Oncogenomics

Detailed characterization of a homozygously deleted region corresponding to a candidate tumor suppressor locus at distal 17p13.3 in human lung cancer

Hiroyuki Konishi1, Miyabi Sugiyama1, Kotaro Mizuno1, Hiroko Saito1, Yasushi Yatabe2, Toshitada Takahashi3, Hirotaka Osada1 and Takashi Takahashi*,1

1Division of Molecular Oncology, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-Ku, Nagoya 464-8681, Japan; 2Department of Anatomic and Molecular Diagnostic Pathology, Aichi Cancer Center Hospital, 1-1 Kanokoden, Chikusa-Ku, Nagoya 464-8681, Japan; 3Division of Immunology, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-Ku, Nagoya 464-8681, Japan

17p13.3 is one of the chromosomal regions most function (Knudson, 1971). Allelic loss is therefore frequently affected by allelic loss in a variety of human considered to be a hallmark of chromosomal regions neoplasms including lung cancer. A number of loss of harboring tumor suppressor genes. Cytogenetic and heterozygosity (LOH) analyses have suggested the molecular analyses of lung cancers have revealed existence of a tumor suppressor gene at 17p13.3, distal frequent occurrence of multiple chromosomal deletions to the p53 locus at 17p13.1. In the present study, we including 3p, 5q, 8p, 9p, 11p, 13q, 17p, 18q and 22q characterized a homozygous deletion at 17p13.3 in a small (Minna et al., 1997). cell lung cancer cell line by constructing a bacterial It has been shown that 17p is among the most artificial chromosome (BAC) contig and a restriction map frequently affected chromosomal regions in lung cancer surrounding the region, as well as by utilizing publicly (Yokota et al., 1987). We and others have clearly available draft sequences. We defined the breakpoint, provided evidence indicating that the p53 gene at assigned and analysed two known genes, 14-3-3e and 17p13.1 is a genuine molecular target for the frequent CRK, and a novel gene LOST1 within or at the end of the 17p deletions in lung cancer (Takahashi et al., 1989, homozygous deletion of about 170 kb in size. Marked 1992). Interestingly, however, our detailed deletion reduction of LOST1 expression was detected in 69% (11/ mapping analysis of 17p with special reference to the 16) of lung cancer specimens by quantitative real-time p53 gene status revealed the presence of an independent, RT–PCR, while significant DNA hypermethylation was commonly deleted region(s)at 17p13.3, suggesting the observed at the 50 end of the LOST1 gene in four of six existence of an additional tumor suppressor gene(s) lung cancer cell lines with negligible LOST1 expression. distal to the p53 gene (Konishi et al., 1998). Frequent We also show here that a polymorphic marker D17S1174, allelic loss at 17p13.3 independent of the p53 locus has which resides within the homozygous deletion, was also been reported to be common in a variety of other apparently located in the middle of the minimum LOH human malignancies including breast, ovarian and region, providing further supportive evidence for the hepatocellular carcinomas as well as neural tumors presence of a tumor suppressor gene(s) in this region. (Coles et al., 1990; Fujimori et al., 1991; Biegel et al., Oncogene (2003) 22, 1892–1905. doi:10.1038/sj.onc.1206304 1992; Cogen et al., 1992; Saxena et al., 1992; Phillips et al., 1996; Schultz et al., 1996). Keywords: homozygous deletion; 17p13.3; tumor sup- Our subsequent search for a clue to the isolation of a pressor gene; lung cancer; LOST1 tumor suppressor gene from this region resulted in the identification of homozygous deletion of the 14-3-3e gene in two small cell lung cancer (SCLC)cell lines, ONCOGENOMICS Introduction which had been established from distinct metastatic sites of the same patient at different treatment periods Accumulating evidence suggests that a series of genetic (Konishi et al., 2002). Consistent with the notion that alterations such as activation of oncogenes and inactiva- the 14-3-3 family proteins play a role in imposing G2 tion of tumor suppressor genes are involved in the arrest in response to DNA damage through cdc2 pathogenesis of lung cancer (Minna et al., 1997). inactivation, we were able to show that introduction of According to Knudson’s two-hit theory, tumor sup- 14-3-3e into the lung cancer cell line with 14-3-3e pressor genes are inactivated by a recessive mutation in homozygous deletion significantly restored G2 check- one allele and loss of the other allele retaining wild-type point impairment. Together with prevalent perturbation of the G1 and M phase checkpoints in lung cancers *Correspondence: T Takahashi (Takahashi et al., 1989, 1999; Nomoto et al., 1999), the Received 9 September 2002; revised 4 December 2002; accepted 6 involvement of the 14-3-3e gene within the homozygous December 2002 deletions was of interest. However, the possibility that a Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1893 systematic, detailed investigation of the homozygously the databases. These two BACs widely overlapped each deleted region may yield an additional gene that plays a other, while 818O24 harbored the CRK gene as well as crucial role in the pathogenesis of lung cancer was not 14-3-3e. Further investigation of these BACs using precluded. public databases, however, showed uncertainty with In the present study, we constructed a bacterial regard to their chromosomal location, that is, the artificial chromosome (BAC)contig and a restriction assignment of these two BACs had been repeatedly map surrounding the homozygously deleted region, and changed through revisions to the public databases. defined the breakpoints. One novel and two known Although 14-3-3e and CRK had originally been assigned genes were found within the homozygous deletion and centromeric to ABR in 17p13.3, data from NCBI built analysed in detail. We also provide further supportive 28 mapped these BACs and genes at 17q21.32, whereas evidence for the presence of a tumor suppressor gene(s) according to NCBI built 29 they were located at close in this region as well as in a more centromeric region proximity to the distal end of 17p. A large physical map between D17S5 and D17S379. of 17p13.3 by McHale et al. (2000)spanned the frequent LOH region at 17p13.3 but did not contain 14-3-3e and CRK. To define precisely the genomic context surrounding Results the homozygous deletion, we constructed our own Construction and characterization of a contig covering the genomic contig covering the homozygous deletion. homozygous deletion at 17p13.3 Polymerase chain reaction (PCR)screening for yeast artificial chromosome (YAC)/BAC DNA pool libraries We previously screened 65 lung cancer cell lines with the were performed using three STS markers including 14-3- aid of nine markers mapped within the commonly 3e exon 1, CRK exon 3 and D17S1174, which yielded deleted region at 17p13.3, and found that a sequence- five YAC and three BAC clones mapped at the tagged site (STS)corresponding to the 14-3-3e gene was homozygously deleted region. The insert end sequences involved in a homozygously deleted region in two SCLC for each YAC/BAC clone were determined and utilized cell lines (Konishi et al., 2002). The two cell lines, ACC- to generate STS markers. All five YAC clones were LC-48 and ACC-LC-52, were originated from distinct apparently chimeric, but they were nevertheless useful to metastatic sites of the same patient at the different align BAC clones as shown in Figure 1a. BLAST search treatment periods, suggesting the occurrence of the for each marker on this contig revealed that 818O24 and homozygous deletion before metastasis in vivo.We 241L15, sources of the NCBI sequence information, initiated this study by carrying out the BLAST search were just next to the previously unconnected three against the National Center for Biotechnology Informa- BACs, that is, 407I21 (AF322450.1), RPCI-11-107N19 tion (NCBI)Nucleotide Sequence Databases using (AC006405.1)and RPCI-11-433M14 (AC068936.2) complementary DNA (cDNA)sequence of 14-3-3e. This (Figure 1a). These three BACs harbored MYO1C, in silico screening resulted in the identification of two SKIP, PITPN, SREC, RILP and PRP8, all of which BAC clones, RPCI-11-818O24 (818O24; AC032044.9) had consistently been assigned to 17p13.3. In addition, and RPCI-11-241L15 (241L15; AC025071.3)via the YAC/BAC end markers on our contig were surveyed by high throughput genomic sequences (HTGS)section of PCR using the Human Chromosome 17 Regional

