MOLECULAR CARCINOGENESIS
HOXA9 Inhibits Migration of Lung Cancer Cells and its Hypermethylation is Associated With Recurrence in Non-Small Cell Lung Cancer Jung-Ah Hwang,1 Bo Bin Lee,2 Yujin Kim,2 Seung-Hyun Hong,1 Young-Ho Kim,3 Joungho Han,4 Young Mog Shim,5 Chae-Yeong Yoon,1 Yeon-Su Lee,1* and Duk-Hwan Kim2** 1
Cancer Genomics Branch, Research Institute, National Cancer Center, Goyang-si, Korea Department of Molecular Cell Biology, Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Suwon, Korea 3 Department of Internal Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea 4 Department of Pathology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea 5 Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea 2
This study was aimed at understanding the clinicopathological significance of HOXA9 hypermethylation in non-small cell lung cancer (NSCLC). HOXA9 hypermethylation was characterized in six lung cancer cell lines, and its clinicopathological significance was analyzed using methylation-specific PCR in 271 formalin-fixed paraffin-embedded tissues and 27 freshfrozen tumor and matched normal tissues from 298 NSCLC patients, and Ki-67 expression was analyzed using immunohistochemistry. The promoter region of HOXA9 was highly methylated in six lung cancer cell lines, but not in normal bronchial epithelial cells. The loss of expression was restored by treatment of the cells with a demethylating agent, 5-aza-20 -deoxycytidine (5-Aza-dC). Transient transfection of HOXA9 into H23 lung cancer cells resulted in the inhibition of cell migration but not proliferation. Conversely, sequence-specific siRNA-mediated knockdown of HOXA9 enhanced cell migration. The mRNA levels of HOXA9 in 27 fresh-frozen tumor tissues were significantly lower than in matched normal tissues (P < 0.0001; Wilcoxon signed-rank test). HOXA9 hypermethylation was found in 191 (70%) of 271 primary NSCLCs. HOXA9 hypermethylation was not associated with tumor size (P ¼ 0.12) and Ki-67 proliferation index (P ¼ 0.15). However, patients with HOXA9 hypermethylation had poor recurrence-free survival (hazard ratio ¼ 3.98, 95% confidence interval ¼ 1.07–17.09, P ¼ 0.01) in never-smokers, after adjusting for age, sex, tumor size, adjuvant therapy, pathologic stage, and histology. In conclusion, the present study suggests that HOXA9 inhibits migration of lung cancer cells and its hypermethylation is an independent prognostic factor for recurrence-free survival in never-smokers with NSCLC. © 2014 Wiley Periodicals, Inc.
Key words: HOXA9; hypermethylation; lung cancer; recurrence; migration
INTRODUCTION Lung cancer is the most common cause of cancerrelated death in the world. Despite advances in the early detection and treatment of lung cancer in the past two decades, the prognosis for patients with the disease continues to remain dismal, with overall 5-yr survival rate steady at 10% to 15% [1]. The poor prognosis of lung cancer patients results primarily from micrometastasis in lymph nodes, which occurs in over two-thirds of patients at the time of diagnosis, and partially from the high rate of recurrence after surgical resection [2]. Thus, the development of biomarkers for the identification of patients who are at high risk of poor prognosis is clearly imperative. In recent years, a number of epigenetic molecules have been developed for the purpose. The homeobox is a sequence of about 180 nucleotides encoding the homeodomain which can bind DNA in a sequence-specific manner and is known to be present in all eukaryotic species. It is estimated that ß 2014 WILEY PERIODICALS, INC.
Abbreviations: NSCLC, non-small cell lung cancer; 5-Aza-dC, 5-aza20 -deoxycytidine; HDF, human dermal fibroblast; MS-HRM, methylation-sensitive high-resolution melting; MSP, methylation-Specific PCR; RFs, recurrence-free survival. Grant sponsor: National Cancer Center (NCC); Grant numbers: 1110140-1; 1110100-2; Grant sponsor: Korea Healthcare technology R&D Project, Ministry for Health, Welfare & Family Affairs; Grant number: A084947AA202308N100010A; Grant sponsor: National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST); Grant number: 2011-0029138; Grant sponsor: Republic of Korea *Correspondence to: 323 Ilsan-ro Ilsandong-gu, Functional Genomics Branch, Research Institute, National Cancer Center, Goyang-si, Korea 410-769. **Correspondence to: #50 Ilwon-dong, Kangnam-Ku, Samsung Biomedical Research Institute, Rm B155, Seoul, Korea 135-710. Received 7 December 2013; Revised 7 April 2014; Accepted 16 April 2014 DOI 10.1002/mc.22180 Published online in Wiley Online Library (wileyonlinelibrary.com).
2
HWANG ET AL.
the human genome has at least 200 homeobox genes including the HOX family. Homeobox genes encode transcription factors that play essential roles in embryonic development and differentiation of adult cells. HOX genes are known to play an essential role in lung development and are expressed in normal human adult lung. In mammals, the HOX genes are organized into clusters named A, B, C, and D on four separate chromosomes [3]. The HOXA cluster contains 12 genes (11 HOX genes and EVX1) and is located in a 155-kb-long genomic region on chromosome 7p15–7p14.2 [4]. Most of the HOXA promoters contain highly dense CpG islands, and its methylation is integral to the control of HOXA gene expression. Recently, HOXA9 hypermethylation in lung cancer has been reported by several groups [4–9]. For example, HOXA9 showed significant methylation of CpG islands in stage I squamous cell carcinomas [4], and altered methylation of the HOXA9 promoter was found in paraffin-embedded sputum from patients with non-small cell lung cancer (NSCLC) [7]. However, the clinicopathological significance of HOXA9 hypermethylation remains the subject of active investigation. Aberrant methylation of normally unmethylated CpG islands at the promoter region of tumor suppressor genes is associated with transcriptional silencing of the genes, playing a crucial role in malignant transformation and tumor progression. In this study we characterized HOXA9 hypermethylation in vitro to gain better insight into the role of the gene in NSCLC and further investigated the association between HOXA9 hypermethylation in paraffinembedded tissues from 271 NSCLCs and clinicopathological parameters. MATERIALS AND METHODS Cell Culture Six human lung cancer cell lines (H23, H226, H460, H520, H1650, A549), a human bronchial epithelial cell line (HBE135-E6E7), and a human dermal fibroblast (HDF) were obtained from the American Type Culture Collection (Manassas, VA). The lung cancer cells and the HBE135-E6E7 and HDF cells were grown in RPMI 1640 (Cellgro, Mediatech Inc. Herndon, VA) and DMEM (WelGENE, KOREA), respectively, supplemented with 10% heat-inactivated fetal bovine serum (Hyclone, Logan, UT) and 1% antibiotic-antimycotic (Gibco, New York, NY), and were maintained at 378C in an atmosphere of 5% CO2. Tissue Samples Twenty-seven fresh-frozen tumor tissues, as well as 271 paraffin-embedded tumor tissues, were obtained from 298 NSCLC patients who underwent curative resection at the Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Seoul, Korea, between August 1994 and February 2006. Molecular Carcinogenesis
Post-operative follow-up for survival or recurrence was conducted as previously described [10]. Pathological TNM stage was determined according to the guidelines of the American Joint Committee on Cancer [11]. Written informed consent for the use of tissues, as approved by the Institutional Review Board at Samsung Medical Center, was obtained from each patient before surgery. Genomic DNA Extraction and Sodium Bisulfite Modification Hematoxylin and Eosin (H&E) staining of freshfrozen and paraffin-embedded tissues was performed, and areas corresponding to a tumor were microdissected manually under a microscope, and all containing at least 75% neoplastic tissue were used in this study. Genomic DNA from cultured cells and tissues was extracted using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) and DNeasy Tissue kit (Qiagen, Valencia, CA), respectively, according to the manufacturer’s instructions. One microgram of genomic DNA was modified by sodium bisulfite using the EZ DNA Methylation-Gold Kit (ZYMO Research, Orange, CA), according to the manufacturer’s protocol. Quantitative Analysis of Methylation using EpiTYPER Methylation status of individual CpG at the promoter of HOXA9 was quantitatively analyzed in six lung cancer cell lines, HBE135-E6E7 cells, and HDF cells using the EpiTYPERTM assay (Sequenom, San Diego, CA). Primers were designed using EpiDesigner software (Supplementary Table S1). After PCR amplification, amplicons were treated with shrimp alkaline phosphatase, in vitro transcribed, cleaved by RNaseA, and subjected to MALDI-TOF Mass Spectrometry in order to determine the methylation status. The results were analyzed using EpiTYPERTM ver. 1.0 software. Methylation-Sensitive High-Resolution Melting (MS-HRM) Assay Methylation status of HOXA9 promoter was also validated using the methylation-sensitive highresolution melting (MS-HRM) assay. Ten nanogram of bisulfite-treated genomic DNA was amplified for the assay. Primers which cover from upstream 2 kb of the ATG site to the intron 1 region in the HOXA9 gene were designed using SEQUENOM EpiDesigner software (Supplementary Table S2). MS-HRM was performed on LC-480 real-time PCR machine, and methylation status was analyzed by Gene Scanning software (Roche, Basel, Switzerland). Methylation change in response to 5-Aza-20 -deoxycytidine (5Aza-dC; Sigma–Aldrich, St. Louis, MO) was also analyzed using MS-HRM assay. Six lung cancer cell lines were incubated with 10 mM 5-Aza-dC for a total of 72 h. Cells were then harvested and washed in ice-cold PBS and genomic DNA was isolated.
