Identification and Characterization of a Novel MelanomaTumor ...

Human Cancer Biology

Identification and Characterization of a Novel MelanomaTumor Suppressor Gene on Human Chromosome 6q21 Jackie Mei-Wah Fung,1 Ross Smith,3 Melissa A. Brown,3 Sze Hang Lau,1 Dan Xie,4 George K. Lau,2 and Xin-Yuan Guan1,4

Abstract

Purpose: By characterizing a complex chromosome rearrangement involving 6q and 17p in melanoma cell line UACC-930, we isolated a candidate tumor suppressor gene at 6q21, named prenyl diphosphate synthase subunit 2 (PDSS2), which was interrupted by an inversion breakpoint. The purpose of this study was to determine the tumor-suppressive potential of PDSS2 in the development of melanoma. Experimental Design: To isolate the rearranged 6q in UACC-930 cells, a bacterial artificial chromosome clone (RP1-67A8) covering the breakpoint at 6q21 was digested with HindIII and each DNA fragment was used as the probe for the breakpoint in Southern blotting. The HindIII fragment probe covering the breakpoint was then used to screen an EcoRI-digested DNA library generated from UACC-930. To characterize the tumor-suppressive potential of PDSS2, PDSS2 was stably transfected into a highly tumorigenic melanoma cell line, UACC-903. The tumor-suppressive function of PDSS2 was shown by both in vitro and in vivo assays. The differential expression of PDSS2 in benign nevi and primary melanoma samples was also studied. Results: Down-regulation of PDSS2 was observed in 59 of 87 (67.8%) primary melanomas, which was significantly higher than that in benign nevi (7 of 66, 10.6%; P < 0.001). In addition, an overexpression of the PDSS2 in UACC-903 cells could inhibit tumor cell growth, decrease the colony-forming ability in soft agar, and totally abrogate the tumorigenicity of UACC-903 in nude mice. Conclusions: Our results support the proposal that PDSS2 is a novel tumor suppressor gene that plays an important role in the development of malignant melanoma.

Melanoma is the most serious type of skin cancer (1). Genetic alterations in melanoma have been widely studied and the deletion of 6q is one of the most common chromosomal changes (2 – 4). The reversion of the tumorigenic phenotype of melanoma cells by introducing a normal human chromosome 6 provides the first biological evidence that chromosome 6 contains a tumor suppressor gene (TSG) associated with melanoma pathogenesis (5). The abnormality of 6q has also been detected in other malignancies, including malignant lymphoma (6), ovarian (7), and prostate carcinomas (8).

Several studies showed that 6q21 is the most frequent deleted region, suggesting the existence of a TSG at 6q21 (2, 3, 9). Although a couple of candidate TSGs at 6q, including AIM1 at 6q21 (10) and SOD2 at 6q25.3 (11), has been studied, the underlying mechanisms of these genes in melanoma development are unclear. In addition, the tumor-suppressive role of SOD2 is controversial. Miele et al. (12) found that SOD2 could not suppress tumorigenicity or metastasis of human melanoma C8161 cells. Therefore, more effort in looking for new candidate TSGs at 6q related to melanoma pathogenesis is justified. In our previous study, we found a reciprocal-like translocation between 6q21 and 17p13 in the melanoma cell line UACC930 (9). The translocation was further characterized by chromosome microdissection and fluorescence in situ hybridization, and a complex chromosomal rearrangement was detected that includes an inversion of 6q [inv(6)(q21q27)] and a translocation between the inverted 6q and 17p [t(6;17)(q27;p13)]. A bacterial artificial chromosome clone (RP1-67A8) spanning the inversion breakpoint at 6q21 was isolated by fluorescence in situ hybridization screening and loss of one or more copies at 6q21 was detected in 10 of 12 melanoma cell lines (9). This result suggests that the bacterial artificial chromosome clone may contain a TSG related to the melanoma development, which was interrupted by the inversion breakpoint. In the present study, we further characterized the breakpoints on the rearranged 6q in UACC-930 and found a candidate TSG, named prenyl diphosphate synthase subunit 2

