Oncogene (2003) 22, 445–450 & 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

DLC-1 inhibits human breast cancer cell growth and in vivo tumorigenicity

Bao-Zhu Yuan1,3,4, Xiaoling Zhou1,4, Marian E Durkin1, Drazen B Zimonjic1, Katrin Gumundsdottir2, Jorunn E Eyfjord2, Snorri S Thorgeirsson1 and NicholasC Popescu* ,1

1Laboratory of Experimental Carcinogenesis, Center for Cancer Research, National Cancer Institute, Bethesda, MD 20892, USA; 2Icelandic Cancer Society, Reykjavik, Iceland

The human DLC-1 (deleted in liver cancer 1) gene was et al., 1998). Determination of the DLC-1 cDNA cloned from a primary human sequence showed that it is the human homologue of (HCC) and mapped to the 8p21–22 region rat p122, which hasbeen found to act asa GAP for frequently deleted in common human cancers and RhoA, and to stimulate the hydrolysis of phosphati- suspected to harbor tumor suppressor . DLC-1 was dylinositol 4,5-bisphosphate by -d1 found to be deleted or downregulated in a significant (Homma and Emori, 1995). Both the DLC-1 and p122 number of HCCs. We expanded our investigations to sequences have a Rho GAP domain (Homma and other cancers with recurrent deletions of 8p22, and in this Emori, 1995) and at least two other potential functional study examined alterations of DLC-1 in primary human motifs, a sterile alpha motif domain and a StAR- breast tumors, human breast, colon, and prostate tumor related lipid-transfer domain (Ponting and Aravind, cell lines. Genomic deletion of DLC-1 was observed in 1999). Overexpression of p122 in cultured cells 40% of primary breast tumors, whereas reduced or resulted in the loss of stress fibers and the undetectable levels of DLC-1 mRNA were seen in 70% detachment of cells from the substratum, implicating of breast, 70% of colon, and 50% of prostate tumor cell DLC-1/p122 in the signal transduction pathways that lines To see whether DLC-1 expression affects cell growth regulate cell morphology and adhesion (Sekimata et al., and tumorigenicity, two breast carcinoma cell lines 1999). lacking the expression of endogenous gene were trans- The DLC-1 gene waslocalized on chromosome fected with the DLC-1 cDNA. In both cell lines, DLC-1 8p21–22, a region of loss of heterozygosity (LOH) in a transfection caused significant growth inhibition and number of human cancers such as prostate, colon, reduction of colony formation. Furthermore, introduction breast, ovarian, liver, lung, bladder, and head and neck of the DLC-1 cDNA abolished the in vivo tumorigenicity cancer (reviewed in Arbieva et al., 2000; Xu et al., 2001). in nude mice, suggesting that the DLC-1 gene plays a role Similarly, DNA copy-number losses in this region were in breast cancer by acting as a bona fide tumor suppressor identified by comparative genomic hybridization (CGH) gene. in tumorsof the breast, liver, colon, lung, ovary, bladder Oncogene (2003) 22, 445–450. doi:10.1038/sj.onc.1206064 and prostate, as well as in osteosarcomas, mesothelio- mas, B-cell lymphomas and mantle cell lymphoma Keywords: ; breast cancer; colon (Martinez-Climent et al., 2001; Zitzelsberger et al., 1997). cancer; prostate cancer; tumorigenicity; metastasis transfer into colon and prostate tumor cell linesprovided functional evidence for the presence of tumor suppressor genes on chromosome 8p21–22 (Gus- Introduction tafson et al., 1996; Cabeza-Arvelaiz et al., 2001). Deletion of the DLC-1 gene wasfound in human The human DLC-1 (deleted in liver cancer 1) gene was primary liver tumorsand in HCC-derived cell lines isolated from a primary human hepatocellular carcino- (Yuan et al., 1998; Ng et al., 2000). In normal adult ma (HCC) by representational difference analysis (Yuan tissues the DLC-1 gene is widely expressed, but in primary HCC and HCC cell linesitsexpressioniseither *Correspondence: NC Popescu, Laboratory of Experimental, Carci- reduced or absent (Yuan et al., 1998; Ng et al., 2000). nogenesis, Center for Cancer Research, National Cancer Institute, Genomic deletion or promoter hypermethylation are Bethesda, MD 20892, USA; responsible for aberrant expression of DLC-1 in HCC, E-mail: [email protected] as well as in breast, colon, and prostate cancer cells, 3Current address: Toxicology and Molecular Biology Branch, Health while mutationsare rather infrequent in colon and Effects, Laboratory Division, National Institute for Occupational ovarian cancer and also in HCC (Wilson et al., 2000; Safety and Health, Morgantown, WV 26505, USA Yuan et al., 2000; Yuan et al., 2002, manuscript in 4These authors contributed equally to this work Received 19 July 2002; revised 18 September 2002; accepted 24 preparation). Transfection of the DLC-1 cDNA into September 2002 two HCC cell lineswith homozygousdeletionsof the DLC-1 gene suppresses in vivo tumorigenicity B-Z Yuan et al 446 gene caused a strong inhibition of cell growth (Ng et al., levelswere decreasedin another eight cell lines(HCT- 2000). 15, DLD-1, HT-29, SW1417, SW480, SW948, LS 180, To obtain further evidence that DLC-1 functionsasa SW48) compared to the normal colon cell line CCD tumor suppressor gene, we expanded our investigations 33Co (Figure 1b). In two out of four prostate cancer cell to include other typesof cancer with frequent deletions lines (LNCaP and PC-3M), no expression of DLC-1 was in the chromosome 8p22 region. In this study, we observed (Figure 1c). examined alterationsof DLC-1 in human breast, colon, To determine whether the DLC-1 locuswasrecur- and prostate tumor cell lines, and in primary breast rently deleted in breast cancer, we compared by South- tumors. We also tested the effect of the DLC-1 gene on ern blot hybridization the copy number of the DLC-1 cell growth and tumorigenicity, by transfecting with the gene in 15 matched primary breast cancer/normal breast DLC-1 cDNA three breast cancer cell lines with no DNA samples. Six of the 15 tumor DNA samples expression of the endogenous gene. DLC-1 transfer into displayed deletion of the DLC-1 gene. A representative tumor cellscausedboth inhibitionsof the cell growth Southern blot of BgllI-digested genomic DNA from two and of the in vivo tumorigenicity, thusindicating that this primary breast cancers that reveals deletion of the gene acts as a tumor suppressor gene in breast cancer. DLC-1 gene isshown(Figure 1d). To examine whether transfection of the DLC-1 cDNA affectscell growth, two breastcancer cell lines Results that lacked expression of the endogenous gene, MDA- 468 and MDA-MB-361, were transfected with the The expression of DLC-1 mRNA was examined in 17 pLXSN/DLC-1 plasmid that directs the expression of breast, 17 colon, and four prostate cancer cell lines. Six a Flag-tagged DLC-1 cDNA under the control of the breast cancer cell lines (UACC-893, UACC-812, ZR- MoMuLV 50 LTR promoter. Cellstransfected with 15B, SK-BR-3, MDA-MB-468, and MDA-MB-361) pLXSN/DLC-1 expressed DLC-1 48 h after transfec- showed no expression of the gene, while in six other tion, asindicated by RT-PCR, but not cellstransfected breast cancer lines (MDA-MB-330, BT-549, BT-483, with pLXSN alone (data not shown). After DLC-1 ZR-75-1, BT-20, BT-474), DLC-1 mRNA appeared to transfection, the cells were harvested and counted at be downregulated compared to the MCF-10F normal 48 h intervalsfor 10 days.In both MDA-468 and MDA- breast cell line (Figure 1a). Among the colon cancer MB-361 cells, significantly slower doubling rates were lines, four cell lines (LS 147T, SW1116, T84, SW403) observed in cells transfected with DLC-1 compared to showed no expression of the gene, while DLC-1 mRNA vector alone (Figure 2).