3
HOXA9 AND RECURRENCE IN NSCLC
Quantitative Real-Time PCR The mRNA level of HOXA9 was analyzed using quantitative real-time PCR. Two microgram of total RNA was extracted from cultured cells using RNeasy Mini kit (Qiagen, Germany), and cDNA was synthesized using the SuperScriptTM First-Strand Synthesis System for use with the RT-PCR (Invitrogen, Carlsbad, CA) kit, according to the manufacturer’s instructions. Quantitative real-time PCR was performed with the SYBR Green I PCR Master Mix (QIAGEN, Penzberg, Germany) on a LightCycler1 480 Real-Time PCR System (Roche, Germany). Experiments were performed in triplicate, and the mean of the three experiments was used as a relative quantitation value. The Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was used for control and normalization. PCR primers are described in Supplementary Table S3. Transient Transfection Assay To determine the effect of HOXA9 on cell migration and proliferation in vitro, full-length HOXA9 cDNA was prepared by PCR amplification using EcoRItagged forward (50 -CCCGAATTCTAATTTCCGTGGGTCGGGCCG-30 ) and BamHI-tagged reverse (50 -CCCGGATCCGAAGGCGGAAGGGGGACGGA-30 ) primers. pAcGFP-C1-HOXA9 construct was generated by cloning HOXA9 cDNA in-frame into a pAcGFP1-C1 vector (CLONTECH, Oxford, UK). For the migration assay, H23 cells were seeded in six well plates at a concentration of 1 105 cells and incubated overnight to approximately 60–80% confluence. The cells were then transfected with 1 mg of a pAcGFP1-C1 constitutively expressing HOXA9 or with an empty vector using LipofectaminTM 2000 (Life Technologies, Carlsbad, CA) according to the manufacturer’s instructions. After 48 h of transfection, the expression of HOXA9 was confirmed by immunoblotting according to standard procedures, using primary antibodies directed to GFP (sc-9996; Santa Cruz Biotechnology, Santa Cruz, CA) and a-tubulin (Sigma–Aldrich). The migration was analyzed using 6.5 mm Transwell1 with an 8 mm pore size (Corning Costar, Lowell, MA), as described previously [12]. The absorbance was measured at 564 nm using VERSAmax microplate reader (Molecular Devices, Sunnyvale, CA). For the cell proliferation assay, 1 103 of H23 lung cancer cells were seeded in a 96 well plate after transfection, and cell proliferation activity was measured every 24 h with MTT assay. siRNA-Mediated Gene Silencing HCT116 cells were transiently tranfected with HOXA9-specific small interfering RNA (50 -TGCTGAGAATGAGAGCGGC-30 [13]) or non-targeting control (NC) siRNA (Cat. No. 10272810, Qiagen, Germany) at a final concentration of 25 nM. At 48 h after transfection, cells were harvested and 5 104 cells were subjected to migration assay. HOXA9 expression Molecular Carcinogenesis
was confirmed by quantitative real-time PCR and western blot analysis using HOXA9 primary antibody (sc-17155; Santa Cruz Biotechnology) 48 h after transfection. Immunohistochemistry Immunohistochemical analysis of Ki-67 was determined as described previously [14]. The fraction of Ki67-positive cells (Ki-67 proliferation index) was defined as the percentage of nuclei that stained positively with a monoclonal anti-Ki-67 antibody (clone MIB-1, DAKO, Carpinteria, CA). Methylation-Specific Polymerase Chain Reaction (MSP) The methylation statuses of HOXA9 in the 271 formalin-fixed paraffin-embedded tissues were determined using MSP, with two sets of primers: one for the unmethylated promoter and the other for the methylated promoter, as described previously [15]. The primer sequences are listed in Supplementary Table S4. Statistical Analysis The differences in clinicopathological characteristics and methylation statuses of the HOXA9 were analyzed using the t-test (or Wilcoxon rank sum test) and the Chi-squared test (or Fisher’s exact test) for continuous and categorical variables, respectively. Following univariate analysis, multivariate logistic regression analysis was performed to calculate the odds ratio (OR) and to determine the relationship between HOXA9 hypermethylation and the covariates found to be significant in univariate analysis after adjusting for potential confounding factors. The effect of HOXA9 hypermethylation on overall survival and recurrence-free survival (RFS) was estimated by Kaplan–Meier survival curves, and the significance of differences in RFS between the two groups was evaluated by the log-rank test. Cox proportional hazards regression model was used to estimate the hazard ratios of independent predictor variables, after adjusting potential confounders. RESULTS Promoter Methylation and Expression of HOXA9 In Vitro The locations of CpGs that were analyzed by EpiTYPERTM, MS-HRM, and MSP are indicated in Figure 1A. The methylation statuses of HOXA9 were analyzed using EpiTYPERTM assay in six lung cancer cell lines, a normal bronchial epithelial cell line (HBE135-E6E7), and a human dermal fibroblast (HDF) cell line. The promoter sequence of HOXA9 was obtained from TESS (http://www.cbil.upenn.edu/cgibin/tess/tess), and methylation statuses of 60 CpGs at the promoter region of HOXA9 were analyzed quantitatively using EpiTYPERTM. Most CpGs at the promoter of HOXA9 were found to be hypermethylated in the lung cancer cells (Figure 1B). Methylation
4
HWANG ET AL.
Figure 1. Analysis of HOXA9 methylation in vitro. (A) Gene map shows location of CpGs that were studied by different methods in this study. (B) The methylation status of HOXA9 was evaluated using the EpiTYPERTM assay in six lung cancer cell lines (H23, H226, H460, H520, H1650, and A549), a bronchial epithelial cell line (HBE135-E6E7), and a normal human dermal fibroblast (HDF). Levels of methylation are depicted in color change on s continuous scale from red (0%
methylated) to light yellow (100% methylated). X-axis and Y-axis indicate CpG sites and cell lines, respectively. (C) Methylation status of 15 fragments covering HOXA9 promoter and intron 1 was further analyzed using MS-HRM assay in six lung cancer cell lines (blue lines) and HDF cells (red line). (D) The mRNA levels of HOXA9 were analyzed by real-time PCR in six lung cancer cell lines and HDF cells. Error bars indicate one standard deviation.
status was further validated using MS-HRM assay (Figure 1C). A total of 15 fragments covering the promoter and intron 1 were amplified in six lung cancer cell lines, and the melting curve was analyzed using GeneScanning software (Roche, Swiss). The HDF cell line was used as a control because normal bronchial epithelial cell (HBE135-E6E7 cells) was slightly methylated at the promoter region of HOXA9. MS-HRM assay showed a substantial dissociation of the melting curve between six lung cancer cell lines and normal HDF cell line in most areas. HOXA9 mRNA expression was analyzed using quantitative real-time PCR (Figure 1D). The mRNA levels of six lung cancer cell lines were downregulated compared to HDF cells, suggesting that HOXA9 hypermethylation may be responsible for transcriptional downregulation of HOXA9 gene.
(Figure 2B) after treatment of lung cancer cells with 10 mM 5-Aza-dC for 72 h. Re-expression of HOXA9 in response to 5-Aza-dC was minimal in H226 cells, but other cell lines showed a substantial increase at the mRNA levels of HOXA9. A fragment (HRM-004) at the promoter region of HOXA9 was amplified to study demethylation of HOXA9, and melting curve was analyzed by GeneScanning software (Roche, Swiss). There was a significant dissociation of melting curves in response to 5-Aza-dC in six lung cancer cell lines (Figure 2B), suggesting that demethylation of the HOXA9 promoter may be associated with re-expression of the silenced HOXA9 gene.
5-Aza-dC Induced Demethylation and Re-Expression of Silenced HOXA9 To validate that the downregulation of HOXA9 in lung cancer cells was dependent on hypermethylation of the gene, we analyzed the re-expression and demethylation of silenced gene using quantitative real-time PCR (Figure 2A) and MS-HRM assay Molecular Carcinogenesis
HOXA9 Regulated Cell Migration in Lung Cancer Cells To investigate the function of HOXA9 in tumorigenesis, cell migration and cell proliferation was analyzed in H23 cells induced by transient transfection of GFP-tagged HOXA9. An anti-GFP antibody was used to detect the expression of HOXA9 by western blot analysis (Figure 3A). Cell migration was significantly reduced in H23 cells transfected with pAcGFP-C1-HOXA9 (P ¼ 0.04; Figure 3B). Cell proliferation was also analyzed in the H23 cells after transfection of the GFP-fusion constructs. However,
HOXA9 AND RECURRENCE IN NSCLC
5
Figure 2. Effect of 5-Aza-dC on demethylation and re-expression of silenced HOXA9. (A) Re-expression of silenced HOXA9 was examined by real-time PCR in six lung cancer cell lines after treatment of the cells with 10 mM 5-Aza-dC for 72 h. Error bars indicate one standard deviation. (B) Demethylation of silenced HOXA9 was analyzed using MS-HRM assay in the lung cancer cell lines. Figures show representative examples of demethylation in six cell lines. Red and blue lines indicate the statuses with and without treatment of 5-Aza-dC, respectively.
