to detection were performed in a Rotor-Gene 6000 Real-Time PCR system dissolve the dark blue crystals. The optical density value of the dissolved (Corbett Research, Sydney, NSW, Australia). On the basis of HOXC6, solute was measured using a Microplate Autoreader at a wavelength of MDR-1 and glyceraldehyde 3-phosphate dehydrogenase cDNA sequences 570 nm. information (GenBank accession numbers: NM_004503.3, NM_000927.4 and NM_002046.3, respectively), two pairs of gene-specific primers were designed. The reaction mixture contained 10 ng of cDNA diluted Plasmid constructs and subcloning in 2.5 ml of DEPC-treated water, 5 ml Power SYBR Green PCR Master Subcloning was performed as previously described, with some modifica- Mix (2 Â ; Applied Biosystems), and 2 ml of gene-specific primers (final tions.29 The progressive deletion fragments of the MDR-1 promoter region concentration 50 nM each), in a final reaction volume of 10 ml. The were generated using PCR with human genomic DNA as the template and sequences of the real-time PCR primers were as follows: HOXC6: forward were subcloned into the KpnI and NheI sites of the pGL3 vector. Both the 50-CACCGCCTATGATCCAGTGAGGCA-30 and reverse 50-GCTGGAACTGAAC full-length and deleted MDR-1 promoter sequences were PCR-amplified ACGACATTCTC-30, MDR-1: forward 50-GACTGTCAGCTGCTGTCTGGGCAA-30 using the primers listed below. Forward primers: pGL3-MDR-1-3001 bp,

Table 1. Specific primers for HOX and MDR genes

Gene Sense Antisense Annealing PCR Size Location temperature cycle (bp) (1C) number

HOXA1 50-CACTCATATGGACAGGAGC-30 50-TGCGCTGGAGAAGATGTCTCCG-30 62 30 125 Exon I, II HOXA2 50-AACACACAGCTTCTAGAGCTGG-30 50-TCCTCCTCTACTTTCTCGGAG-30 62 30 224 Exon I, II HOXA3 50-GAAAGAGTTCCACTTCAACCGC-30 50-GTGAGCTTGGGTGCTTCCTG-30 62 30 707 Exon I, II HOXA4 50-CTCCAACTACATCGAGCCC-30 50-CATGGATCTTCTTCATCCAGG-30 60 30 580 Exon I, II HOXA5 50-ACTCATTTTGCGGTCGCTATCC-30 50-CGACGCTGAGATCCATGCC-30 60 30 149 Exon I, II HOXA6 50-AGCACTCCATGACGAAGGC-30 50-CGGCGGTTCTGGAACCAG-30 60 30 257 Exon I, II HOXA7 50-ACTTCTTGCTCCTTTGCTCC-30 50-AGGTCCTGAAGACCGCATCC-30 62 30 305 Exon I, II HOXA10 50-TCCCTGGGCAATTCCAAAGGTG-30 