Figure 1 STS-based map of the homozygous deletion region at 17p13.3. (a)The STS markers used for mapping are shown at the top. YACs and BACs (closed boxes), identified by PCR screening as well as BACs (open boxes) whose draft sequences are available in the NCBI database as collections of unordered pieces, are indicated in relation to the STS markers. The extent of the homozygous deletion is shown at the bottom. The map is not drawn to scale. (b)Duplex PCR analysis showing no amplification of CRK exon 3 or D17S1174 despite robust amplification of p53 in ACC-LC-48 and ACC-LC-52

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1894 Mapping Panel, confirming their localization at 17p13.3 these predicted exons as probes yielded no cDNA (data not shown). clones. To further define the extent of the homozygous We also searched for UniGene clusters, which reside deletion, we carried out PCR examinations for all of on the BAC 818O24 draft sequence, using the NCBI the STS markers on our contig map using genomic MapViewer program. One of seven UniGene clusters on DNAs (gDNAs)from ACC-LC-48 and ACC-LC-52 as 818O24, Hs. 146059, appeared to represent 30 end of an templates. The result of the PCR examination revealed unknown transcriptional unit, while others agreed with the involvement of all six exons of the 14-3-3e gene, exonic or intronic sequences of either 14-3-3e or CRK. CRK exon 3 and microsatellite marker D17S1174 within We purchased three cDNA clones corresponding to Hs. the homozygous deletion in ACC-LC-48 and ACC-LC- 146059 (C20774, C44602 and C44924 in GeneConnec- 52, whereas CRK exons 1 and 2 were excluded (Figure tion Discovery Clones; Stratagene)and determined the 1a and b). sequence. On further assessment by means of the exon- connection PCR, 50 rapid amplification of cDNA ends (RACE)and RNA protection assay, this transcriptional Restriction enzyme mapping and size estimation of the unit was shown to consist of three exons with a putative homozygous deletion open reading frame (ORF)encoding 177 amino DNA sequences of 818O24 and 241L15 covering the acids (Figure 3a, b). We designated this novel homozygous deletion had been made available as gene LOST1 (LOcated at Seventeen-p-Thirteen point collections of 10–30 unordered pieces accompanied by three 1). unknown lengths of gaps depending on the time points Simple Modular Architecture Research Tool during the course of this study, and there are still many (SMART)predicted two transmembrane domains gaps of uncertain length. We thus conducted restriction close to the C-terminus of LOST1 (Figure 3c), while enzyme mapping for CITB-105A24 (105A24), GS1- hydrophobicity analysis consistently showed the pre- 224gO1 (224gO1)and CITB-442O5 (442O5),with five sence of two hydrophobic peaks at the regions rare-cutting restriction enzymes including MluI, NruI, annotated as transmembrane domains (data not shown). NotI, NheI and SalI using cDNA probes for 14-3-3e and Protein–protein BLAST search against the NCBI CRK (Figure 2). Also, we generated STS markers at the Peptide Sequence Databases identified an unknown MluI and NotI sites on 442O5 (Table 1). As a result, the human protein, namely ‘protein for MGC:9889’ size of the homozygous deletion could be estimated as (AAH11405.1)as the only registration with homology approximately 170 kb. Consequently, MYOIC, SKIP to LOST1 (Figure 3c). Strong homology was detected at and PITPN were demonstrated to reside on 105A24 and AA101–AA176 in the C-terminal region of LOST1, 224g01, but these genes were excluded from the which included the two transmembrane domains (45% homozygously deleted region (Figure 2). of identity; E value ¼ 2 Â 10À4). We performed 30 RACE for LOST1 and found two messenger RNA isoforms in agreement with the existence of two polyadenylation Search for a novel gene at the homozygous deletion region signals (AATAAA)(Figure 3a).A relatively high To investigate as yet unidentified transcriptional units in expression was observed in muscular organs such as the homozygous deletion, we conducted gene predic- the heart, smooth muscle and skeletal muscle, while tions based on the draft sequence of 818O24 using LOST1 was also modestly expressed in the brain and GENSCAN, Grail, FGENESH and HMMgene. the bronchus (Figure 4a). Southern blot analysis Although four exons (termed as PE1-PE4)of a potential demonstrated rearranged bands in ACC-LC-48 and novel gene(s)were predicted, screening against an adult ACC-LC-52, indicating that LOST1 had been disrupted lung cDNA library and a fetal brain cDNA library with by the homozygous deletion (Figure 4b).

Figure 2 Restriction map of the BAC clones. Mapping was conducted for a small BAC contig composed of three clones (105A24, 442O5 and 224g01). The directions of the BACs are shown with ‘T7’ and ‘SP6’ at each end. Six genes (PITPN, SKIP, MYO1C, CRK, 14-3-3e and LOST1)as well as the homozygous deletion are shown in relation to the BAC contig. A scale bar is provided in kb. M, MluI; Nr, NruI; No, NotI; Nh, NheI and S, SalI. Cen, centromere; Tel, telomere

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1895 Table 1 Details of STS markers generated in the present study STS marker Primer position Primer pair Product size Annealing Mapping to (bp) temperature (1C) 17p13.3 614C6 righta Insideb TACAGAGTTATCCCGCCATC 196 55 Yc Outsided GTCACCGGAAAAACAAAGCA 614C6 lefte Inside ATATTCATAGTTTTTCAATA 100 55 Nf Outside TTGTTAGCACAGACATCTAC 636G8 right Inside ATGTCAACTTAAATGTATTG 195 55 N Outside CTTCTGGATAGAGACCTGCA 660E6 right Inside TCACCTTCTGAGACAACAAA 181 55 N Outside AAGCTCCAACAATAGATTAA 660E6 left Inside CATGGGAAGGAGAGACTGAC 96 55 N Outside CACTGCAAGGCTAGCTACAT 664A4 right Inside TACTCTATGGTTTGCCTTTG 165 55 N Outside TTAGTTCTCCAGAAGCCAA 664A4 left Inside AGCAAATGGTTATCAAGGTA 152 55 Y Outside CATGTCATCCTTCTTTAATG 697G11 right Inside ACTGCTGGTAAGAAGCCCTA 180 55 N Outside TGTACAACATCTATGGGAAC 697G11 left Inside GAGACTGGCTGTCCACTTAG 112 55 Y Outside CCTAAGATAAGTGTGTATTT 105A24 T7g Inside CAAATTGTACTTGAATGCTT 126 50 Y Outside AATCAGTTTTTTAAAGGAGC 224g01 T7 Inside TAAACGAGTAGTATTGCTCC 238 52 Y Outside ATGACAATCAACATTCTGCC 105A24 SP6h Inside CATGCACACACCTTGTTTCA 128 55 Y (224g01 SP6)i Outside CAACGTTAAATGCATTGTGG 442O5 T7 Inside CCCATGAAATAACTGAGGAC 113 58 Y Outside TATTCATTCCTTGGAGGTTG 442O5 SP6 Inside GGCGTGAGCCACCTCACCCA 111 60 Y Outside CTATCTACCTCACAAGGAGT 442O5-MluIj GTTGGGCTGGGAGTCTTATT 103 56 Y ACCTGTGACGGGATTATCGA 442O5-NotIk CCCTGGAAGTGTAAGCGGTC 125 59 Y TTGGGACGACGCCTGGTGAT aRight end of insert of a yeast artiﬁcial chromosome (YAC). bInside primer (a primer distant from insert end). cSTS marker mapped at 17p13.3 by somatic cell hybrid analysis. dOutside primer (a primer nearer to insert end). eLeft end of insert of a YAC. fSTS marker excluded from 17p13.3 by somatic cell hybrid analysis. gT7 end of insert of a bacterial artiﬁcial chromosome (BAC). hSP6 end of insert of a BAC. iSP6 ends of inserts of 105A24 and 224g01 are identical. jSTS marker around the MluI site on 442O5. kSTS marker around the NotI site on 442O5.

