¹Department of Pathology, USC/Norris Comprehensive Cancer Center, USC School of Medicine, Los Angeles, California 90033, USA

²Department of Urology, USC/Norris Comprehensive Cancer Center, USC School of Medicine, Los Angeles, California 90033, USA

^aAuthor for correspondence

Multiple distinct regions of chromosome 6 are frequently affected by losses of heterozygosity in primary human ovarian carcinomas. We introduced a normal human chromosome 6 into HEY and SKOV-3 ovarian carcinoma cell lines using microcell-mediated chromosome transfer techniques to further investigate the role of this chromosome in ovarian tumorigenesis. The exogenous chromosome was stably propagated in the recipient cells based on fluorescence in situ hybridization (FISH) analyses with a chromosome 6 painting probe. The tumorigenicity of HEY and SKOV-3 cells was completely suppressed after transfer of chromosome 6, but not after transfer of a chromosome 11q13-qter fragment used as control. Using 46 polymorphic microsatellite markers, the region bounded by D6S1649 and D6S1564 was found to be commonly deleted in HEY: chromosome 6 tumorigenic revertant clones. The boundaries of the commonly deleted region could be further narrowed down to a 2 cM (based on the Whitehead genetic map) or 0.36 megabase (based on gdb mapping data) region between D6S1637 and D6S1564 after transferring the exogenous chromosome from revertants into mouse L cells and performing allelic deletion mapping studies against this mouse background. We conclude that this region contains a tumor suppressor gene important for the control of ovarian tumor development.

tumor suppressor; chromosome 6; ovarian cancer; chromosome transfer

Introduction

Structural abnormalities involving chromosome 6 are among the most frequently reported chromosomal alterations in human ovarian carcinomas. Wake et al. (1980) first demonstrated a translocation between chromosomes 6 and 14 in the papillary serous subtype of ovarian tumors. Deletions and translocations involving different portions of chromosome 6q were subsequently reported by cytogenetic analyses of primary ovarian carcinomas and ovarian cancer cell lines (Buick et al., 1985; Deger et al., 1997; Sheer et al., 1987; Tibiletti et al., 1996; Trent and Salmon, 1981; Whang-Peng et al., 1984). Frequent losses of heterozygosity (LOH) on this chromosome were also observed in primary ovarian tumors (Cliby et al., 1993; Ehlen and Dubeau, 1990; Foulkes et al., 1993; Lee et al., 1990). Distinct regions of chromosome 6 were found to be targeted by losses of heterozygosity in several allelic deletion mapping studies (Cooke et al., 1996; Foulkes et al., 1993; Orphanos et al., 1995; Saito et al., 1992, 1996; Wan et al., 1994). It thus appears that different regions of this chromosome may contain one or several tumor suppressor gene(s) important for the control of ovarian tumor development. However, functional evidence for this hypothesis was not provided by these cytogenetic and molecular genetic analyses. In addition, the current candidate regions for the suspected suppressor genes are too large to allow their isolation by positional cloning techniques.

We introduced a normal chromosome 6 into the highly tumorigenic HEY and SKOV-3 ovarian cancer cell lines in the present study in order to develop a functional system to facilitate the exact localization and isolation of one or several of the suspected tumor suppressor genes on this chromosome. These cell lines were selected because they carried karyotypic abnormalities on chromosome 6q (Buick et al., 1985). We first sought to verify our hypothesis that transfer of a normal chromosome 6 would reduce or abolish tumorigenicity in these cells. We then isolated clones that regained their tumorigenic potential and examined the exogenous chromosome 6 for allelic deletions in these tumorigenic revertants, reasoning that such reversion was due to loss of a tumor suppressor gene which controlled the tumorigenic phenotype on this chromosome.