cell proliferation was not inhibited in H23 cells induced by pAcGFP-C1-HOXA9 (Figure 3C). In contrast, target sequence-specific siRNA-mediated knockdown of HOXA9 in HCT116 cells reduced the expression of HOXA9 at the level of mRNA (Figure 3D) and protein (Figure 3E) and also led to a significant increase in cell migration (P ¼ 0.0003; Figure 3F). Based on these observations, it is likely that HOXA9 may play a role by inhibiting cell migration in tumorigenesis of lung cancer. Clinicopathological Characteristics of HOXA9 Hypermethylation Before analyzing the clinical significance of HOXA9 hypermethylation, we first compared expression levels of HOXA9 between 27 fresh-frozen tumor and matched normal tissues using quantitative real-time PCR. The mRNA levels of HOXA9 were found to be Molecular Carcinogenesis
significantly lower in tumor tissues than in matched normal tissues (P < 0.0001; Wilcoxon signed-rank test; Figure 4A). We next analyzed the methylation status of the gene using methylation-specific PCR (Figure 4B) in the 271 formalin-fixed paraffinembedded tissues. The relationship between clinicopathological characteristics and HOXA9 hypermethylation is described in Table 1. HOXA9 hypermethylation was found in 191 (70%) of 271 patients studied. HOXA9 hypermethylation was not associated with patient age, sex, tumor size, exposure to tobacco smoke, differentiation, pathologic stage, histology, and tumor recurrence. Multivariate logistic regression analysis was performed to find an independent risk factor for HOXA9 hypermethylation after controlling for potential confounding factors, but no association was found between HOXA9 hypermethylation and covariates (Supplementary Table S5). To further
6
HWANG ET AL.
Figure 3. The effect of HOXA9 on cell migration and proliferation. (A) To analyze the effect of HOXA9 on cell migration and proliferation, H23 lung cancer cells were transfected with pAcGFP-C1-HOXA9 or empty vector. After transfection, the expression of HOXA9 was confirmed by western blot using primary antibody directed to GFP (sc-9996; Santa Cruz Biotechnology). (B) H23 cells transfected with pAcGFP-C1-HOXA9 were re-seeded in 6.5 mm Transwell1 with 8 mm pore size (Corning). The cells were induced to migrate into 1% of FBS containing media in 24-well plate. After 48 h each transwell insert was stained by 1% crystal violet and destained with 10% acetic acid. The absorbance was
measured at 564 nm using VERSAmax microplate reader (Molecular devices). (C) H23 cells transfected with pAcGFP-C1-HOXA9 or empty vector were seeded in 96-well plate, and cell proliferation activity was measured every 24 h with MTT assay. Y-axis indicates absorbance measured at 570 nm using VERSAmax microplate reader (Molecular devices), and the data are presented as mean standard error (SE) of eight experiments. (D and E) Knockdown of HOXA9 in HCT116 cells was evaluated by quantitative real-time PCR (D) and immunoblotting (E). (F) HCT116 cells were transfected with non-targeting control (NC) siRNA and HOXA9 siRNA, and cell migration was measured as (B).
understand the effect of HOXA9 hypermethylation on cell growth in vivo, we analyzed Ki-67 proliferation index using immunohistochemistry (Figure 4C) and found that HOXA9 hyermethylation was not associated with Ki-67 proliferation index in 317 patients (P ¼ 0.15; Figure 4D). The association of HOXA9 hypermethylation with the Ki-67 index also was not statistically significant in adenocarcinoma (P ¼ 0.17) and squamous cell carcinoma (P ¼ 0.78).
effect of HOXA9 hypermethylation on RFS was confounded by smoking status due to the interaction between smoking status and HOXA9 hypermethylation (Breslow–Day test of homogeneity, P ¼ 0.002). Therefore, data was stratified according to the smoking status for the analysis of RFS. HOXA9 hypermethylation was significantly associated with the RFS in never-smokers (P ¼ 0.009; Figure 5A) but not in former smokers (P ¼ 0.83; Figure 5B) and current smokers (P ¼ 0.58; Figure 5C). Cox proportional hazards analysis also showed that neversmokers with HOXA9 hypermethylation had 3.98 times (95% CI ¼ 1.06–17.09, P ¼ 0.01) poor hazard ratio of recurrence than those without, after adjusting for age, sex, tumor size, adjuvant therapy, pathologic stage, and histology.
Survival Analysis Finally, the relationship between HOXA9 hypermethylation and patient’s survival was investigated in the 271 NSCLCs with a median follow-up period of 4.9 yr. HOXA9 hypermethylation was not associated with overall survival (data not shown). However, the Molecular Carcinogenesis
HOXA9 AND RECURRENCE IN NSCLC
7
Figure 4. Analysis of HOXA9 hypermethylation in primary NSCLC. (A) The mRNA levels of HOXA9 were compared between 27 fresh-frozen tumor and matched normal tissues using real-time PCR. (B) HOXA9 hypermethylation was analyzed in paraffin-embedded tissues from 271 NSCLC patients using methylation-specific PCR. Patient identification numbers are indicated, and “Pos” represents the positive controls for the methylated (M) and unmethylated (U) alleles. Negative control
without DNA template was included for each PCR. (C) Expression of Ki67 was analyzed using immunohistochemistry. Figures show representative examples of positive expression of Ki-67 in adenocarcinoma (left) and squamous cell carcinoma (right). (D) Ki-67 proliferation index was compared according to methylation status of HOXA9. The “adenoca” and “squamous” represent adenocarcinoma and squamous cell carcinoma, respectively. Error bars indicate one standard deviation.
DISCUSSION
cancer cells in vitro and analyzed the methylation status of HOXA9 in 271 paraffin-embedded tumor tissues. In this study, siRNA-mediated knockdown of HOXA9 resulted in a significant increase in cell migration, and HOXA9 hypermethylation was associated with recurrence after surgery. The prognostic significance of HOXA9 hypermethylation has been reported in a variety of tumors. For example, a scoring of quantified methylation levels of five genes (HOXA6, HOXA9, PENK, UPK3A, and IGF2BP1) is associated with recurrence in meningioma [16], and HOXA9 hypermethylation in urine specimen is associated with bladder cancer recurrence [17]. In addition, HOXA9 hypermethylaton is also associated with high grade serous ovarian carcinoma [18] and with mortality in noninfant patients with neuroblastoma [19]. Recently, Sandoval et al. [9] reported that hypermethylation of five genes (HIST1H4F, PCDHGB6, NPBWR1, ALX1, and HOXA9) was significantly associated with shorter RFS in stage I NSCLC. In contrast to Sandoval’s finding [9], the present study showed no association between HOXA9 hypermethylation and poor RFS in stage I NSCLC, suggesting that additional alteration of other genes in addition to HOXA9 is required to affect the RFS of a patient with stage I NSCLC. The relationship between recurrence and HOXA9 hypermethylation in this study was confounded by smoking status: HOXA9 hypermethylation occurred at a similar frequency in never-smokers, former and current smokers, but the effect of HOXA9 hypermethylation on poor RFS was found only in never-smokers. Studies have suggested that lung cancers occurring in never-smokers may be a distinctive disease due to the unique molecular profiles [20–22]. Definite differences in the epidemiology and the biology underlying the pathogenesis and behavior of lung cancer exist between neversmokers and smokers. However, it is not clear how HOXA9 hypermethylation affects patient survival. In this study, transient transfection of pAcGFP-C1-
To understand the clinicopatholgocial significance of HOXA9 hypermethylation in lung cancer, we characterized HOXA9 hypermethylation in lung Table 1. Clinicopathological Characteristics (N ¼ 271) HOX9 hypermethylation Variables Agea Tumor size (cm)a Pack-years (smoking)a Sex Women Men Smoking status Never Former current Histology Adenocarcinoma Squamous cell carcinoma Others Differentiationb Well Moderately Poorly Undifferentiated Stage I II III IV Recurrence No Yes
No (N ¼ 80)
Yes (N ¼ 191)
P-value
59 10 3.3 1.9 31 27
60 11 3.7 2.1 33 26
0.92 0.12 0.52
25 55
48 143
0.30
20 19 41
54 38 99
0.73
46 30
92 86
4
13
0.37c
16 41 10 1
31 92 31 6
0.75c
51 13 15 1
108 35 43 5
0.77c
52 28
104 87
0.11
Values indicate mean standard deviation. Differentiation data are missing for 53 patients. c Based on Fisher’s exact test. a
b
Molecular Carcinogenesis
8
HWANG ET AL.
Figure 5. Analysis of recurrence-free survival. The effect of HOXA9 hypermethylation on recurrence-free survival was analyzed using logrank test in never-smokers (A), former smokers (B), and current smokers (C). The yellow and blue lines indicate groups with and without hypermethylation of HOXA9, respectively.