Authors’ Affiliations: Departments of 1Clinical Oncology and 2Medicine, The University of Hong Kong, Pokfulam, Hong Kong Special Administrative Region, China; 3School of Molecular and Microbial Sciences,The University of Queensland, St. Lucia, Queensland, Australia; and 4State Key Laboratory of Oncology in Southern China, Cancer Center, SunYat-Sen University, Guangzhou, China Received 6/11/08; revised 10/23/08; accepted 10/25/08. Grant support: Leung Kwok Tze Foundation, Sun Yat-Sen University ‘‘Hundred Talents Program’’ (85000-3171311), and The Major State Basic Research Program of China (2006CB910104). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Xin-Yuan Guan, Department of Clinical Oncology, The University of Hong Kong, Room 56, 10/F Laboratory Block, 21 Sassoon Road, Pokfulam, Hong Kong Special Administrative Region, China. Phone: 852-28199782; Fax: 852-28169126; E-mail: xyguan@ hkucc.hku.hk. F 2009 American Association for Cancer Research. doi:10.1158/1078-0432.CCR-08-1472

www.aacrjournals.org

797

Clin Cancer Res 2009;15(3) February 1, 2009


(Clontech). The open reading frame of PDSS2 flanked by the primers 5¶-TACTGTAAGTTCCTCCTCTGC and 5¶-AATTTTGTTTGTAGCAGTTCC were amplified using PCR. Cycle sequencing reactions were set up using BigDye Terminator v3.1 Cycle Sequencing kits (Applied Biosystems) according to the manufacturer’s instructions. Quantitative PCR. cDNA was subjected to quantitative PCR with a SYBR Green PCR kit (Applied Biosystems) using primers 5¶-GAATCAGGTAGTGTCAGAGG and 5¶-GAGGCTATTCCAGCTGTCATG for PDSS2 , whereas 5¶-CTCTTAGCTGAGTGTCCCGC and 5¶-CTGATCGTCTTCGAACCTCC for 18S rRNA. The amplification protocol consisted of incubations at 95jC for 15 s, 60jC for 1 min, and 72jC for 1 min for 40 cycles. Quantification was done using the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). All quantitative PCRs were done in triplicate. The C T of each gene of interest [CT(gene of interest) test] in each sample was normalized with control 18S rRNA [C T(18S) test] for RNA amount variation and calibrator for plate-to-plate variation using the following formula: DC T(test) = C T(gene of interest) test - C T(18S) test. Relative expression level was presented as the relative fold change and calculated using the following formula: 2-DDC T = [C T(gene of interest) - C T(18S)] test - [C T(gene of interest) - C T(18S)] calibrator. Southern, Northern, and Western blot analysis. For Southern blot analysis, 10 Ag of genomic DNA from tumor cell lines were digested with HindIII and EcoRV, fractionated on a 0.8% agarose gel, transferred to a nylon membrane (GE Healthcare), and hybridized with 32P-labeled probes overnight at 42jC. For Northern blot analysis, 10 Ag of total RNA from each sample were fractionated on a 1% agarose gel, transferred to a nylon membrane, and hybridized with 32P-labeled probes overnight at 65jC. For Western blot analysis, 10 Ag of protein extract were separated by 10% SDS-PAGE, transferred to a polyvinylidene difluoride Hybond-P membrane (GE Healthcare), and detected by a mouse monoclonal antibody specific against PDSS2 (1:1,000; Wolwo Biotech) and a goat polyclonal antibody against h-actin (1:2,000; Santa Cruz Biotechnology). The antibody detects a single band at 37 kDa, which is the estimated size of PDSS2 protein. Tissue microarray and immunohistochemistry. Archived paraffin blocks from 99 melanomas and 82 benign nevi were collected from the Cancer Center of Sun Yat-Sen University (Guangzhou, China). The tissue microarray block was constructed according to the method described previously (15). Immunohistochemistry was done using standard antigen-antibody methods with the anti-PDSS2 antibody. As a negative control, the primary antibody was replaced with blocking serum in PBS. The antibody could specifically detect PDSS2 expression (cytoplasmic staining) in most benign nevi and some of melanoma cells. In addition, PDSS2 expression could be observed in all endothelial cells of blood vessels in both benign nevi and melanoma tissues, which was used as a positive control. PDSS2 antibody was also used to study PDSS2 expression in hepatocellular carcinomas and the results showed that cytoplasmic expression of PDSS2 could be detected in most nontumorous liver cells but only in f50% hepatocellular carcinoma specimens (data not shown). Absent or very weak expression of PDSS2 was considered as negative staining. Moderate to strong cytoplasmic staining of PDSS2 was considered as positive. Lost samples and unrepresentative samples were classified as noninformative and excluded from data analysis. Establishment of stably PDSS2-expressing UACC-903 cells. The open reading frame of PDSS2 was cloned into the expression vector pcDNA 3.1(+) (Invitrogen) using primers 5¶-CCATGAACTTTCGGCAGCTGC and 5¶-AATTTTGTTTGTAGCAGTTCC and then used to transfect a melanoma cell line UACC-903. Clones stably expressing PDSS2 were selected with 600 Ag/mL G418 (Invitrogen). A blank vector was also transfected into UACC-903 cells (903-vec) as a control. In vitro tumorigenic assays. Soft agar assays were carried out by suspending 1 104 cells in 0.4% Seaplaque agar (BioWhittaker Molecular Applications) and seeding onto solidified 0.6% agar in six-well plate. Colonies that were at least four times as large as the

