Chromosome 3p14 homozygous deletions and sequence analysis of FRA3B
Chromosome 3p14 homozygous deletions and sequence analysis of FRA3BFerenc Boldog1,[Dagger], Robert M. Gemmill1,[Dagger], James West1, Misi Robinson1,Linda Robinson1, Efang Li1, Joelle Roche1,+, Sean Todd1, Barbara Waggoner1, Ron Lundstrom2, Jan Jacobson1, Michael R. Mullokandov3, Harold Klinger3 and Harry A. Drabkin1,*
1Division of Medical Oncology, University of Colorado Health Sciences Center, 4200 E. 9th Ave., Denver, CO 80262, USA, 2Eleanor Roosevelt Institute, 1899 Gaylord, Denver, CO 80206, USA and 3Department of Molecular Genetics, Albert Einstein College of Medicine, 1300 Morris Park Ave., New York, NY 10461, USA
Received September 16, 1996;Revised and Accepted November 27, 1996
Loss of heterozygosity (LOH) involving 3p occurs in many carcinomas but is complicated by the identification of four distinct homozygous deletion regions. One putative target, 3p14.2, contains the common fragile site, FRA3B, a hereditary renal carcinoma-associated 3;8 translocation and the candidate tumor suppressor gene, FHIT. Using a ~300 kb cosmid/[lambda] contig, we identified homozygous deletions in cervix, breast, lung and colorectal carcinoma cell lines. The smallest deletion (CC19) was shown not to involve FHIT coding exons and no DNA sequence alterations were present in the transcript. We also detected discontinuous deletions as well as deletions in non-tumor DNAs, suggesting that FHIT is not a selective target. Further, we demonstrate that some reported FHIT aberrations represent normal splicing variation. DNA sequence analysis of 110 kb demonstrated that the region is high in A-T content, LINEs and MER repeats, whereas Alu elements are reduced. We note an intriguing similarity in repeat sequence composition between FRA3B and a 152 kb segment from the Fragile-X region. We also identified similarity between a FRA3B segment and a small polydispersed circular DNA. In contrast to the selective loss of a tumor suppressor gene, we propose an alternative hypothesis, that some putative targets including FRA3B may undergo loss as a consequence of genomic instability. This instability is not due to DNA mismatch repair deficiency, but may correlate in part with p53 inactivation.
LOH involving 3p occurs frequently in carcinomas of the lung, kidney, cervix, breast and other epithelial neoplasms (1 -7 ). However, 3p loss is complex, involving at least four distinct homozygously deleted regions (8 -12 ). One of the most frequently lost regions is 3p14, especially in cervical carcinomas (13 ). This region is also of interest since it contains the site of a hereditary renal carcinoma-associated translocation, t(3;8)(p14.2;q24.1) (14 ), and is the location of the most inducible common fragile site in the genome, FRA3B (15 ). We previously reported cloning of the 3;8 translocation breakpoint (16 ) and demonstrated by fluorescence in situ hybridization that FRA3B was further telomeric (17 ). These studies also suggested that breaks in FRA3B occur over a region instead of at a single site. While searching for genes in this region, we identified a homozygous deletion in the cervical carcinoma cell line, HeLa, involving marker D3S1300.
We developed a ~300 kb cosmid/[lambda] contig within FRA3B containing D3S1300. Probes from the region detected frequent homozygous deletions in cervical, lung, colorectal and breast carcinoma cell lines. Cervical carcinomas, which are associated with papilloma virus infection and p53 inactivation (18 ), were most frequently deleted and the smallest deletion occurred in cell line CC19. During our investigations, Ohta et al. reported identification of a candidate tumor suppressor gene, FHIT, which spanned the t(3;8) breakpoint and was deleted in various carcinoma cell lines (19 ). However, our analysis indicates that FHIT is not the target of these deletions. We also observed that 3p14 deletions tend not to occur in tumors with deficiencies in DNA mismatch repair.
aDefective refers to a reported deficiency in mismatch repair or microsatellite instability.bOverall, deletions (5) or rearrangements (1) were identified in 6/12 colorectal carcinoma DNAs. In addition to those listed, deletions occurred in COLO-205 and COLO-320. No deletions were identified in SKCO-1, CaCo-2, SW-1417 and T-84. cRepresents all cervical carcinoma samples examined.dPersonal communication from Dr Garret M. Hampton.NT = not tested.
Following identification of the HeLa deletion, additional cervical carcinoma lines were examined using probes spanning a ~300 kb cosmid/phage contig from the FRA3B region (Fig. 1 A and B). Figure 2 A shows an example using cosmid c136C3. Deletions were evident in DNAs from MS751, SIHA and Caski (lanes 2, 5 and 7) whereas CC19 (lane 6) shows missing and altered bands (arrow). Altogether, homozygous deletions were detected in seven out of eight lines (87.5%). Similar hybridizations were performed using DNAs from 12 colon tumor lines, and deletions or rearrangements were seen in 50% (Fig. 2 B). Other homozygous deletions (Fig. 1 C) were detected in three lung (NCI-H1648, NCI-H460 and NCI-H892) and one breast carcinoma (MDA-231) cell lines. The smallest (~40 kb) was in the cervical carcinoma line CC19. This line was characterized further by DNA sequence analysis and FHIT gene expression (see below). Two (HeLa and MDA-231) contained discontinuous deletions, which was surprising since a single deletion should have been sufficient to inactivate a tumor suppressor gene. No deletions were detected in five renal carcinoma lines (KRC/Y, CAKI-1, CAKI-2, ACHN and KV6).