Results

Transfer of a normal chromosome 6 into selected human ovarian carcinoma cell lines

We selected two ovarian carcinoma cell lines, HEY and SKOV-3, which both contain karyotypic abnormalities in chromosome 6q, as recipients in chromosome transfer experiments using mouse : human chromosome 6 microcell hybrids as chromosome donor. A chromosome 11q13-qter fragment was transferred into each cell line as control. Figure 1 shows fluorescence in situ hybridization studies of HEY cells with a chromosome 6 painting probe before and after transfer of chromosome 6 to verify that the recipient cells had indeed acquired an extra copy of this chromosome. As anticipated, two hybridization signals were obtained with HEY cells before being subjected to the chromosome transfer procedures. The presence of a third hybridization signal after these procedures confirmed the presence of an exogenous chromosome 6 copy (Figure 1). An additional copy of this chromosome could similarly be demonstrated in hypertetraploid SKOV-3 cells following chromosome 6 transfer procedures (results not shown). The presence of an extra chromosome 11q13-qter fragment in control cells was verified by Southern blotting using probes specific for this chromosomal region (results not shown).

Consequences of chromosome 6 or 11q transfers on the in vitro kinetics, cloning efficiency, and tumorigenicity of HEY and SKOV-3 ovarian carcinoma cell lines

We compared the doubling time and cloning efficiency of HEY and SKOV-3 cells before and after transfer of either an intact chromosome 6 or of a chromosome 11q fragment in order to evaluate the phenotypic consequences of such transfers. Transfer of chromosome 6 resulted in reduced growth rates in both cell lines (Table 1). In addition, both cell lines lost their anchorage-independent growth abilities after transfer of this chromosome, in spite of the high cloning efficiencies of the parental cell lines (Table 1). In contrast, neither growth kinetics nor cloning efficiencies were affected by transfer of the chromosome 11q13-qter fragment (Table 1).

We next examined the consequences of the above chromosome transfers on the tumorigenic potential of HEY and SKOV-3 cells. Subcutaneous inoculation of 100 000 cells from either of the two parental cell lines into nude mice resulted in visible tumors within 2 weeks (Table 1). None of the mice inoculated with a similar number of either HEY : chromosome 6 or SKOV-3 : chromosome 6 hybrids formed tumors during the 3 month observation period (Table 1). Transfer of the chromosome 11q13-qter fragment showed no effect on tumorigenicity (Table 1).

Isolation and characterization of tumorigenic revertant clones

Five nude mice were inoculated each with 100´10⁶ HEY/chromosome 6 hybrid cells at high passage in an attempt to isolate revertant subclones that regained their tumorigenic potential. Given that the parental HEY cells could readily form tumors in these animals when as little as 0.1´10⁶ cells were inoculated, we reasoned that inoculation of 100´10⁶ hybrid cells should result in the development of tumors if we assumed a reversion rate of around 0.1%. Such a reversion rate appeared reasonable given that the cells were no longer under the selective pressure of neomycin after inoculation. Tumors were indeed observed in all five mice. Fourteen different subclones were isolated from tissue culture explants of these tumors. Amplification of microsatellite polymorphisms assigned to various regions of chromosome 6, which allowed distinction between alleles from the exogenous versus endogenous chromosomes, revealed that most of these revertant subclones had lost the entire copy of the exogenous chromosome 6. Five subclones, however, had retained this copy. These five different subclones, called respectively HEY/Chr6/R5a, HEY/Chr6/R5b, HEY/Chr6/R5c, HEY/Chr6/R5d, and HEY/Chr6/R5e, were analysed by fine allelic deletion mapping studies to better localize the deletion site (Table 2). The smallest deletion was seen in clones HEY/Chr6/R5b and HEY/Chr6/R5c. It involved the D6S311 locus and was bounded by D6S1649 and D6S1564, corresponding to a 3.3 cM fragment according to recent Whitehead mapping data. Autoradiographs demonstrating deletion of the exogenous D6S311 locus and retention of the exogenous D6S1649 and D6S1564 loci in a non-tumorigenic revertant clone are shown in Figure 2.

Only 30 of the 46 markers examined are shown in Table 2 for simplicity. The additional markers, which were from the 6q26-q27 region, showed no interstitial deletions in any of the clones examined.

Several tumorigenic revertant clones were also isolated from SKOV-3 : chromosome 6 cells. However, analysis of the donor chromosome in these revertants using the same panel of markers as for HEY : chromosome 6 cells revealed that the entire donor chromosome had been deleted, precluding further allelic deletion mapping studies based on this cell line.