Translational Relevance The incidence of melanoma has increased markedly over the past 40 years inWestern countries. Deletion of chromosome 6q is one of the most frequently chromosomal alterations in melanoma, suggesting the presence of tumor suppressor gene(s) at 6q. In this article, we identified a novel candidate tumor suppressor gene, prenyl diphosphate synthase subunit 2 (PDSS2), which was interrupted by an inversion breakpoint in a melanoma cell line, UACC930. Our study showed that PDSS2 was frequently down-regulated in primary melanomas and transfection of PDSS2 into a highly tumorigenic melanoma cell line UACC-903 could completely reverse the tumorigenic phenotypes. PDSS2 encodes an essential enzyme involved in the coenzyme Q10 (CoQ10) biosynthetic pathway. Low CoQ10 level has been reported in different malignancies, including melanoma. Supplement of CoQ10 has been used in clinics for cancer treatment. We believe that this study provides a new insight in the melanoma development, which may lead to finding out new therapeutic targets in the future.

(PDSS2), which was interrupted by the 6q21 breakpoint. The sequence of PDSS2 (AF254956) was firstly submitted to National Cancer for Biotechnology Information by the authors in 2000 and was then named as C6orf210 by Human Genome Organization. Recently, C6orf210 was found to be important in determining the length of the side chain of ubiquinone in mammals and was named as PDSS2 (13). Mutation of PDSS2 had only been reported in Leigh syndrome with nephropathy and coenzyme Q10 (CoQ10) deficiency (14). In this article, the authors, for the first time, showed the tumor-suppressive ability of PDSS2 by both in vivo and in vitro assays. The results indicated that PDSS2 may play an important role in the development of melanoma.

Materials and Methods Cell lines. The melanoma cell lines including UACC-325, UACC457, UACC-827, UACC-903, UACC-930, UACC-1022, UACC-1227, and UACC-1237 were obtained from the Tissue Culture Core Facility at the Arizona Cancer Center (Tucson, AZ). The normal human epidermal melanocyte, HEMn-LP, isolated from the lightly pigmented adult skin was obtained from Cascade Biologics, Inc. All the melanoma cell lines were cultured in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen). The HEMn-LP line was cultured in medium 254 (Cascade Biologics) plus 1% human melanocyte growth supplement (Cascade Biologics). Construction of a genomic DNA library for breakpoint screening. To isolate breakpoints in UACC-930, genomic DNA from the melanoma cell line UACC-930 was digested with EcoRI for the construction of DNA library according to the manufacturer’s instructions (ZAP Express Predigested Gigapack Cloning kit, Stratagene). DNA fragments spanning breakpoints were selected by two-round screening using 32 P-labeled specific probes. Mutation analysis of PDSS2. RNA from cell culture was extracted using Trizol (Invitrogen) or an RNeasy Mini kit (Qiagen) according to the manufacturer’s instructions. Two micrograms of total RNA were used to synthesize cDNA using the reverse transcription-PCR kit