Figure 1 Northern blot analysis of DLC-1 expression in human cancer cell lines and Southern blot analysis of the DLC-1 gene in primary breast carcinomas. (a). Northern blot of total RNA from the indicated human breast cancer cell lines and the normal breast cell line MCF-10F. The blot washybridized to the DLC-1 probe, then strippedand rehybridized to a GAPDH cDNA probe. ( b) Northern blot of RNA from human colon cancer cell linesand the normal colon line CCD33Co, hybridized to the DLC-1 and GAPDH probesasin A. ( c) Northern blot of RNA from human prostate cancer cell lines hybridized sequentially to the DLC-1 and GAPDH probes. (d) Representative Southern blot of BglII-digested genomic DNA from two primary breast tumors that show deletion of the DLC-1 gene (IC06T, IC09T) and from peripheral blood of the same patients (IC06B and IC09B, respectively). After hybridization to the DLC-1 probe, the blot wasstripped and rehybridized to a b-actin probe. The sizes of the bands detected by the probesare indicated

Oncogene DLC-1 gene suppresses in vivo tumorigenicity B-Z Yuan et al 447 dramatic reduction (over 50%) in the number of G418-resistant colonies was observed in DLC-1 transfected compared to vector alone transfected cul- turesin both MDA-468 and MDA-MB-361 cells (Figure 3). To further assess the inhibitory effects of DLC-1, we tested the tumorigenicity of the 3 DLC-1-transfected breast tumor cell lines in Balb/c athymic nude mice. The G418-resistant MDA-486 and MDA-MB cells trans- fected with pLXSN/DLC-1, or the vector alone were harvested, assayed for viability by trypan-blue staining, counted and injected into nude mice. Animalsthat received MDA-468 DLC-1 or MDA-MB-361 DLC-1- transfected cells remained tumor-free for the length of the experiment of over 3 months. However, animals injected with MDA-486 or MDA-MB-361 cellstrans- fected with the vector alone developed tumorsduring the first 3 weeks of the experiment, which continued to grow progressively (Table 1). Histopathological analysis of the tumorswasconsistentwith mammary adenocar- cinoma.The third line tested, UACC-893, although derived from an infiltrating ductal carcinoma, did not produce tumors. In order to establish the copy number of the DLC-1 gene in the breast cancer cell lines used for in vivo tumorigenicity studies, chromosome metaphases and Figure 2 Growth inhibitory effect of DLC-1 gene on breast interphase nuclei from MDA-MB-468 and MDA-MB- carcinoma cell lines. MDA-MB- 468 and MDA-MB-361 were transfected with DLC-1 and vector only. The cells were harvested 361 cellswere hybridized in situ with the DLC-1 and counted at 2-day intervals for 10 days. Results represent the genomic probe. Both MDA-MB-468 and MDA-MB- mean of three independent experiments (bars, s.d.) 361 linesare aneuploid with 60–67 and 55–61 chromo- somes per cell, respectively. FISH analysis showed one copy of DLC-1 in majority of MDA-MB-361 cells, and To test the effect of DLC-1 on colony formation, cells predominantly two copiesof the gene in MDA-MB-468 transfected with DLC-1 and control vector were cells. However, an additional copy above the average cultured in G418 for 2 weeksand the colony number wasdetected only in MDA-MB-361 DLC-1-transfected was counted on crystal-violet-stained dishes. A cells, very likely resulting from duplication of the

Figure 3 Inhibition of colony formation in MDA-468 and MDA-MB-361 control cellsand cellstransfectedwith DLC-1. Stably transfected MDA-468 and MDA-MB-361 were seeded and cultured for 2 weeks in media containing G418. Petri dishes were stained with crystal violet solution and the number of colonies was counted in cells transfected with DLC-1 and with vector alone. The histogram shows the colony formation efficiency