Further fine scale deletion mapping using human/mouse somatic cell hybrids

The number of polymorphic markers assigned to the 3.3 cM candidate region is currently limited, complicating more precise localization of the deleted chromosome segment in the above revertants. In addition, many of the available markers were non-informative in our allelic deletion mapping studies because the corresponding alleles in the exogenous and endogenous copies of chromosome 6 were of equal sizes. We therefore transferred the exogenous chromosome 6 from revertant clones into mouse L cells, reasoning that human loci in this chromosome would be readily detected against a mouse genetic background. Fine-scale deletion mapping against such background not only confirmed the presence of a deletion at D6S311, but also showed retention of D6S1637 in the exogenous chromosome (Figure 3). This marker was mapped to 148.2 cM from the top of the chromosome 6 linkage group, very close to D6S311 in the Whitehead database. However, D6S311 is 0.22 megabase distal to D6S1637 based on the mapping information in gdb. The human transcript map also places D6S311 distal to D6S1637. We favor the latter ordering because putting D6S311 proximal to D6S1637 would imply that two small deletions occurred on each side of D6S1637 in HEY/Chr6/R5a (Table 2). Assuming that this order is correct and combining the results of Table 2 and Figure 3, the deleted region in tumorigenic revertants HEY/Chr6/R5b and HEY/Chr6/R5c is bounded by D6S1637 and D6S1564. This region is 2 cM on the Whitehead map and 0.36 megabase based on mapping data in gdb. If D6S311 was proximal to D6S1637, a possibility which we cannot rule out with certainty, the candidate region would be bounded by D6S1649 and D6S1637. The latter region is 1.3 cM on the Whitehead map and 0.83 megabase in gdb.

Discussion

Our results clearly show that introduction of a normal chromosome 6 into selected highly tumorigenic ovarian carcinoma cell lines can suppress their tumorigenic potential as well as their anchorage independence in vitro, which are two important manifestations of the malignant phenotype. Evidence that these phenotypic changes were not the mere consequence of manipulations associated with the chromosome transfer protocols but were truly controlled by chromosome 6-specific DNA sequences comes from two observations. First, similar changes were not seen when a chromosome 11q fragment was transferred into the same recipient cells instead of a chromosome 6. Second, loss of only a small (0.36 MB) portion of the exogenous chromosome 6 in the hybrid cell lines resulted in regaining both, tumorigenic and anchorage independent abilities, further attesting to the association of these phenotypic properties with specific DNA sequences. These results provide functional evidence for the hypothesis, originally based on numerous cytogenetic and molecular genetic studies, that chromosome 6 harbors one or more tumor suppressor gene(s) important for the control of ovarian carcinoma development.

Chromosome 6 was previously shown to suppress tumorigenicity and metastasis in melanoma cells (Trent et al., 1990; Welch et al., 1994) as well as tumorigenicity in breast carcinoma cells (Negrini et al., 1994). This suggests that its role in tumor suppression is not restricted to tumors of ovarian origin, but may be important for a large variety of different cancer types. Indeed, deletions or losses of heterozygosity affecting chromosome 6 were reported in cancers from a large number of different organs including carcinomas of the breast (Devilee et al., 1991; Sheng et al., 1996), endometrium (Tibiletti et al., 1997), and prostate (Cooney et al., 1996), as well as in small cell lung carcinomas (Merlo et al., 1994), lymphomas and leukemias (Gerard et al., 1997; Menasce et al., 1994), and melanomas (Walker et al., 1994).