Mutational analysis of the three genes within the 15 adenocarcinomas, 11 squamous cell carcinomas and homozygous deletion one large cell carcinoma)by RT–PCR–single-strand conformation polymorphism (SSCP)analysis followed We conducted mutational analysis of genes mapped by sequencing of the shifted bands. Six breast cancer cell within the homozygous deletion. The entire ORF of the lines and seven hepatocellular carcinoma cell lines were 14-3-3e and CRK genes were analysed in 28 lung cancer additionally included in the examination. In the 14-3-3e cell lines (14 SCLCs, ﬁve adenocarcinomas, ﬁve gene, an identical sequence variation at codon 252 with squamous cell carcinomas and four large cell carcino- no amino acid substitution was detected in a squamous mas)as well as in 40 lung cancer specimens (13 SCLCs, cell carcinoma cell line QG56 and an adenocarcinoma

" Figure 3 Molecular cloning and characterization of the LOST1 gene. (a)Nucleotide and deduced amino acid sequences of the LOST1 cDNA. The furthest 50 end of the cDNA clones obtained by the 50-RACE procedure is indicated as nucleotide 1. Underlines indicate two polyadenylation signals. Vertical bars coupled with solid arrowheads indicate exon–intron boundaries. LOST1 is located at the telomeric edge of the homozygous deletion, and nucleotide sequence downstream from the vertical bar with a closed circle is lost in ACC-LC-48 and ACC-LC-52. (b)Schematic presentation of genomic organization of the LOST1 gene along with its nucleotide sequences at the exon–intron boundary. Solid boxes represent untranslated regions. (c)Homology of predicted LOST1 protein to unknown human protein AAH11405. The sequence alignment was carried out with ClustalW. Identical amino acids are in black; conserved amino acids are shaded in gray. Two open boxes indicate the putative transmembrane domains predicted by SMART. The amino acid position for each of the proteins are indicated on the right

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1896

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1897

Figure 4 (a)Detection of LOST1 transcripts by RT–PCR in various fetal and adult tissues. Expression of LOST1 is detectable in various adult organs including lung (bronchus), brain, adrenal gland as well as muscular tissues such as smooth and skeletal muscle. (b) Southern blot analysis of LOST1 gene in human lung cancer cell lines, showing rearranged bands in ACC-LC-48 and ACC-LC-52. A total of 5 mg of gDNAs digested with EcoRI were used specimen (Table 2). As for the CRK gene, frequent and Table 2 Single nucleotide polymorphisms identified within the coding rare silent polymorphisms were detected at codons 17 exons of 14-3-3e, CRK and LOST1 and 38, respectively. In a hepatocellular carcinoma cell Substitution line PLC/PRF5, nucleotide substitutions were detected at codon 68 (CGC to CGG, no amino acid change). Gene Codon Nucleotide Amino acid Mutational analysis of the entire ORF of the LOST1 14-3-3e 252 GAC Aspartic acid gene was conducted by direct sequencing following PCR GAT Aspartic acid amplification of each exon using gDNAs from 41 lung CRK 17 CGG Arginine AGG Arginine cancer cell lines (16 SCLCs, 13 adenocarcinomas, seven 38 CGG Arginine squamous cell carcinomas and five large cell carcino- CGA Arginine mas)as well as 40 lung cancer cases (10 SCLCs, 15 68 CGC Arginine adenocarcinomas, 12 squamous cell carcinomas and CGG Arginine LOST1 3 CAC Histidine three large cell carcinomas). This examination identified TAC Tyrosine several SNPs, all of which were located at the N- 15 CCA Proline terminal half of exon 1 (Table 2). All the variant alleles TCA Serine detected in the tumor tissues were also present in the 18 GCC Alanin corresponding normal lungs. No somatic mutations ACC Threonine 20 TCC Serine were identified in any of the three genes, indicating that TTC Phenylalanine these genes are not frequent targets for somatic 34 GAG Glutamic acid mutations in lung cancer. GAC Aspartic acid 57 GGC Glycine AGC Serine Expression analysis of the three genes within the homozygous deletion tissues (three SCLCs, six adenocarcinomas, six squa- We analysed the expression of the three genes mapped mous cell carcinomas and one large cell carcinoma). The within the homozygous deletion using 16 lung cancer expression of the 14-3-3e gene and the CRK gene in lung

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1898 cancer tissues was examined by Northern blot analysis, specimens prompted us to examine the DNA and showed levels similar to those in normal lung methylation status of LOST1. Genomic DNAs tissues. Interestingly, significantly reduced, negligible from six lung cancer cell lines with negligible expression of LOST1 was observed in 11 of 16 (69%)of LOST1 expression, as well as those from two lung cancer tissues (two of three SCLCs, four of six normal lung tissues, were examined. Following adenocarcinomas, four of six squamous cell carcinomas sodium bisulfite conversion, genomic DNAs encom- and in the large cell carcinoma)in the analysis by RT– passing the 50 end of the LOST1 gene were amplified PCR, as well as by quantitative real-time RT–PCR and cloned for sequence determination. Four of the six analysis (Figure 5). lung cancer cell lines demonstrated increased DNA methylation at CpG sites (67, 80, 83 and 83%, respectively), whereas normal lung tissues, which Methylation analysis of the LOST1 gene express LOST1 showed methylation at only 33–37% The identification of negligible expression of the (10/30 and 11/30, respectively)of the CpG sites LOST1 gene in a significant fraction of lung cancer examined (Figure 6).