Oncogene DLC-1 gene suppresses in vivo tumorigenicity B-Z Yuan et al 448 Table 1 Inhibition of in vivo tumorigenicity by DLC-1 gene in breast cancer cell lines

Cell line Number of cells inoculations Latency (days) Tumor size (L Â W)a mm Number of tumors/ Number of inoculated MDA-MB-468 (Vector) 1 Â 106 25 6.0 Â 8.0 5/5 MDA-MB-468 (DLC-1) 1 Â 10 90 N/A 0/5 MDA-MB-468 (Vector) 2 Â 106 25 10.0 Â 12.0 5/5 MDA-MB-468 (DLC-1) 2 Â 106 90 N/A 0/5 MDA-MB-361 (Vector) 2 Â 106 20 1.8 Â 2.2 6/8 90 7.5 Â 7.5 MDA-MB-361 (DLC-1) 2 Â 106 90 N/A 0/9

aTumor size was measured with a linear caliper 2 times per week for up to 3 months. L isthe length of tumor and W isthe width of tumor.

Figure 4 DLC-1 copy number visualized by FISH in control and DLC1-transfected MDA-MB-361 cells. (a) One FISH signal located on an apparently normal copy of chromosome 8 was observed in a majority of the MDA-MB-361 control cells. (b) DLC-1-transfected cells had two FISH signals on a new rearranged chromosome

existing copy because of a chromosome rearrangement et al., 2000). Although they contain one or two copies triggered by transfection (Figure 4). of DLC-1, they do not express the gene because of promoter hypermethylation (Yuan et al., 2002). Both cell linesformed tumorsin nude mice, but introduction of the DLC-1 cDNA abolished their in vivo tumor- Discussion igenicity. An association of chromosome 8p loss and metastasis In this study we have demonstrated that the DLC-1 gene hasearlier been demonstrated by CGH in human HCC isdeleted in 40% of primary breasttumorsand (Qin et al., 1999). Similarly, LOH at 8p21–22 in HCC, downregulated or not expressed in 70% of breast, breast, and prostate carcinomas was associated with an 70% of colon, and 50% of prostate tumor cell lines invasive tumor phenotype and metastasis (Yaremko examined. Examining the effectsof thisgene on in vitro et al., 1998; Jenkins et al., 1998; Qin et al., 2001). cell growth and in vivo tumorigenicity, we have also Interestingly, two isogenic breast cancer cell lines with shown that the transfection of the DLC-1 cDNA into diametrically opposite metastatic capabilities were two breast carcinoma cell lines that lack DLC-1 derived from MDA-MB-435 cells. DLC-1 and two expression caused a significant inhibition of cell growth other tumor suppressor genes, DPC4 (Smad-4) and in vitro and colony formation. MDA-MB-468 and BARD1, were identified asdownregulated in the MDA-MB-361 lineswere isolatedfrom patientswith nonmetastatic relative to the metastatic human breast metastatic breast adenocarcinoma and exhibit high-level cancer cell line (Euer et al., 2002). By Affymetrix amplification of the ErbB-2 oncogene (Zimonjic microarray analysis, DLC-1 was