Figure 4. (A) RT-PCR analysis of FHIT expression in tumor cell lines. Primers designed to amplify the coding region of FHIT were used to amplify cDNAs prepared from five cervical carcinoma lines, one renal carcinoma line (KRC/Y) and an E1A-transformed embryonal kidney line (E293), as indicated. Normal products of 638 bp (arrow) were observed in three cell lines, including CC19 with a 40 kb homozygous deletion. A normal sized product was also observed in MS751, although it is amplified weakly. In (B), evidence is presented for alternative splicing of FHIT mRNA in normal tissues. RT-PCR analysis of FHIT expression in fetal brain, adult kidney, E293 cells (whole cells and cytoplasmic fraction), lung tumor A549 and kidney carcinoma cell line KRC/Y were performed as in (A). Both normal sized products (arrow marked N) and smaller products (left arrow) were observed.
In cervical and colorectal carcinoma cell lines, we observed a trend of inverse correlation between the presence of a 3p14 deletion and reported microsatellite instability (22 -26 ). However, for some lines, we were unable to discern the replication error (RER) status from the literature. Therefore, we subcloned selected lines and tested 10 clones each with up to six microsatellite loci (D3S1300, D3S1210, D3S1286, D3S1233 and AFM320yb5). Instability was accepted if new bands appeared from two or more loci, although there were no cases of only one alteration. Results for samples where we have information on both DNA mismatch repair and 3p14 deletion are shown in Table 1 . While the number of RER+ lines is small, both 3p14 deletions and the RER phenotype are usually discordant. One simple conclusion is that defects associated with microsatellite instability alone are not responsible for the observed deletions. We also observed the highest incidence of deletions in cervical carcinoma lines, where p53 alterations appear very common (27 ,28 ). This may relate to the reduced frequency of deletions observed in RER+ colorectal tumors (see Discussion).
To examine the role of the FHIT gene in these deletions, we derived primers (FHIT-5' and FHIT-3') for a one-stage RT-PCR amplification of the entire coding region (exons 5-9). An expected 638 bp product was amplified after 35 cycles from embryonal kidney 293 cells (Fig. 4 A, lane 7). This was used as a probe against the contig which showed that only a single FHIT exon, located adjacent to D3S1300, was present (Fig. 1 B). Comparison of our map with that described by Ohta et al. (19 ) allowed us to conclude that this was exon 5 (also confirmed by DNA sequencing, not shown) and demonstrated that several tumor deletions exclusively affected large introns (Fig. 1 C). This occurred in CC19 and SIHA (cervix), NCI-H1648 and NCI-H460 (lung) and MDA-231 (breast).
Analysis of FHIT in the tumor lines shown in Figure 4 A demonstrated that 3/6, including CC19, contained normal sized bands (arrow). To characterize further the FHIT product from CC19, four independent isolates were sequenced. Each contained only normal sequences comprising coding exons 5-9. Thus, the homozygous deletion had no apparent effect on FHIT mRNA, suggesting either a different target gene or unselected genomic instability. Additionally, a faint normal sized product was seen in MS751 (Fig. 4 A, lane 5), which by Southern blot contains a homozygous deletion including exon 5. A possible explanation is that the deletion is heterogeneous within the cell population suggesting it occurred during culture.
The cervical carcinoma lines C33A and SIHA contained multiple smaller RT-PCR products with no detectable wild-type product. While C33A had an exon 5 deletion that could explain one smaller band (Fig. 1 C), the multiplicity of products suggested alternative splicing. The SIHA deletion does not affect an exon although there may be non-recognized discontinuous or overlapping bi-allelic deletions. Despite these obvious differences, the RT-PCR products appeared identical (Fig. 4 A, lanes 2 and 3). To examine this, we amplified FHIT coding exons from RNA or cDNAs prepared from normal (fetal brain, adult kidney), immortalized (E293) and tumor (A549, KRC/Y) samples (Fig. 4 B). While normal FHIT (arrow marked N) was observed in each, a smaller product was also seen which predominated in the normal fetal brain cDNA library. When this was cloned and sequenced, exon 8 containing the conserved histidine triad motif and possible zinc-binding site (29 ) was missing (not shown). This variation had been reported to represent an aberration in squamous cell carcinomas of the head and neck (30 ). Additional larger products were identified from the adult kidney cDNA library although these have not been characterized. Thus, alternative splicing definitely occurs in normal tissues.
Since CC19 contained the smallest deletion, we initiated large-scale DNA sequencing studies to identify new genes or structural features that might be responsible for the genomic instability. Six genome equivalents of sequence were obtained from c81E9, c120E3, c136C3 and [lambda]33 and the data assembled. Gaps were closed using directed primers and a 1.3 kb clone gap between c136C3 and [lambda]33 was closed by PCR amplification using [lambda]48 (Fig. 1 B) as template. This resulted in 110.4 kb of contiguous sequence (GenBank U66722) at an accuracy of ~99.6%. Predicted restriction maps from the assembled sequence also matched those determined experimentally. Sequence analysis (shown schematically in Fig. 5 ) included similarity searches, repetitive sequence identification and exon prediction.
Figure 5. Assembled sequence data totaling 110 435 bp. GeneMark and GRAIL2 predicted exons are indicated (filled boxes = direct strand; open boxes = complementary strand). Repetitive sequences homologous to LINE (L1) and Alu elements are designated by boxes filled with diagonal lines and shading, respectively. Other repetitive elements are labeled. These largely consist of MER elements (medium reiterated repeats) and LTRs (long terminal repeats). Information about particular repetitive elements can be found in Repbase (ftp://ncbi.nlm.nih.gov/repository/repbase). Unique features of the sequence are indicated on the lowest row.