Figure 5 Expression analysis of the three genes within the homozygous deletion. (a)Northern blot analysis showing ubiquitous expressions of both 14-3-3e and CRK in lung cancer specimens. (b)RT–PCR detection of LOST1 in lung cancer specimens. LOST1 expression was undetectable in 69% (11/16)of lung cancer specimens. ( c)Quantitative real-time RT–PCR analysis of LOST1 expression in lung cancer specimens, conﬁrming data of RT–PCR analysis in (b). Results are shown as LOST1 expression relative to that in normal lung #1 after normalization to the 18S ribosomal RNA expression in each case. SCLC, small cell lung cancer; AD, adenocarcinoma; SQ, squamous cell carcinoma; LA, large cell carcinoma

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1899

Figure 6 Increased DNA methylation at the LOST1 gene locus in lung cancer cell lines showing negligible LOST1 expression. (a)RT– PCR examination of LOST1 in lung cancer cell lines with negligible expression of LOST1. N.C., negative control. (b)Quantitative real- time RT–PCR analysis of LOST1 expression in lung cancer cell lines, conﬁrming the data from RT–PCR analysis in (a)Expression data are normalized and shown as in Figure 5c. (c)DNA methylation status of the LOST1 gene in lung cancer cell lines as well as in normal lung tissues, showing association between increased DNA methylation and marked reduction of LOST1 expression

Isolation and sequencing of breakpoints of the be very close to the telomeric end of the homozygous homozygous deletion deletion, were utilized for PCR with gDNAs, resulting in the robust ampliﬁcation of PCR products of 1918 bp To further localize the centromeric border of the exclusively in ACC-LC-48 and ACC-LC-52. Duplex homozygous deletion, we generated ﬁve STS markers PCR using a primer pair encompassing the homozygous within CRK intron 2. The PCR examinations revealed deletion as well as that of p53 indicated that ACC-LC- that only the most centromeric marker was retained in 48 and ACC-LC-52 possess interstitial deletions ACC-LC-48. Thus, the forward primer for this STS (Figure 7a). Sequence analysis of both ends of the marker and the PE1-F primer, which was presumed to homozygous deletion revealed the presence of a

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1900

Figure 7 Identification and analysis of breakpoints of the homozygous deletion. (a)Duplex PCR with the aid of a primer pair specific to genomic sequences just outside the homozygous deletion as well as p53 specific primers. Robust amplification with a primer pair encompassing the fused breakpoints indicates the presence of interstitial deletions in ACC-LC-48 and ACC-LC-52. H.D., homozygous deletion; N.C., negative control. (b)Nucleotide sequences surrounding centromeric and telomeric breakpoints in ACC-LC-48. A 2 bp microhomology is indicated in black. Two pairs of short inverted repeats each at breakpoints are indicated by underlines and directions are shown by open or closed arrows. Asterisks (*)indicate the identities of their nucleotides

potential 2 bp short direct repeat as well as a set of and/or at least one or more markers mapped between inverted repeats at both sides of the breakpoints p53 and the 17p13.3 markers (Figure 8). Consequently, (Figure 7b), suggesting that either of these homologous two chromosomal regions surrounding D17S1174 and repeats may have played a role in the generation of this D17S379 were found to be two independent minimum interstitial deletion. deleted regions, while five cases (22, 23, 49, 66 and 100) showing small deletions harboring either D17S1174 or D17S379. Confinement of the minimum deleted regions LOH study at D17S1174 in combination with other to small regions surrounding D17S1174 or D17S379 17p13.3 markers in cases of primary lung cancer provided further support for the presence of tumor We assessed LOH status specifically of a genomic region suppressor genes involved in lung carcinogenesis. involved in the homozygous deletion in 100 cases of primary lung cancer. D17S1174 was chosen for this study, since this marker was found to be the only Discussion microsatellite marker within 818O24 in the computa- tional search in the NCBI UniSTS database. The result Detailed analyses of homozygous deletions have pro- was analysed with reference to the LOH data for nine vided important clues to the isolation of a number of other polymorphic markers at 17p, which had been tumor suppressor genes including RB, p16, FHIT, obtained in our previous study (Konishi et al., 1998). RASSF1, Smad4 and PTEN/MMAC1 (Friend et al., D17S1174 showed LOH in 57% of the cases (90% in 1986; Lee et al., 1987; Kamb et al., 1994; Nobori et al., SCLCs, 48% in adenocarcinomas, 58% in squamous 1994; Hahn et al., 1996; Ohta et al., 1996; Li et al., 1997; cell carcinomas, and 75% in large cell carcinomas), Steck et al., 1997; Dammann et al., 2000; Burbee et al., being the most frequently affected marker at 17p13.3 2001). In the current study, we characterized a homo- along with D17S5 (56%). In addition, D17S1174 in case zygous deletion at 17p13.3 by analysing YAC and BAC 94 and D17S5 in case 64 showed apparent retention of clones covering the region as well as by utilizing publicly heterozygosity flanked by LOH at closely mapped loci, a available draft sequences. We also assigned one potential indication of the presence of homozygous novel and two known genes within or at the end deletions (Cairns et al., 1994). of the homozygous deletion of about 170 kb in size. Detailed inspection of the LOH analysis data LOH analysis with primary lung cancer specimens identified 31 cases with LOH at 17p13.3 despite revealed that a polymorphic marker D17S1174, which apparent retention of heterozygosity at the p53 locus resides within the homozygous deletion, was apparently

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1901

Figure 8 Extent of small LOHs conﬁned to 17p13.3 in a panel of 100 cases of lung cancer. The extent is indicated with gray boxes. The data in this ﬁgure were generated by combining the present data with D17S1174, which was found here to be contained in the homozygous deletion, as well as data from our previous study. AD, adenocarcinoma; SQ, squamous cell carcinoma; LA, large cell carcinoma. Two representative cases exhibiting LOH exclusively at the D17S1174 locus are shown on the right. Open and solid arrowheads refer to larger and smaller alleles, respectively. N, normal; T, tumor

located in the middle of the minimum LOH region, markers in lung cancers (Konishi et al., 1998). The other supporting the notion that the small chromosomal frequently affected region is a subtelomeric region of region harboring the homozygous deletion may contain 17p13.3 encompassing D17S34, D17S926 and D17S824, an important gene in terms of the molecular pathogen- which corresponds to the minimum deleted regions in esis of human lung cancer. several types of neoplasms such as neurogenic tumors 17p13.3 is known to be a chromosomal region and hepatocellular carcinoma (Biegel et al., 1992; Cogen frequently affected in a variety of neoplasms. A number et al., 1992; Saxena et al., 1992; Zhao et al., 2001), while of LOH studies have demonstrated the existence of a ﬁne deletion mapping study in breast cancer has commonly deleted regions at 17p13.3, which is indepen- demonstrated complex distribution of commonly af- dent of and distal to the p53 locus at 17p13.1 (Coles et al., fected regions often involving either D17S5 locus or the 1990; Fujimori et al., 1991; Biegel et al., 1992; Cogen subtelomeric 17p13.3 region (Stack et al., 1995). Several et al., 1992; Saxena et al., 1992; Phillips et al., 1996; candidate genes, which may account for the high Schultz et al., 1996; Konishi et al., 1998). However, only frequency of LOH, have been reported including HIC- two homozygous deletions have been found in this 1, OVCA1/DPH2L and ROX/Mnt around D17S5 region in any type of human cancers (Phillips et al., 1996; (Wales et al., 1995; Schultz et al., 1996; Hurlin et al., Haataja et al., 1997). Our report presented here 1997; Meroni et al., 1997)as well as HCCS1 recently represents the third such example and the ﬁrst detailed isolated from a region near D17S926 (Zhao et al., 2001). molecular analysis of a homozygous deletion at 17p13.3 LOH analysis of a series of 100 cases of lung cancer in in combination with systematic screening of the genes this study demonstrated that the D17S1174 locus within residing in this region of great interest. the homozygous deletion was affected at the highest Two genomic regions at 17p13.3 have been reported frequency among the 17p markers we have thus far thus far to harbor potential tumor suppressor genes in examined, while three of the 100 cases exhibited LOH different types of malignancies. A region around D17S5 exclusively at D17S1174. We note that this chromoso- (YNZ22)had previously been shown to exhibit the mal region has not been delineated thus far as a highest frequency of LOH in ovarian cancer (Phillips frequently affected minimum region of LOH, which is et al., 1996; Schultz et al., 1996), while D17S5 was also located roughly 1 Mb apart from the two regions found to be one of the most frequently affected 17p13.3 described above. Taken together, these data suggest