HOXA9 into H23 lung cancer cells did not affect cell proliferation, and HOXA9 hypermethylation was not associated with tumor size and the Ki-67 proliferation index in 271 primary NSCLCs. Accordingly, HOXA9 hypermethylation may affect patient survival through other mechanism rather than cell growth. The molecular mechanisms underlying the high rates of recurrence in patients with HOXA9 hypermethylation are not clear, but it may have resulted from the failure to block migration of residual cancer cells after surgery. Cell migration and invasion are crucial steps in tumor metastasis that occurs through specific molecular interactions among tumor cells, extracellular matrix and stroma cells. Further study is required Molecular Carcinogenesis
to clearly understand how HOXA9 hypermethylation contributes to tumor cell migration. HOXA9 is frequently deregulated in a variety of human cancers, in which it acts as a tumor suppressor or as an oncogene. Although the mechanisms underlying these differential functions remain incompletely understood, HOXA9 seems to exert its function by interacting with different kinds of proteins in a tissue-specific manner. For example, the oncogenic potential of HOXA9 has clearly been implicated in leukemia. HOXA9 is known to increase endothelial cell migration and tube formation by stimulating the expression of EphB4 and to play an important role for the nucleoporin gene NUP98 by existing as a chimeric NUP98/HOXA9 transcript in human myeloid leukemia cells [13,23]. In addition, HOXA9 acts as a transcriptional activator of E-selectin that promotes tumor metastasis by enhancing the adhesion of circulating tumor cells to endothelial cells [24]. HOXA9 is also activated by a phosphoinostide 3-kinase (PI3K) pathway in human glioblastoma multiforme and increases cell proliferation and inhibits apoptosis [25]. In contrast to the oncogenic role, several studies suggest a tumor suppressor role of HOXA9 in human cancer. For example, HOXA9 is known to promote the differentiation of breast cancer cells and inhibits the aggression of breast cancer cells by modulating expression of BRCA1 [26]. The function of HOXA9 in lung cancer has also exhibited conflicting results. In this study, transient transfection of HOXA9 to H23 lung cancer cells inhibited tumor cell migration, and the suppression of HOXA9 by siRNA enhanced the migration activity of cancer cells, thereby suggesting that HOXA9 may act as a tumor suppressor in lung cancer by playing a role for regulating the migration activity of cancer cells. However, Plowright et al. [27] recently reported that inhibition of HOX function using the HXR9 peptide, an 18-amino-acid peptide consisting of hexapeptide sequence that can bind to a cofactor PBX and nine C-terminal arginine residues, resulted in apoptosis in vitro and reduced the growth of tumor in vivo. Although HOX proteins have functions that are independent of DNA-binding proteins, which act as co-factors such as PBX, MEIS, and PREP etc., many aspects of the functions of HOX are mostly dependent on interaction with co-factors [28]. The co-factors function by modifying the DNA-binding specificity of HOX proteins in vitro and by influencing the regulation of transcription. Accordingly, complex and diverse interaction between HOX proteins and their DNA-binding partners may affect the functions of HOX proteins in cancer cells. This study had several limiting actors. First, silencing by DNA methylation usually involves methylbinding domain proteins binding to the DNA and recruiting histone deacetylases (HDACs). Therefore, the DNA binding and HDACs recruiting to the plasmid and the nucleosome assembling on a
9
HOXA9 AND RECURRENCE IN NSCLC
transfected plasmid needed to be studied for understanding the mechanism underlying transcriptional silencing by HOXA9 hypermethylation. Second, HOXA9 was associated with tumor cell migration, but its downstream targets are still unknown. Delineation of the target genes regulated by HOXA9 is required. Third, we analyzed methylation status of HOXA9 using MSP in 271 formalin-fixed paraffinembedded tissues. Employing quantitative assay (COBRA, Q-MSP, other methods) instead of MSP may have provided more convincing data regarding the DNA methylation status of the gene. In conclusion, the present study suggests that HOXA9 hypermethylation may be an independent prognostic indicator of poor RFS in never-smokers and that HOXA9 may contribute to NSCLC by being involved in cell migration rather than cell proliferation.
11. 12.
13.
14.
15.
16.
ACKNOWLEDGMENTS The authors wish to thank Eun-Kyung Kim and JinHee Lee for data collection and management, and Hoon Suh and Seo Kyu Park for sample collection. This work was supported by grants from the National Cancer Center (NCC 1110140-1 and 1110100-2), the Korea Healthcare technology R&D Project, Ministry for Health, Welfare&Family Affairs (A084947AA202308N100010A), and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 20110029138), Republic of Korea.
20.
REFERENCES
21.
1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2008. CA Cancer J Clin 2005;58:71–96. 2. Kim JS, Kim JW, Han J, Shim YM, Park J, Kim DH. Cohypermethylation of p16 and FHIT promoters as a prognostic factor of recurrence in surgically resected stage I non-small cell lung cancer. Cancer Res 2006;66:4049–4054. 3. Krumlauf R. Hox genes in vertebrate development. Cell 1994; 78:191–201. 4. Rauch T, Wang Z, Zhang X, et al. Homeobox gene methylation in lung cancer studied by genome-wide analysis with a microarray-based methylated CpG island recovery assay. Proc Natl Acad Sci USA 2007;104:5527–5532. 5. Bibikova M, Lin Z, Zhou L, et al. High-throughput DNA methylation profiling using universal bead arrays. Genome Res 2006;16:383–393. 6. Nelson HH, Marsit CJ, Christensen BC, et al. Key epigenetic changes associated with lung cancer development: Results from dense methylation array profiling. Epigenetics 2012;7: 559–566. 7. Olaussen KA, Soria JC, Park YW, et al. Assessing abnormal gene promoter methylation in paraffin-embedded sputum from patients with NSCLC. Eur J Cancer 2002;41:2112–2119. 8. Selamat SA, Galler JS, Joshi AD, et al. DNA methylation changes in atypical adenomatous hyperplasia, adenocarcinoma in situ, and lung adenocarcinoma. PLoS ONE 2011;6: e21443. 9. Sandoval J, Mendez-Gonzalez J, Nadal E, et al. A prognostic DNA methylation signature for stage I non-small-cell lung cancer. J Clin Oncol 2013;31:4140–4147. 10. Kim JS, Lee H, Kim H, et al. Promoter methylation of retinoic acid receptor beta 2 and the development of second primary
Molecular Carcinogenesis
17.
18.
19.
22. 23.
24. 25.
26. 27.
28.
lung cancers in non-small-cell lung cancer. J Clin Oncol 2004;22:3443–3450. Edge SB, Byrd DR, Compton CC, Fritz AF, Greene FL, Trotti A, editors. Lung AJCC Cancer Staging Manual. 7th edition. New York: Springer; 2010. p. 253–270. Hwang JA, Kim Y, Hong SH, et al. Epigenetic inactivation of heparan sulfate (glucosamine) 3-o-sulfotransferase 2 in lung cancer and its role in tumorigenesis. PLoS ONE 2013;8: e79634. Bruhl T, Urbich C, Aicher D, Acker-Palmer A, Zeiher AM, Dimmeler S. Homeobox A9 transcriptionally regulates the EphB4 receptor to modulate endothelial cell migration and tube formation. Circ Res 2004;94:743–751. Kim JS, Han J, Shim YM, Park J, Kim DH. Aberrant methylation of H-cadherin (CDH13) promoter is associated with tumor progression in primary nonsmall cell lung carcinoma. Cancer 2005;104:1825–1833. Herman JG, Graff JR, Myöhänen S, Nelkin BD, Baylin SB. Methylation-specific PCR: A novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 1996;93:9821– 9826. Kishida Y, Natsume A, Kondo Y, et al. Epigenetic subclassification of meningiomas based on genome-wide DNA methylation analyses. Carcinogenesis 2012;33:436–441. Reinert T, Borre M, Christiansen A, Hermann GG, Ørntoft TF, Dyrskjøt L. Diagnosis of bladder cancer recurrence based on urinary levels of EOMES, HOXA9, POU4F2, TWIST1, VIM, and ZNF154 hypermethylation. PLoS ONE 2012;7:e46297. Montavon C, Gloss BS, Warton K, et al. Prognostic and diagnostic significance of DNA methylation patterns in high grade serous ovarian cancer. Gynecol Oncol 2012;124:582– 588. Alaminos M, Davalos V, Cheung NK, Gerald WL, Esteller M. Clustering of gene hypermethylation associated with clinical risk groups in neuroblastoma. J Natl Cancer Inst 2004;96: 1208–1219. Subramanian J, Govindan R. Molecular genetics of lung cancer in people who have never smoked. Lancet Oncol 2008;9:676–682. Rudin CM, Avila-Tang E, Harris CC, et al. Lung cancer in never smokers: Molecular profiles and therapeutic implications. Clin Cancer Res 2009;15:5646–5661. Yano T, Miura N, Takenaka T, et al. Never-smoking nonsmall cell lung cancer as a separate entity: Clinicopathologic features and survival. Cancer 2008;113:1012–1018. Nakamura T, Largaespada DA, Lee MP, et al. Fusion of the nucleoporin gene NUP98 to HOXA9 by the chromosome translocation t(7;11)(p15;p15) in human myeloid leukaemia. Nat Genet 1996;12:154–158. Bandyopadhyay S, Ashraf MZ, Daher P, Howe PH, DiCorleto PE. HOXA9 participates in the transcriptional activation of Eselectin in endothelial cells. Mol Cell Biol 2007;27:4207–4216. Costa BM, Smith JS, Chen Y, et al. Reversing HOXA9 oncogene activation by PI3K inhibition: Epigenetic mechanism and prognostic significance in human glioblastoma. Cancer Res 2010;70:453–462. Gilbert PM, Mouw JK, Unger MA, et al. HOXA9 regulates BRCA1 expression to modulate human breast tumor phenotype. J Clin Invest 2010;120:1535–1550. Plowright L, Harrington KJ, Pandha HS, Morgan R. HOX transcription factors are potential therapeutic targets in nonsmall-cell lung cancer (targeting HOX genes in lung cancer). Br J Cancer 2009;100:470–475. Moens CB, Selleri L. Hox cofactors in vertebrate development. Dev Biol 2006;291:193–206.
