Physical and Transcript Map of the Region Between D6S264 and D6S149 on Chromosome 6Q27, the Minimal Region of Allele Loss in Sporadic Epithelial Ovarian Cancer

Physical and transcript map of the region between D6S264 and D6S149 on chromosome 6q27, the minimal region of allele loss in sporadic epithelial ovarian cancer

Ying Liu1, Gracy Emilion1, Andrew J Mungall2, Ian Dunham2, Stephan Beck2, Valerie G Le Meuth-Metzinger3, Andrew N Shelling4, Francis ML Charnock2 and Trivadi S Ganesan*,1

1ICRF Molecular Oncology Laboratories, John Radclie Hospital, Headington, Oxford, OX3 9DS, UK; 2Sanger Centre, Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SA, UK; 3INRA-UMR Genetique Animale, 65 Rue de Saint Brieuc, 35042 Rennes CEDEX, France; 4Research Centre in Reproductive Medicine, Department of Obstetrics and Gynaecology, National Women's Hospital, Auckland, New Zealand

We have previously shown a high frequency of allele loss chromosomal arm 6q initially de®ned the minimal region at D6S193 (62%) on chromosomal arm 6q27 in ovarian of allele loss between D6S149 and D6S193 (1.9 cM) in tumours and mapped the minimal region of allele loss one tumour (Figure 1) (Saito et al., 1992, 1996). between D6S297 and D6S264 (3 cM). We isolated and Subsequent studies have shown increased frequency of mapped a single non-chimaeric YAC (17IA12, 260 ± allele loss on 6q around the same region, although a 280 kb) containing D6S193 and D6S297. A further minimal region was not de®ned (Orphanos et al., 1995; extended bacterial contig (between D6S264 and D6S149) Wan et al., 1994). Based on the analysis of 56 malignant has been established using PACs and BACs and a ovarian tumours we showed previously that the minimal transcript map has been established. We have mapped region of allele loss on 6q27 is between D6S297 and six new markers to the YAC; three of them are ESTs D6S264 (3 cM) (Figure 1) (Chenevix-Trench et al., 1997; (WI-15078, WI-8751, and TCP10). We have isolated Cooke et al., 1996b). The allele loss was observed in all three cDNA clones of EST WI-15078 and one clone types of epithelial ovarian cancer. The maximal contains a complete open reading frame. The sequence frequency of allele loss occurred at D6S193 (62%) and shows homology to a new member of the ribonuclease D6S297 (52%). Three tumours showed loss of D6S193 family. The other two clones are splice variants of this only, while retaining ¯anking markers thus suggesting new gene. The gene is expressed ubiquitously in normal that the putative tumour suppressor gene was close to tissues. It is expressed in 4/8 ovarian cancer cell lines by D6S193. Thus the extended region of allele loss taking Northern analysis. The gene encodes for a 40 kDa the two previous studies into account suggests that the protein. Direct sequencing of the gene in all the eight putative gene lies within D6S264 ± D6S149 (approxi- ovarian cancer cell lines did not identify any mutations. mately 7.4 cM). Clonogenic assays were performed by transfecting the Further evidence that there might be a putative full-length gene in to ovarian cancer cell lines and no tumour suppressor gene around D6S193 is provided by suppression of growth was observed. FISH studies using YACs on direct metaphase spreads Oncogene (2002) 21, 387 ± 399 DOI: 10.1038/sj/onc/ from fresh ovarian tumours which suggested that this 1205067 change might be important even in early ovarian tumours (Tibiletti et al., 1996, 1998). It has also been Keywords: ovarian cancer chromosome 6q27 shown that the same region is implicated in a subset of lymphomas and breast cancer (Gaidano et al., 1992; Menasce et al., 1994; Rodriguez et al., 2000; Tibiletti et al., 2000). Introduction To identify the gene on chromosomal band 6q27 implicated in the pathogenesis of ovarian cancer, we Chromosome 6 has been implicated to contain a putative undertook a positional cloning approach involving tumour suppressor gene important in the pathogenesis of isolation of yeast arti®cial chromosomes (YACs), P1 ovarian cancer, both by karyotypic analysis and allele derived arti®cial chromosomes (PACs), bacterial arti- loss studies (Shelling et al., 1995). The analysis of 70 ®cial chromosomes (BACs) and construction of a malignant ovarian tumours using cosmids mapping to physical and transcript map around the key polymorphic markers. We have identi®ed seven genes in the interval between D6S264 and D6S149. We present the detailed analysis of a gene located telomeric to D6S193 *Correspondence: TS Ganesan; E-mail: [email protected] Received 26 March 2001; revised 5 October 2001; accepted 12 and D6S297, a member of the ribonuclease family October 2001 (RNASE6PL). Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 388

Figure 1 Combined genetic, physical and transcript map. The genetic map shown is as previously published (Cooke et al., 1996a). The Genethon map is indicated above with the integrated map below. The integrated map is shown with more markers than that on the Genethon map. The distance between markers is shown in centrimorgans (cM). The YAC corresponds to 17IA12 (260 ± 280 kb). Restriction enzymes are as shown. The STS markers D6S193, D6S297, D6S1585 and D6S133 were present in the YAC. The entire bacterial contig from D6S264 until D6S149 is shown with a gap between BAC 178P20 and PAC 431P23. The direction of the arrows of the PAC correspond to orientation. The sequence BAC 178P20 is incomplete although there is overlap with PAC 366N23. The orientation of PAC 431P23 has not yet been determined in relation to the contig. The key STS markers are shown in the bacterial contig. The genes identi®ed within the region are drawn to scale in relation to the bacterial contig and shown below the contig. TCP10 is shown twice as exons 8 and 9 are present in an antisense orientation in the PAC 366N23. The AF-6 gene is present in part in PAC 431P23

were evaluated by PCR and three ESTs (WI-8751, WI- Results 15078 and TCP10) were identi®ed to be located in the YAC, in addition to the STS markers D6S193 and Genetic and physical mapping D6S297. We had previously constructed an integrated genetic Simultaneously, PACs and BACs were isolated from map of chromosome 6q24 ± 6q27 using the Genethon whole genome libraries (Ioannou et al., 1994; Shizuya map as a backbone (Figure 1) (Cooke et al., 1996a). et al., 1992) for all the markers in this region and were This map improved the number and order of markers assembled into a contig by STS content and within 6q24 ± 6q27 compared to the Genethon map. ®ngerprinting (Mungall et al., 1997). The orientation We initially screened the ICRF, CEPH and ICI YAC of the minimal overlapping set of bacterial clones in libraries by hybridization and PCR with markers from relation to the genetic map and YAC map is shown in this region. Several YACs were isolated and character- Figure 1. The PAC RP1-167A14 contained STSs ized for sequence tagged site (STS) content, size and stSG11162, stSG17631 that were not present in the integrity using FISH. Three YACs were identi®ed YAC 17IA12. This suggested that there was an internal which contained the relevant markers D6S193 and deletion in the YAC corresponding to these markers. D6S297 from the ICI YAC library. One non-chimaeric The bacterial contig is not yet complete as there is still YAC 17IA12 (ICI YAC library) was shown to contain a gap between RP11-178P20 and the PAC that D6S193, D6S297, D6S1585 and D6S133. A detailed contains the marker D6S149 and part of the gene long-range restriction map of this YAC (260 ± 280 kb) AF6 (RP3-431P23). was constructed and an overlapping contig of cosmids In order to facilitate the identi®cation of genes was made. Thirty ESTs/STS mapping in the region within this region between D6S264 ± D6S149, sequen-