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1902 the existence of a potential tumor suppressor gene in 1 possesses several lines of features, which are in line with lung cancer in this genomic region. its role as a tumor suppressor gene, including reduced In this study, we conducted an extensive search for expression because of tumor-specific hypermethylation genes contained within the homozygous deletion, result- and inhibition of cell proliferation (Wales et al., 1995). ing in the identification of three genes. A novel gene, To date, to the best of our knowledge, 15 solid tumor named LOST1, was shown to be disrupted at the middle cases or cell lines carrying homozygous deletions have of exon 1 by the homozygous deletion, resulting in the been sequenced and reported (Canning and Dryja, 1989; perturbation of its coding capacity. Our preliminary Tadokoro et al., 1992; Murata et al., 1994; Pomykala examination failed to identify somatic mutations within et al., 1994; Kohno et al., 1996; Inoue et al., 1997; the ORF of LOST1 as well as in those of two other genes Sekido et al., 1998; Nishioka et al., 2000; Shigemitsu residing in the homozygous deletion, that is, 14-3-3e and et al., 2001; Tamura et al., 2002). As a result, short direct CRK. Interestingly, however, significantly reduced ex- repeats were found at the breakpoints in seven cases, pression of the LOST1 gene was detected in 69% (11/16) while repetitive sequences such as long interspersed of lung cancer specimens, suggesting that loss of LOST1 nuclear elements (LINEs)and Alu sequences appeared expression might play a role in the pathogenesis of lung to be involved in five cases. There is one report cancers. Our methylation analysis suggested the poten- describing potential involvement of inverted repeats. tial relation between loss of LOST1 expression and In the present case, 2 bp microhomology was found at aberrantly increased hypermethylation of the LOST1 the breakpoints, suggesting the possibility of double- gene locus in a subset of lung cancer cell lines. Although strand break-induced repair in the process generating the silencing mechanism for cell lines without LOST1 interstitial deletion (Chen and Kolodner, 1999). Alter- hypermethylation remains to be elucidated, it is of note natively, two pairs of short inverted repeats encompass- that abnormal chromatin conformation alone has been ing each end of the homozygous deletion may have shown to be responsible for loss of TGF-b type II mediated the interstitial deletion by forming a stem loop receptor expression at least in a fraction of lung cancer at the points of breakage. cell lines (Osada et al., 2001). It is also possible that The human genome project is about to complete expression of a gene(s)within or close to the homo- sequencing the whole human genome. During the course zygous deletion including LOST1 may be regulated by a of this study, sequence information of the genetic region transcriptional regulatory element such as an enhancer of special interest to us was also considerably accumu- residing in this homozygous deletion, while subtle lated. However, sequences of the majority of BACs alterations of such an element might have been over- assigned still consist of unordered pieces of fragments looked because of our ORF-focused search for muta- with gaps of uncertain length. In addition, the order of tions. Alternatively, considering that the lack of 14-3-3e genes contained in the contigs had been changed appeared to be responsible for the defective G2 repeatedly through revisions to the database. Our study checkpoint in ACC-LC-48 (Konishi et al., 2002), it is provides supplementary information, which firmly possible that 14-3-3e may have played a primary role in establishes the order of genes in this region as cen- the clonal expansion of lung cancer cells carrying the PITPN-SKIP-MYO1C-CRK-14-3-3e-LOST1-tel. homozygous deletion. It has not been formally pre- In conclusion, together with our previous observa- cluded that the chromosomal region surrounding tion, this study provides further evidence for the D17S1174 but outside the homozygously deleted region presence of a putative tumor suppressor gene within a might contain a gene with frequent mutations, and that close proximity to the homozygous deletion at 17p13.3. 17p13.3 might harbor multiple lung cancer-related genes The findings described here should serve as an im- like 3p21.3 with multiple candidate tumor suppressor portant foundation for future efforts toward the genes including RASSF1A, FUS1, SEMA3B and H37/ identification of the putative tumor suppressor gene at RBM5 (Dammann et al., 2000; Lerman and Minna, 17p13.3, molecular cloning of which should shed light 2000; Burbee et al., 2001; Kondo et al., 2001; Tomizawa on the molecular mechanisms of lung carcinogenesis, et al., 2001; Ji et al., 2002; Oh et al., 2002). It is of note considering that LOH in this region has been shown to that candidate genes thus far identified at 3p21.3 show occur earlier than that of the p53 locus (Konishi et al., low frequencies of somatic mutations with frequently 1998; Yatabe et al., 2000). reduced expression owing to hypermethylation at the promoter regions (Lerman and Minna, 2000). Taken together, further functional analysis of LOST1 as well as Materials and methods a search for genes in the surrounding region will prove interesting and may reveal intriguing features of LOST1 Tissue specimens and lung cancer cell lines and an as yet unrevealed tumor suppressor candidate(s) in terms of lung carcinogenesis. Lung cancer specimens together with uninvolved lung tissues An additional minimum LOH region at 17p13.3 was were collected from surgical or necropsy cases (14 SCLCs, 54 adenocarcinomas, 28 squamous cell carcinomas and four large also demonstrated to exist around D17S379. It is notable cell carcinomas). Adult and fetal tissues from various types of that a putative tumor suppressor gene, HIC-1, resides organs were obtained during necropsy and elective abortion, between D17S379 and the above-mentioned D17S5 locus respectively. All tissues were quickly frozen in liquid nitrogen (Wales et al., 1995). Although frequent somatic muta- and stored at À801C until analysis. A total of 65 lung cancer tions in tumor DNA have not been detected so far, HIC- cell lines (32 SCLCs, 19 adenocarcinomas, seven squamous cell