Overall, the sequence is AT-rich (61.1%) and very depleted in CG dinucleotides. No identities to known genes were seen. GRAIL2 predicted 15 exons of which six had moderate (<0.4), five had good (0.4-0.6) and four had excellent (>0.6) scores. Four putative exons were coincident with repeat sequences and two others (positions 8.0 and 71.5 kb) were coincident with Genemark predicted coding segments. Genemark (31 ) was utilized in order to implement a matrix for higher A-T content regions. While 61 potential coding segments were identified, no significant similarities were observed. Many predicted coding segments clearly occurred within LINE and MER elements, and others not directly within repeats nevertheless demonstrated similarity to LINEs. Four putative exons were clustered near position 107 kb, one of which showed perfect identity with a partially sequenced cDNA, EST-N70372. However, Northern analysis failed to identify a transcript from this region (data not shown) and the cDNA sequence included a portion of the LINE element at position 108 kb. Based on end sequences from the cDNA clone and insert length, the cDNA was co-linear with genomic DNA. Further, the 3' end of the cDNA was coincident with a poly(A) tract in genomic DNA, all suggesting this represented an unprocessed transcript, or more likely resulted from false priming of contaminating DNA. BLASTN searches determined two additional regions with similarity to non-annotated cDNAs (positions 58.2 and 64.3 kb). However, neither showed similarity to known genes nor were directly superimposed on predicted exons. That these sequences were observed adjacent to, rather than superimposed on, predicted exons could be due to conservative prediction algorithms used which underestimate the extent of many exons. Alternatively, the homologies could be due to infrequent repeats. We note the presence of five remaining putative exons, denoted in Figure 5 , having high probability scores and which do not overlap repeats.
Analysis with Pythia (32 ) showed that 20.2% of the sequenced region is comprised of known repeats, a level comparable with other regions chosen for comparison (Table 2 ). However, in this 110.4 kb, LINE and MER elements make up the bulk of repeats. Intriguingly, the repeat composition is very similar to a 152 kb sequence from the Fragile-X region (33 ) and differs from arbitrarily selected segments in 3p21.3 and 4p16.3 (Table 2 ). In particular, [lambda]33 (from position 96 to 110.4 kb) within the CC19 deletion is nearly identical in LINE element composition to the Fragile-X region. Similarly, both Fragile-X and FRA3B have a low level of predicted coding regions. Speculatively, these similarities may influence the observed tendency of both regions to undergo breakage. However, there are obvious differences between the two sites, notably the presence of a triplet repeat and methylated CpG island in Fragile-X (34 ) and the common versus rare nature of FRA3B.
Position 38 kb contains a cluster of terminal deletion breakpoints induced by aphidicolin treatment (GenBank U46001) further linking our sequence to FRA3B. Flanking these aphidicolin breakpoints is a long polypurine tract and two extended variable dinucleotide repeats. The telomeric breakpoint of the CC19 deletion lies within a 1.0 kb region at position 76 kb between L1 and MER/LTR elements. Within the CC19 deletion region, near position 86 kb, is an HPV16 integration site from a cervical carcinoma (35 ). It is of note that this integration was associated with an interstitial deletion (35 ,36 ) which we now know would not affect FHIT coding sequences. Interestingly, a very significant similarity (1e-77) to a small polydispersed circular (spc) DNA was observed at 106 kb (GenBank X96885). Figure 6 shows a FASTA alignment over 585 bp. The region is 72.1% identical and contains interspersed blocks having up to 89% identity (e.g., spcDNA bp 121-207). This spcDNA, which appears non-repetitive, was isolated from a tuberous sclerosis-associated angiofibroma (I. Hinkel-Schreiner, Ph.D. Thesis). Characteristics of spcDNAs include derivation from chromosomal sequences (37 ), association with clustered repeats [such as [beta]-satellites and other clustered elements although a single family member may be predominantly involved (38 )] and elevation in conditions associated with genomic instability such as Fanconi's anemia (39 ). spcDNAs are also increased by DNA-damaging agents (37 ) and inhibitors of DNA and protein synthesis including the fragile site inducer aphidicolin (40 ). Given the limited number of spcDNAs that have been sequenced, this similarity may be biologically important.
Using a ~300 kb cosmid/[lambda] contig, located ~150 kb telomeric to the 3;8 translocation breakpoint, we have identified homozygous deletions in various carcinoma cell lines that overlap the most inducible common fragile site in the genome, FRA3B. From various aphidicolin-induced breakpoints and plasmid or viral integration sites (Fig. 1 ), FRA3B represents a region rather than a single site. Studies by Wilke et al. (36 ) and Smith et al. (41 ) indicate that some clustering of breakpoints may occur. However, their rearrangements were induced by aphidicolin in a single chromosome 3-containing hybrid and, unlike the interstitial deletions we observed in tumor and non-tumor samples, appear to represent terminal breaks. Where we have defined the boundaries for the carcinoma-associated deletions accurately, one or both are contained within the FRA3B region.