Our ultimate goal in this study was to better define the chromosomal region(s) associated with suppression of tumorigenicity in ovarian carcinomas in order to facilitate the eventual isolation of the specific gene(s) controlling such suppression. This was achieved by obtaining subclones of HEY : chromosome 6 hybrids that regained their tumorigenic and anchorage-independent properties. Examination of the integrity of the exogenous chromosome 6 in these revertants showed that either the entire chromosome or a portion of it had been lost in each case. Further characterization of the partial deletions showed that the deleted chromosomal segment overlapped with a segment previously shown to be affected by losses of heterozygosity in primary ovarian tumors (Cooke et al., 1996; Foulkes et al., 1993; Orphanos et al., 1995; Saito et al., 1992, 1996; Wan et al., 1994). The smallest deletion which could be defined by this approach extended from D6S1649 to D6S1564. The fact that most of the known polymorphisms within this region were non-informative because the corresponding alleles on the exogenous chromosome were indistinguishable from those on the endogenous copies complicated attempts at further narrowing this candidate region. We therefore transferred the endogenous chromosome 6 from selected revertants into mouse L cells, reasoning that human alleles would be readily identifiable against a mouse genetic background. Indeed, this approach led to the demonstration that the chromosomal deletion associated with regaining of tumorigenic potential involved D6S311 and was within a segment bounded by D6S1637 proximally and D6S1564 distally. We conclude that this region, which is within bands 6q24-q25, contains a tumor suppressor gene important for controlling the development of ovarian carcinomas as well as probably several other tumor types. This conclusion was based on the assumption that D6S311 is distal to D6S1637. The order of these two loci is not clear because they are both said to be located at 148.2 cM on the Whitehead map. We favor the interpretation that D6S311 is the most distal of these two markers because placing it proximal would imply that two small deletions had occurred concomitantly in the tumorigenic revertant clones HEY/Chr6/R5b and HEY/Chr6/R5c, one on each side of D6S1637. This order also agrees with the information in gdb. However, the possibility that D6S311 is proximal cannot be ruled out. This would place the candidate region between D6S1649 and D6S1637. The segment between D6S1637 and D6S1564, which is our favored candidate region, is 2 cM on the Whitehead map and 0.36 megabase in gdb. The alternate candidate region between D6S1649 and D6S1637 is 1.3 cM on the Whitehead map and 0.83 megabase according to the information in gdb.

The region between D6S1649 and D6S1637 contains the hLOT1/hZac (Abdollahi et al., 1997; Varrault et al., 1998) gene according to the Whitehead map. Although the integrity and the state of expression of this gene in our HEY cell hybrids is not clear, the proximity of our candidate region to this gene underscores the possibility that it is indeed a strong candidate tumor suppressor gene important for the control of ovarian tumorigenesis. Several cDNA sequences have also been mapped to the candidate region including several unidentified transcripts as well as a few better characterized genes. The potential role of these cDNAs and ESTs in ovarian tumorigenesis is presently unclear.

HEY and SKOV-3 cells, which were used as chromosome recipients in our studies, were both reported to contain interstitial deletions within the 6q21-q23 chromosomal segment (Buick et al., 1985). Although it is the presence of these karyotypic abnormalities that led to our choice of these cells for our experiments, it is now clear that the gene(s) responsible for controlling tumorigenic potential in our studies is(are) not within this region. This conclusion is based on the fact that our 2 cM candidate region is assigned to chromosomal bands 6q24-q25, which is more distal than the above deletions, as well as to the fact that many of the polymorphic markers assigned to the 6q21-q23 region were found to be heterozygous in HEY cells during the course of our allelic deletion mapping studies (not shown). The possibility that another locus within 6q21-q23 also harbors a gene functioning as tumor suppressor for ovarian carcinoma cannot be ruled out. Indeed, the current evidence suggests that chromosome 6 contains several tumor suppressor genes important in these tumors (Banga et al., 1997; Cooke et al., 1996; Foulkes et al., 1993; Orphanos et al., 1995; Saha et al., 1995; Saito et al., 1992, 1996; Tibiletti et al., 1998; Wan et al., 1994). More recently, Sandhu et al. (1996) provided functional evidence for another candidate region on chromosome 6q important in ovarian tumorigenesis. These authors reported on the suppression of in vitro immortality after introduction of a normal chromosome 6 in SKOV-3 cells. They showed that the locus controlling this phenotypic change was located within 6q14-q21, which is clearly proximal to our candidate region. They concluded that this region contained a gene controlling senescence which they called SEN 6A. Given that none of the tumorigenic revertants from SKOV-3 : chromosome 6 hybrids showed interstitial deletions in the exogenous chromosome using our panel of markers, the exact location of the chromosomal segment controlling tumorigenicity in these cells is not known and could possibly be within the region reported by Sandhu et al., 1996. However, the phenotypic changes that we observed after our chromosome transfer procedures were different than those reported by Sandhu et al., 1996 as our SKOV-3 : chromosome 6 hybrids showed loss of tumorigenicity but retained their in vitro immortality. One possible explanation for these apparent discrepancies may be that the donor chromosome used in our studies was different than that used by Sandhu et al., 1996.