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1903 carcinomas, two adenosquamous cell carcinomas and five Cloning of LOST1 gene large cell carcinomas)were analysed in this study. Three cDNA clones corresponding to Hs. 146059 (LOST1) were purchased from Stratagene. Exon-connection PCR was YAC/BAC map construction performed to extend the sequence of LOST1 exon 1 toward 50 direction with either of the following three forward primers Identification of YAC/BAC clones was conducted by PCR generated on the provisional exon 1: LOST1-exon1F1: 50- screening of CEPH ‘A’ Human YAC DNA Pools (ResGen), CGCATTCTTCTCTCTGCTCT; LOST1-exon1F2: 50- CITB Human BAC DNA Pools Release IV (ResGen), and GAACTCAGCTGGCTTGAGAA and LOST1-exon1F3: 50- Down-to-the-Well Human BAC DNA Pools Release I TGTCCCAGCCTGGAGCTGCA, together with a reverse (Genome Systems, Inc.)according to the manufacturers’ primer PE4-R. instructions. The STS markers used for the screening were 50 and 30 RACE were also carried out to identify each end of 14-3-3e exon 1, CRK exon 3 and D17S1174. YAC manipula- the LOST1 gene essentially as described by Frohman (Froh- tion was performed according to standard procedures, and man, 1993). cDNA synthesis for generating a template for 50 BACs were prepared following the instructions of the RACE of LOST1 was conducted using LOST1-exon1R3: 50- suppliers. YAC/BAC insert end sequences were determined GCCTTGGTGAGGAGTATCTC. 50 RACE was then carried by direct sequencing (BACs)or inverse PCR (YACs and out using a forward primer RACE-B: 50-GGAT(C) in BACs)to generate new STS markers corresponding to each of 13 combination with reverse primers, LOST1-exon1R2: 50- their ends. YAC/BAC end markers were analysed by PCR TGAAGGAAACTCGGACTGCA and LOST1-exon1R1: 50- using the Human Chromosome 17 Regional Mapping Panel GTGCTGCTGAGAGCACAGAT for the first and second (NIGMS Human Genetic Mutant Cell Repository)to confirm rounds of PCR, respectively. With regard to 30 RACE their localization at 17p13.3. Alignment of YAC/BAC clones of LOST1, a primer Qt: 50-CACGCGTCGTCTAGTAC- was done by PCR using markers corresponding to exons of the GAC(T) was used for cDNA template synthesis, and two 14-3-3e and CRK genes, D17S1174 and the YAC/BAC insert 17 rounds of PCR were carried out with either LOST1-exon3F1: end sequences. 50-GGACCAGGTTTGGGAATACA and LOST1-exon3F2: PCR amplification of STS markers corresponding to the end 50-TTTGTCCCATCAGGACCCTC, or LOST1-exon3F3: 50- of YAC/BAC inserts as well as to regions harboring 442O5- CCTGGAGGAGAAAGCCATTC. The reverse primer used MluI and 442O5-NotI were carried out with the aid of primer for 30 RACE was Qo: 50-CACGCGTCGTCTAGTACGAC. pairs described in Table 1. STS markers for exons were In RNase protection assay, a 489 bp fragment amplified by designed to include intron and/or untranslated regions in PCR with a primer pair PE2-F and LOST1-exon1R1 was addition to the ORF sequence. Detailed information regarding subcloned into pBluescriptII SK(À), and used as a template PCR conditions is available upon request. for generating riboprobes by in vitro transcription. In vitro transcription from T7 promoter and RNase protection assay Restriction enzyme mapping were performed using RiboQuant RPA Kit (Pharmingen) according to the manufacturer’s instructions. DNAs derived from three BACs were digested with restriction enzymes, processed for pulsed-field gel electrophoresis (PFGE) using clamped homogeneous electric field (CHEF)-DR II Southern blot and Northern blot analyses apparatus (BioRad). Agarose gels (1%) were run at 180 V in Extraction of gDNA and total RNA from cell lines and tissues 0.5 Â TBE at 141C for 15 h with a pulse time from 1 to 5 s. were conducted according to standard procedures. Southern After electrophoresis, the gels were blotted and the filters were blot analysis was performed using 5 mg of gDNAs digested hybridized with cDNA probes as well as BAC insert end with EcoRI. A LOST1 cDNA probe for Southern blot probes according to the standard protocol. hybridization spanning the entire ORF was generated by For subcloning of the sequence around the MluI site on RT–PCR with the aid of a primer pair, LOST1-exon1F2 442O5, a HindIII fragment, which was shown to contain an and LOST1-exon3R1: 50-TTTCTTGTCAGATGGGGGCT. MluI site in double digestion with HindIII and MluI, was Northern blot analysis was performed using 10 mg of total isolated and inserted into pBluescriptII SK(À). Subcloning of RNA according to standard procedures. CRK cDNA probe the NotI site on 442O5 was carried out in the same manner. was generated with CRK-exon3F2: 50-CGTCTGCTGGAT- CAACAGAA and CRK-exon3R2: CAGGGTAAGCAGTG- cDNA library screening TGTAAC. The 14-3-3e cDNA probe used in this study was identical to that described previously (Konishi et al., 2002). Human Lung 50-STRETCH cDNA Library (Clontech)as well as Human Fetal Brain 50-STRETCH PLUS cDNA Library Expressional analysis of LOST1 (Clontech)was screened with predicted exon probes. Genomic regions, which were predicted as exons by a minimum of two Expression of LOST1 was analysed by using both the RT– in four publicly available on-line programs, GENSCAN, Grail PCR and quantitative real-time RT–PCR techniques. Oligo- (Ver. 3.1), FGENESH (Ver. 1.1) and HMMgene (Ver. 1.1), nucleotide primers used were LOST1-exon1F4: 50-TATGCC- were considered as predicted exons (PE1–PE4). PE1–PE4 were CAAGACCAAGAAGC and PE3-R. RT–PCR amplification amplified from gDNA using four sets of primer pairs: PE1-F: with RNAs from various organs was carried out with 35 cycles 50-GGGAATGGGGAGGAAGAATG and PE1-R: 50- (95, 63 and 721C, each for 30 s)after the initial denaturation TCAGGATCCTGAGCCTTTCT; PE2-F: 50-GTCCAGCAT- step (951C for 3 min)in the presence of 5% DMSO. For the TTCCCAACAGA and PE2-R: 50-CTGGAACAAAGAC- analysis of the LOST1 expression in lung cancer specimens, CAGCACT; PE3-F: 50-TCCTGTTCTCTTCACTGCTC and RT–PCR was carried out under the same conditions, except PE3-R: 50-CTGCTTTGTAACAGAGGGCA and PE4-F: 50- for amplification with 40 instead of 35 cycles. Quantitative ATGAAGCTGACCCACAGCCT and PE4-R: 50-CTGTGA- real-time RT–PCR analysis was conducted with the aid of ABI AGTTGACGGTCACG. The resultant PCR products (PE1– Prism 7900HT (Applied Biosystems)and SYBR Green PCR PE4: 1174, 500, 190 and 1375 bp in size)were cloned into Master Mix (Applied Biosystems)according to the manufac- plasmids and prepared as probes for further analysis. turer’s instructions. Amplification was performed with 50