Our initial hypothesis was that the smallest deletion (CC19) would contain elements of a tumor suppressor gene. During our studies, Ohta et al. (19 ) identified the FHIT gene with reported abnormalities in RT-PCR products. However, FHIT has similarity to a yeast di-adenosine hydrolase which would represent an unexpected function for a tumor suppressor gene. Our results indicate that FHIT is not the target of these deletions. First, the CC19 deletion does not involve FHIT coding sequences (Fig. 1 B and C) and, based on RT-PCR and cDNA sequence analysis, the coding portion of the FHIT transcript is normal. Second, we have observed deletions in genomic DNAs from non-tumor sources (Fig. 3 A). The somatic cell hybrid 2A3CT was formed by spontaneous terminal deletion at 3p21.3 from a non-tumor-derived chromosome 3 hybrid, UCTP2A3 (11 ). However, 2A3CT also acquired a 3p14 interstitial deletion overlapping an aphidicolin-induced breakpoint, a pSV2neo integration into FRA3B (20 ) and the telomeric borders of several carcinoma-associated deletions. We also identified an overlapping deletion in an unselected subclone from YAC 74B2. While neither the hybrid nor YAC 3p14 DNA segments are in a `normal' background, they clearly are unselected from a tumorigenic standpoint. We note that `hotspots' of recombination in human DNA can be maintained in a yeast background (42 ), thus it is not unreasonable that unstable regions may behave similarly or more so. Third, discontinuous deletions appear common in this region (i.e. HeLa, MDA231), both from our analysis and from that reported by Ohta et al. (19 ). Multiple deletions might be expected if there were no common target gene and if the region was unstable. Fourth, we have observed that FHIT undergoes alternative splicing in normal tissues (Fig. 4 B) which explains some previously reported abnormal PCR products (19 ,30 ,43 ). Thiagalingam et al. (44 ) recently reported lack of FHIT involvement in colorectal carcinomas and suggested that PCR artifacts might be responsible for some previously observed alterations (19 ,43 ). It also seems likely that many deletions would have been missed in their study since only a few markers were tested. One of the possible features that suggested FHIT could be a tumor suppressor gene was that it crossed the hereditary renal carcinoma-associated 3;8 breakpoint (19 ). However, we found no alterations in RT-PCR products from five renal carcinoma cell lines. Moreover, Bugert et al. (manuscript submitted) have observed normal FHIT transcripts and no point mutations in a large series of renal cancers. Thus, FHIT does not appear to be involved in renal carcinoma. With respect to other possible target genes, from our sequencing studies we identified a 100% identity to two expressed sequence tags from a liver/spleen library. However, we determined that both clones were identical, were not expressed using a commercial Northern blot (Clontech), were co-linear with genomic DNA containing a poly(A) tract corresponding to their 3' end, encoded no significant open reading frame and overlapped a partial LINE element.
While other tumor suppressor genes may exist within FRA3B, an alternative possibility is that the deletions are due to primary genomic instability affecting a particularly susceptible region. This hypothesis is consistent with several of the observations reported here including their discontinuity and occurrence in non-tumor cell lines. By using numerous probes, we were able to identify a high frequency of homozygous deletions, especially in cervical carcinomas where p53 inactivation is very common (18 ). In this regard, it is interesting that we observed an inverse correlation between 3p14 deletions and microsatellite instability (RER+). Importantly, p53 mutations, which have been shown to destabilize the genome (45 ,46 ), appear infrequent in RER+ colorectal carcinomas and gastric tumors [(47 ,48 ) and P. Cottu, presented at Cancer and the Cell Cycle, Lausanne, Switzerland 1996]. Thus, these findings would be consistent, at least in part, with 3p14 deletions resulting from the genomic instability which accompanies p53 inactivation. We note that although CC19 is HPV negative, it expresses a mutant p53 protein (53 ).
Figure 6. FASTA alignment of 585 bp of spcDNA clone (X96885) and FRA3B.
What have we learned from the DNA sequence analysis to date? First, is that the region is high in A-T content with frequent LINEs and MER repeats, and is conversely low in Alu sequences and confirmed genes. In contrast to the reported rare folate-sensitive sites which are associated with expanded CGG repeats (49 ,50 ), FRA3B does not contain an expanded triplet repeat or methylated CpG island; nor did we identify any telomeric repeats which have been suggested as a possible cause of breaks (51 ). However, we do note that over the 110.4 kb region there is an overall repeat sequence similarity to 152 kb from Fragile-X. In vitro, expanded CGG repeats have been shown to inhibit DNA replication (52 ). Whether or not specific sequences within FRA3B may have a similar effect on replication awaits experimental testing. Our discovery of a strong spcDNA homology in FRA3B may therefore not be coincidental. While we do not know if the spcDNA site is a primary cause of FRA3B instability or simply a marker for this property, the DNA sequences reported should provide the means to test this.
The sequences used for comparison were derived from Xq27.3 (contains the Fragile-X region, GenBank L29074) and 4p16.3 (in the region of Huntington's disease, GenBank Z69837). The sequence from 3p21.3 was downloaded from ftp://genome.wustl.edu/pub/gsc1/sequence/st.louis/human/shotgun/3/H_LUCA14.seq.