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s web-site.
HOXA9 Inhibits Migration of Lung Cancer Cells and its Hypermethylation is Associated With Recurrence in Non-Small Cell Lung Cancer Jung-Ah Hwang,1 Bo Bin Lee,2 Yujin Kim,2 Seung-Hyun Hong,1 Young-Ho Kim,3 Joungho Han,4 Young Mog Shim,5 Chae-Yeong Yoon,1 Yeon-Su Lee,1* and Duk-Hwan Kim2** 1
Cancer Genomics Branch, Research Institute, National Cancer Center, Goyang-si, Korea Department of Molecular Cell Biology, Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Suwon, Korea 3 Department of Internal Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea 4 Department of Pathology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea 5 Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea 2
This study was aimed at understanding the clinicopathological significance of HOXA9 hypermethylation in non-small cell lung cancer (NSCLC). HOXA9 hypermethylation was characterized in six lung cancer cell lines, and its clinicopathological significance was analyzed using methylation-specific PCR in 271 formalin-fixed paraffin-embedded tissues and 27 freshfrozen tumor and matched normal tissues from 298 NSCLC patients, and Ki-67 expression was analyzed using immunohistochemistry. The promoter region of HOXA9 was highly methylated in six lung cancer cell lines, but not in normal bronchial epithelial cells. The loss of expression was restored by treatment of the cells with a demethylating agent, 5-aza-20 -deoxycytidine (5-Aza-dC). Transient transfection of HOXA9 into H23 lung cancer cells resulted in the inhibition of cell migration but not proliferation. Conversely, sequence-specific siRNA-mediated knockdown of HOXA9 enhanced cell migration. The mRNA levels of HOXA9 in 27 fresh-frozen tumor tissues were significantly lower than in matched normal tissues (P < 0.0001; Wilcoxon signed-rank test). HOXA9 hypermethylation was found in 191 (70%) of 271 primary NSCLCs. HOXA9 hypermethylation was not associated with tumor size (P ¼ 0.12) and Ki-67 proliferation index (P ¼ 0.15). However, patients with HOXA9 hypermethylation had poor recurrence-free survival (hazard ratio ¼ 3.98, 95% confidence interval ¼ 1.07–17.09, P ¼ 0.01) in never-smokers, after adjusting for age, sex, tumor size, adjuvant therapy, pathologic stage, and histology. In conclusion, the present study suggests that HOXA9 inhibits migration of lung cancer cells and its hypermethylation is an independent prognostic factor for recurrence-free survival in never-smokers with NSCLC. © 2014 Wiley Periodicals, Inc.
Key words: HOXA9; hypermethylation; lung cancer; recurrence; migration
INTRODUCTION Lung cancer is the most common cause of cancerrelated death in the world. Despite advances in the early detection and treatment of lung cancer in the past two decades, the prognosis for patients with the disease continues to remain dismal, with overall 5-yr survival rate steady at 10% to 15% [1]. The poor prognosis of lung cancer patients results primarily from micrometastasis in lymph nodes, which occurs in over two-thirds of patients at the time of diagnosis, and partially from the high rate of recurrence after surgical resection [2]. Thus, the development of biomarkers for the identification of patients who are at high risk of poor prognosis is clearly imperative. In recent years, a number of epigenetic molecules have been developed for the purpose. The homeobox is a sequence of about 180 nucleotides encoding the homeodomain which can bind DNA in a sequence-specific manner and is known to be present in all eukaryotic species. It is estimated that ß 2014 WILEY PERIODICALS, INC.
Abbreviations: NSCLC, non-small cell lung cancer; 5-Aza-dC, 5-aza20 -deoxycytidine; HDF, human dermal fibroblast; MS-HRM, methylation-sensitive high-resolution melting; MSP, methylation-Specific PCR; RFs, recurrence-free survival. Grant sponsor: National Cancer Center (NCC); Grant numbers: 1110140-1; 1110100-2; Grant sponsor: Korea Healthcare technology R&D Project, Ministry for Health, Welfare & Family Affairs; Grant number: A084947AA202308N100010A; Grant sponsor: National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST); Grant number: 2011-0029138; Grant sponsor: Republic of Korea *Correspondence to: 323 Ilsan-ro Ilsandong-gu, Functional Genomics Branch, Research Institute, National Cancer Center, Goyang-si, Korea 410-769. **Correspondence to: #50 Ilwon-dong, Kangnam-Ku, Samsung Biomedical Research Institute, Rm B155, Seoul, Korea 135-710. Received 7 December 2013; Revised 7 April 2014; Accepted 16 April 2014 DOI 10.1002/mc.22180 Published online in Wiley Online Library (wileyonlinelibrary.com).
2
HWANG ET AL.
the human genome has at least 200 homeobox genes including the HOX family. Homeobox genes encode transcription factors that play essential roles in embryonic development and differentiation of adult cells. HOX genes are known to play an essential role in lung development and are expressed in normal human adult lung. In mammals, the HOX genes are organized into clusters named A, B, C, and D on four separate chromosomes [3]. The HOXA cluster contains 12 genes (11 HOX genes and EVX1) and is located in a 155-kb-long genomic region on chromosome 7p15–7p14.2 [4]. Most of the HOXA promoters contain highly dense CpG islands, and its methylation is integral to the control of HOXA gene expression. Recently, HOXA9 hypermethylation in lung cancer has been reported by several groups [4–9]. For example, HOXA9 showed significant methylation of CpG islands in stage I squamous cell carcinomas [4], and altered methylation of the HOXA9 promoter was found in paraffin-embedded sputum from patients with non-small cell lung cancer (NSCLC) [7]. However, the clinicopathological significance of HOXA9 hypermethylation remains the subject of active investigation. Aberrant methylation of normally unmethylated CpG islands at the promoter region of tumor suppressor genes is associated with transcriptional silencing of the genes, playing a crucial role in malignant transformation and tumor progression. In this study we characterized HOXA9 hypermethylation in vitro to gain better insight into the role of the gene in NSCLC and further investigated the association between HOXA9 hypermethylation in paraffinembedded tissues from 271 NSCLCs and clinicopathological parameters. MATERIALS AND METHODS Cell Culture Six human lung cancer cell lines (H23, H226, H460, H520, H1650, A549), a human bronchial epithelial cell line (HBE135-E6E7), and a human dermal fibroblast (HDF) were obtained from the American Type Culture Collection (Manassas, VA). The lung cancer cells and the HBE135-E6E7 and HDF cells were grown in RPMI 1640 (Cellgro, Mediatech Inc. Herndon, VA) and DMEM (WelGENE, KOREA), respectively, supplemented with 10% heat-inactivated fetal bovine serum (Hyclone, Logan, UT) and 1% antibiotic-antimycotic (Gibco, New York, NY), and were maintained at 378C in an atmosphere of 5% CO2. Tissue Samples Twenty-seven fresh-frozen tumor tissues, as well as 271 paraffin-embedded tumor tissues, were obtained from 298 NSCLC patients who underwent curative resection at the Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Seoul, Korea, between August 1994 and February 2006. Molecular Carcinogenesis
Post-operative follow-up for survival or recurrence was conducted as previously described [10]. Pathological TNM stage was determined according to the guidelines of the American Joint Committee on Cancer [11]. Written informed consent for the use of tissues, as approved by the Institutional Review Board at Samsung Medical Center, was obtained from each patient before surgery. Genomic DNA Extraction and Sodium Bisulfite Modification Hematoxylin and Eosin (H&E) staining of freshfrozen and paraffin-embedded tissues was performed, and areas corresponding to a tumor were microdissected manually under a microscope, and all containing at least 75% neoplastic tissue were used in this study. Genomic DNA from cultured cells and tissues was extracted using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) and DNeasy Tissue kit (Qiagen, Valencia, CA), respectively, according to the manufacturer’s instructions. One microgram of genomic DNA was modified by sodium bisulfite using the EZ DNA Methylation-Gold Kit (ZYMO Research, Orange, CA), according to the manufacturer’s protocol. Quantitative Analysis of Methylation using EpiTYPER Methylation status of individual CpG at the promoter of HOXA9 was quantitatively analyzed in six lung cancer cell lines, HBE135-E6E7 cells, and HDF cells using the EpiTYPERTM assay (Sequenom, San Diego, CA). Primers were designed using EpiDesigner software (Supplementary Table S1). After PCR amplification, amplicons were treated with shrimp alkaline phosphatase, in vitro transcribed, cleaved by RNaseA, and subjected to MALDI-TOF Mass Spectrometry in order to determine the methylation status. The results were analyzed using EpiTYPERTM ver. 1.0 software. Methylation-Sensitive High-Resolution Melting (MS-HRM) Assay Methylation status of HOXA9 promoter was also validated using the methylation-sensitive highresolution melting (MS-HRM) assay. Ten nanogram of bisulfite-treated genomic DNA was amplified for the assay. Primers which cover from upstream 2 kb of the ATG site to the intron 1 region in the HOXA9 gene were designed using SEQUENOM EpiDesigner software (Supplementary Table S2). MS-HRM was performed on LC-480 real-time PCR machine, and methylation status was analyzed by Gene Scanning software (Roche, Basel, Switzerland). Methylation change in response to 5-Aza-20 -deoxycytidine (5Aza-dC; Sigma–Aldrich, St. Louis, MO) was also analyzed using MS-HRM assay. Six lung cancer cell lines were incubated with 10 mM 5-Aza-dC for a total of 72 h. Cells were then harvested and washed in ice-cold PBS and genomic DNA was isolated.