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1904 cycles (951C for 15 s, 651C for 30 s and 721C for 30 s), after the TAGAGTGTGGGCTGCCATGG and LOST1-intron2R: 50- initial denaturation step (951C for 10 min), and followed by TTCCTCCTCTCCAAGGAGCT for exon 2 and LOST1- dissociation curve analysis. Expression of LOST1 was normal- intron2F: 50-ATCCTCCAGTCAGGGACTTG and LOST1- ized to expression of 18S ribosomal RNA, which was obtained exon3R1 for exon 3. by real-time RT–PCR with oligonucleotide primers: 50- AATCAGGGTTCGATTCCGGA and 50-CCAAGATCCA- ACTACGAGCT. Definition of the breakpoints of the homozygous deletion Primer pairs for five STS markers within CRK intron 2 were Methylation analysis of the LOST1 gene designed referring to the draft sequence of BAC 818O24. Detailed information will be provided upon request. PCR DNA methylation analysis based on the sodium bisulfite amplification of the 1918 bp region of the fused breakpoints conversion technique was performed as described previously owing to the homozygous deletion in ACC-LC-48 and ACC- (Osada et al., 2001). Oligonucleotide primers used for the PCR LC-52 was carried out with the aid of the following primers: amplification after bisulfite conversion were 50- CRK-intron2F1: 50-GATGTCCCCTGTGTAGTGAA and GTTTGAGTTTTTTTGGGAAGAGG and 50-ACTAAA- PE1-F. ACCTACCTACTACCCC. The resultant PCR products encompassing the 50 end of the LOST1 transcripts shown in Figure 3a harbored five CpG sites, and corresponded to Duplex PCR nucleotides 191840–191666 of AC032044.5 of BAC 818O24. Duplex PCR amplifications were performed using gDNA, followed by electrophoresis on 3% agarose gels. The following Mutation search oligonucleotide primers were used for amplification: CRK- In RT–PCR–SSCP analysis, the ORF of 768 bp (14-3-3e)and exon3F: 50-ACAGTTTTGCTGACAGATGGG and CRK- 915 bp (CRK)were divided into three and four fragments, exon3R: 50-GGAATGGAGCAGTTCAGTCTA; p53-in- respectively. The primer pairs utilized to amplify three tron6F: 50-TGCTTGCCACAGGTCTCC and p53-intron7R: 14-3-3e gene fragments were as follows: 14-3-3e-fragment1F: 50-ATCGGTAAGAGGTGGGCC; D17S1174-F: 50-TGCA- 50-AGACCCTGCCGGAGCCGCTG, 14-3-3e-fragment1R: CAGGTACCCCAGAACT and D17S1174-R: 50-CCATTA- 50-GCTCAGTCTCAACCATTTGCCG; 14-3-3e-fragment2F: CCTCAGTAAGTGAAC; p53-intron7F: 50-TGGGAGTA- 50-CAAGCTAAAAATGATTCGGGA, 14-3-3e-fragment2R: GATGGAGCCT and p53-exon8R: 50-TCTCCCAGGACAG- 50-GAGAGCAAGACCTAAGCGAA; 14-3-3e-fragment3F: GCACAAA; CRK-intron2BP: 50-CAGAGCTGTAAGAAG- 50-ACTTCCACCAACGCATCCTA, 14-3-3e-fragment3R: 50- CTCCT and LOST1-exon1F2; and p53-exon7F: 50-GTAA- TTTCTCTTGTTGGCTTATGTC. Primer pairs used for the CAGTTCCTGCATGGGC and p53-exon8R. amplification of four CRK fragments were as follows: CRK-fragment1F: 50-GGCGGCGGCACCCCAGCGTT, CRK- fragment1R: 50-TGGGCGGGCGACGGTGGCAC; CRK-frag- ment2F: 50-TACATCATCAACAGCAGCGG, CRK-fragment Acknowledgements 2R: 50-CCCGGATTCTCAAGATGTCT; CRK-fragment The GenBank accession number for the LOST1 cDNA is AB 3F: 50-GATCTTCCCTTTAAGAAAGG, CRK-fragment3R: 090231. This work was supported in part by a Grant-in-Aid 50-CCATTCTGGAGGTTAGGGAG; CRK-fragment4F: 50- for Scientific Research on Priority Areas from the Ministry of AACCCAGCGTCAACACTCCG, CRK-fragment4R: 50-CCC- Education, Culture, Sports, Science and Technology of Japan, ATCTGTCAGCAAAACTG. LOST1 was analysed by direct and a Grant-in-Aid for the Second Term Comprehensive 10- sequencing of each exon using the following primers: LOST1- Year Strategy for Cancer Control from the Ministry of Health exon1F2 and PE1R for exon 1; LOST1-intron1F: 50- and Welfare, Japan.

References

Biegel JA, Burk CD, Barr FG and Emanuel BS. (1992). Cancer Friend SH, Bernards R, Rogelj S, Weinberg RA, Rapaport Res., 52, 3391–3395. JM, Albert DM and Dryja TP. (1986). Nature, 323, 643–646. Burbee DG, Forgacs E, Zo¨ chbauer-Mu¨ ller S, Shivakumar L, Frohman MA. (1993). Methods Enzymol., 218, 340–356. Fong K, Gao B, Randle D, Kondo M, Virmani A, Bader S, Fujimori M, Tokino T, Hino O, Kitagawa T, Imamura T, Sekido Y, Latif F, Milchgrub S, Toyooka S, Gazdar AF, Okamoto E, Mitsunobu M, Ishikawa T, Nakagama H, Lerman MI, Zabarovsky E, White M and Minna JD. (2001). Harada H, Yagura M, Matsubara K and Nakamura Y. J. Natl. Cancer Inst., 93, 691–699. (1991). Cancer Res., 51, 89–93. Cairns P, Tokino K, Eby Y and Sidransky D. (1994). Cancer Haataja L, Raffel C, Ledbetter DH, Tanigami A, Petersen D, Res., 54, 1422–1424. Heisterkamp N and Groffen J. (1997). Cancer Res., 57, Canning S and Dryja TP. (1989). Proc. Natl. Acad. Sci. USA, 32–34. 86, 5044–5048. Hahn SA, Schutte M, Hoque AT, Moskaluk CA, da Costa Chen C and Kolodner RD. (1999). Nat. Genet., 23, 81–85. LT, Rozenblum E, Weinstein CL, Fischer A, Yeo CJ, Cogen PH, Daneshvar L, Metzger AK, Duyk G, Edwards MS Hruban RH and Kern SE. (1996). Science, 271, 350–353. and Shefﬁeld VC. (1992). Am. J. Hum. Genet., 50, 584–589. Hurlin PJ, Queva C and Eisenman RN. (1997). Genes Dev., 11, Coles C, Thompson AM, Elder PA, Cohen BB, Mackenzie 44–58. IM, Cranston G, Chetty U, Mackay J, Macdonald M, Inoue H, Ishii H, Alder H, Snyder E, Druck T, Huebner K and Nakamura Y, Hoyheim B and Steel CM. (1990). Lancet, Croce CM. (1997). Proc. Natl. Acad. Sci. USA, 94, 14584– 336, 761–763. 14589. Dammann R, Li C, Yoon JH, Chin PL, Bates S and Pfeifer Ji L, Nishizaki M, Gao B, Burbee D, Kondo M, Kamibayashi GP. (2000). Nat. Genet., 25, 315–319. C, Xu K, Yen N, Atkinson EN, Fang B, Lerman MI,