In conclusion, our results provide functional evidence for the presence of a tumor suppressor gene important for the control of ovarian tumorigenesis within a small region of chromosome 6 previously suspected to harbor such a gene based on loss of heterozygosity distributions. Isolation of this gene and the subsequent elucidation of its role in cancer suppression should provide important insights into the mechanisms of development of ovarian carcinomas, which constitute one of the deadliest tumors of women.

Materials and methods

Cell lines

The SKOV-3 human ovarian carcinoma cell line was obtained from American Type Culture Collection (ATCC # HTB77). HEY cells were obtained from Dr Ronald Buick, University of Toronto (Buick et al., 1985). Mouse/human hybrid MCH262A1.D6 cells containing a human chromosome 6 were obtained from Dr Eric Stanbridge (University of California at Irvine). Hdm-18, containing an intact human chromosome 11, was obtained from Dr Keith Fournier from the Fred Hutchinson Cancer Center (Lugo et al., 1987). The human chromosome from each of these cell lines was transferred into L-M (TK^-) mouse cells (ATCC # CCL 1.3) before fusion with recipient human cells. A portion of the chromosome 11, which extended from 11pter to 11q13, was lost in this process (results not shown).

Microcell-mediated chromosome transfer and analyses of transferred chromosome

Microcell-mediated chromosome transfers of chromosomes 6 or 11q into human ovarian carcinoma cell lines were performed according to published procedures (Lugo et al., 1987) with only slight modifications. Briefly, donor cells were treated with 0.15 g/ml colcemid (Sigma, St. Louis, MO, USA, cat # D-7385) for 48 h. The resulting micronuclei were enucleated by treatment with 2 g/ml cytochalasin B (Sigma, cat # C6762) followed by centrifugation. The microcells were sequentially filtered through 8, 5 and 3 m polycarbonate filters (Nucleopore Corp., Pleasanton, CA, USA) to obtain populations containing primarily single chromosomes. One ml of filtered microcells resuspended in serum-free medium was added to recipient cells and agglutinated in 1 ml of 200 g/ml phytohemagglutinin P (DIFCO, Detroit, MI, USA, cat # 3110-57-3). The phytohemagglutinin solution was then aspirated and 1 ml of 50% polyethylene glycol 1500 (Koch-Light, Ltd., Haverhill, Suffolk, UK, cat # 00-14806/C) was added for 60 s, followed by immediate and rapid rinses in culture medium. The hybrid cells were selected for resistance to 500 g/ml neomycin (GIBCO/BRL, cat # 11811-031). Exogenous chromosomes 6 carrying interstitial deletions were transferred from HEY and into mouse L TK- cells using similar protocols except that the filtration step was omitted and 30 g/ml 5-bromo-deoxyuridine (Sigma, cat # B9285) was added to the fused cells in addition to neomycin in order to eliminate intact donor cells.

FISH studies with chromosome painting probes

The chromosome 6 painting probe (Coatasome^TM 6 total chromosome probe) was obtained from ONCOR (Gaithersburg, MD, USA). The conditions for hybridization and post-hybridization washes were as recommended by the manufacturer.