Cell lines. Breast, colon and cervical carcinoma lines were obtained from the American Type Culture Collection. The cervical carcinoma cell line, CC19, was established as described (53 ). Lung tumor lines were obtained through the Colorado Lung Cancer SPORE Tissue Culture Core Laboratory. Normal cell lines included the human lymphoblastoid cell lines TL8229 and AG4103 and the E1A-immortalized human embryonic kidney cell line, E293. The somatic cell hybrids have been described previously (54 ). Libraries. The gridded chromosome 3-specific cosmid library (55 ) was obtained from Lawrence Livermore National Laboratories. The 850A6 YAC subclone library was previously described (16 ). Fetal brain and adult kidney cDNA libraries were obtained from Clontech. DNA and RNA isolations. DNA was isolated from cell lines by standard methods. Cosmid DNAs were isolated using alkaline lysis; preparations used for DNA sequence analysis were purified further by CsCl gradient centrifugation. DNA was isolated from single or pooled phage clones by the Grossberger method (56 ). RNA was isolated from cell lines when the cultures reached 90% of confluency using the RNA-STAT-60 kit from Tel-Test, Inc. (Friendswood, TX). Library screening, contig assembly and hybridization analysis. The gridded cosmid library was spotted onto filters at high density (1536 clones per filter) and hybridized using standard techniques with a 370 kb MluI fragment derived from YAC 74B2. The 850A6 YAC phage library (16 ) was screened using the same probe. Resulting clones were assembled into a contiguous segment by hybridization analysis using end probes and total inserts followed by analysis with the software tool SEGMAP. Complete and partial digestion analyses (57 ) were used to restriction map the central 170 kb of the contig prior to DNA sequence analysis.
Biotinylated primers were obtained from Research Genetics, Huntsville, AL. PCR amplifications were performed with 40 ng of template DNA utilizing hot start and touch down procedures. After separation on denaturing polyacrylamide gels, PCR products were detected using the New England BioLabs Phototopetm Detection Kit. Alleles were scored by visual inspection of band patterns.
RT-PCR was performed using primers FHIT-5' (5'-CTCGAA- TTCTTAGACCCTCTATAAAAGC-3') and FHIT-3' (5'-CTGATTCAGTTCCTCTTG-3') derived from non-coding exons 4 and 10, respectively. First strand synthesis was accomplished with 1-3 mg of total RNA and the Superscript II kit (Life Technologies). Subsequent amplification utilized one-fifth of the reverse transcriptase reaction together with the FHIT primers. Standard PCR conditions of 94oC denaturation (1 min), 55oC annealing (1 min) and 72oC elongation for 35 cycles were employed. PCR products were subcloned into the EcoRI site of pBluescript II SK+ using an introduced EcoRI site present in the FHIT-5' primer and a natural site located 21 bp downstream of the FHIT stop codon.
Cosmid and phage clones were sequenced using a random shotgun subcloning and end-sequencing strategy. Clone DNA was sonicated and size selected by LMP-agarose gel electrophoresis. Recovered fragments of 1-2 kb were end-repaired with Klenow fragment of Escherichia coli DNA polymerase I and T4 DNA polymerase, ligated into the phosphatased EcoRV site of pBluescript II and transformed into E.coli DH10B. AmpR/[beta]-Gal- subclones were grown in 3 ml of TB for isolation of sequencing templates. To eliminate vector sequences from the subclone library, phage inserts were amplified by long-range PCR. Inserts were gel purified, 32P labeled and hybridized to subclone libraries. Positive subclones were sequenced using an ABI 373 or 377 and the ABI Prism dye terminator cycle sequencing kit. Chromatograms were transferred to a SUN workstation, analyzed and assembled using PHRED and PHRAP (from Dr Phil Green). Gaps were closed by primer walking. Based on independently obtained overlapping contigs, as well as analysis of cosmid vector sequences in some subclones, sequencing accuracy was at least 99.6%.
Assembled sequences were analyzed for database similarities by BLASTN and BLASTX, searching the nr and dbest databases with default parameters. Strongly similar sequences (P <e-30) were retrieved from GenBank for further analysis. Homologies to repeat sequences were found using Pythia (32 ). Potential exons were identified using GRAIL2 (58 ) and GeneMark (31 ) programs. GeneMark was run using 5th order matrices trained on sequences with GC content similar to the sequences checked. Sequence alignments were prepared using the FASTA program from the Wisconsin Package (GCG).
This work was supported by DAMD17-94-J-4391 (HAD, RMG), CA51395 (HPK, MRM) and the DNA sequencing studies were supported by HG000358 (H.A.D., R.M.G.). We gratefully acknowledge support from the Colorado Cancer League (2531718) to F.B. Additional Core support was provided through the University of Colorado Cancer Center (CA 46934) and by Dr Wilbur Franklin through the Lung Cancer Spore.