50-AAGTTGGCTGTGAGCTCC-30 62 30 268 Exon I, II HOXA11 50-CTTACTACGTCTCGGGTCCAG-30 50-CTGTCCGAACTTGAAGTTGC-30 62 30 525 Exon I, II HOXB1 50-TGAACTCCTTCTTAGAGTACCC-30 50-CCCTCGCTTGCATAGCTGTC-30 66 30 110 Exon I, II HOXB2 50-ACGCAGCTGCTGGAACTGGAG-30 50-TGCGCCCTCTAAGCGAACGGC-30 66 30 354 Exon I, II HOXB3 50-CCCCTGGATGAAAGAGTCGAGG-30 50-TGGAACTCCTTCTCCAGCTCC-30 62 30 213 Exon I, II HOXB4 50-GTCGACCCCAAGTTCCCTCCATG-30 50-CGTGCTCACGTGAACTTTGCGC-30 62 30 307 Exon I, II HOXB6 50-GTTACCAGACGCTGGAGCTGG-30 50-AGCTGAGACGCGCTGAGCAG-30 60 30 174 Exon I, II HOXB7 50-AATATCCAGCCTCAAGTTCGG-30 50-TTGAAGGAACTCGGCTCGAG-30 60 30 219 Exon I, II HOXB8 50-TTTCGTCAACTCACTGTTCTCC-30 50-CTTGCGGGCGCATCCAGGGG-30 62 30 413 Exon I, II HOXB13 50-ATGCCACCTTGGATGGAGC-30 50-CTCCAAAGTAACCATAAGGCAC-30 66 30 310 Exon I, II HOXC5 50-GAACCAGTTACACGCGCTACC-30 50-TCTTCCACTTCATCCTGCGG-30 60 30 158 Exon I, II HOXC6 50-CAGATCTACTCGCGGTACCAG-30 50-TCCTCTTCTGTCTCTTCCCGC-30 60 30 256 Exon I, II HOXC8 50-AGCATGAGCTCCTACTTCGTC-30 50-ATGGAAACATGAGGCTGGGAG-30 60 30 423 Exon I, II HOXC9 50-TGCCCCTACACCAAGTACCAGAC-30 50-CTGGGTAGGGTTTAGGACTGCTC-30 56 30 326 Exon I, II HOXC10 50-GCCTGGAACAACCTGTTGG-30 50-GTCTTGCTAATCTCCAGGC-30 60 30 305 Exon I, II HOXC11 50-TCCAACCTCTATCTGCCCAG-30 50-ATGAGGATCTCGGTGACGGTGG-30 64 30 256 Exon I, II HOXC12 50-ATGGGCGAGCATAATCTCCTG-30 50-TCGGATTCCAGCGACTGGC-30 58 30 185 Exon I, II HOXC13 50-TCGCACAACGTGAACCTGCAGC-30 50-ACCACCGACACGTCCAGGTAGC-30 64 30 250 Exon I, II HOXD1 50-CTACCTGGAGTACGTGTCATGC-30 50-GCGGCACCAGGTTCGTAGGC-30 60 30 278 Exon I, II HOXD3 50-AACTCAGAGCAGCAGCCACCAC-30 50-TTGCTGATGGTGGCTGAGGAGC-30 59 30 143 Exon I, II HOXD4 50-ACTCCAAGTATGTGGACCCC-30 50-CCTCCGAAAGGCTGCTCACC-30 60 30 165 Exon I, II HOXD8 50-CTTCGTGAACCCGCTGTACTCC-30 50-ATCCACGGAAACATTTGAGAAG-30 59 30 217 Exon I, II HOXD9 50-GCCCTCAGCTTGCAGCGACCA-30 50-GCATTTCTCCTTGCTCATCTT-30 59 30 220 Exon I, II HOXD10 50-TACACCAACCAGCAATTGGC-30 50-CTCGCATCCGCTTCTCTCGG-30 62 30 233 Exon I, II HOXD11 50-GACTTTCCCCTACTCTTCCAACC-30 50-GGGAGAAGCTCGCGTTGCATGG-30 60 30 290 Exon I, II HOXD12 50-GTGAGCGCAGTCTCTACAGAGC-30 50-TTAGCCTGCTCTTCGGGTCCG-30 62 30 150 Exon I, II