3
HOXA9 AND RECURRENCE IN NSCLC
Quantitative Real-Time PCR The mRNA level of HOXA9 was analyzed using quantitative real-time PCR. Two microgram of total RNA was extracted from cultured cells using RNeasy Mini kit (Qiagen, Germany), and cDNA was synthesized using the SuperScriptTM First-Strand Synthesis System for use with the RT-PCR (Invitrogen, Carlsbad, CA) kit, according to the manufacturer’s instructions. Quantitative real-time PCR was performed with the SYBR Green I PCR Master Mix (QIAGEN, Penzberg, Germany) on a LightCycler1 480 Real-Time PCR System (Roche, Germany). Experiments were performed in triplicate, and the mean of the three experiments was used as a relative quantitation value. The Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was used for control and normalization. PCR primers are described in Supplementary Table S3. Transient Transfection Assay To determine the effect of HOXA9 on cell migration and proliferation in vitro, full-length HOXA9 cDNA was prepared by PCR amplification using EcoRItagged forward (50 -CCCGAATTCTAATTTCCGTGGGTCGGGCCG-30 ) and BamHI-tagged reverse (50 -CCCGGATCCGAAGGCGGAAGGGGGACGGA-30 ) primers. pAcGFP-C1-HOXA9 construct was generated by cloning HOXA9 cDNA in-frame into a pAcGFP1-C1 vector (CLONTECH, Oxford, UK). For the migration assay, H23 cells were seeded in six well plates at a concentration of 1 105 cells and incubated overnight to approximately 60–80% confluence. The cells were then transfected with 1 mg of a pAcGFP1-C1 constitutively expressing HOXA9 or with an empty vector using LipofectaminTM 2000 (Life Technologies, Carlsbad, CA) according to the manufacturer’s instructions. After 48 h of transfection, the expression of HOXA9 was confirmed by immunoblotting according to standard procedures, using primary antibodies directed to GFP (sc-9996; Santa Cruz Biotechnology, Santa Cruz, CA) and a-tubulin (Sigma–Aldrich). The migration was analyzed using 6.5 mm Transwell1 with an 8 mm pore size (Corning Costar, Lowell, MA), as described previously [12]. The absorbance was measured at 564 nm using VERSAmax microplate reader (Molecular Devices, Sunnyvale, CA). For the cell proliferation assay, 1 103 of H23 lung cancer cells were seeded in a 96 well plate after transfection, and cell proliferation activity was measured every 24 h with MTT assay. siRNA-Mediated Gene Silencing HCT116 cells were transiently tranfected with HOXA9-specific small interfering RNA (50 -TGCTGAGAATGAGAGCGGC-30 [13]) or non-targeting control (NC) siRNA (Cat. No. 10272810, Qiagen, Germany) at a final concentration of 25 nM. At 48 h after transfection, cells were harvested and 5 104 cells were subjected to migration assay. HOXA9 expression Molecular Carcinogenesis
was confirmed by quantitative real-time PCR and western blot analysis using HOXA9 primary antibody (sc-17155; Santa Cruz Biotechnology) 48 h after transfection. Immunohistochemistry Immunohistochemical analysis of Ki-67 was determined as described previously [14]. The fraction of Ki67-positive cells (Ki-67 proliferation index) was defined as the percentage of nuclei that stained positively with a monoclonal anti-Ki-67 antibody (clone MIB-1, DAKO, Carpinteria, CA). Methylation-Specific Polymerase Chain Reaction (MSP) The methylation statuses of HOXA9 in the 271 formalin-fixed paraffin-embedded tissues were determined using MSP, with two sets of primers: one for the unmethylated promoter and the other for the methylated promoter, as described previously [15]. The primer sequences are listed in Supplementary Table S4. Statistical Analysis The differences in clinicopathological characteristics and methylation statuses of the HOXA9 were analyzed using the t-test (or Wilcoxon rank sum test) and the Chi-squared test (or Fisher’s exact test) for continuous and categorical variables, respectively. Following univariate analysis, multivariate logistic regression analysis was performed to calculate the odds ratio (OR) and to determine the relationship between HOXA9 hypermethylation and the covariates found to be significant in univariate analysis after adjusting for potential confounding factors. The effect of HOXA9 hypermethylation on overall survival and recurrence-free survival (RFS) was estimated by Kaplan–Meier survival curves, and the significance of differences in RFS between the two groups was evaluated by the log-rank test. Cox proportional hazards regression model was used to estimate the hazard ratios of independent predictor variables, after adjusting potential confounders. RESULTS Promoter Methylation and Expression of HOXA9 In Vitro The locations of CpGs that were analyzed by EpiTYPERTM, MS-HRM, and MSP are indicated in Figure 1A. The methylation statuses of HOXA9 were analyzed using EpiTYPERTM assay in six lung cancer cell lines, a normal bronchial epithelial cell line (HBE135-E6E7), and a human dermal fibroblast (HDF) cell line. The promoter sequence of HOXA9 was obtained from TESS (http://www.cbil.upenn.edu/cgibin/tess/tess), and methylation statuses of 60 CpGs at the promoter region of HOXA9 were analyzed quantitatively using EpiTYPERTM. Most CpGs at the promoter of HOXA9 were found to be hypermethylated in the lung cancer cells (Figure 1B). Methylation
4
HWANG ET AL.
Figure 1. Analysis of HOXA9 methylation in vitro. (A) Gene map shows location of CpGs that were studied by different methods in this study. (B) The methylation status of HOXA9 was evaluated using the EpiTYPERTM assay in six lung cancer cell lines (H23, H226, H460, H520, H1650, and A549), a bronchial epithelial cell line (HBE135-E6E7), and a normal human dermal fibroblast (HDF). Levels of methylation are depicted in color change on s continuous scale from red (0%
methylated) to light yellow (100% methylated). X-axis and Y-axis indicate CpG sites and cell lines, respectively. (C) Methylation status of 15 fragments covering HOXA9 promoter and intron 1 was further analyzed using MS-HRM assay in six lung cancer cell lines (blue lines) and HDF cells (red line). (D) The mRNA levels of HOXA9 were analyzed by real-time PCR in six lung cancer cell lines and HDF cells. Error bars indicate one standard deviation.
status was further validated using MS-HRM assay (Figure 1C). A total of 15 fragments covering the promoter and intron 1 were amplified in six lung cancer cell lines, and the melting curve was analyzed using GeneScanning software (Roche, Swiss). The HDF cell line was used as a control because normal bronchial epithelial cell (HBE135-E6E7 cells) was slightly methylated at the promoter region of HOXA9. MS-HRM assay showed a substantial dissociation of the melting curve between six lung cancer cell lines and normal HDF cell line in most areas. HOXA9 mRNA expression was analyzed using quantitative real-time PCR (Figure 1D). The mRNA levels of six lung cancer cell lines were downregulated compared to HDF cells, suggesting that HOXA9 hypermethylation may be responsible for transcriptional downregulation of HOXA9 gene.
(Figure 2B) after treatment of lung cancer cells with 10 mM 5-Aza-dC for 72 h. Re-expression of HOXA9 in response to 5-Aza-dC was minimal in H226 cells, but other cell lines showed a substantial increase at the mRNA levels of HOXA9. A fragment (HRM-004) at the promoter region of HOXA9 was amplified to study demethylation of HOXA9, and melting curve was analyzed by GeneScanning software (Roche, Swiss). There was a significant dissociation of melting curves in response to 5-Aza-dC in six lung cancer cell lines (Figure 2B), suggesting that demethylation of the HOXA9 promoter may be associated with re-expression of the silenced HOXA9 gene.
5-Aza-dC Induced Demethylation and Re-Expression of Silenced HOXA9 To validate that the downregulation of HOXA9 in lung cancer cells was dependent on hypermethylation of the gene, we analyzed the re-expression and demethylation of silenced gene using quantitative real-time PCR (Figure 2A) and MS-HRM assay Molecular Carcinogenesis
HOXA9 Regulated Cell Migration in Lung Cancer Cells To investigate the function of HOXA9 in tumorigenesis, cell migration and cell proliferation was analyzed in H23 cells induced by transient transfection of GFP-tagged HOXA9. An anti-GFP antibody was used to detect the expression of HOXA9 by western blot analysis (Figure 3A). Cell migration was significantly reduced in H23 cells transfected with pAcGFP-C1-HOXA9 (P ¼ 0.04; Figure 3B). Cell proliferation was also analyzed in the H23 cells after transfection of the GFP-fusion constructs. However,
HOXA9 AND RECURRENCE IN NSCLC
5
Figure 2. Effect of 5-Aza-dC on demethylation and re-expression of silenced HOXA9. (A) Re-expression of silenced HOXA9 was examined by real-time PCR in six lung cancer cell lines after treatment of the cells with 10 mM 5-Aza-dC for 72 h. Error bars indicate one standard deviation. (B) Demethylation of silenced HOXA9 was analyzed using MS-HRM assay in the lung cancer cell lines. Figures show representative examples of demethylation in six cell lines. Red and blue lines indicate the statuses with and without treatment of 5-Aza-dC, respectively.