1 Ogawa, O., Kakehi, Y., Ogawa, K., Koshiba, M., Sugiyama, T. and Yoshida, O. (1991) Allelic loss at chromosome 3p characterizes clear cell phenotype of renal cell carcinoma. Cancer. Res., 51, 949-953.MEDLINE Abstract
2 Devilee, P., van den Broek, M., Kuipers-Dijkshoorn, N., Kolluri, R., Khan, P.M., Pearson, P.L. and Cornelisse, C.J. (1989) At least four different chromosomal regions are involved in loss of heterozygosity in human breast carcinoma. Genomics, 5, 554-560.MEDLINE Abstract
3 Hibi, K., Takahashi, T., Yamakawa, K., Ueda, R., Sekido, Y., Ariyoshi, Y., Suyama, M., Takagi, H. and Nakamura, Y. (1992) Three distinct regions involved in 3p deletion in human lung cancer. Oncogene, 7, 445-449.MEDLINE Abstract
4 Lothe, R.A., Fossa, S.D., Stenwig, A.E., Nakamura, Y., White, R., Borresen, A.L. and Brogger, A (1989) Loss of 3p or 11p alleles is associated with testicular cancer tumors. Genomics, 5, 134-138.MEDLINE Abstract
5 Yang-Feng, T.L., Han, H., Chen, K.C., Li, S.B., Claus, E.B., Carcangiu, M.L., Chambers, S.K., Chambers, J.T. and Schwartz, P.E. (1993) Allelic loss in ovarian cancer. Int. J. Cancer, 54, 546-551.MEDLINE Abstract
6 Kohno, T., Takayama, H., Hamaguchi, M., Takano, H., Yamaguchi, N., Tsuda, H., Hirohashi, S., Vissing, H., Shimizu, M., Oshimura, M. et al. (1993) Deletion mapping of chromosome 3p in human uterine cervical cancer. Oncogene, 8, 1825-1832MEDLINE Abstract
7 Maestro, R., Gasparotto, D., Vukosavljevic, T., Barzan, L., Sulfaro, S. and Boiocchi, M. (1993) Three discrete regions of deletion at 3p in head and neck cancers. Cancer Res., 53, 5775-5779.MEDLINE Abstract
8 Yamakawa, K., Takahashi, T., Horio, Y., Murata, Y., Takahashi, E., Hibi, K., Yokoyama, S., Ueda, R. and Nakamura, Y. (1993) Frequent homozygous deletions in lung cancer cell lines detected by a DNA marker located at 3p21.3-p22. Oncogene, 8, 327-330.MEDLINE Abstract
9 Drabkin, H.A., Mendez, M.J., Rabbitts, P.H., Varkony, T., Bergh, J., Schlessinger, J., Erickson, P. and Gemmill, R.M. (1992) Characterization of the submicroscopic deletion in the small- cell lung carcinoma (SCLC) cell line U2020. Genes, Chromosomes Cancer, 5, 67-74.MEDLINE Abstract
10 Rabbitts, P., Bergh, J., Douglas, J., Collins, F. and Waters, J. (1990) A submicroscopic homozygous deletion at the D3S3 locus in a cell line isolated from a small cell lung carcinoma.. Genes, Chromosomes Cancer, 2, 231-238.MEDLINE Abstract
11 Roche, J., Boldog, F., Robinson, M., Robinson, L., Varella-Garcia, M., Swanton, M., Waggoner, B., Fishel, R., Franklin, W., Gemmill, R. and Drabkin, H. (1996) Distinct 3p21.3 deletions in lung cancer and identification of a new human semaphorin. Oncogene, 12, 1289-1297.MEDLINE Abstract
12 Lisitsyn, N.A., Lisitsina, N.M., Dalbagni, G., Barker, P., Sanchez, C.A., Gnarra, J., Linehan, W.M., Reid, B.J. and Wigler, M.H. (1995) Comparative genomic analysis of tumors: detection of DNA losses and amplification. Proc. Natl Acad. Sci. USA, 92, 151-155.MEDLINE Abstract
13 Yokota, J., Tsukada, Y., Nakajima, T., Gotoh, M., Shimosato, Y., Mori, N., Tsunokawa, Y., Sugimura, T. and Terada, M. (1989) Loss of heterozygosity on the short arm of chromosome 3 in carcinoma of the uterine cervix. Cancer Res., 49, 3598-3601.MEDLINE Abstract
14 Cohen, A.J., Li, F.P., Berg, S., Marchetto, D.J., Tsai, S., Jacobs, S.C. and Brown, R.S. (1979) Hereditary renal-cell carcinoma associated with a chromosomal translocation. N. Engl. J. Med., 301, 592-595.MEDLINE Abstract
15 Glover, T.W., Berger, C., Coyle, J. and Echo, B. (1984) DNA polymerase alpha inhibition by aphidicolin induces gaps and breaks at common fragile sites in human chromosomes. Hum. Genet., 67, 136-142.MEDLINE Abstract
16 Boldog, F.L., Gemmill, R.M., Wilke, C.M., Glover, T.W., Nilsson, A.S., Chandrasekharappa, S.C., Brown, R.S., Li, F.P. and Drabkin, H.A. (1993) Positional cloning of the hereditary renal carcinoma 3;8 chromosome translocation breakpoint. Proc. Natl Acad. Sci. USA, 90, 8509-8513.MEDLINE Abstract
17 Wilke, C.M., Guo, S.W., Hall, B.K., Boldog, F., Gemmill, R.M., Chandrasekharappa, S.C., Barcroft, C.L., Drabkin, H.A. and Glover, T.W. (1994) Multicolor FISH mapping of YAC clones in 3p14 and identification of a YAC spanning both FRA3B and the t(3;8) associated with hereditary renal cell carcinoma. Genomics, 22, 319-326.
18 Milde-Langosch, K., Albrecht, K., Joram, S., Schlechte, H., Giessing, M. and Loning, T. (1995) Presence and persistence of HPV infection and p53 mutation in cancer of the cervix uteri and the vulva. Int. J. Cancer, 63, 639-645.MEDLINE Abstract
19 Ohta, M., Inoue, H., Cotticelli, M.G., Kastury, K., Baffa, R., Palazzo, J., Siprashvili, Z., Mori, M., McCue, P., Druck, T., Croce, C.M. and Huebner, K. (1996) The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell, 84, 587-597.
20 Rassool, R.V., Le Beau, M.M., Shen, M.-L., Neilly, M.E., Espinosa, R, III, Ong, S.T., Boldog, F., Drabkin, H., McCarroll, R. and McKeithan, T.W. (1996) Direct cloning of DNA sequences from the common fragile site region at chromosome band 3p14.2. Genomics, in press,
21 Rassool, F.V., McKeithan, T.W., Neilly, M.E., van Melle, E., Espinosa, R. and LeBeau, M.M. (1991) Preferential integration of marker DNA into the chromosomal fragile site at 3p14: an approach to cloning fragile sites. Proc. Natl Acad. Sci. USA, 88, 6657-6661.MEDLINE Abstract
22 Bhattacharyya, N.P., Ganesh, A., Phear, G., Richards, B., Skandalis, A. and Meuth, M. (1996) Molecular analysis of mutations in mutator colorectal carcinoma cell lines. Hum. Mol. Genet., 4, 2057-2064.