Oncogene Homozygous deletion at 17p13.3 in lung cancer H Konishi el al 1905 Roth JA and Minna JD. (2002). Cancer Res., 62, Osada H, Tatematsu Y, Masuda A, Saito T, Sugiyama M, 2715–2720. Yanagisawa K and Takahashi T. (2001). Cancer Res., 61, Kamb A, Gruis NA, Weaver-Feldhaus J, Liu Q, Harshman K, 8331–8339. Tavtigian SV, Stockert E, Day RSI, Johnson BE and Phillips NJ, Ziegler MR, Radford DM, Fair KL, Steinbrueck Skolnick MH. (1994). Science, 264, 436–440. T, Xynos FP and Donis-Keller H. (1996). Cancer Res., 56, Knudson Jr AG. (1971). Proc. Natl. Acad. Sci. USA, 68, 606–611. 820–823. Pomykala HM, Bohlander SK, Broeker PL, Olopade OI and Kohno T, Otsuka T, Inazawa J, Abe T and Yokota J. (1996). Dı´ az MO. (1994). Mol. Cell Biol., 14, 7604–7610. DNA Res., 3, 421–424. Saxena A, Clark WC, Robertson JT, Ikejiri B, Oldﬁeld EH Kondo M, Ji L, Kamibayashi C, Tomizawa Y, Randle D, and Ali IU. (1992). Cancer Res., 52, 6716–6721. Sekido Y, Yokota J, Kashuba V, Zabarovsky E, Kuzmin I, Schultz DC, Vanderveer L, Berman DB, Hamilton TC, Lerman M, Roth J and Minna JD. (2001). Oncogene, 20, Wong AJ and Godwin AK. (1996). Cancer Res., 56, 6258–6262. 1997–2002. Konishi H, Nakagawa T, Harano T, Mizuno K, Saito H, Sekido Y, Ahmadian M, Wistuba II, Latif F, Bader S, Wei Masuda A, Matsuda H, Osada H and Takahashi T. (2002). MH, Duh FM, Gazdar AF, Lerman MI and Minna JD. Cancer Res., 62, 271–276. (1998). Oncogene, 16, 3151–3157. Konishi H, Takahashi T, Kozaki K, Yatabe Y, Mitsudomi T, Shigemitsu K, Sekido Y, Usami N, Mori S, Sato M, Horio Y, Fujii Y, Sugiura T, Matsuda H, Takahashi T and Takahashi Hasegawa Y, Bader SA, Gazdar AF, Minna JD, Hida T, T. (1998). Oncogene, 17, 2095–2100. Yoshioka H, Imaizumi M, Ueda Y, Takahashi M and Lee WH, Bookstein R, Hong F, Young LJ, Shew JY and Lee Shimokata K. (2001). Oncogene, 20, 4249–4257. EY. (1987). Science, 235, 1394–1399. Stack M, Jones D, White G, Liscia DS, Venesio T, Casey G, Lerman MI and Minna JD. (2000). Cancer Res., 60, Crichton D, Varley J, Mitchell E, Heighway J and 6116–6133. Stantibanez-Koref M. (1995). Hum. Mol. Genet., 4, Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, 2047–2055. Puc J, Miliaresis C, Rodgers L, McCombie R, Steck PA, Pershouse MA, Jasser SA, Yung WK, Lin H, Ligon Bigner SH, Giovanella BC, Ittmann M, Tycko B, Hibshoosh AH, Langford LA, Baumgard ML, Hattier T, Davis T, Frye H, Wigler MH and Parsons R. (1997). Science, 275, C, Hu R, Swedlund B, Teng DH and Tavtigian SV. (1997). 1943–1947. Nat. Genet., 15, 356–362. McHale JC, McKie AB, Tarttelin EE and Inglehearn CF. Tadokoro K, Fujii H, Ohshima A, Kakizawa Y, Shimizu K, (2000). Cytogenet. Cell Genet., 88, 225–229. Sakai A, Sumiyoshi K, Inoue T, Hayashi Y and Yamada M. Meroni G, Reymond A, Alcalay M, Borsani G, Tanigami A, (1992). Oncogene, 7, 1215–1221. Tonlorenzi R, Nigro CL, Messali S, Zollo M, Ledbetter DH, Takahashi T, Carbone D, Takahashi T, Nau MM, Hida T, Brent R, Ballabio A and Carrozzo R. (1997). EMBO J., 16, Linnoila I, Ueda R and Minna JD. (1992). Cancer Res., 52, 2892–2906. 2340–2343. Minna JD, Sekido Y, Fong KM and Gazdar AF. (1997). Takahashi T, Haruki N, Nomoto S, Masuda A, Saji S, Osada Cancer Principles & Practice of Oncology, Vol. 1. DeVita H and Takahashi T. (1999). Oncogene, 18, 4295–4300. VTJ, Hellman S and Rosenberg SA (eds). Lippincott-Raven: Takahashi T, Nau MM, Chiba I, Birrer MJ, Rosenberg RK, Philadelphia, pp. 849–857. Vinocour M, Levitt M, Pass H, Gazdar AF and Minna JD. Murata Y, Tamari M, Takahashi T, Horio Y, Hibi K, (1989). Science, 246, 491–494. Yokoyama S, Inazawa J, Yamakawa K, Ogawa A, Tamura K, Miwa W, Maruyama T, Sekiya T and Murakami Takahashi T and Nakamura Y. (1994). Hum. Mol. Genet., Y. (2002). J. Hum. Genet., 47, 348–354. 3, 1341–1344. Tomizawa Y, Sekido Y, Kondo M, Gao B, Yokota J, Roche J, Nishioka M, Kohno T, Takahashi M, Niki T, Yamada T, Drabkin H, Lerman MI, Gazdar AF and Minna JD. (2001). Sone S and Yokota J. (2000). Oncogene, 19, 6251–6260. Proc. Natl. Acad. Sci. USA, 98, 13954–13959. Nobori T, Miura K, Wu DJ, Lois A, Takabayashi K and Wales MM, Biel MA, el Deiry W, Nelkin BD, Issa JP, Carson DA. (1994). Nature, 368, 753–756. Cavenee WK, Kuerbitz SJ and Baylin SB. (1995). Nat. Med., Nomoto S, Haruki N, Takahashi T, Masuda A, Koshikawa T, 1, 570–577. Takahashi T, Fujii Y, Osada H and Takahashi T. (1999). Yatabe Y, Konishi H, Mitsudomi T, Nakamura S and Oncogene, 18, 7180–7183. Takahashi T. (2000). Am. J. Pathol., 157, 985–993. Oh JJ, West AR, Fishbein MC and Slamon DJ. (2002). Cancer Yokota J, Wada M, Shimosato Y, Terada M and Sugimura T. Res., 62, 3207–3213. (1987). Proc. Natl. Acad. Sci. USA, 84, 9252–9256. Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo Zhao X, Li J, He Y, Lan F, Fu L, Guo J, Zhao R, Ye Y, He J, Siprashvili Z, Mori M, McCue P, Druck T, Croce CM and M, Chong W, Chen J, Zhang L, Yang N, Xu B, Wu M, Wan Huebner K. (1996). Cell, 84, 587–597. D and Gu J. (2001). Cancer Res., 61, 7383–7387.

Oncogene