cell proliferation was not inhibited in H23 cells induced by pAcGFP-C1-HOXA9 (Figure 3C). In contrast, target sequence-specific siRNA-mediated knockdown of HOXA9 in HCT116 cells reduced the expression of HOXA9 at the level of mRNA (Figure 3D) and protein (Figure 3E) and also led to a significant increase in cell migration (P ¼ 0.0003; Figure 3F). Based on these observations, it is likely that HOXA9 may play a role by inhibiting cell migration in tumorigenesis of lung cancer. Clinicopathological Characteristics of HOXA9 Hypermethylation Before analyzing the clinical significance of HOXA9 hypermethylation, we first compared expression levels of HOXA9 between 27 fresh-frozen tumor and matched normal tissues using quantitative real-time PCR. The mRNA levels of HOXA9 were found to be Molecular Carcinogenesis
significantly lower in tumor tissues than in matched normal tissues (P < 0.0001; Wilcoxon signed-rank test; Figure 4A). We next analyzed the methylation status of the gene using methylation-specific PCR (Figure 4B) in the 271 formalin-fixed paraffinembedded tissues. The relationship between clinicopathological characteristics and HOXA9 hypermethylation is described in Table 1. HOXA9 hypermethylation was found in 191 (70%) of 271 patients studied. HOXA9 hypermethylation was not associated with patient age, sex, tumor size, exposure to tobacco smoke, differentiation, pathologic stage, histology, and tumor recurrence. Multivariate logistic regression analysis was performed to find an independent risk factor for HOXA9 hypermethylation after controlling for potential confounding factors, but no association was found between HOXA9 hypermethylation and covariates (Supplementary Table S5). To further
6
HWANG ET AL.
Figure 3. The effect of HOXA9 on cell migration and proliferation. (A) To analyze the effect of HOXA9 on cell migration and proliferation, H23 lung cancer cells were transfected with pAcGFP-C1-HOXA9 or empty vector. After transfection, the expression of HOXA9 was confirmed by western blot using primary antibody directed to GFP (sc-9996; Santa Cruz Biotechnology). (B) H23 cells transfected with pAcGFP-C1-HOXA9 were re-seeded in 6.5 mm Transwell1 with 8 mm pore size (Corning). The cells were induced to migrate into 1% of FBS containing media in 24-well plate. After 48 h each transwell insert was stained by 1% crystal violet and destained with 10% acetic acid. The absorbance was
measured at 564 nm using VERSAmax microplate reader (Molecular devices). (C) H23 cells transfected with pAcGFP-C1-HOXA9 or empty vector were seeded in 96-well plate, and cell proliferation activity was measured every 24 h with MTT assay. Y-axis indicates absorbance measured at 570 nm using VERSAmax microplate reader (Molecular devices), and the data are presented as mean standard error (SE) of eight experiments. (D and E) Knockdown of HOXA9 in HCT116 cells was evaluated by quantitative real-time PCR (D) and immunoblotting (E). (F) HCT116 cells were transfected with non-targeting control (NC) siRNA and HOXA9 siRNA, and cell migration was measured as (B).
understand the effect of HOXA9 hypermethylation on cell growth in vivo, we analyzed Ki-67 proliferation index using immunohistochemistry (Figure 4C) and found that HOXA9 hyermethylation was not associated with Ki-67 proliferation index in 317 patients (P ¼ 0.15; Figure 4D). The association of HOXA9 hypermethylation with the Ki-67 index also was not statistically significant in adenocarcinoma (P ¼ 0.17) and squamous cell carcinoma (P ¼ 0.78).
effect of HOXA9 hypermethylation on RFS was confounded by smoking status due to the interaction between smoking status and HOXA9 hypermethylation (Breslow–Day test of homogeneity, P ¼ 0.002). Therefore, data was stratified according to the smoking status for the analysis of RFS. HOXA9 hypermethylation was significantly associated with the RFS in never-smokers (P ¼ 0.009; Figure 5A) but not in former smokers (P ¼ 0.83; Figure 5B) and current smokers (P ¼ 0.58; Figure 5C). Cox proportional hazards analysis also showed that neversmokers with HOXA9 hypermethylation had 3.98 times (95% CI ¼ 1.06–17.09, P ¼ 0.01) poor hazard ratio of recurrence than those without, after adjusting for age, sex, tumor size, adjuvant therapy, pathologic stage, and histology.
Survival Analysis Finally, the relationship between HOXA9 hypermethylation and patient’s survival was investigated in the 271 NSCLCs with a median follow-up period of 4.9 yr. HOXA9 hypermethylation was not associated with overall survival (data not shown). However, the Molecular Carcinogenesis
HOXA9 AND RECURRENCE IN NSCLC
7
Figure 4. Analysis of HOXA9 hypermethylation in primary NSCLC. (A) The mRNA levels of HOXA9 were compared between 27 fresh-frozen tumor and matched normal tissues using real-time PCR. (B) HOXA9 hypermethylation was analyzed in paraffin-embedded tissues from 271 NSCLC patients using methylation-specific PCR. Patient identification numbers are indicated, and “Pos” represents the positive controls for the methylated (M) and unmethylated (U) alleles. Negative control
without DNA template was included for each PCR. (C) Expression of Ki67 was analyzed using immunohistochemistry. Figures show representative examples of positive expression of Ki-67 in adenocarcinoma (left) and squamous cell carcinoma (right). (D) Ki-67 proliferation index was compared according to methylation status of HOXA9. The “adenoca” and “squamous” represent adenocarcinoma and squamous cell carcinoma, respectively. Error bars indicate one standard deviation.
DISCUSSION
cancer cells in vitro and analyzed the methylation status of HOXA9 in 271 paraffin-embedded tumor tissues. In this study, siRNA-mediated knockdown of HOXA9 resulted in a significant increase in cell migration, and HOXA9 hypermethylation was associated with recurrence after surgery. The prognostic significance of HOXA9 hypermethylation has been reported in a variety of tumors. For example, a scoring of quantified methylation levels of five genes (HOXA6, HOXA9, PENK, UPK3A, and IGF2BP1) is associated with recurrence in meningioma [16], and HOXA9 hypermethylation in urine specimen is associated with bladder cancer recurrence [17]. In addition, HOXA9 hypermethylaton is also associated with high grade serous ovarian carcinoma [18] and with mortality in noninfant patients with neuroblastoma [19]. Recently, Sandoval et al. [9] reported that hypermethylation of five genes (HIST1H4F, PCDHGB6, NPBWR1, ALX1, and HOXA9) was significantly associated with shorter RFS in stage I NSCLC. In contrast to Sandoval’s finding [9], the present study showed no association between HOXA9 hypermethylation and poor RFS in stage I NSCLC, suggesting that additional alteration of other genes in addition to HOXA9 is required to affect the RFS of a patient with stage I NSCLC. The relationship between recurrence and HOXA9 hypermethylation in this study was confounded by smoking status: HOXA9 hypermethylation occurred at a similar frequency in never-smokers, former and current smokers, but the effect of HOXA9 hypermethylation on poor RFS was found only in never-smokers. Studies have suggested that lung cancers occurring in never-smokers may be a distinctive disease due to the unique molecular profiles [20–22]. Definite differences in the epidemiology and the biology underlying the pathogenesis and behavior of lung cancer exist between neversmokers and smokers. However, it is not clear how HOXA9 hypermethylation affects patient survival. In this study, transient transfection of pAcGFP-C1-
To understand the clinicopatholgocial significance of HOXA9 hypermethylation in lung cancer, we characterized HOXA9 hypermethylation in lung Table 1. Clinicopathological Characteristics (N ¼ 271) HOX9 hypermethylation Variables Agea Tumor size (cm)a Pack-years (smoking)a Sex Women Men Smoking status Never Former current Histology Adenocarcinoma Squamous cell carcinoma Others Differentiationb Well Moderately Poorly Undifferentiated Stage I II III IV Recurrence No Yes
No (N ¼ 80)
Yes (N ¼ 191)
P-value
59 10 3.3 1.9 31 27
60 11 3.7 2.1 33 26
0.92 0.12 0.52
25 55
48 143
0.30
20 19 41
54 38 99
0.73
46 30
92 86
4
13
0.37c
16 41 10 1
31 92 31 6
0.75c
51 13 15 1
108 35 43 5
0.77c
52 28
104 87
0.11
Values indicate mean standard deviation. Differentiation data are missing for 53 patients. c Based on Fisher’s exact test. a
b
Molecular Carcinogenesis
8
HWANG ET AL.
Figure 5. Analysis of recurrence-free survival. The effect of HOXA9 hypermethylation on recurrence-free survival was analyzed using logrank test in never-smokers (A), former smokers (B), and current smokers (C). The yellow and blue lines indicate groups with and without hypermethylation of HOXA9, respectively.