Gene Sense Antisense Size (bp)

MDR-1 50-CGTAATGCTGACGTCATCGCTGGTTT-30 50-CCAAGGGCTAGAAACAATAGTGAAAAC-30 538 MRP-1 50-CGTCAGTGGCATGAGGATCAAGACC-30 50-AATGTGCACAAGGCCACCAGAAAGG-30 553 ABCG2 50-CGACAGCTTCCAATGACCTGAAGG-30 50-CCACGGATAAACTGAGTTCCAACC-30 514 GAPDH 50-CCAAGGTCATCCATGACAACT-30 50-GTCATACCAGGAAATGAGCTTG-30 485

Abbreviations: ABCG2, breast-cancer resistance protein; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; MDR, multidrug resistance; MRP1, MDRprotein1.

& 2013 Macmillan Publishers Limited Oncogene (2013) 3339 – 3349 HOXC6 regulates MDR-1 expression K-J Kim et al 3348 50-GAAGTGCTTGGCAGTTCCTGATAAATGA-30; pGL3-MDR-1-2859 bp, 50-AAG Molecular Biochemicals, Indianapolis, IN, USA) according to the manufac- TTGGTGAAATGGCTACATG-30; pGL3-MDR-1-1966 bp, 50-GAAACCCAGTCTCT turer’s instructions. The cells were harvested 24 h after transfection. Total ACGAAAAA-30; pGL3-MDR-1-1391 bp, 50-GTTGGCAGTAAATATGGAAGGA-30. cell lysates were separated by SDS–PAGE and analyzed by western blot, as Reverse primer 50-AGCTCATTAGCCAAATGCATGAGCCTC-30. described above. The MDR-1-deleted mutants were created using MDR-1-2859 bp as a template with the following primers. Forward primers: pGL3-MDR-1-P1, 50-ACCTGATGCTCAAGATTGTAGA-30; pGL3-MDR-1-P2, 50-GAGATCATAGGC Intracellular rhodamine 123 accumulation ACAAATAAGA-30; pGL3-MDR-1-P3, 50-ACAGTGAACAATGCTGTACACT-30; After HOXC6 siRNA transfection, the cells were incubated for 3 h with pGL3-MDR-1-P4, 50-AAGAGGCCGAGCTGTAGCTCACGCCT-30.Reverseprimer culture medium containing 6.5 mmol/l of the fluorescent MDR-1 substrate, 50-AGCTCATTAGCCAAATGCATGAGCCTC-30. All deletion and mutant con- rhodamine 123 (Rho123). Verapamil, a known inhibitor of MDR-1- structs were verified by DNA sequencing. dependent transport activity, was used as a positive control. Subsequently, the adherent cells were washed five times with ice-cold PBS, and the dye was extracted by incubating each dish with 2.5 ml of n-butanol for 10 min MDR-1 promoter-luciferase reporter assay at room temperature. The Rho123 concentrations were measured The pGL3-3067 MDR-1 promoter construct served as a template to fluorometrically. generate the pGL3-2922, pGL3-2032 and pGL3-1457 deletion constructs. We also used the pGL3-2922 MDR-1 promoter construct as the template to generate a deletion construct in which the four putative HOXC6-binding Flow cytometric cell cycle analysis and Annexin V-FITC/ propidium sites were deleted (pGL3-P1, pGL3-P2, pGL3-P3 and pGL3-P4). Luciferase iodide (PI) double staining reporter assays were performed as described previously.30 The cells were For cell cycle analysis, the cells were harvested and fixed with 70% ethanol grown in 12-well plates and cotransfected with MDR-1 promoter-luciferase for 1 h at 4 1C. After washing with cold PBS, the cells were incubated with plasmid DNA and either pCDNA3.1-HOXC6 or HOXC6 siRNA. After 48 h, cell DNase-free RNase and PI at 37 1C for 30 min. The specific binding of lysates were analyzed for luciferase activities. The luciferase reporter assays Annexin V-FITC/PI was performed by incubating the cells for 15 min at were performed using the Dual-Luciferase Reporter Assay System room temperature in a binding buffer (10 mM HEPES (pH 7.4), 140 mM NaCl (Promega, Madison, WI, USA) following the manufacturer’s instructions. and 2.5 mM CaCl2) containing saturating concentrations of Annexin V-FITC Each experiment was performed in triplicate and repeated twice. and PI. After incubation, the cells were pelleted and analyzed using a FACScan analyzer (Beckman Coulter Inc, Fullerton, CA, USA). Quantitative evaluation for HOXC6 expression FaDu cells were transiently transfected with pcDNA3-HOXC6 and allowed Caspase-3 activity assays to grow for an additional 48 h. FaDu cells were harvested by trypsinization The activity of caspase-3 was determined using the ApoAlert Caspase and fixed in 4% paraformaldehyde for 10 min at 25 1C. The cells were Profiling Plate (Clontech, Mountain View, CA, USA) according to the permeabilized with methanol for 2 min and incubated for 2 h with HOXC6 manufacturer’s protocol. The release of fluorochrome aminomethyl antibody (1:100). Goat anti-rabbit IgG-FITC (Vector Laboratories Ins, comarin (AMC) was analyzed at an excitation of 380 nm and an emission Burllingame, CA, USA) was used as secondary antibody. After washing of 460 nm using a multiplate fluorescence spectrophotometer. The data and resuspending, the cells were immediately measured by flow cytometry. are displayed as the mean±s.e. of three independent experiments performed in duplicate and expressed in arbitrary fluorescence units/mg Chromatin immunoprecipitation Assay of protein. The FaDu/PTX cells were grown to nearly 80% confluence and cross-linked with formaldehyde (Sigma, St Louis, MO, USA) at room temperature for Statistical analysis 10 min. The cross-linked chromatin was prepared with a commercial ChIP Statistical analysis was performed with the data obtained from three assay kit (EZ-Magna ChIP; Millipore, Billerica, MA, USA) and immunopre- independent experiments. The data are represented as the mean±s.e.m. cipitated using 4 mg of normal rabbit anti-IgG (Santa Cruz Biotechnology, A P-value o0.05 was considered significant. Santa Cruz, CA, USA) or 4 mg of anti-HOXC6 antibody (Santa Cruz Biotechnology). The HOXC6-binding site was PCR-amplified using the input DNA or DNA isolated from the precipitated chromatin as the CONFLICT OF INTEREST template in combination with primers flanking the putative HOXC6- The authors declare no conflict of interest. binding sites in the MDR-1 promoter. The primer sets used for the amplification of the MDR-1 promoter regions were obtained from BIONEER Co (Daejeon, Korea). The forward primer sequence was 50-GAGATCATAGG CACAAATAAGA-30, and the reverse primer sequence was 50-AGTGTACAG ACKNOWLEDGEMENTS CATTGTTCACTGT-30. This research was supported by National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No. R13-2008-010-00000-0; Western blot analysis No. 2011-0029091). The cell pellet was dissolved in a 1% cell lysate buffer and centrifuged for 10 min at 12 000 r.p.m. The protein concentrations were determined using REFERENCES the BCA Protein Assay Kit (Pierce, Rockgord, IL, USA). The proteins were separated on a 10% SDS–PAGE gel and transferred to a polyvinylidene 1 Abate-Shen C. Deregulated homeobox gene expression in cancer: cause or fluoride (PVDF) membrane followed by western blot analysis. A solution consequence? Nat Rev Cancer 2002; 2: 777–785. of 5% non-fat dried milk in TBS containing 0.1% Tween-20 was used 2 Veraksa A, Del Campo M, McGinnis W. Developmental patterning genes and their to block non-specific binding. The membrane was subsequently incu- conserved functions: from model organisms to humans. Mol Genet Metab 2000; bated with an anti-HOXC6 mouse monoclonal antibody (Santa Cruz 69: 85–100. Biotechnology) or an anti-poly (ADP-ribose) polymerase antibody (Santa 3 Friedmann Y, Daniel CA, Strickland P, Daniel CW. Hox genes in normal and Cruz Biotechnology). After incubation, the blots were extensively washed neoplastic mouse mammary gland. Cancer Res 1994; 54: 5981–5985. in TBS containing 0.1% Tween-20. For detection, the ECL kit (Amersham 4 Castronovo V, Kusaka M, Chariot A, Gielen J, Sobel M. Homeobox genes: potential Life Sciences, Arlington Heights, IL, USA) was used according to the candidates for the transcriptional control of the transformed and invasive phe- manufacturer’s instructions. notype. Biochem Pharmacol 1994; 47: 137–143. 5 Chen H, Sukumar S. HOX genes: emerging stars in cancer. Cancer Biol Ther 2003; 2: 524–525. siRNA interference 6 Bodey B, Bodey Jr B , Siegel SE, Kaiser HE. Immunocytochemical detection of the The HOXC6 siRNA was obtained in the form of Silencer select validated homeobox B3, B4, and C6 gene products in breast carcinomas. Anticancer Res siRNA (Applied Biosystems). The sense sequence of the HOXC6 siRNA was 2000; 20: 3281–3286. 50-CUCGUUCUCGGCUUGUCUA (dTdT)-30, and the antisense sequence was 7 Castronovo V, Kusaka M, Chariot A, Gielen J, Sobel M. Homeobox genes: potential 50-UAGACAAGCCGAGAACGAG (dTdT)-30. The cells were transfected with candidates for the transcriptional control of the transformed and invasive siRNA (20 nM) using X-tremeGENE siRNA Transfection Reagent (Roche phenotype. Biochem Pharmacol 1994; 47: 137–143.