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 389 cing of the entire contig was undertaken. Except for testis (data not shown) we did not characterize this the BAC RP11-178P20, which has not yet been gene further in ovarian cancer (Islam et al., 1993). sequenced to completion, linear overlapping sequence is available from PAC RP4-655C5 until the PAC RP3- WI-15078 The third EST which was identi®ed and 366N23. mapped to the YAC (WI-15078, IMAGE clone 174311, size 1049 bp) was sequenced in its entirety. We used this clone as a probe and isolated two larger partial Transcript map cDNA clones (IMAGE clones 199297 1332 bp, 246630 Several methods were used to identify coding sequences 1195 bp) after screening of cDNA libraries. The three within the region. They included homology search of clones were completely sequenced in both strands. the GenBank/EMBL databases using BLAST algo- Comparison and analysis of the cDNA sequences of all rithms and a suite of exon prediction programme to the three clones suggested that there was an open identify genes (Altschul et al., 1990). In addition, reading frame in clone 246630 of 242 amino acids that experiments were also performed to assess ESTs encoded for a protein of 29 kDa. A search of the mapped within the interval between D6S264 ± D6S149 database showed that the sequence was homologous to and so far none have yielded new bona ®de genes. The the family of ribonucleases (Hime et al., 1995). This order of genes identi®ed so far is D6S264 ± p90Rsk-3 ± family is characterized by two conserved motifs RNASE6PL ± FOP ± CCR6 ± GPR31 ± TCP10 ± HUnc93 ± `IHGWLP' and `KHGTC' that span the active site of D6S149 ± AF6. the enzyme (Kurihara et al., 1996). This gene is the human equivalent and was the ®rst reported member of this family (Trubia et al., 1997). A comparison of the Candidate genes amino-acid sequence of the human protein with that of WI-8751 Initially, we focused on the three ESTs that other species is shown (Figure 3a). had been mapped to the YAC before the PAC contig and the sequence data were available. The cDNA clone Intron-exon structure of the ®rst EST WI-8751 (IMAGE clone 68587, size 1.1 kb) was obtained, sequenced and used to screen We compared the full-length cDNA sequence of clone cDNA libraries for obtaining full-length cDNA clones. 246630 against the genomic sequence contained within After extensive screening of several cDNA libraries (12) PAC RP3-505P2 and BAC RP11-514012. This showed we were unable to isolate any full-length clones. The that the gene is coded by nine exons spread over 27 kb full-length sequence of the cDNA clone of EST WI- (Figure 3b, Table 2). Comparison of the other two 8751 did not contain an open reading frame. Further, cDNA clones (199297 and 174311) suggested that these on Northern analysis of multiple tissues using this two were splice variants of the same gene. Interestingly, clone as probe we were unable to identify any both clones 199297 and 174311 contained unique exons transcript (data not shown). On comparison with the (5a, 791 bp; exon 5b, 46 bp) which are not present in sequence contained in BAC RP11-514O12 that over- the cDNA clone 246630. In clone 199297, exon 5a has lapped with PAC RP3-505P2, it was clear that the an in-frame stop codon. This exon has sequence which entire coding sequence of EST WI-8751 was contained is identical to exon 5b in clone 174311, the intervening as one linear sequence without any intron ± exon sequence and exon 5 of clone 246630. The size of the boundaries. We also did not identify any promoter full-length cDNA for clone 199297 was 1954 bp. The sequences in the genomic sequence 5' of the cDNA. predicted peptide of this splice variant contains 124 Although it is possible that the EST WI-8751 is amino acids. In clone 174311, the novel exon 5b is expressed at low levels in a speci®c tissue, it is most followed by exon 6a and exon 5 is spliced out. Exon 6a likely to be a pseudogene. contains sequence identical to exon 6 of clone 246630, followed by intervening sequence which has an in- TCP10 The second EST which we mapped to the frame stop codon. The size of the full-length cDNA for YAC was a previously identi®ed gene TCP10 (T complex protein 10A) (Blanche et al., 1992). Interest- ingly, on the bacterial contig the gene mapped quite far Table 1 Intron ± Exon structure of TCP10 away on BAC RP11-517H2. As the size of the YAC was 260 ± 280 kb, and did not contain markers within Exons Nucleotide 5' splice site 3' splice site PACs RP4-505P2 and RP1-167A14, this suggested that 1 1 ± 76 tgaggGGGGAGG ACCACgtgagg there was a large deletion within the YAC 17IA12 2 77 ± 221 aaccagGATGCTG ATGCCGgtgag corresponding to these sequences. This gene (26 kb) is 3 222 ± 437 atgacagTCATTAC CACTCTGgtaccaa 4 438 ± 575 ttttagGCATTGG TCCAAAGgtgac encoded by nine exons and they are present on BAC 5 576 ± 734 ttttcagCCAATGA GAAAGgtgacct RP11-517H2. Unusually, the last two exons are also 6 735 ± 846 gccccagATTATG GGAGCTGgtgagca present in PAC RP3-366N23 which is nearly 150 kb 7 847 ± 948 tttttagACAGAA GGCAGGtaggtg away (Figure 2). Further, exon 8 is telomeric to exon 9 8 949 ± 1066 tctgcagAAAAATTCT TGGAAAGgtcag and is transcribed in the antisense orientation. The 9 1067 ± 2190 cctctcagGAACGG AAAAgaattc intron-exon structure is as shown in Table 1. As the The exon location corresponds to TCP10 cDNA sequence (Accession expression of this gene was exclusively limited to the number: U03399). Intron sequences are shown in lower case