798


PDSS2 in Melanoma original single cell were counted at day 21. Triplicate independent experiments were done. Cell growth rate was measured by 2,3-bis[2-methoxy-4-nitro-5sulfophenyl]-2H-tetrazolium-5-carboxanilide inner salt (XTT) assay. Briefly, cells were seeded onto 96-well plate at a density of 2 103 per well supplemented with 2% fetal bovine serum for 1 to 5 d. The cell growth rate was detected using cell proliferation XTT kit (Roche) according to the manufacturer’s instructions. Triplicate independent experiments were done. Tumorigenicity in nude mice. Tumor xenograft experiments were carried out to test tumorigenicity of PDSS2-expressing cells and control cells. Approximately 2.5 106 cells were injected s.c. into the left (903vec) and right (PDSS2-expressing cells) flanks of 4- to 6-wk-old nude mice (5 mice for 903-C1, 5 mice for 903-C2, 5 mice for 903-C3, and 15 mice for 903-C4). Tumor formation in nude mice was monitored over a 30-d period. Statistical analysis. The correlation between the expression of PDSS2 in benign nevi and melanomas was evaluated by Fisher’s exact test with P < 0.05 considered significant. The association between PDSS2 expression and cell proliferation, anchorage independent growth, tumor formation in nude mice and was assessed by two-tailed unpaired Student’s t test with P < 0.05 designated as statistically significant.

cells, a bacterial artificial chromosome clone (RP1-67A8) covering the breakpoint at 6q21 was digested with HindIII and each DNA fragment was used as the probe for the breakpoint by Southern blot analysis. Using this approach, a 7.9-kb rearranged HindIII fragment was detected by a 2.7-kb fragment probe (Fig. 1A). The 2.7-kb fragment probe was then used to screen an EcoRI-digested DNA library generated from UACC-930. Two rearranged fragments were isolated and sequenced. Results showed that they spanned the translocation breakpoint 6q21(proximal)/17p13(distal) and the inversion breakpoint 6q21(d)/6q22(d), respectively (Fig. 1B). We then designed probes detecting 6q22(p) and 17p13(p) parts to screen the library and isolated DNA fragments covering the inversion breakpoint at 6q22(p)/ 6q27(d) and translocation breakpoint 6q27(p)/17p13(p), respectively (Fig. 1B). These results revealed that the complex chromosome rearrangement in UACC-930 includes two inversions at 6q and a translocation between the inverted 6q and 17p (Fig. 1C). Sequencing analysis showed that the first intron of PDSS2 was interrupted by the breakpoint at 6q21. PDSS2 gene. The PDSS2 gene spans f270 kb on chromosome 6 and contains eight coding exons. The 3,553-bp PDSS2 cDNA encodes a protein containing 399 amino acids with a predicted molecular size of 44 kDa. Sequencing analysis predicts that PDSS2 contains a Trans_IPPS_HT domain (Fig. 2A). This conserved domain includes all-trans-isoprenyl

Results Mapping translocation breakpoints in UACC-930 cells. To characterize breakpoints of the rearranged 6q in UACC-930

Fig. 1. Characterization of breakpoints of a complex chromosome rearrangement in UACC-930. A, detection of DNA fragment covering breakpoint at 6q21by Southern blot analysis. Rearranged 7.9-kb HindIII and 8-kb BstXI fragments were detected in UACC-930 cells compared with normal placenta (N). B, A 2.7-kb HindIII fragment containing the 6q21breakpoint was used to isolate breakpoints by screening a genomic DNA library generated from UACC-930 (p, proximal; d, distal). Two rearranged DNA fragments covering 6q21(p)/17p13(d) and 6q22(d)/6q21(d) breakpoints were isolated. Using the same strategy, the remaining two breakpoints 6q22(p)/ 6q27(d) and 17p13(p)/6q27(p) were identified. DNA sequences flanking each breakpoint are listed above or below the DNA fragment. During the rearrangement, 8 bp was lost at 6q21breakpoint (red) and 3 bp was inserted at the 6q22(p)/6q27(d) breakpoint (black). C, a summary of the complex rearrangement in UACC-930. The rearrangement includes two inversions at 6q and a translocation involving the inverted 6q and 17p13.