Oncogene (2013) 3339 – 3349 & 2013 Macmillan Publishers Limited HOXC6 regulates MDR-1 expression K-J Kim et al 3349 8 Fujiki K, Duerr EM, Kikuchi H, Ng A, Xavier RJ, Mizukami Y et al. Hoxc6 is over- 20 Szakacs G, Annereau JP, Lababidi S, Shankavaram U, Arciello A, Bussey KJ et al. expressed in gastrointestinal carcinoids and interacts with JunD to regulate tumor Predicting drug sensitivity and resistance: profiling ABC transporter genes in growth. Gastroenterology 2008; 135: 907–916. cancer cells. Cancer Cell 2004; 6: 129–137. 9 Bodey B, Bodey BJR, Siegel SE, Luck JV, Kaiser HE. Homeobox B3, B4, and C6 gene 21 Deeley RG, Cole SP. Substrate recognition and transport by multidrug resistance product expression in osteosarcomas as detected by immunocytochemistry. protein 1 (ABCC1). FEBS Lett 2006; 580: 1103–1111. Anticancer Res 2000; 20: 2717–2721. 22 Litman T, Druley TE, Stein WD, Bates SE. From MDR to MXR: new understanding of 10 Miller GJ, Miller HL, van Bokhoven A, Lambert JR, Werahera PN, Schirripa O et al. multidrug resistance systems, their properties and clinical significance. Cell Mol Aberrant HOXC expression accompanies the malignant phenotype in human Life Sci 2001; 58: 931–959. prostate. Cancer Res 2003; 63: 5879–5888. 23 Cho S, Lu M, He X, Ee PL, Bhat U, Schneider E et al. Notch1 regulates the 11 Garcia-Gasca A, Spyropoulos DD. Differential mammary morphogenesis along the expression of the multidrug resistance gene ABCC1/MRP1 in cultured cancer cells. anteroposterior axis in Hoxc6 gene targeted mice. Dev Dyn 2000; 219: 261–276. Proc Natl Acad Sci USA 2011; 108: 20778–20783. 12 Ramachandran S, Liu P, Young AN, Yin-Goen Q, Lim SD, Laycock N et al. Loss of 24 Ernst T, Hergenhahn M, Kenzelmann M, Cohen CD, Bonrouhi M, Weninger A et al. HOXC6 expression induces apoptosis in prostate cancer cells. Oncogene 2005; 24: Decrease and gain of gene expression are equally discriminatory markers for 188–198. prostate carcinoma: a gene expression analysis on total and microdissected 13 McCabe CD, Spyropoulos DD, Martin D, Moreno CS. Genome-wide analysis of the prostate tissue. Am J Pathol 2002; 160: 2169–2180. homeobox C6 transcriptional network in prostate cancer. Cancer Res 2008; 68: 1988–1996. 25 De Vita G, Barba P, Odartchenko N, Givel JC, Freschi G, Bucciarelli G et al. 14 Grishina IB, Kim SY, Ferrara C, Makarenkova HP, Walden PD. BMP7 inhibits Expression of homeobox-containing genes in primary and metastatic colorectal branching morphogenesis in the prostate gland and interferes with Notch sig- cancer. Eur J Cancer 1993; 29: 887–893. naling. Dev Biol 2005; 288: 334–347. 26 Tiberio C, Barba P, Magli MC, Arvelo F, Le Chevalier T, Poupon MF et al. HOX gene 15 Ricort JM, Binoux M. Insulin-like growth factor-binding protein-3 activates a expression in human small-cell lung cancers xenografted into nude mice. Int J phosphotyrosine phosphatase. Effects on the insulin-like growth factor signaling Cancer 1994; 58: 608–615. pathway. J Biol Chem 2002; 277: 19448–19454. 27 Svingen T, Tonissen KF. Hox transcription factors and their elusive mammalian 16 Van der Geer P, Hunter T, Lindberg RA. Receptor protein-tyrosine kinases and gene targets. Heredity 2006; 97: 88–96. their signal transduction pathways. Annu Rev Cell Biol 1994; 10: 251–337. 28 Jones FS, Holst BD, Minowa O, De Robertis EM, Edelman GM. Binding and tran- 17 Lin Y, Liu G, Zhang Y, Hu YP, Yu K, Lin C et al. Fibroblast growth factor receptor 2 scriptional activation of the promoter for the neural cell adhesion molecule by tyrosine kinase is required for prostatic morphogenesis and the acquisition of strict HoxC6 (Hox-3.3). Proc Natl Acad Sci USA 1993; 90: 6557–6561. androgen dependency for adult tissue homeostasis. Development 2007; 134: 723–734. 29 Kim SA, Yoon JH, Ahn SG. Heat shock factor 4a (HSF4a) represses HSF2 18 Gottesman MM, Fojo T, Bates SE. Multidrug resistance in cancer: role of ATP expression and HSF2-mediated transcriptional activity. J Cell Physiol 2012; 227: dependent transporters. Nat Rev Cancer 2002; 2: 48–58. 1–6. 19 Haimeur A, Conseil G, Deeley RG, Cole SP. The MRP-related and BCRP/ABCG2 30 Guo C, Ding J, Yao L, Sun L, Lin T, Song Y et al. Tumor suppressor gene Runx3 multidrug resistance proteins: Biology, substrate specificity and regulation. Curr sensitizes gastric cancer cells to chemotherapeutic drugs by downregulating Drug Metab 2004; 5: 21–53. Bcl-2, MDR-1 and MRP-1. Int J Cancer 2005; 116: 155–160.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

& 2013 Macmillan Publishers Limited Oncogene (2013) 3339 – 3349