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 390

Figure 2 Genomic structure of TCP10. Schematic diagram showing exon structure of TCP10 cDNA related to genomic bacterial clones. `cds' represents translated sequence. The ®rst eight exons and part of exon 9 (up to 1490 of TCP10 sequence) is present in BAC 517H2. The exons 8 and 9 are also present in PAC 366N23 but in reverse orientation

clone 174311 was 1671 bp. The predicted peptide from band of approximately 4 kb was detected in the cell this splice variant contains 240 amino acids. Alignment lines OAW42 and SKOV-3, by all three probes, this of the amino-acid sequence from all of the three splice probably does not represent a true transcript but variants is shown in Figure 3c. Thus, the two splice hybridization to the 28S RNA which is of the same variants 199297 and 174311 produce truncated pro- size. Expression of the splice variant 246630 was teins. The splice variant 199297 was con®rmed by RT ± subsequently con®rmed by the more sensitive RT ± PCR using primers speci®c to exon 5a (Figure 3d). We PCR in all of these eight cell lines (Figure 4c). were unable to con®rm the existence of the second In order to determine the size of the protein encoded splice variant 174311 by using primers speci®c for exon by the full-length clone 246630, we transfected a HA- 5b and 6a by RT ± PCR on mRNA from dierent epitope tagged construct in Cos-7 cells. This showed a tissues. fusion protein of 40 kDa (Figure 4d). The molecular weight of the fusion protein was higher than that predicted (29 kDa) from the open reading frame. This Expression suggests that post-translational modi®cations may We examined the expression of the three splice variants account for the increase in the size of the mature protein. in various tissues using the full-length probe for clone 246630 and the original partial cDNA clones for Clonogenic assays 199297 and 174311 which contain unique sequences speci®c to these cDNAs. All of the three splice variants As the RNASE6PL gene was close to the markers are expressed in all tissues with maximal expression in D6S193 and D6S297 and had variable expression in pancreas and lymphocytes. The major message is cell lines, we sequenced the coding region in its entirety 1.35 kb. In addition, in clone 174311, a larger in all of the eight cell lines. The sequence did not show transcript of 2.5 kb is expressed in heart and muscle. any variation compared to the wild type sequence. To There were minor larger transcripts detected with clone further assess whether the RNASE6PL gene was a 199297 which are probably due to cross reactivity with potential candidate tumour suppressor gene, we another species. There was variation in the level of the transfected the cDNA (HA epitope tagged 246630 major transcript detected by all three cDNA's in the clone) into the four ovarian cancer cell lines (UC101, dierent tissues, particularly by cDNA clones 174311 PEO1, OAW28 and 41M) and assessed colony and 246630. All of the three splice variants were formation in the presence of the selectable marker expressed in the ovary (Figure 4a). G418. In the two cell lines PEO1 and 41M, where there Expression of all of the three splice variants were was relatively little expression of the gene on Northern examined in a panel of eight ovarian cancer cell lines analysis, there was no dierence in the number of by Northern analysis. There was variable expression of colonies compared to the vector transfected clones the major transcript of 1.35 kb in four (OAW28, (Figure 5). In the other two cell lines where the OAW42, SKOV-3, UC101) of the eight cell lines and RNASE6PL gene was expressed, there was no on prolonged exposure at a lower level in the dierence in the number of colonies when compared remaining cell lines (Figure 4b). Although a higher to vector transfected controls either.

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 391 between D6S264 and D6S149 was 7.4 cM. To identify Other candidate genes within the internal the putative tumour suppressor gene on 6q27, we D6S264 ± D6S149 undertook a positional cloning approach. Although FOP The gene FOP (FGFR1 Oncogene Partner) was our initial strategy was based on YACs, chimaerism identi®ed to be wholly contained within the PAC RP1- proved a particular diculty in assembling an over- 167A14. It was isolated as a fusion partner of a rare lapping contig within this interval. We identi®ed one translocation in leukaemia (Popovici et al., 1999). The YAC 17IA12, which on FISH was non-chimaeric and gene is expressed in the ovary (data not shown) as contained D6S193 and D6S297. However, with the previously published and was uniformly expressed in availability of bacterial genomic libraries (PAC and all the ovarian cell lines (Figure 6). In addition, BAC), the establishment of a physical contig was transfection of the full-length cDNA into ovarian pursued using these resources. To accelerate the cancer cell lines and evaluation in a clonogenic assay identi®cation of genes within this interval, we initially showed there was no suppression of growth compared mapped all ESTs within 6q27 in the database to the to vector control (data not shown). We have not YAC 17IA12 and subsequently to the PACs. We then analysed the gene further for mutations. undertook direct sequencing of the entire PAC/BAC contig as part of the genome project. The entire contig CCR-6 and GPR31 These two genes have been from D6S264 until D6S149 has one gap between RP11- isolated previously (Baba et al., 1997; Greaves et al., 178P20 and RP3-431P23. We estimate this to be of 1997; Liao et al., 1997a; Zingoni et al., 1997). They one-two PAC/BAC length. The overall linear sequence encode for chemokine receptors and are expressed available within this interval is approximately 1.1 Mb. primarily in lymphoid cells by Northern analysis. Comparison of the YAC map with that of the order of Further, they are contained in the BAC RP11-517H2 STS/genes on the bacterial contig suggested that there as one linear sequence without any introns, a was an internal deletion of sequence within the YAC characteristic of other chemokine receptors. Although corresponding to the PAC RP1-167A14. it was previously reported that CCR-6 is coded by two The primary method of identi®cation of genes was exons (Liao et al., 1997a), comparison of the cDNA by comparison with GenBank/EMBL databases. In sequence with the genomic sequence showed the addition, suites of programs were used to predict absence of any introns. In the absence of their coding sequences. Overall, seven genes were identi®ed expression in the ovary (data not shown) we have within this interval. In addition, several EST matches not analysed these genes further in ovarian cancer. have been identi®ed but we have not been able to show The human homologue for Unc-93 which is that any of them represent bona ®de genes experimen- contained within PAC RP3-366N23 was examined in tally. detail (Liu et al., manuscript submitted) in ovarian As the RNASE6PL gene was located telomeric to cancer. Similarly, p90 Rsk-3 which is contained within D6S193, it was investigated in some detail. Members of PACs RP3-497J21 and RP1-168L15 is the basis of a the Rh/T2/S-Glycoprotein ribonuclease gene family are separate report (Emilion et al., manuscript submitted). found predominantly in fungi, plants, and bacteria, AF-6 which is partly contained within RP3-431P23 has where they have been implicated in functions such as the been previously examined by FISH and analysed for phosphate-starvation response and self-incompatibility. mutations in ovarian cancer (Saha et al., 1995; Saito et Self-incompatibility, the ability that stops a plant's al., 1996). It maps outside the interval of deletion at pollen from fertilizing its own pistel, is controlled by a chromosome 6q27 and no mutations have been single multi-allelic locus in Nicotiana alata (Anderson et identi®ed in ovarian tumours. al., 1989; Ioerger et al., 1990; McClure et al., 1989). In Nicotiana alata, this locus encodes proteins termed the stylar glycoproteins (S-glycoprotein) which exhibit Discussion homology with two of the most studied families of extracellular fungal RNAses T2 and Rh, from Aspergillus The clue to the location of a putative tumour oryzae and Rhizopus niveus (McClure et al., 1989). suppressor gene at 6q27 was primarily provided by Therefore this gene family is termed as Rh/T2/S- allele loss studies using polymorphic markers on Glycoprotein ribonuclease. These `stylar RNAses', serve chromosomal arm 6q (Cooke et al., 1996b). This was as a plant immune system, aiding in maintaining genetic supported by FISH studies performed directly on diversity and strength of the species (Sassa et al., 1996). metaphase spreads from ovarian tumours using YACs The ®rst member of this family found in invertebrates (Tibiletti et al., 1996, 1998). Precise delineation of the was the DmRNAse-66B gene in the Drosophila melano- minimal interval of allele loss was shown to be between gaster genome in 1995 (Hime et al., 1995). Ribonucleases D6S264 and D6S297 by us previously, which over- (RNAses) are a group of enzymes of varying speci®ty to lapped with that between D6S193 and D6S149 (Saito control the RNA population post-transcriptionally. et al., 1992). Thus, the extended interval was Once regarded as ubiquitous `housekeeping enzymes', established to be between D6S264 and D6S149. The they are now recognized to control processes ranging high frequency of allele loss at D6S193 was con®rmed from splicing to organogenesis (Schein, 1997). It has by analysis of a dierent set of samples from Australia been known that both plant and mammalian RNAses (Chenevix-Trench et al., 1997). The predicted distance are produced in response to stress (Ye and Droste, 1996),

