Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Oncogene (1998) 17, 1321 ± 1326 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 http://www.stockton-press.co.uk/onc SHORT REPORT Up-regulation of CIR1/CROC1 expression upon cell immortalization and in tumor-derived human cell lines Libin Ma1,4, Stacey Broom®eld2, Carol Lavery1, Stanley L Lin3, Wei Xiao2 and Silvia Bacchetti1 1 Cancer Research Group, Department of Pathology, HSC-4H30, McMaster University, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada; 2Department of Microbiology, University of Saskatchewan, 107 Wiggins Road, Saskatoon, SK S7N 5E5, Canada; 3Department of Psychiatry, UMDNJ-Robert Wood Johnson Medical School, 675 Hoes Lane, Piscataway, New Jersey 08854, USA Acquisition of the immortal phenotype by tumor cells represents an essential and potentially rate-limiting step in tumorigenesis. To identify changes in gene expression that are associated with the early stages of cell immortalization, we compared genetically matched pairs of pre-immortal and immortal human cell clones by mRNA dierential display. Two transcripts, denoted CIR1 and CIR2, were identi®ed which were up-regulated in immortal cells. Sequence analysis revealed CIR1 to be identical to the recently cloned CROC1/UEV-1 gene, whereas CIR2 corresponds to an as yet uncharacterized 1.2 kb mRNA. A 5 ± 6-fold elevation in CIR1/CROC1 expression and a 2 ± 3-fold elevation in CIR2 expression were observed in SV40-transformed human enbryonic kidney cells immediately following proliferative crisis, suggesting a potential role for these genes in immortalization. Expression of CIR1/CROC1 was found to be elevated also in a variety of immortal human tumorderived cell lines, as compared to their normal tissue counterparts. These results are compatible with induction of CIR1/CROC1 being an early event in the acquisition of immortality and with a role for this gene in the immortal phenotype of tumor cells. Keywords: CIR1; CROC1; UEV-1; immortalization; tumorigenesis dierential display Normal somatic human cells have a ®nite replicative life span and are ultimately subject to an irreversible arrest of cell division, called cellular senescence (Hay¯ick and Moorhead, 1961; Goldstein, 1990), whereas many human tumor cells display an unlimited cell division capacity when grown in vitro, i.e. they exhibit an immortal phenotype (Stamps et al., 1992). Acquisition of immortality by pre-neoplastic or neoplastic cells is critical to continued tumor growth and metastatic spread in neoplasia. Transformation of human cells in vitro, by DNA tumor viruses or other agents, can eciently abrogate senescence and confer an extended life span (Girardi et al., 1965; Shay et al., 1991; Ozer et al., 1996). However, transformed cells eventually enter a proliferative crisis, from which rare clonal populations of immortal cells arise (Girardi et Correspondence: S Bacchetti Received 21 February 1998; revised 21 April 1998; accepted 21 April 1998 al., 1965). Cell culture models indicate that in order to become immortal human cells must overcome the barriers of both senescence and crisis (or Mortality stages 1 and 2) (Shay et al., 1991). Genetic analysis has suggested that immortalization is the result of loss of gene function(s) that are essential for senescence. Cell hybrids between immortal and normal cells revealed the recessive nature of immortalization, and pair-wise cell fusion analysis between tumor-derived and immortalized human cells suggested the existence of dierent complementation groups (Smith and Pereira-Smith, 1996). Re-introduction of individual human chromosomes into immortal cells can result in growth inhibition and the re-appearance of a senescent-like morphology (Smith and Pereira-Smith, 1996). In some cases, more than one chromosome has been shown capable of abrogating the immortal phenotype in a given cell line (Sasaki et al., 1994), suggesting multiple pathways of cell senescence and immortalization. On the other hand, transfection of oncogenes into cells which undergo senescence has been shown to confer immortality (Edwards, 1993; Harrington et al., 1994), indicating that de novo or elevated expression of speci®c genes could also be involved in immortalization. Moreover, elevated expression of immortalization-related genes is needed for either the prevention of the decline in, or the maintenance of proliferative capacity. For example, immortality is usually associated with the acquisition of telomerase activity which compensates with the proliferation-related decline in telomere length (Counter et al., 1992; Kim et al., 1994; Bacchetti, 1996; Harley and Sherwood, 1997). Identi®cation of such genes and their regulation can be expected to provide insight into mechanisms of cell immortalization and aging. Several laboratories, using dierent approaches such as subtractive hybridization, have identi®ed a number of genetic alterations in immortal cells (Imai and Takano 1992; Satoh et al., 1994; Pardinas et al., 1997), most of them of unknown function in immortalization. These studies focused primarily on the identi®cation of genes that are lost or down-regulated after immortalization. We have used mRNA dierential display (Liang and Pardee, 1992; Liang et al., 1993) to compare gene expression in genetically matched preimmortal and immortal human cells. We report the identi®cation of two transcripts, designated CIR1 and CIR2 (for Cell Immortalization-Related genes), whose expression is up-regulated in immortal cells. While CIR2 represents a novel cDNA, CIR1 is identical to CIR1/CROC1 over-expression in immortal cells L Ma et al 1322 the recently cloned gene, CROC1 (also called UEV-1), which was originally identi®ed on the basis of its capacity to activate c-fos gene transcription (Rothofsky and Lin, 1997), and as a transcript that was downregulated upon dierentiation of the human HT-29-M6 colon carcinoma cell line (Sancho et al., 1998). The location of this gene at 20q13.2 (Sancho et al., 1998), a