HOXA9 into H23 lung cancer cells did not affect cell proliferation, and HOXA9 hypermethylation was not associated with tumor size and the Ki-67 proliferation index in 271 primary NSCLCs. Accordingly, HOXA9 hypermethylation may affect patient survival through other mechanism rather than cell growth. The molecular mechanisms underlying the high rates of recurrence in patients with HOXA9 hypermethylation are not clear, but it may have resulted from the failure to block migration of residual cancer cells after surgery. Cell migration and invasion are crucial steps in tumor metastasis that occurs through specific molecular interactions among tumor cells, extracellular matrix and stroma cells. Further study is required Molecular Carcinogenesis
to clearly understand how HOXA9 hypermethylation contributes to tumor cell migration. HOXA9 is frequently deregulated in a variety of human cancers, in which it acts as a tumor suppressor or as an oncogene. Although the mechanisms underlying these differential functions remain incompletely understood, HOXA9 seems to exert its function by interacting with different kinds of proteins in a tissue-specific manner. For example, the oncogenic potential of HOXA9 has clearly been implicated in leukemia. HOXA9 is known to increase endothelial cell migration and tube formation by stimulating the expression of EphB4 and to play an important role for the nucleoporin gene NUP98 by existing as a chimeric NUP98/HOXA9 transcript in human myeloid leukemia cells [13,23]. In addition, HOXA9 acts as a transcriptional activator of E-selectin that promotes tumor metastasis by enhancing the adhesion of circulating tumor cells to endothelial cells [24]. HOXA9 is also activated by a phosphoinostide 3-kinase (PI3K) pathway in human glioblastoma multiforme and increases cell proliferation and inhibits apoptosis [25]. In contrast to the oncogenic role, several studies suggest a tumor suppressor role of HOXA9 in human cancer. For example, HOXA9 is known to promote the differentiation of breast cancer cells and inhibits the aggression of breast cancer cells by modulating expression of BRCA1 [26]. The function of HOXA9 in lung cancer has also exhibited conflicting results. In this study, transient transfection of HOXA9 to H23 lung cancer cells inhibited tumor cell migration, and the suppression of HOXA9 by siRNA enhanced the migration activity of cancer cells, thereby suggesting that HOXA9 may act as a tumor suppressor in lung cancer by playing a role for regulating the migration activity of cancer cells. However, Plowright et al. [27] recently reported that inhibition of HOX function using the HXR9 peptide, an 18-amino-acid peptide consisting of hexapeptide sequence that can bind to a cofactor PBX and nine C-terminal arginine residues, resulted in apoptosis in vitro and reduced the growth of tumor in vivo. Although HOX proteins have functions that are independent of DNA-binding proteins, which act as co-factors such as PBX, MEIS, and PREP etc., many aspects of the functions of HOX are mostly dependent on interaction with co-factors [28]. The co-factors function by modifying the DNA-binding specificity of HOX proteins in vitro and by influencing the regulation of transcription. Accordingly, complex and diverse interaction between HOX proteins and their DNA-binding partners may affect the functions of HOX proteins in cancer cells. This study had several limiting actors. First, silencing by DNA methylation usually involves methylbinding domain proteins binding to the DNA and recruiting histone deacetylases (HDACs). Therefore, the DNA binding and HDACs recruiting to the plasmid and the nucleosome assembling on a
9
HOXA9 AND RECURRENCE IN NSCLC
transfected plasmid needed to be studied for understanding the mechanism underlying transcriptional silencing by HOXA9 hypermethylation. Second, HOXA9 was associated with tumor cell migration, but its downstream targets are still unknown. Delineation of the target genes regulated by HOXA9 is required. Third, we analyzed methylation status of HOXA9 using MSP in 271 formalin-fixed paraffinembedded tissues. Employing quantitative assay (COBRA, Q-MSP, other methods) instead of MSP may have provided more convincing data regarding the DNA methylation status of the gene. In conclusion, the present study suggests that HOXA9 hypermethylation may be an independent prognostic indicator of poor RFS in never-smokers and that HOXA9 may contribute to NSCLC by being involved in cell migration rather than cell proliferation.
11. 12.
13.
14.
15.
16.
ACKNOWLEDGMENTS The authors wish to thank Eun-Kyung Kim and JinHee Lee for data collection and management, and Hoon Suh and Seo Kyu Park for sample collection. This work was supported by grants from the National Cancer Center (NCC 1110140-1 and 1110100-2), the Korea Healthcare technology R&D Project, Ministry for Health, Welfare&Family Affairs (A084947AA202308N100010A), and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 20110029138), Republic of Korea.
20.
REFERENCES
21.
1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2008. CA Cancer J Clin 2005;58:71–96. 2. Kim JS, Kim JW, Han J, Shim YM, Park J, Kim DH. Cohypermethylation of p16 and FHIT promoters as a prognostic factor of recurrence in surgically resected stage I non-small cell lung cancer. Cancer Res 2006;66:4049–4054. 3. Krumlauf R. Hox genes in vertebrate development. Cell 1994; 78:191–201. 4. Rauch T, Wang Z, Zhang X, et al. Homeobox gene methylation in lung cancer studied by genome-wide analysis with a microarray-based methylated CpG island recovery assay. Proc Natl Acad Sci USA 2007;104:5527–5532. 5. Bibikova M, Lin Z, Zhou L, et al. High-throughput DNA methylation profiling using universal bead arrays. Genome Res 2006;16:383–393. 6. Nelson HH, Marsit CJ, Christensen BC, et al. Key epigenetic changes associated with lung cancer development: Results from dense methylation array profiling. Epigenetics 2012;7: 559–566. 7. Olaussen KA, Soria JC, Park YW, et al. Assessing abnormal gene promoter methylation in paraffin-embedded sputum from patients with NSCLC. Eur J Cancer 2002;41:2112–2119. 8. Selamat SA, Galler JS, Joshi AD, et al. DNA methylation changes in atypical adenomatous hyperplasia, adenocarcinoma in situ, and lung adenocarcinoma. PLoS ONE 2011;6: e21443. 9. Sandoval J, Mendez-Gonzalez J, Nadal E, et al. A prognostic DNA methylation signature for stage I non-small-cell lung cancer. J Clin Oncol 2013;31:4140–4147. 10. Kim JS, Lee H, Kim H, et al. Promoter methylation of retinoic acid receptor beta 2 and the development of second primary
Molecular Carcinogenesis
17.
18.
19.
22. 23.
24. 25.
26. 27.
28.
lung cancers in non-small-cell lung cancer. J Clin Oncol 2004;22:3443–3450. Edge SB, Byrd DR, Compton CC, Fritz AF, Greene FL, Trotti A, editors. Lung AJCC Cancer Staging Manual. 7th edition. New York: Springer; 2010. p. 253–270. Hwang JA, Kim Y, Hong SH, et al. Epigenetic inactivation of heparan sulfate (glucosamine) 3-o-sulfotransferase 2 in lung cancer and its role in tumorigenesis. PLoS ONE 2013;8: e79634. Bruhl T, Urbich C, Aicher D, Acker-Palmer A, Zeiher AM, Dimmeler S. Homeobox A9 transcriptionally regulates the EphB4 receptor to modulate endothelial cell migration and tube formation. Circ Res 2004;94:743–751. Kim JS, Han J, Shim YM, Park J, Kim DH. Aberrant methylation of H-cadherin (CDH13) promoter is associated with tumor progression in primary nonsmall cell lung carcinoma. Cancer 2005;104:1825–1833. Herman JG, Graff JR, Myöhänen S, Nelkin BD, Baylin SB. Methylation-specific PCR: A novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 1996;93:9821– 9826. Kishida Y, Natsume A, Kondo Y, et al. Epigenetic subclassification of meningiomas based on genome-wide DNA methylation analyses. Carcinogenesis 2012;33:436–441. Reinert T, Borre M, Christiansen A, Hermann GG, Ørntoft TF, Dyrskjøt L. Diagnosis of bladder cancer recurrence based on urinary levels of EOMES, HOXA9, POU4F2, TWIST1, VIM, and ZNF154 hypermethylation. PLoS ONE 2012;7:e46297. Montavon C, Gloss BS, Warton K, et al. Prognostic and diagnostic significance of DNA methylation patterns in high grade serous ovarian cancer. Gynecol Oncol 2012;124:582– 588. Alaminos M, Davalos V, Cheung NK, Gerald WL, Esteller M. Clustering of gene hypermethylation associated with clinical risk groups in neuroblastoma. J Natl Cancer Inst 2004;96: 1208–1219. Subramanian J, Govindan R. Molecular genetics of lung cancer in people who have never smoked. Lancet Oncol 2008;9:676–682. Rudin CM, Avila-Tang E, Harris CC, et al. Lung cancer in never smokers: Molecular profiles and therapeutic implications. Clin Cancer Res 2009;15:5646–5661. Yano T, Miura N, Takenaka T, et al. Never-smoking nonsmall cell lung cancer as a separate entity: Clinicopathologic features and survival. Cancer 2008;113:1012–1018. Nakamura T, Largaespada DA, Lee MP, et al. Fusion of the nucleoporin gene NUP98 to HOXA9 by the chromosome translocation t(7;11)(p15;p15) in human myeloid leukaemia. Nat Genet 1996;12:154–158. Bandyopadhyay S, Ashraf MZ, Daher P, Howe PH, DiCorleto PE. HOXA9 participates in the transcriptional activation of Eselectin in endothelial cells. Mol Cell Biol 2007;27:4207–4216. Costa BM, Smith JS, Chen Y, et al. Reversing HOXA9 oncogene activation by PI3K inhibition: Epigenetic mechanism and prognostic significance in human glioblastoma. Cancer Res 2010;70:453–462. Gilbert PM, Mouw JK, Unger MA, et al. HOXA9 regulates BRCA1 expression to modulate human breast tumor phenotype. J Clin Invest 2010;120:1535–1550. Plowright L, Harrington KJ, Pandha HS, Morgan R. HOX transcription factors are potential therapeutic targets in nonsmall-cell lung cancer (targeting HOX genes in lung cancer). Br J Cancer 2009;100:470–475. Moens CB, Selleri L. Hox cofactors in vertebrate development. Dev Biol 2006;291:193–206.
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s web-site.