799



Fig. 2. Expression of PDSS2 in tissues and melanomas. A, predicted protein structure of PDSS2. B, expression of PDSS2 in human tissues was detected by Northern blot analysis. h-Actin was used as loading control. C, expression of PDSS2 was detected by Western blotting in melanoma cell lines and a normal melanocyte (HEMn-LP). h-Actin was used as loading control. D, examples of PDSS2 expression in benign nevus (left) and primary melanoma (right) detected by immunostaining. Magnification, 200. Cytoplasmic expression of PDSS2 (brown) was observed in nevus but not in melanoma cells.

diphosphate synthase, which catalyzes the synthesis of an isoprenoid side chain of a ubiquinone molecule by the addition of isopentenyl diphosphate molecules to farnesyl diphosphate or geranylgeranyl diphosphate in multiple steps to synthesize CoQ10. Northern blot analysis was used to assess the expression pattern of PDSS2 in human tissues. Results indicated that PDSS2 expresses two transcripts: 3.5 and 1.2 kb (Fig. 2B). The 3.5-kb transcript was expressed strongly in heart, prostate, and testis; moderately in brain, kidney, liver, lung, spleen, duodenum, esophagus, pancreas, thymus, and thyroid; and weakly in colon, muscle, small intestine, salivary gland, and uterus and was absent in stomach, peripheral blood lymphocyte, and urinary bladder (Fig. 2B). Expression of the 1.2-kb transcript was detected in heart, kidney, liver, pancreas, prostate, testis, thymus, and thyroid (Fig. 2B). PDSS2 expression in melanoma cell lines was tested by Western blotting using a monoclonal anti-PDSS2 antibody, which showed that PDSS2 was up-regulated in UACC-457 and UACC-1022 and down-regulated in UACC-930 and UACC-1227 when compared with that expressed in the melanocyte HEMn-LP (Fig. 2C). Similar results were also obtained by using quantitative PCR (data not shown).

Mutation analysis. The open reading frame of PDSS2 was analyzed by sequencing in melanoma cell lines UACC-325, UACC-457, UACC-827, UACC-903, UACC-930, UACC-1022, UACC-1227, and UACC-1237. No mutation was found in these melanoma cell lines. Down-regulation of PDSS2 in primary melanomas. To look for evidence for a role for PDSS2 in melanoma development, PDSS2 expression in primary melanomas was determined by immunohistochemical analysis of a tissue microarray containing 99 melanomas and 82 benign nevi. Results showed that absent or very weak expression of PDSS2 in tumor tissue (negative staining) was detected in 59 of 87 (67.8%) informative melanomas and 7 of 66 (10.6%) informative nevi (Fig. 2D; Table 1). Therefore, a significant association of PDSS2 down-regulation and melanoma development (P < 0.001, Fisher’s exact test) was observed. In vitro tumor-suppressive ability of PDSS2. Because UACC930 cells do not form colonies in soft agar or tumors in nude mice, another melanoma cell line (UACC-903) was chosen for functional analysis of PDSS2. In previous studies, Trent et al. (5) showed that introducing a normal human chromosome 6 into UACC-903 cells could effectively reverse their tumorigenic phenotype. To determine the tumor-suppressive ability of

Table 1. Expression of PDSS2 in 82 benign nevi and 99 melanomas Cases

Nevi Melanomas

P*

PDSS2 expression

Informative/total case

Positive

Weak/negative

66/82 87/99

59 (89.4%) 28 (32.2%)

7 (10.6%) 59 (67.8%)