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 392 and pathogenic agents, particularly viruses (Walter et al., The gene encoding for a novel ribonuclease 1996). RNAse active in greenhouse wheat increases (RNASE6PL) was unusual in several aspects. Firstly, during senescence, as does RNAse activity in human as previously reported, it is the ®rst human homo- tissues (Francesconi et al., 1984). The role of RNAses in logue of an extra-cellular ribonuclease observed the mammalian immune system has not yet been primarily in plants (Trubia et al., 1997). The predicted investigated. protein has two motifs, which are completely

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 393

Figure 3 (a) Alignment of sequences of the ribonuclease gene family across species. The peptide sequence of the clone 246630 was compared with peptides from other species. The two conserved motifs `IHGLWP' and `KHGTC' are highlighted. Alignment was performed using PileUP and displayed using MacBoxshade. Accession numbers of all sequences are shown. (b) Schematic diagram showing the exon distribution and the alternative splice variants of the RNASE6PL gene. Exons in all of the splice variants are shown as rectangles. The exons are shaded in black (phase 1) stippled (phase 2) and in grey (phase 3). In clone 246630 there are nine exons and the primers used to amplify for RT ± PCR are indicated. M1 refers to the ®rst motif `IGHLWP' and M2 to the second motif `KGHTC'. Individual exon sizes are shown below. In clone 199297, a novel exon 5a (791 bp) is present with an in-frame stop codon. The remaining exons (6 ± 9) are also present in the cDNA. The primers used to amplify 199297 cDNA are indicated. In clone 174311, a novel exon 5b (46 bp) and exon 6a (1004 bp) is present and the rest of the exons (7 ± 9) are absent. Exon 5a in clone 199297 comprises of identical sequence present in 5b with intervening sequence and exon 5. Exon 6a comprises of exon 6 and subsequent intervening sequence. (c) Alignment of the predicted peptide sequence of three splice variants. The identical sequences between all three splice variants are shown in red which comprises of sequences translated from the ®rst four exons. The amino acid identity between splice variants 199297 and 174311 is shown in blue which comprises of sequence translated from sequence present in exons 5a and exon 5b. Alignment was performed using Pileup and displayed using MacBoxshade. (d) Expression of splice variant (199297). Ethidium bromide-stained agarose gel of ampli®cation products from the RT ± PCR. The template was the total mRNA of eight ovarian cancer cell lines (2 mg each) and the primers were designed speci®c for 199297 cDNA (199297F1 and 246630R2). Top, a fragment of correct size (644 bp) was obtained by RT ± PCR in the indicated cell lines. Bottom, as a positive control for cDNAs' integrity, a portion of GAPDH coding region was ampli®ed by RT ± PCR using the RNAs from the same panel of cell lines (2 mg each). Lane (a) was the negative control for the RT ± PCR without any template. Lane (b) was the negative control without reverse transcriptase

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 394

Figure 4 (a) Expression of RNASE6PL in normal adult tissues. Northern blots containing poly(A)+ mRNA (2 mg each lane) from adult tissues (Clontech) hybridized to the three splice variants respectively. The membranes were washed at high stringency and exposed to Fuji ®lm at 7708C for 4 ± 7 days. One major transcript of 1.35 kb was observed when the blots were hybridized with the three cDNA probes 199297, 246630 or 174311, while an additional transcript of 2.4 kb was detected with hybridization to 174311 (marked by arrow). The position of RNA size standard is indicated for each blot. (b) Expression of RNASE6PL in ovarian cancer cell lines. Northern blots of total RNA from eight ovarian cancer cell lines (10 mg each) prepared in the same manner were hybridized to the three splice variants respectively. The membranes were hybridized, washed and exposed to Fuji ®lm at 7708C for 4 ± 7 days. The expression pattern of these three splice variants in these eight cell lines were similar in that: the major transcript of 1.35 kb was most abundant in cell lines OAW42 and SKOV3, less abundant in cell lines OAW28 and UC101, and not detectable in