23 Boyer, J.C., Umar, A., Risinger, J.I., Lipford, J.R., Kane, M., Yin, S., BarrettJ.C., Kolodner, R.D. and Kunkel, T.A. (1995) Microsatellite instability, mismatch repair deficiency, and genetic defects in human cancer cell lines. Cancer Res., 55, 6063-6070.MEDLINE Abstract
24 da Costa, L.T., Liu, B., el-Deiry, W., Hamilton, S.R., Kinzler, K.W., VogelsteinB., Markowitz, S., Willson, J.K., de la Chapelle, A., Downey, K.M. et al. (1995) Polymerase delta variants in RER colorectal tumours [letter]. Nature Genet., 9, 10-11.MEDLINE Abstract
25 Umar, A., Boyer, J.C., Thomas, D.C., Nguyen, D.C., Risinger, J.I., Boyd, J., Ionov, Y., Perucho, M. and Kunkel, T.A. (1994) Defective mismatch repair in extracts of colorectal and endometrial cancer cell lines exhibiting microsatellite instability. J. Biol. Chem., 269, 14367-14370.MEDLINE Abstract
26 Larson, A.A., Kern, S., Sommers, R.L., Yokota, J., Cavenee, W.K. and Hampton, G.M. (1996) Analysis of replication error (RER+) phenotypes in cervical carcinoma. Cancer Res., 56, 1426-1431.MEDLINE Abstract
27 Johnson, E.S., Ma, P.C., Ota, I.M. and Varshavsky, A. (1995) A proteolytic pathway that recognizes ubiquitin as a degradation signal. J. Biol. Chem., 270, 17442-17456.MEDLINE Abstract
28 Band, V., Dalal, S., Delmolino, L. and Androphy, E.J. (1993) Enhanced degradation of p53 protein in HPV-6 and BPV-1 E6-immortalized human mammary epithelial cells. EMBO J., 12, 1847-1852.MEDLINE Abstract
29 Lima, C.D., Klein, M.G., Weinstien, I.B. and Hendrickson, W.A. (1996) Three-dimensional structure of human protein kinase C interacting protein 1, a member of the HIT family of proteins. Proc. Natl Acad. Sci. USA, 93, 5357-5362.MEDLINE Abstract
30 Virgilio, L., Shuster, M., Gollin, S.M., Veronese, M.L., Ohta, M., Huebner, K. and Croce, C.M. (1996) FHIT gene alterations in head and neck squamous cell carcinomas. Proc. Natl Acad. Sci. USA, 93, 9770-9775.MEDLINE Abstract
31 Borodovsky, M. and McIninch, J. (1993) GenMark: parallel gene recognition for both DNA strands. Comput. Chem., 17, 123-133.
32 Milosavljevic, A. (1994) InImperial Cancer Research Fund Handbook of Genome Analysis. Vol. 1, Blackwell Scientific Publications.
33 Verkerk, A.J., Pieretti, M., Sutcliffe, J.S., Fu, Y.H., Kuhl, D.P., Pizzuti, A., Richards, S., Victoria, M.F., Zhang, F.P. et al. (1991) Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome. Cell, 65, 905-914.MEDLINE Abstract
34 Kremer, E.J., Pritchard, M., Lynch, M., Yu, S., Holman, K., Baker, E., Warren, S. T., Sutherland, G.R. and Richards, R.I. (1991) Mapping of DNA instability at the fragile X to a trinucleotide repeat sequence p(CCG)n. Science, 252, 1711-1714.MEDLINE Abstract
35 Wagatsuma, M., Hashimoto, K. and Matsukura, T. (1990) Analysis of integrated human papillomavirus type 16 DNA in cervical cancers: amplification of viral sequences together with cellular flanking sequences. J. Virol., 64, 813-821.MEDLINE Abstract
36 Wilke, C.M., Hall, B.K., Hoge, A., Paradee, W., Smith, D.I. and Glover, T.W. (1996) FRA3B extends over a broad region and contains a spontaneous HPV16 integration site: direct evidence for the coincidence of viral integration sites and fragile sites. Hum. Mol. Genet., 5, 187-195.MEDLINE Abstract
37 Cohen, S. and Lavi, S. (1996) Induction of circles of heterogeneous sizes in carcinogen-treated cells: two-dimensional gel analysis of circular DNA molecules. Mol. Cell. Biol., 16, 2002-2014.MEDLINE Abstract
38 Misra, R., Shih, A., Rush, M., Wong, E. and Schmid, C.W. (1987) Cloned extrachromosomal circular DNA copies of the human transposable element THE-1 are related predominantly to a single type of family member. J. Mol. Biol.,196, 233-243.MEDLINE Abstract
39 Motejlek, K., Schindler, D., Assum, G. and Krone, W. (1993) Increased amount and contour length distribution of small polydisperse circular DNA (spcDNA) in Fanconi anemia. Mutat. Res., 293, 205-214.MEDLINE Abstract
40 Sunnerhagen, P., Sjoberg, R.M. and Bjursell, G. (1989) Increase of extrachromosomal circular DNA in mouse 3T6 cells on perturbation of DNA synthesis: implications for gene amplification. Somat. Cell. Mol. Genet., 15, 61-70.MEDLINE Abstract
41 Paradee, W., Mullins, C., He, Z., Glover, T., Wilke, C., Opalka, B. and Schutte, J. (1995) Precise localization of aphidicolin-induced breakpoints on the short arm of human chromosome 3. Genomics, 27, 358-361.MEDLINE Abstract
