Oncogene (2001) 20, 2956 ± 2964 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc Negative regulation of G1/S transition by the candidate bladder tumour suppressor gene DBCCR1 Hiroyuki Nishiyama1, Jason H Gill1, Eva Pitt1, Wendy Kennedy1 and Margaret A Knowles*,1 1ICRF Clinical Centre, St. James's University Hospital, Beckett Street, Leeds, LS9 7TF, UK Deletion of all or part of chromosome 9q is the most and/or 9q and one of these genes might represent a common genetic alteration in all stages and grades of `gatekeeper' for urothelial cells (Kinzler and Vogelstein, bladder cancer. DBCCR1 (deleted in bladder cancer 1996). Detailed LOH studies of chromosome 9q in chromosome region candidate 1) maps to the chromosome TCC have indicated that there are several candidate region 9q32-33, a candidate tumour suppressor locus for tumour suppressor loci (Cairns et al., 1993; Dalbagni et bladder cancer. Although no mutations of DBCCR1 have al., 1993; Olumi et al., 1990; Simoneau et al., 1999), been detected in bladder tumours, expression of DBCCR1 one of which is at 9q32-33 (Habuchi et al., 1995, 1997; is silenced by promoter hypermethylation in 50% of Nishiyama et al., 1999b). The putative tumour bladder cancer cell lines analysed. Here we sought to suppressor gene in this region has been designated provide functional evidence to authenticate DBCCR1 as a DBC1 (deleted in bladder cancer 1). Using YAC and tumour suppressor using gene-transfer methods. Exogen- PAC contigs spanning the DBC1 region, we identi®ed a ous expression of DBCCR1 protein or an HA epitope- candidate tumour suppressor gene, DBCCR1 (deleted tagged fusion protein, HA-DBCCR1 in NIH3T3 cells and in bladder cancer chromosomal region candidate 1) human bladder tumour cell lines resulted in suppression of (Habuchi et al., 1998; Nishiyama et al., 1999a). proliferation. Cell cycle analyses in NIH3T3 cells DBCCR1 encodes a putative protein of 761 amino revealed that DBCCR1-mediated growth inhibition was acids with a predicted mass of 88.7 kDa. Mutation due to an increase in the number of cells in the G1 phase analyses of the coding region and Southern blot of the cell cycle. The levels of apoptosis were not altered. analyses detected neither somatic mutations nor gross These results demonstrate a role for DBCCR1 in cell genetic alterations in primary TCC. Although tran- cycle control, thereby supporting the hypothesis that this scripts of DBCCR1 are expressed in multiple normal is the tumour suppressor gene targeted by 9q32-33 human tissues including urothelium, its expression is deletion in bladder cancer. Oncogene (2001) 20, suppressed by hypermethylation of the 5' CpG island 2956 ± 2964. in 50% of bladder cancer cell lines investigated, suggesting that DBCCR1 may be the target gene Keywords: transitional cell carcinoma; DBCCR1; cell (Habuchi et al., 1998). The function of DBCCR1 cycle remains elusive since it has no signi®cant homology with known amino acid sequences and as yet no predicted structure. To address its candidacy as a Introduction tumour suppressor gene, we have used gene transfer methods to investigate the eects of DBCCR1 protein Transitional cell carcinoma (TCC) of the urothelium of expression on cell proliferation and apoptosis. the bladder, ureter and renal pelvis is amongst the commonest of human cancers with 12 000 new cases in the UK and 54 000 in the US per annum (Parker et al., Results 1997; Oce for National Statistics, 1996). In TCC, loss of heterozygosity (LOH) on chromosome 9q and/or 9p Polymorphisms of DBCCR1 is one of the most frequent genetic alterations (450%) and is detected consistently in all stages and grades of Previous studies identi®ed three silent polymorphisms tumours (Cairns et al., 1993; Dalbagni et al., 1993; in the coding sequence of DBCCR1 (Habuchi et al., Olumi et al., 1990; Wu et al., 1991) indicating that this 1997); T 1036 C (Ser -4Ser), C 2044 A (Ile -4Ile), and may represent an early event in tumour development. T 2642 C (Leu -4Leu). Further sequencing of cDNA This observation indicates that there are likely to be clones, human DNAs from the blood of normal important tumour suppressor genes for TCC on 9p volunteers and TCC tumour patients resulted in the identi®cation of four further single base changes; T 1459 A (Arg -4Ser), G 1491 A (Arg -4His), A 1727 G (Thr -4Ala), T 1759 C (Cys -4Cys). Analysis of *Correspondence: MA Knowles Received 7 September 2000; revised 16 January 2001; accepted 26 these changes in samples of normal volunteers and February 2001 cDNA clones revealed that three of the changes were Functional analysis of DBCCR1 H Nishiyama et al 2957 common polymorphisms; T 1459 A (33% of samples), supported by analysis of NIH3T3 cells expressing a A 1727 G (25%), and T 1759 C (60%). The fourth DBCCR1-GFP fusion in which the protein was also single base change (G 1491 A) was found only in expressed only in the cytoplasm (Figure 1d,e). cDNA clone ICRFp507K12270 (Habuchi et al., 1997) but not in 23 normal human genomic samples and six Inhibition of NIH3T3 cell proliferation by DBCCR1 cDNA clones analysed. For functional analyses, we used a cDNA sequence containing none of these The eect of DBCCR1 on cell proliferation was polymorphisms (wildtype) and a sequence containing investigated using NIH3T3 cells stably transfected with G 1491 A and A 1727 G (polymorphic). either epitope tagged HA-DBCCR1, non-tagged DBCCR1, antisense-DBCCR1, HA-DBCCR1-P (con- tains both G1491A and A1727G sequence variants), or Molecular weight of DBCCR1 protein vector alone (control). Cells were maintained in culture Based on the predicted protein sequence of DBCCR1, for 10 days following transfection before cell number the calculated mass was 88.7 kDa (Habuchi et al., and colony morphology were examined. Comparison 1997). To verify this, DBCCR1 cDNA was fused with of HA-DBCCR1 transfected cells with those trans- an HA epitope tag at the 5'-end, transfected into fected with vector alone (control) indicated that the NIH3T3 cells and analysed by Western blotting using majority of HA-DBCCR1 transfectant colonies were an anti-HA antibody. The relative mobility of HA- smaller in size than controls and showed altered DBCCR1 was shown to be approximately 100 kDa morphology (Figure 2a,b). In addition, the number of (Figure 1a, left). Since this was larger than the viable cells in HA-DBCCR1 transfectant populations predicted size, DBCCR1 was fused with green ¯uor- were about fourfold lower than those in control escent protein (GFP) at the 3'-end, transfected and cultures (Figure 2c). Similarly, transfection of non- examined by Western blotting with an anti-GFP tagged DBCCR1 and DBCCR1-P resulted in growth antibody. This detected DBCCR1-GFP at 130 kDa suppression compared to controls (Figure 2c). Con- and GFP alone at 30 kDa (Figure 1a, right), con®rm- versely, transfection with antisense-DBCCR1 did not ing the relative mobility of DBCCR1 to be 100 kDa. result in growth suppression (Figure 2c). Large colonies that were obtained following HA-DBCCR1 transfection were morphologically similar to controls Subcellular localization of DBCCR1 protein but when picked and expanded, did not show The subcellular localization of DBCCR1 was investi- expression of the HA-DBCCR1 fusion protein by gated using transient transfectants of NIH3T3 expres- Western blotting. Smaller colonies were also picked but sing HA-DBCCR1. Immuno¯uorescence studies less than half of these could be expanded and only half showed that DBCCR1 protein was localized primarily of those which were expanded showed HA-DBCCR1 in the cytoplasm (Figure 1b,c). This observation was expression (data not shown). Figure 1 Expression of DBCCR1.(a) Western blot analysis of NIH3T3 cells transfected with HA-DBCCR1 or DBCCR1-GFP fusion protein constructs. Ten mg of protein lysates were resolved by SDS ± PAGE, and detected with anti-HA antibody (left). Empty vector (pMKIT/HA) was transfected as a control (HA). DBCCR1-GFP was immunoprecipitated and analysed by Western blotting using anti-GFP (right). HA-DBCCR1 (white arrowhead); DBCCR1-GFP (black arrowhead); GFP (arrow). (b±e) Subcellular localization of exogenous DBCCR1. (b) Expressed HA-DBCCR1 was detected by immuno¯uorescence using anti-HA antibody followed by a FITC-labelled secondary antibody. (c) Phase contrast micrograph of the cells shown in (b). (d) DBCCR1- GFP transfectants. (e) Nuclei of cells shown in (d) stained using Hoechst 33258. White bar: 10 mm Oncogene Functional analysis of DBCCR1 H Nishiyama et al 2958 Figure 2 Suppression of NIH3T3 cell proliferation. (a) Methylene blue-stained colonies of HA-DBCCR1 transfectants after 9 days of G418-selection. Cells were transfected with either HA-DBCCR1 or control vector (HA). (b) Phase contrast micrograph of HA- DBCCR1 and control transfectants (HA) after 7 days of G418 selection. Bar: 50 mm. (c) Total cell numbers after 9 days of G418 selection. HA, control vector; DBCCR1, vector containing DBCCR1; HA-DBCCR1, HA-tagged DBCCR1; HA-DBCCR1P, HA- tagged polymorphic DBCCR1; AS, antisense orientated DBCCR1. (d) Reduction in the number of DBCCR1 transfectants with time. Cells were harvested at the times indicated post-transfection and analysed by ¯ow cytometry. The percentages of GFP- positive, co-transfected cells were measured. Each experiment was performed in triplicate, and a typical result depicted. Data was shown as mean+s.d. DBCCR1-induced growth suppression of NIH3T3 DBCCR1/HA-DBCCR1 transfectants demonstrated no cells was also demonstrated using a transient co- dierence in cell number compared to controls 24 h transfection system. In this assay, an expression after transfection, but a dramatic decrease in the construct containing GFP targeted to the endoplasmic proportion of ER-GFP positive cells was observed at 4 reticulum (ER-GFP) was co-transfected with either days post-transfection compared to controls (Figure 2d).