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 395 conserved in evolution suggesting that they are However, it was recently reported that RNASE6PL functionally important. Indeed, the motifs, IHGWLP might be the putative gene for senescence and tumour and KGHTC have been shown to be the catalytic site suppressor gene in ovarian cancer (Acquati et al., by resolution of the tertiary structure (Kurihara et al., 2001). Although the authors found loss of expression 1996). The two histidines in these motifs are in 30% of ovarian tumours and 75% of cell lines, the completely conserved suggesting that there is func- expression was not completely absent but reduced tional conservation across species. Secondly, the two which was similar to our results. Further they did not splice variants are unusual in that the translation perform RT ± PCR, a more sensitive method of proceeds into the intervening sequence in frame and analysing expression of RNASE6PL. Similar to our terminates with an in-frame stop codon. The expres- results, they did not identify mutations within the sion of the three splice variants is broadly similar in coding region of the gene either. Instead of clonogenic that a major transcript of 1.3 kb is present in all assays, they established stable cell lines by transfection tissues although of varying levels. However, there was of RNASE6PL (HEY4, an ovarian cancer cell line a larger transcript expressed in the heart and skeletal which shows 10% expression of this gene compared to muscle (2.5 kb) by the splice variant 174311. The size that in normal ovarian tissue; SG10G, an ovarian of the cDNA clone 246630 correlates with the major cancer cell line has lost chromosome 6; XP12ROSV, a message of 1.35 kb. However, the correlation of SV40-immortalized ®broblast cell line having no mRNA message size with clones 199297 and 174311 expression of the gene; pRPc, a mouse cell line derived is less clear. Although only four of the cell lines had from a tumour induced in nude mice by the human/ expression of all three transcripts, longer exposure mouse monochromosomal hybrid harbouring a dele- revealed faint bands in the remaining cell lines. The tion at 6q27). Examination of their data did not show expression of the full-length cDNA encoded by clone statistically signi®cant dierence between propagated 246630 was con®rmed by the more sensitive RT ± live colonies and control (in HEY4, 60.2%, vs 100%; PCR in all of the cell lines. The full-length transcript in SG10G, 72.1% vs 100.0%; in XP12ROSV, 10.3% vs was expressed in an epitope tagged vector and 66.7%; and in pRPc, 86.2% vs 100%), except possibly encoded for a protein of approximately 40 kDa. The in XP12ROSV. Therefore, the results from this increase in size compared to that predicted is clonogenic assay was again similar to our results in probably accounted by post-translational modi®ca- that the colony formation was reduced, but not tion. To exclude the possibility that the gene might be signi®cantly, after the transfection of this gene when mutated in ovarian cancer cell lines, we sequenced the compared to the vector control in ovarian cancer cell coding region in its entirety and it was completely lines. We did not perform nude mice tumorigenicity normal. The clonogenic assays provide further sup- experiments as our clonogenic assays did not suggest port that this gene is unlikely to be a putative tumour that RNASE6PL inhibited cell growth. Further, even suppressor gene. in their experiments there was no signi®cant reduction

Table 2 Intron ± Exon structure of RNASE6 PL Exon Exon* Exon size (bp) location 5' splice site 3' splice site

1 448 1 tgggagcaac/GCGACTG CGCCTGCG/gtgagttacc 2 61 457 gttttaacag/TGACAACC TATGCGAG/gtgagtgttt 3 56 518 ttctccctag/AAAATTCA GGACTATG/ggcaagtgtc 4 58 574 gaatttccag/GCCCGATA AGAGATTAAG/gtattgactt 5 71 632 taatgcatag/GATCTTTT CGCTTCTG/gtaagtccca 6 114 703 atcctcacag/GAAGCATG CTCAACAG/gtgggtgcgc 7 46 817 taaattttag/TGTGCTTCTA ACTACCAA/gtaagtcttg 8 75 863 ttaaaatcag/GTTGCAGAT CAAGCCAG/gttagacagt 9 268 938 ttacttttag/GATGAGGAAG GCAGCTTTCT/3'UTR

*The exon location corresponds to the 246630 cDNA sequence. Intron sequences are shown in lower case

cell lines 41M, 59M, PEO1 and PEO4. (c) Expression of the RNASE6PL gene in eight ovarian cancer cell lines by RT ± PCR. Ethidium bromide-stained agarose gel of ampli®cation products from RT ± PCR. The template was the total RNA of eight ovarian cancer cell lines (2 mg each) and the primers were designed from the coding sequence of 246630 (246630F2 and R2). Top, a fragment of correct size (292 bp) was obtained by RT ± PCR in the indicated cell lines with the ®rst lane as a template negative control. Bottom, as a positive control for RNA integrity, a portion of GAPDH coding region was ampli®ed by RT ± PCR using the RNAs from the same panel of cell lines (2 mg each). (d) Expression of RNASE6PL in Cos cells. The ORF of 246630 was subcloned into an epitope tagged mammalian expression vector (pMH) (Boehringer Mannheim). Cos-7 cells were transiently transfected with the 246630-pMH construct, and pMH vector respectively, immunoprecipitated with anti-HA antibody and detected with HA-HRP. A speci®c band around 40 kDa (arrow) was detected in the cells transfected with 246630-pMH, but not in the cells transfected with pMH vector

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 396 366N23. This might represent a duplication of this gene. This remains to be veri®ed, as there is a gap in the contig after RP3-366N23. Of the remaining genes, CCR6 and GPR31 were not expressed in the ovary as previously published (data not shown) (Baba et al., 1997; Greaves et al., 1997; Liao et al., 1997b; Zingoni et al., 1997). FOP, which was isolated as a translocation partner in a rare type of leukaemia (Popovici et al., 1999), was expressed uniformly in all the ovarian cell lines. Further, in the clonogenic assay, there was no suppression in the growth of ovarian cancer cell lines upon transfection of full-length FOP cDNA. This suggests that it is unlikely to be a putative tumour suppressor gene. It is possible, however, that missense mutations may occur in ovarian tumours which do not interrupt translation. Figure 5 Clonogenic assay. Ovarian cancer cell lines UC101, In ovarian cancer, only P53 and PTEN have been OAW28, PEO1 and 41M were analysed. In all the four cell lines, shown to be mutated (Obata et al., 1998; Shelling et there was no signi®cant dierence in the number of colonies al., 1995). P53 mutations are observed in all subtypes between transfection of vector and wild type RNASE6PL cDNA. Open bar represents vector transfected cells and hatched vector of cancer and are more common in advanced stages of plus RNASE6PL cDNA. Standard error bars are shown. The the disease. PTEN, in contrast is mutated in only 20% results are the mean of three separate experiments of endometrioid type of epithelial ovarian cancer. The majority of genes identi®ed as potential tumour suppressor genes in ovarian cancer either by positional cloning or dierential mRNA display have not been shown to harbour mutations in ovarian cancer although they do suppress growth in clonogenic assays (Liu and Ganesan, 2001). The mechanism of loss of expression of the putative gene has been either by methylation of the promoter or imprinting. Indeed, the genes important in hereditary forms of ovarian cancer such as BRCA1, BRCA2 or the mismatch repair genes have not been shown to be mutated frequently in sporadic ovarian cancer, thus deviating from the classical model (Katso et al., 1997). However, BRCA1 is not expressed in a proportion of tumours due to methylation of the promoter (Catteau et al., 1999; Mancini et al., 1998). Thus mechanisms other than Figure 6 Expression of FOP. Twenty mg of total mRNA from a mutations might be more common in the inactivation panel of ovarian cell lines was separated on a formaldehyde gel, transferred and probed with full-length cDNA probe for FOP of tumour suppressor genes in ovarian tumours. This (Popovici et al., 1999). The expected transcript is as shown by nearly completed integrated physical and transcript arrow and both transcripts are expressed in all cell lines map in a 1 Mb region at chromosome 6q27 has been instrumental in evaluation of candidate tumour suppressor genes important in ovarian cancer.