42 Klein, S., Zenvirth, D., Sherman, A., Ried, K., Rappold, G. and Simchen, G. (1996) Double-strand breaks on YACs during yeast meiosis may reflect meiotic recombination in the human genome. Nature Genet., 13, 481-484.MEDLINE Abstract
43 Sozzi, G., Veronese, M.L., Negrini, M., Baffa, R., Cotticelli, M.G., Inoue, H., Pilotti, S., De Gregorio, L., Pastorino, U., Pierotti, M. A., OhtaM., Huebner, K. and Croce, C.M. (1996) The FHIT gene 3p14.2 is abnormal in lung cancer. Cell, 85, 17-26.MEDLINE Abstract
44 Thiagalingam, S., Lisitsyn, N.A., Hamaguchi, M., Wigler, M.H., Willson, J.K.V., Markowitz, S.D., Leach, F.S., Kinzler, K.W. and Vogelstein, B. (1996) Evaluation of the FHIT gene in colorectal cancers. Cancer Res., 56, 2936-2939.MEDLINE Abstract
45 Yin, Y., Tainsky, M.A., Bischoff, F.Z., Strong, L.C. and Wahl, G.M. (1992) Wild-type p53 restores cell cycle control and inhibits gene amplification in cells with mutant p53 alleles. Cell, 70, 937-948.MEDLINE Abstract
46 Livingstone, L.R., White, A., Sprouse, J., Livanos, E., Jacks, T. and Tlsty, T.D. (1992) Altered cell cycle arrest and gene amplification potential accompany loss of wild-type p53. Cell, 70, 923-935.MEDLINE Abstract
47 Renault, B., Calistri, D., Buonsanti, G., Nanni, O., Amadori, D. and Ranzani, G.N. (1996) Microsatellite instability and mutations of p53 and TGF-beta RII genes in gastric cancer. Hum. Genet., in press. MEDLINE Abstract
48 Wu, C., Akiyama, Y., Imai, K., Miyake, S., Nagasaki, H., Oto, M., Okabe, S., Iwama, T., Mitamura, K. and Masumitsu, H. (1994) DNA alterations in cells from hereditary non-polyposis colorectal cancer patients. Oncogene, 9, 991-994.MEDLINE Abstract
49 Pintado, E., de Diego, Y., Hmadcha, A., Carrasco, M., Sierra, J. and Lucas, M. (1995) Instability of the CGG repeat at the FRAXA locus and variable phenotypic expression in a large fragile X pedigree. J Med. Genet, 32, 907-908.MEDLINE Abstract
50 Nancarrow, J.K., Holman, K., Mangelsdorf, M., Hori, T., Denton, M., Sutherland, G.R. and Richards, R.I. (1995) Molecular basis of p(CCG)n repeat instability at the FRA16A fragile site locus. Hum. Mol. Genet., 4, 367-372.MEDLINE Abstract
51 Bouffler, S., Silver, A., Papworth, D., Coates, J. and Cox, R. (1993) Murine radiation myeloid leukaemogenesis: relationship between interstitial telomere-like sequences and chromosome 2 fragile sites. Genes, Chromosomes Cancer, 6, 98-106.MEDLINE Abstract
52 Usdin, K. and Woodford, K.J. (1995) CGG repeats associated with DNA instability and chromosome fragility form structures that block DNA synthesis in vitro. Nucleic Acids Res., 23, 4202-4209.
53 Mullokandov, M.R., Kholodilov, N.G., Atkin, N.B., Burk, R.D., Johnson, A.B. and Klinger, H.P. (1996) Genomic alterations in cervical carcinoma: losses of chromosome heterozygosity and human papilloma virus tumor status. Cancer Res., 56, 197-205.MEDLINE Abstract
54 Drabkin, H.A., Wright, M., Jonsen, M., Varkony, T., Jones, C., Sage, M., Gold, S. and Erickson, P. (1990) Development of a somatic cell hybrid mapping panel and molecular probes for human chromosome 3. Genomics, 8, 435-446.
55 Ioannou, P.A., Amemiya, C.T., Garnes, J., Kroisel, P.M., Shizuya, H., Chen, C., Batzer, M.A. and de Jong, P.J. (1994) A new bacteriophage P1-derived vector for the propagation of large human DNA fragments. Nature Genet., 6, 84-89.MEDLINE Abstract
56 Grossberger, D. (1987) Minipreps of DNA from bacteriophage lambda. Nucleic Acids Res., 15, 6737.MEDLINE Abstract
57 Sambrook, J. Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
58 Xu, Y., Einstein, J.R., Mural, R.J., Shah, M. and Uberbacher, E.C. (1994) Proceedings of the 2nd International Conference on Intelligent Systems for Molecular Biology. Vol. 1, AAAI Press, pp. 376-384.
Sequencing of the CC19 deletion region has been completed with the finalization of lambda 94, a phage proximal to lambda 33. This 18 810 bp sequence maintains the same characteristics as the balance of FRA3B. It is 63% AT and 27% is comprised of known repeats with the L1 type elements predominating.
+Present address: CNRS URA1172, IBMIG, Universite de Poitiers, 40 Av. du Recteur Pineau, 86022 Poitiers Cédex, France
}These authors contributed equally to this work
This page is maintained by OUP admin. Last updated Fri Jan 31 11:45:11 GMT 1997. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996