Isolation of tumorigenic revertants from non-tumorigenic hybrids

Tumorigenic revertants were obtained from two different methods. The first method took advantage of the fact that the recipient ovarian carcinoma cell lines lost their ability to grow in soft agar after transfer of an exogenous chromosome 6. Hybrids containing such an exogenous chromosome tagged with a neomycin resistance gene were plated on soft agar in the absence of neomycin. Revertant colonies were recovered and cultured on plastic dishes and selected for neomycin resistance. All revertants examined that had been isolated by this method had lost the entire exogenous chromosome 6 in spite of the fact that they retained the neomycin resistance trait, suggesting that the resistance gene had become incorporated into an endogenous chromosome via recombination. The second method involved injecting a larger number of hybrid cells (about 100´10⁶ cells per mouse) into nude mice in order to directly select for tumorigenic revertants. Loss of the gene controlling tumorigenic potential in the hybrid cells needed to occur in 0.1% of the innoculum or more in order for tumorigenic revertants to be recovered because as little as 100 000 of the parental HEY or SKOV-3 cells readily form tumors in nude mice. The revertant tumors were explanted and cultured in the presence of neomycin in order to enrich for cells that had retained at least part of the exogenous chromosome.

Tumorigenicity studies

We inoculated 0.1 - 100´10⁶ cells resuspended into 0.1 ml of serum-free culture medium into the flanks of 3 - 4 week old female nude mice (nu/nu). Tumors usually developed within 2 weeks. All grossly visible tumors were examined histologically in order to verify their authenticities. No cell line was scored as non-tumorigenic unless mice inoculated with such cells were observed for at least 12 weeks following inoculation.

Fine scale deletion mapping studies of the exogenous chromosome 6 in revertant clones

Matched genomic DNA from HEY or SKOV-3 ovarian carcinoma cell lines, non-tumorigenic hybrids, and revertants were analysed for allelic losses by enzymatic amplification of microsatellite polymorphisms assigned to chromosome 6. This allowed for further verification that new alleles had indeed been transferred in the non-tumorigenic hybrid cells as well as for the detection of missing alleles in the revertant clones. The 46 chromosome 6-specific microsatellite markers used were either described before (Wan et al., 1994), taken from published genetic linkage maps (Dib et al., 1996; Hudson et al., 1995; Orphanos et al., 1994; Volz et al., 1994), or taken from on-line genetic databases including the Whitehead Institute (http://www-genome.wi.mit.edu/), gdb (http://gdbwww. gdb.org:80/gdb/gdbtop.html), Sanger's map (http://www. sanger.ac.uk/cgi-bin/rhtop?chr=6), and the human transcript map (http://www.ncbi.nlm.nih.gov/SCIENCE96/). The order of D6S1637 and D6S311 is critical to our conclusions. Both markers are said to be located at 148.2 cM in the Whitehead map. However, D6S311 is distal to D6S1637 on the gdb map as well as on the human transcript map. Since our own mapping data obtained from the distribution of allelic losses from the exogenous chromosome 6 in tumorigenic revertants (see Table 2) is more compatible with the latter, we favor the gdb and human transcript mapping order but cannot completely rule out that D6S311 is the more proximal marker.

Acknowledgements

The authors would like to thank Dr Eric Stanbridge for providing us with the chromosome 6 mouse:human microcell hybrid, Dr Jane Fountain for helpful comments and advice throughout the course of this work, and Michelle Luo for help in preparing the manuscript. This study was supported by grant RO1 CA51167 from the US National Cancer Institute.

Abdollahi A, Roberts D, Godwin AK, Schultz DC, Sonoda G, Testa JR and Hamilton TC. (1997). Oncogene 14, 1973-1979. MEDLINE

Banga SS, Kim S-H, Hubbard K, Dasgupta T, Jha KK, Patsalis P, Hauptschein R, Gamberi B, Dalla-Favera R, Kraemer P and Ozer HL. (1997). Oncogene 14, 313-321. MEDLINE

Buick RN, Pullano R and Trent JM. (1985). Cancer Res. 45, 3668-3676. MEDLINE

Cliby W, Ritland S, Hartmann L, Dodson M, Halling KC, Keeney G, Podratz KC and Jenkins RB. (1993). Cancer Res. 53, 2393-2398. MEDLINE

Cooke IE, Shelling AN, Le Meuth VG, Charnock ML and Ganesan TS. (1996). Genes, Chromosomes & Cancer 15, 223-233.