Oncogene DLC-1 gene suppresses in vivo tumorigenicity B-Z Yuan et al 449 elevated 4.5-fold in the nonmetastatic cell line relative to Materials and methods the metastatic human breast cancer cell line grown in vitro. Thiswasconfirmed by real-time quantitative Cell lines PCR, which found that DLC-1 mRNA was20 times Most of the breast and colon tumor cell lines as well as normal more abundant in the nonmetastatic cell line. Similar human breast and colon cells were purchased from the analysis of xenotransplanted tumors revealed a 3.7-fold American Type Culture Collection. Three prostate carcinoma increase in DLC-1 expression in the nonmetastatic cell cell lines, LNCaP, SP3031 and PS-3M, were kindly provided line (Nicholson B, Goodison S and Tarin D, manuscript by Dr Sen Pathak (The University of Texas, MD Anderson in preparation). Thisstrikingdifference in DLC-1 Cancer Center, Houston, TX, USA). Cells were cultured in expression relative to tumor phenotype raises the DMEM/F12 or RPMI1640 media with 10% fetal bovine possibility that downregulation or loss of the DLC-1 serum and antibiotics. gene is involved in the metastatic process. This assump- tion iscompatible with the ability of the Rho family Northern blot analysis proteinsto remodel the actin cytoskeleton, leading to Total RNA wasextracted from cell linesusingTRIzol reagent tumor cell migration and invasion (Yoshioka et al., (Life Technologies, Inc., Grand Island, NY, USA). In all, 1998). 20 mg of total RNA of each sample was resolved on 1% Although the Rho branch of the small GTPase family agarose/MOPS/formaldehyde gels, transferred to a nylon 32 wasidentified in human cancer more than 15 yearsago, membrane, and hybridized to a P-labeled human DLC-1 only in the past few years has evidence accumulated full-length cDNA in QuikHyb buffer (Stratagene). The blot waswashedtwice at room temperature in 1.0 Â SSC/0.1% implicating Rho signaling in malignant trans- SDS for 30 min each, once in 0.1 Â SSC/0.1% SDS at 621C for formation and metastasis (Boetter and van Aelst, 2002; 30 min, then exposed to a phosphorimager plate, and the Jaffe and Hall, 2002). DLC-1 gene deregulation in image analyzed by ImageQuant ersion 3.3 software (Molecular cancer cellsisone of several examplesof the involve- Dynamics, Sunnyvale, CA, USA). After stripping, the blot was ment of Rho GAP genesin human cancer development. hybridized to GAPDH cDNA probe. Among the others are GRAF, a tumor suppressor gene, which is deleted or mutated in myelodysplastic syn- Southern blot analysis dromes, and the p190-A gene, which is located in a Genomic DNAsextracted from human primary breastcancer region of deletion and recombination in gliomasand tissue derived from patients in Iceland and peripheral blood astrocytomas (Borkhardt et al., 2000; Tikoo et al., of the same patient were digested with BglII, separated on 2000). 1% agarose gels, and blotted onto nylon membranes. The Despite the large number of complex genomic membrane wasfirsthybridized to the 32P-labeled L7-3 probe, a changes in breast cancer, gain of 1q and loss of 8p 500 bp genomic fragment of the DLC-1 gene originally isolated and 13 are represented in over 90% of the cases by representative difference analysis from a human primary (Tirkkonen et al., 1998). Significantly, two genescoding HCC tissue (Yuan et al., 1998), then stripped and rehybridized for novel DLC-1-related proteinshave been identified in to a 32P-labeled b-actin cDNA probe. Hybridization, washing, the human cDNA and genomic DNA sequence data- and detection were performed asdescribedabove for Northern bases, one mapping to chromosome 13q14 (GenBank blots. AL049801) and another at chromosome Xq13 (KIAA0189 gene, Nagase et al., 1996; GenBank Construction of DLC-1 cDNA expression vector, transfection, D80011). It is possible that simultaneous deletion or and assays for cell proliferation and colony formation inactivation of the DLC-1 gene on 8p and itsisoform on The 3.3 kb full-length DLC-1 open reading frame starting from 13q, both regionsof recurrent lossof DNA copy the ATG codon to the TGA stop codon (nt 348–3600 of number and LOH in both breast and several other GenBank AF035119) wasamplified by PCR from human cancers, may contribute in concert to tumor develop- placental cDNA. The primersincluded BamHI linkers, and at the 50 end the forward primer incorporated the Flag antigen ment and metastasis. 0 0 In summary, our current observations provide sequence, 5 GAC TAC AAG GAC GAC GAT GAC AAG 3 , to serve as a tag for expression of the protein. After digestion evidence that DLC-1 playsa role in cancer development with BamHI, the PCR product wasinsertedinto the BamHI by acting as a bona fide tumor suppressor gene that site of the multiple cloning site of the pLXSN retrovirus can inhibit the proliferation and tumorigenicity of expression vector (CLONTECH, Palo Alto, CA, USA). DNA cancer cells. Our preliminary indications are that sequencing and restriction enzyme digestion were used to DLC-1 expression has an inhibitory effect on in vitro confirm that the sequence and orientation of the construct cell growth and in vivo tumorigenicity not only in breast were correct. To assay the effect of DLC-1 on cell growth, cancer cell lines but also in HCC-derived and prostate 2 Â 105 MDA-MB-468 and MDA-MB-361 cellswere seededin carcinoma-derived cell lines. In addition, we have 25 cm flasks and transfected with 5 mg of either pLXSN-DLC-1 isolated and characterized the mouse DLC-1 gene or pLXSN vector alone using cationic lipid reagents. The cells (Yuan et al., 1999; Durkin et al., 2002) and have were trypsinized and counted at the designed time. To test for colony formation, MDA-MB-468 and MDA-MB-361 cells constructed a targeting vector for generating mice grown to 20% confluence in 100 mm tissue culture dishes were with a null allele of the gene by homologousrecombina- transfected with 5 mg of either pLXSN-DLC-1 or pLXSN as tion. This step should shed additional light on the indicated above. After 48 h the transfection medium was function and tumor suppressor status of the DLC-1 replaced with medium containing 250 mg/ml G418 (Geneticin; gene. Life Technologies, Inc). After 2 weeks the G418-resistant