in the number of tumours in mice. The principal results was that tumour xenografts showed loss of expression of the transfected RNASE6PL gene. However, other Materials and methods genes in the region between D6S264 and D6S149 were not examined to evaluate whether there was loss of Cell lines their expression in these tumours. Ovarian cancer cell lines (Laval et al., 1994), and Hela, Cos-7 The remaining ESTs identi®ed on the YAC were WI- cell lines were supplied by Cell Service, Clare Hall, ICRF. 8751 and TCP10. WI-8751 was unlikely to represent a The cells were cultured in the media with 10% FCS and bona ®de gene, as not only expression was absent on grown at 378C in the presence of 5% CO2. conventional Northern blots but also there was no de®ned intron ± exon structure or an open reading Preparation of yeast, YAC DNA blocks and PFGE frame. TCP10 was again not examined further because YAC DNA in yeast cells was prepared in agarose blocks for of its lack of expression in the ovary. It has an unusual FISH and Pulsed Field Gel Electrophoresis (PFGE). Intact intron ± exon structure. Although all the nine exons of chromosomal YAC DNA was generated by the Lithium TCP10 were contained within the BAC RP11-517H2, method (Anand et al., 1990) in which whole yeast cells are the last two exons were also present in the PAC RP3- embedded in agarose blocks before spheroplasting, lysis and

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 397 deproteinization. Two per cent (w/v) low melting gel in 1 M extension (728C, 10 min). Primer sequences for expression of sorbitol and 20 mM EDTA was dissolved by boiling, and 246630 in cell lines were: Forward primer (244630F2) 5'- cooled to 508C in a water bath, and 14 mM of 2-ME was taatagatcttggcccttca-3' and reverse primer 5'-caagggcatctt- added. Block moulds (Pharmacia) were laid out on a glass taaaatctgc-3' (246630R2). Primer sequences for expression plate placed on ice. Yeast cells from 50 ml culture were of 199297 were (199297F1) 5'-cgtcgttggaatcatacag,-3' and pelleted at 1000 r.p.m. for 10 min. The pellet was resus- (246630R2) 5'-caagggcatctttaaaatctgc-3'. pended in 400 ml of solution I (1 M sorbitol, 20 mM EDTA, 14 m 2-ME) and mixed with 500 ml of the molten agarose M Sequencing and dispensed into the slots of the block mould on ice. The set agarose inserts were released in a 50 ml Falcon tube Plasmids and genomic DNA were sequenced using BigDye containing 5 ml of 1 M sorbitol, 20 mM EDTA, 14 mM 2- Terminator Cycle Sequencing Ready Reaction kit (Perkin ME, 10 mM Tris-HCL, pH 7.5, 2 U/ml Zymolyase and Elmer) on the Applied Biosystems model 373 DNA incubated at 378C for 2 h. The agarose inserts were incubated sequencing system (PE Applied Biosystems). One mgof in the lysis solution (1% lithium dodecyl sulphate, 100 mM plasmid DNA, 10 pmol of primer and 4 ml of Terminator mix EDTA, 10 mM Tris-HCL, pH 8.0) at 378C for 1 h and stored were used for a 12 ml sequencing reaction. Primers from both in TE at 48C until loading on a gel for sizing and purifying 5' and 3' sides were used, while primers were also designed YAC DNA fragment from agarose gel using standard from cDNA sequence when necessary. A Thermal Cycler-480 protocol. PFGE was performed according to standard (Perkin Elmer) was used to carry out the 25 cycles of protocol. denaturation (968C, 30 s), annealing (508C, 15 s), and extension (608C, 4 min). The DNA was precipitated, spun down, washed, resuspended in loading buer, denatured, and cDNA library screening loaded onto ABI 373 Sequencher according to manufac- cDNA libraries of human brain, liver, lung (all foetal) and turer's instruction. placenta were available as membranes with immobilized cDNA clones (Resource Centre/Primary Database of the Construction of epitope tagged RNASE6PL German Human Genome Project). cDNA probes were labelled with 32P-dCTP using MegaprimeTM labelling systems Primers (with restriction sites) were designed to amplify the (Amersham). The membranes were prehybridized in 20 ml 246630 open reading frame (ORF). It was subcloned into the Churchill & Gilbert buer (0.22 M NaH2PO4, 0.35 M epitope tagged vector pMH (Boehringer Mannheim) with the Na2HPO4, 7% SDS, 1 mM EDTA) 658C for 2 h and then cDNA in frame with the HA tag at the 3' end. Primer hybridized at 658C with denatured probe in the same buer sequences to amplify the full-length cDNA and clone it in overnight. The membranes were washed with 26SSC/0.5% frame were 5'-ctgaaagcttcaggtcggcaccatgcgc-3' and 5'-tcac- SDS for 15 ± 20 min at 658C twice, with 0.16SSC/0.1% SDS gaattctcgcaatgcttggtctttttaggtgg-3'. for 30 min once at 658C. Immunoprecipitation and immunoblotting Northern analysis Cos-7 cells on 100-mm plate were transfected by the 246630- Total cellular RNA was isolated from 5 ± 106107 cells of ORF-pMH construct with Eectene (QIAGEN) according to each cell line with RNAeasy Midi/Maxi kit (QIAGEN) manufacturer's instructions. After incubation for 48 h at according to manufacturer's protocol. Ten mg of RNA from 378C, transfected cells were lysed by adding 1.0 ml lysis each cell line was separated in 1% denaturing formaldehyde/ buer (1% Triton, 0.5% NP40, 1 mM EDTA in phosphate- agarose gels and immobilized to Hybond-N+ (Amersham) buered saline PBSA, which contains 137 mM NaCl, 2.7 mM membrane by standard capillary blotting in 206SSC over- KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4,) containing freshly night and baking at 808C for 2 h. Membranes were added protease inhibitors (1 mg each of Leupeptin and prehybridized in ExpressionHybTM Hybridisation Solution Pepstin per ml, 10 mg each of Aprotinin and Trypsin per (Clontech) at 688C for 1 ± 2 h. 174311, 1992 ± 97, 246630 ml, and 0.5 mM of Phenylmethylsulfonyl Floride per ml) at cDNA probes were labelled with 32P-dCTP using Mega- 08C for 10 min. Insoluble material was removed by primeTM labelling systems (Amersham) and puri®ed through centrifugation for 30 min at 11 000 r.p.m. at 48C. The cell G-50 Sephadex columns. Membranes were hybridized in lysate was incubated with 1 mg anti-HA antibody (Roche) at ExpressionHybTM Hybridisation Solution (Clontech) at 688C 48C for 3 h and then absorbed with 25 ml of protein G- with recommended amount of denatured probe for overnight agarose beads at 48C overnight. The immune complexes were and washed according to manufacturer's protocol. The washed twice in ice-cold cell lysis buer supplemented with probes used for labelling and hybridization were: full-length protease inhibitors and then in 0.5 M NaCl, supplemented cDNA 246630, partial cDNA of 199297 containing exons 5a with protease inhibitors. The immunoprecipitates were eluted up to 9, partial cDNA of clone 174311 containing exons 5b and denatured by boiling for 5 min in sodium dodecyl and 6a, and for FOP full-length cDNA. sulphate (SDS) sample buer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 5% b-mercaptoethanol, 0.25% bromo- phenol blue). The immunoprecipitates were resolved by 12% RT ± PCR SDS polyacrylamide gel electrophoresis (PAGE) and then RNA isolated from each cell line as described above was used transferred to PVDF membrane (Amersham) according to in a RT ± PCR reaction using QIAGEN One Step RT ± PCR standard protocol. The membrane was blocked for 2 h in Kit (QIAGEN) according to the manufacturer's protocol. All PBSA (5% nonfat milk, 0.1% Tween 20), rinsed, with PBSA the reactions were carried out on a PTC-200 Thermocycler (0.1% Tween 20), and then incubated with Anti-HA- (GRI Lab Care Service) for the reverse transcription (508C, peroxidase (1 : 1000 dilution after reconstitution as 25 u/ml) 40 min), initial PCR activation step (958C, 15 min), 40 cycles (Boehringer Mannheim) for 1 h in PBSA (0.5% nonfat milk, of denaturation (948C, 40 s), annealing (Tm for each pair of 0.1% Tween 20). The membrane was washed in PBSA (0.5% primers, 1 min), extension (728C, 1 min), and 1 cycle of ®nal nonfat milk, 0.1% Tween 20) for 50 min with changing