Cooney KA, Wetzel JC, Consolino CM and Wojno KJ. (1996). Cancer Res. 56, 4150-4153. MEDLINE

Deger RB, Faruqi SA and Noumoff JS. (1997). Cancer Genet. Cytogenet. 96, 166-173. MEDLINE

Devilee P, van Vliet M, van Sloun P, Kuipers Dijkshoorn N, Hermans J, Pearson PL and Cornelisse CJ. (1991). Oncogene 6, 1705-1711. MEDLINE

Dib C, Faure S, Fizames C, Samson D, Drouot N, Vignal A, Millasseau P, Marc S, Hazen J, Seboun E, Lathrop M, Gyapay G, Morissette J and Weissenbach J. (1996). Nature 380, 152-154. MEDLINE

Ehlen T and Dubeau L. (1990). Oncogene 5, 219-223. MEDLINE

Foulkes WD, Ragoussis J, Stamp GW, Allan GJ and Trowsdale J. (1993). Br. J. Cancer 67, 551-559. MEDLINE

Gerard B, Cave H, Guidal C, Dastugue N, Vilmer E and Grandchamp B. (1997). Leukemia 11, 228-232. MEDLINE

Hudson TJ, Stein LD, Gerety SS, Ma J, Castle AB, Silva J, Slonim DK, Baptista R, Kruglyak L and Xu SH. (1995). Science 270, 1945-1954. MEDLINE

Lee JH, Kavanagh JJ, Wildrick DM, Wharton JT and Blick M. (1990). Cancer Res. 50, 2724-2728. MEDLINE

Lugo TG, Handelin B, Killary AM, Housman DE and Fournier RE. (1987). Molec. Cell. Biol. 7, 2814-2820. MEDLINE

Menasce LP, Orphanos V, Santibanez-Koref M, Boyle JM and Harrison CJ. (1994). Genes, Chromosomes & Cancer 10, 286-288.

Merlo A, Gabrielson E, Mabry M, Vollmer R, Baylin SB and Sidransky D. (1994). Cancer Res. 54, 2322-2326. MEDLINE

Negrini M, Sabbioni S, Possati L, Rattan S, Corallini A, Barbanti-Brodano G and Croce CM. (1994). Cancer Res. 54, 1331-1336. MEDLINE

Orphanos V, McGown G, Hey Y, Thorncroft M, Santibanez-Koref M, Russell SE, Hickey I, Atkinson RJ and Boyle JM. (1995). Br. J. Cncer 71, 666-669.

Orphanos V, Santibanez-Koref M, McGown G, Hey Y, Rackstraw C and Boyle JM. (1994). Genomics 20, 301-304. MEDLINE

Saha V, Lillington DM, Shelling AN, Chaplin T, Yaspo ML, Ganesan TS and Young BD. (1995). Genes, Chromosomes & Cancer 14, 220-222.

Saito S, Saito H, Koi S, Sagae S, Kudo R, Saito J, Noda K and Nakamura Y. (1992). Cancer Res. 52, 5815-5817. MEDLINE

Saito S, Sirahama S, Matsushima M, Suzuki M, Sagae S, Kudo R, Saito J, Noda K and Nakamura Y. (1996). Cancer Res. 56, 5586-5589. MEDLINE

Sandhu AK, Kaur GP, Reddy DE, Rane NS and Athwal RS. (1996). Oncogene 12, 247-252. MEDLINE

Sheer D, Sheppard DM, Gorman PA, Ward B, Whelan RD and Hill BT. (1987). Cancer Genet. Cytogenet. 26, 339-349. MEDLINE

Sheng ZM, Marchetti A, Buttitta F, Champeme MH, Campani D, Bistocchi M, Lidereau R and Callahan R. (1996). Br. J. Cancer 73, 144-147. MEDLINE

Tibiletti MG, Bernasconi B, Furlan D, Riva C, Trubia M, Buraggi G, Franchi M, Bolis P, Mariani A, Frigerio L, Capella C and Taramelli R. (1996). Cancer Res. 56, 4493-4498. MEDLINE

Tibiletti MG, Bernasconi B, Taborelli M, Furlan D, Fabbri A, Franchi M, Taramelli R, Trubia M and Capella C. (1997). Br. J. Cancer 75, 1831-1835. MEDLINE