Oncogene DLC-1 gene suppresses in vivo tumorigenicity B-Z Yuan et al 450 colonies were washed twice with PBS, stained with crystal histopathologic examination, pieces of tumor tissue were violet, and counted. fixed in 10% neutral buffered formalin, embedded in paraffin, cut, deparaffinized and stained with hematoxylin Tumorigenicity assay and eosin. MDA-MB-468, UACC 893, and MDA-MB-301 cellsstably FISH A human BAC clone (RP11-465E23), which contains transfected with pLXSN-DLC-1 or vector only that survived the sequence of the entire DLC-1 gene, was identified by a G418 selection were used for this assay. The cells were search of the GenBank High-Throughput Genomic Sequence harvested before inoculation and resuspended in medium database and purchased from Research Genetics. Biotinylated without serum and antibiotics. The cells were washed with BAC DNA washybridized on interphasenuclei from MDA- medium twice, and 1 or 2 Â 106 cellswere inoculated MB-468 and MDA-MB-361 transfected and control cells. subcutaneously at the proximal dorsal midline of 4-week- Detection of the hybridization signal, digital image acquisi- old Balb/c athymic nude mice (NCI breeding facility). Tumor tion, and analysis were carried out as previously described size was measured in two dimensions twice a week. For (Zimonjic et al., 2000).