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 398 solution every 10 min. The proteins were detected by the RP11-366H19:AL353591, RP11-568A1:AL353747, RP3-366N23: Western blotting detection reagents ECLTM (Amersham). AL021331, RP3-431P23:AL009178, RP11-178P20: AL592444. The accession numbers for cDNA sequences are, TCP10:U03399; CCR6: U45984, GPR31:U65402, FOP:Y18046. The accession Clonogenic assay numbers for clones 246630: AJ419865; 199297: AJ419866; The 246630-ORF-pMH construct was transfected into several 174311: AJ419867. ovarian cancer cell lines by Eectene (QIAGEN) according to manufacturer's instructions. After incubation for 48 h at 378C, 105 ±106 transfected cells were trypsinized and seeded to 100- mm dishes. Medium was added with G418 (400 mg/ml to 1200 mg/ml) for surviving selection. Two to four weeks later, Acknowledgments colonies on plates were ®xed with Methanol/Glacial Acid (3 : 1 Dr Iain Goldsmith is acknowledged for synthesis of in volume) and stained with Crystal Violet in H2O (1 mg/ml). oligonucleotides, Mr Reginald Boone and Dr Anna Colonies were counted manually and the results were analysed Richardson for assistance with sequencing. The FOP clone by paired t-test for means for statistical signi®cance. was obtained from Dr MJ Pebusque. YAC 17IA12 was obtained from Dr J Boyle. Gridded YAC and cDNA libraries were obtained from the MRC HGMP centre. We Accession numbers would like to thank all the members of the chromosome 6 project group at Sanger Centre. The work described in this The accession numbers for PAC/BAC sequences RP4- paper is supported by the Imperial Cancer Research Fund, 655C5:AL121956; RP3-427A4:Z98049, RP3-497J21:AL023775, Association of International Cancer Research, Wellbeing RP1-168L15:AL022069, RP11-514012:AL159163, RP4,505P2: and Human mapping/sequencing at the Sanger Centre is AL133458, RP1-167A14:Z94721, RP11-517H2:AL121935, funded by the Wellcome Trust.