Tibiletti MG, Trubia M, Ponti E, Sessa L, Acquati F, Furlan D, Bernasconi B, Fichera M, Mihalich A, Ziegler A, Volz A, Facco C, Riva C, Cremonesi L, Ferrari M and Taramelli R. (1998). Oncogene 16, 1639-1642. MEDLINE

Trent JM and Salmon SE. (1981). Cancer Genet. Cytogenet 3, 279-291. MEDLINE

Trent JM, Stanbridge EJ, McBride HL, Meese EU, Casey G, Araujo DE, Witkowski CM and Nagle RB. (1990). Science 247, 568-571. MEDLINE

Varrault A, Ciani E, Apiou F, Bilanges B, Hoffmann A, Pantaloni C, Bockaert J, Spengler D and Journot L. (1998). Proc. Natl. Acad. Sci. USA 95, 8835-8840. Article MEDLINE

Volz A, Boyle JM, Cann HM, Cottingham RW, Orr HT and Ziegler A. (1994). Genomics 21, 464-472. MEDLINE

Wake N, Hreshchyshyn MM, Piver SM, Matsui S and Sandberg AA. (1980). Cancer Res. 40, 4512-4518. MEDLINE

Walker GJ, Palmer JM, Walters MK, Nancarrow DJ, Parsons PG and Hayward NK. (1994). Int. J. Cancer 58, 203-206. MEDLINE

Wan M, Cofer KF and Dubeau L. (1996). Br. J. Cancer 73, 1398-1400. MEDLINE

Wan M, Zweizig S, D'Ablaing G, Zheng J, Velicescu M and Dubeau L. (1994). Int. J. Oncol. 5, 1043-1048.

Welch DR, Chen P, Miele ME, McGary CT, Bower JM, Stanbridge EJ and Weissman BE. (1994). Oncogene 9, 255-262. MEDLINE

Whang-Peng J, Knutsen T, Douglass EC, Chu E, Ozols RF, Hogan WM and Young RC. (1984). Cancer Genet. Cytogenet. 11, 91-106. MEDLINE

Figure 1 FISH analyses of ovarian carcinoma cell lines before and after transfer of a human chromosome 6. HEY ovarian carcinoma cells were analyzed by FISH using a human chromosome 6 painting probe before (left) and after (right) transfer of an exogenous chromosome using microcell mediated chromosome transfer techniques. The presence of an additional chromosome showing a fluorescent hybridization signal after the chromosome transfer procedures attests to the success of these procedures

Figure 2 Interstitial losses of heterozygosity in tumorigenic revertants involve the D6S311 locus. Genomic DNA was isolated from HEY ovarian carcinoma cells (parental), from the mouse : human microcell hybrids used as chromosome 6 donor (donor), from non-tumorigenic hybrids of HEY cells and microcell donors (hybrid), and from tumorigenic revertants (revertant). The various loci were amplified by PCR in each DNA sample, electrophoresed on polyacrylamide gels, and the radiolabeled PCR products were visualized by autoradiography

Figure 3 Interstitial deletions of the exogenous chromosome 6 in tumorigenic revertants are distal to D6S1637. The exogenous chromosome 6 in tumorigenic revertants (clone b) was transferred into mouse L cells using microcell-mediated chromosome transfer techniques. Genomic DNA was obtained from the parental HEY cells (HEY), from non-tumorigenic HEY : chromosome 6 hyrbids (HEY+transferred chromosome 6), from tumorigenic revertants (tumorigenic revertant), from mouse L cells containing no human chromosome (mouse L cells), from mouse L cells after transfer of the revertant chromosome 6 (mouse L cells+ chromosome 6 transferred from revertant clone), and from the original mouse L cells : chromosome 6 hyrbids used as our chromosome 6 donor cells (original chromosome 6 donor cells). The D6S1637 and D6S311 loci were amplified by PCR in each DNA sample. The PCR products were electrophoresed on 1% agarose gels in the presence of ethidium bromide and visualized under UV light. The arrows indicate the expected position of the D6S1637 and D6S311 products in each respective gel

Table 1 Table 1

Table 2 Table 2