References

Arbieva ZH, Banerjee K, Kim SY, Edassery SL, Maniatis VS, Zhuang ZP. (2001). J. Cancer Res. Clin. Oncol., 127, Horrigan SK and Westbrook CA. (2000). Genome Res., 10, 482–488. 244–257. Sekimata M, Kabuyama Y, Emori Y and Homma Y. (1999). Boetter B and van Aelst L. (2002). Gene, 286, 155–174. Biol. Chem., 274, 17757–17762. Borkhardt A, Bojesen S, Haas OA, Fuchs U, Bartelheimer D, Tikoo A, Czekay S, ViarsC, White S, Heath JK, Arden K and Loncarevic IF, Bohle RM, Harbott J, Repp R, Jaeger U, Maruta H. (2000). Gene 257, 23–31. Viehmann S, Henn T, Korth P, Scharr D and Lampert F. Tirkkonen MTM, Karhu R, Kallioniemi A, Isola J and (2000). Proc. Natl. Acad. Sci. USA, 97, 9168–9173. Kallioniemi OP. (1998). Genes Cancer, 21, Cabeza-Arvelaiz Y, Sepulveda JL, Lebovitz RM, Thompson 177–184. TC and Chinault AC. (2001). Oncogene, 31, 4169–4179. Wilson PJ, McGlinn E, Marsh A, Evans T, Arnold J, Wright Durkin ME, Yuan B-Z, Thorgeirsson SS and Popescu NC. K, Biden K, Young J, Wainwright B, Wicking C (2002), Gene, 288, 119–127. and ChenevixTrench G. (2000). Human Mut., 15, Euer N, Schwirzke M, Evtimova V, Burtscher H, Jarsch M, 156–165. Tarin D and Weidle U. (2002), Anticancer Res., 22, 733–740. Xu J, Zheng SL, HawkinsGA, Faith DA, Kelly B, IsaacsSD, Gustafson CE, Wilson PJ, Lukeis R, Baker E, Woolatt E, Wiley KE, Chang B, Ewing CM, Bujnovszky P, Carpten JD, Annab L, Hawke L, Barrett J and Chenevix-Trench G. Bleecker ER, Walsh PC, Trent JM, Meyers DA and Isaacs (1996).Cancer Res., 56, 5238–5245. WB. (2001). Am. J. Hum. Genet., 69, 341–350. Homma Y and Emori Y. (1995). EMBO J., 14, 286–291. Yaremko ML, Kutza C, Lyzak J, Mick R, Recant WM and Jaffe AB and Hall A. (2002). Adv. Cancer Res., 84, 57–80. Westbrook CA. (1998). Genes Chromosomes Cancer, 21, Jenkins R, Takahashi S, DeLacey K, Bergstralh E and Lieber 189–195. M. (1998). Genes Chromosomes Cancer, 21, 131–143. Yoshioka K, Matsumara F, Akedo H and Itoh K. (1998). J. Martinez-Climent JA, Vizcarra E, Sanchez D, Blesa D, Biol. Chem., 273, 5146–5154. Marugan I, Benet I, Sole F, Rubio-Moscardo F, Terol Yuan BZ, Durkin ME and Popescu NC. (2002). Cancer Genet. MJ, Climent J, Sarsotti E, Tormo M, Andreu E, Sallido M, Cytogenet., in press. Ruiz MA, Prosper F, Siebert R, Dyer MJS and Garcia- Yuan B-Z, Miller MJ, Keck CL, Zimonjic DB, Thorgeirsson Conde J. (2001). Blood, 98, 3479–3482. SS and Popescu NC. (1998). Cancer Res., 58, 2196–2199. Nagase T, Seki N, Ishikawa K, Tanaka and Nomura N. Yuan B-Z, Thorgeirsson SS, Eyfjord J, Gudmundsdottir K (1996). DNA Res., 3, 17–24. and Popescu NC. (2000). Proc. Am. Assoc. Cancer Res, 41, Ng IO, Liang Z-D, Cao L and Lee TK. (2000), Cancer Res., 105. 60, 6581–6581. Yuan BZ, Yang Y, Keck–Waggoner CL, Zimonjic DB, Ponting CP and Aravind L. (1999). Trends Biochem. Sci., 24, Thorgeirsson SS and Popescu NC (1999). Cytogenet. Cell 130–132. Genet., 87, 189–190. Qin LX, Tang ZY, Sham JS, Ma ZC, Ye SL, Zhou XD, Wu Zimonjic DB, Keck-Waggoner CL, Yuan B-Z, KrausMH and ZQ, Trent JM and Guan XY. (1999). Cancer Res., 59, Popescu NC. (2000). Int. J. Oncol., 16, 221–230. 5662–5665. Zitzelsberger H, Lehmann L, Werner M and Bauchinger M. Qin LX, Tang ZY, Ye SL, Liu YK, Ma ZC, Zhou XD, Wu (1997). Histochem. Cell. Biol., 108, 403–417. ZQ, Lin ZY, Sun FX, Tian J, Guan XY, Pack SD and

Oncogene