References

AcquatiF,MorelliC,CinquettiR,BianchiMG,PorriniD, Ioannou PA, Amemiya CT, Garnes J, Kroisel PM, Shizuya Varesco L, Gismondi V, Rochetti R, Talevi S, Possati L, H, Chen C, Batzer MA and de Jong PJ. (1994). Nat. Magnanini C, Tibiletti MG, Bernasconi B, Daidone MG, Genet., 6, 84 ± 89. Shridhar V, Smith DI, Negrini M, Barbanti-Brodano G Ioerger TR, Clark AG and Kao TH. (1990). Proc. Natl. and Taramelli R. (2001). Oncogene, 20, 980 ± 988. Acad.Sci.USA,87, 9732 ± 9735. Altschul SF, Gish W, Miller W, Myers EW and Lipman D. Islam SD, Pilder SH, Decker CL, Cebra-Thomas JA and (1990). J. Mol. Biol., 215, 403 ± 410. Silver LM. (1993). Human Mol. Genet., 2, 2075 ± 2079. Anand R, Riley JH, Butler R, Smith JC and Markham AF. Katso RMT, Manek S, O'Byrne K, Playford MP, Le Meuth (1990). Nucleic Acids Res., 18, 1951 ± 1956. V and Ganesan TS. (1997). Cancer Metas. Rev., 16, 81 ± Anderson MA, McFadden GI, Bernatzky R, Atkinson A, 107. OrpinT,DedmanH,TregearG,FernleyRandClarke Kurihara H, Nonaka T, Mitsui Y, Ohgi K, Irie M and AE. (1989). Plant Cell., 1, 483 ± 491. Nakamura KT. (1996). J. Mol. Biol., 255, 310 ± 320. Baba M, Imai T, Nishimura M, Kakizaki M, Takagi S, Laval S, Butler R, Shelling AN, Hanby AM, Poulsom R and Hieshima K, Nomiyama H and Yoshie O. (1997). J. Biol. Ganesan TS. (1994). Cell Growth Dier., 5, 1173 ± 1183. Chem., 272, 14893 ± 14898. LiaoF,AldersonR,SuJ,UllrichSJ,KreiderBLandFarber Blanche H, Wright LG, Vergnaud G, de Gouyon B, Lauthier JM. (1997a). Biochem. Biophys. Res. Comm., 236, 212 ± V, Silver LM, Dausset J, Cann HM and Spielman RS. 217. (1992). Genomics, 12, 826 ± 828. Liao F, Lee HH and Farber JM. (1997b). Genomics, 40, 175 ± Catteau A, Harris WH, Xu CF and Solomon E. (1999). 180. Oncogene, 18, 1957 ± 1965. Liu Y and Ganesan TS. (2001). Re. Reprod., (in press). Chenevix-TrenchG,KerrJ,HurstT,ShihYC,PurdieD, Mancini DN, Rodenhiser DI, Ainsworth PJ, O'Malley FP, Bergman L, Friedlander M, Sanderson B, Zournazi A, Singh SM, Xing W and Archer TK. (1998). Oncogene,16, Coombs T, Leary JA, Crawford E, Shelling AN, Cooke I, 1161 ± 1169. Ganesan TS, Searle J, Choi C, Barrett JC, Khoo SK and McClure BA, Haring V, Ebert PR, Anderson MA, Simpson Ward B. (1997). Genes Chromosomes Cancer, 18, 75 ± 83. RJ, Sakiyama F and Clarke AE. (1989). Nature, 342, 955 ± Cooke IE, Cox SA, Shelling AN, Le Meuth VG, Spurr NK 957. and Ganesan TS. (1996a). Mamm. Genome, 7, 157 ± 159. Menasce LP, Orphanos V, Santibanez-Koref M, Boyle JM Cooke IE, Shelling AN, Le Meuth VG, Charnock ML and and Harrison CJ. (1994). Genes Chromosomes Cancer, 10, Ganesan TS. (1996b). Genes Chromosomes Cancer, 15, 286 ± 288. 223 ± 233. Mungall AJ, Humphray SJ, Ranby SA, Edwards CA, Francesconi M, Meryn S, Rogan AM, Szalay S, Graninger Heathcott RW, Clee CM, Holloway E, Peck AI, Harrison W and Schmidbauer CP. (1984). Cancer, 53, 1927 ± 1930. P, Green LD, Butler AP, Langford CF, William RG, Gaidano G, Hauptschein RS, Parsa NZ, Ot K, Rao PH, Huckle EJ, Baron L, Smith A, Leversha MA, Ramsey YH, Lenoir G, Knowles DM, Chagnati RS and Dalla-Favera Clegg SM, Rice CM, Maslen GL, Hunt SE, Scott CE, R. (1992). Blood, 80, 1781 ± 1787. Soderlund CA, Theaker AJ, Carter NP, Ross MT, Greaves DR, Wang W, Dairaghi DJ, Dieu MC, Saint-Vis B, Deloukas P, Bentley DR and Dunham I. (1997). DNA Franz-Bacon K, Rossi D, Caux C, McClanahan T, Sequence, 8, 151 ± 154. Gordon S, Zlotnik A and Schall TJ. (1997). J. Exper. Obata K, Morland SJ, Watson RH, Hitchcock A, Chenevix- Med., 186, 837 ± 844. Trench G, Thomas EJ and Campbell IG. (1998). Cancer Hime G, Prior L and Saint R. (1995). Gene, 158, 203 ± 207. Res., 58, 2095 ± 2097.

Oncogene Chromosome 6q27 ovarian cancer tumour suppressor gene Y Liu et al 399 Orphanos V, McGown G, Hey Y, Thorncroft M, Santiba- Tibiletti MG, Bernasconi B, Furlan D, Riva C, Trubia M, nez-Koref M, Russell SE, Hickey I, Atkinson RJ and Buraggi G, Franchi M, Bolis P, Mariani A, Frigerio L, Boyle JM. (1995). Br. J. Cancer, 71, 666 ± 669. Capella C and Taramelli R. (1996). Cancer Res., 56, Popovici C, Zhang B, Gregoire MJ, Jonveaux P, Lafage- 4493 ± 4498. Pochitalo M, Birnbaum D and Pebusque MJ. (1999). Tibiletti MG, Sessa F, Bernasconi B, Cerutti R, Broggi B, Blood, 93, 1381 ± 1389. Furlan D, Acquiti F, Bianchi M, Russo A, Capella C and Rodriguez C, Causse A, Ursule E and Theillet C. (2000). Taramelli R. (2000). Clin. Cancer Res., 6, 1422 ± 1431. Genes Chromosomes Cancer, 27, 76 ± 84. Tibiletti MG, Trubia M, Ponti E, Sessa L, Acquati F, Furlan Saha V, Lillington DM, Shelling AN, Chaplin T, Yaspo ML, D, Bernasconi B, Fichera M, Mihalich A, Ziegler A, Volz GanesanTSandYoungBD.(1995).Genes Chromosomes A, Facco C, Riva C, Cremonesi L, Ferrari M and Cancer, 14, 220 ± 222. Taramelli R. (1998). Oncogene, 16, 1639 ± 1642. Saito S, Saito H, Koi S, Sagae S, Kudo R, Saito J, Noda K Trubia M, Sessa L and Taramelli R. (1997). Genomics, 42, and Nakamura Y. (1992). Cancer Res., 52, 5815 ± 5817. 342 ± 344. Saito S, Sirahama S, Matsushima M, Suzuki M, Sagae S, Walter MH, Liu JW, Wunn J and Hess D. (1996). Eur. J. Kudo R, Saito J, Noda K and Nakamura Y. (1996). Biochem., 239, 281 ± 293. Cancer Res., 56, 5586 ± 5589. Wan M, Zweizig S, D'Ablaing G, Zheng J, Velicescu M and Sassa H, Nishio T, Kowyama Y, Hirano H, Koba T and Dubeau L. (1994). Int. J. Oncol., 5, 1043 ± 1048. Ikehashi H. (1996). Mol. Gen. Genet., 252, 222. Ye ZH and Droste D. (1996). Plant Mol. Biol., 30, 697 ± 709. Schein CH. (1997). Nature Biotech., 15, 529 ± 536. ZingoniA,RocchiM,StorlazziCT,BernardiniG,SantoniA Shelling AN, Cooke IE and Ganesan TS. (1995). Br. J. and Napolitano M. (1997). Genomics, 42, 519 ± 523. Cancer, 71, 521 ± 527. Shizuya H, Birren B, Kim UJ, Mancino V, Slepak T, Tachiiri Y and Simon M. (1992). Proc. Natl. Acad. Sci. USA, 89, 8794 ± 8797.

Oncogene