region that is frequently highly ampli®ed in many types of human cancers (Tanner et al., 1994), suggests that its overexpression may be instrumental to tumorigenesis. Consistent with a role in immortalization, we have observed elevated expression of CIR1/CROC1/UEV1 in several tumor-derived cell lines. Results and discussion Strategy for dierential display of cell immortalizationrelated (CIR) genes To identify cellular genes that are regulated during the process of cell immortalization and potentially contribute to the immortal phenotype, we have utilized dierential display of mRNAs (Liang and Pardee, 1992; Liang et al., 1993) to compare pre-immortal clonal populations of SV40-transformed human embryonic kidney (HEK) cells with their corresponding immortal clones isolated following proliferative crisis (Stewart and Bacchetti, 1991; Counter et al., 1992; S Bacchetti, unpublished results). Such matched pairs of pre-immortal and immortal cells are particularly suitable for the comparison of gene expression, since they were derived from the same transformed cell and thus should be essentially isogenic. Previously, we had shown that in the pre-immortal HA1 clone, telomere length declines with cell division and telomerase remains inactive until the cells reach proliferative crisis, while in immortal HA1 cells telomere length stabilizes and telomerase is activated (Counter et al., 1992; Avilion et al., 1996). Recently, we also found that up-regulation of the mRNA encoding the catalytic subunit of human telomerase, hTERT (Meyerson et al., 1997; Weinrich et al., 1997; Nakamura and Cech, 1998), coincides with the emergence of immortal HA1 cells (Meyerson et al., 1997). Stringent criteria were utilized to de®ne and select immortalization-related genetic alterations. First, we chose to compare pre-immortal and immortal cell populations just prior to and immediately after proliferative crisis, to eliminate from the analysis genetic alterations that occurred upon transformation but are not necessarily related to cell immortalization. To further minimize random changes in gene expression that may arise during growth of transformed cells, we included in the analysis three HA1 subclones (A33, A34 and C12) that were obtained by limiting dilution at very early times after transformation. Like the parental cells, these subclones underwent proliferative crisis and eventually gave rise to immortal cells (S Bacchetti, unpublished results). Lastly, only changes in gene expression that were consistent in all four clones, when analysed by dierential display and/or Northern hybridization, were considered as indicative of potential Cell Immortalization-Related (CIR) genes. Up-regulation of CIR1 and CIR2 expression in immortal cells Using 16 arbitrary primers in combination with three one-base anchored primers (see Figure 1), an average of *100 ± 120 cDNA bands were displayed in each lane. Thus, a total of approximately 5000 cDNAs from each cell population, representing roughly 10 ± 15% of the transcribed genes (Liang and Pardee, 1992), were directly compared on display gels. We identi®ed 16 candidate bands that diered in intensity between preimmortal and immortal HA1 and A33 populations (e.g. see Figure 1). Of these, only two cDNA fragments, designated CIR1 and CIR2 (Figure 1a,b), were consistently up-regulated in immortal cells on display gels and were dierentially expressed in all four cell lines as judged by Northern analysis (Figure 2, compare lanes 1, 3, 5 and 7 to lanes 2, 4, 6 and 8). Dierential expression of the remaining cDNAs could not be con®rmed on Northern blots: these cDNAs either did not produce detectable signal, or were not dierentially expressed, or exhibited dierential expression only in a subset of cell lines (data not shown). Others have also reported a comparable high percentage of false positive signals on display gels (Liang et al., 1993). On display gels, the CIR1 band (*650 bp in size) was visible in pre-immortal HA1 and A33 cell populations but its intensity increased in immortal cells (Figure 1a, lanes 1 and 2 versus 3 and 4). The CIR2 band (*500 bp) was barely visible in pre-immortal cells but became readily detectable after immortalization (Figure 1b, lanes 1, 2 versus 3, 4). Northern blot analysis con®rmed these observations (Figure 2). The CIR1 fragment detected a doublet of mRNA species of approximately 2.0 ± 2.1 kb, which corresponds to two isoforms of the CROC1 gene (see below) and the CIR2 probe hybridized to a single mRNA species of &1.2 kb (Figure 2). Based on quantitation by PhosphorImager analysis and normalization for RNA loading by GAPDH control hybridization, we estimated the increase in expression level upon immortalization to be 5 ± 6-fold for CIR1 and 2 ± 3-fold for CIR2. The extent of up-regulation of CIR1 and CIR2 in immortal cells was reproducibly observed on independent Northern blots (data not shown). CIR1 is identical to CROC1 The CIR1 and CIR2 fragments were cloned (see Figure 3); sequencing of four independent clones of each revealed that they contain identical CIR1 or CIR2 inserts. When used as probes on Northern blots, the cloned fragments detected the same bands and revealed the same pattern of dierential expression in preimmortal versus immortal cell populations (data not shown), thus ensuring the authenticity of the cloned fragments. The sequences of the CIR1 and CIR2 clones are shown in Figure 3. Both fragments were ¯anked by sequences corresponding to the one-base anchored oligo(dT) primer and the arbitrary primer used for dierential display, and putative polyadenylation signal sequences were identi®ed in both cases. Homology searches through the NIH GenBank and the EMBL databases, using the BLAST and FASTA programs, revealed no signi®cant homology with any published genes in the case of the CIR2 sequence, although a CIR1/CROC1 over-expression in immortal cells L Ma et al number of human expressed sequence tags (ESTs) of unknown function that are identical or almost identical to CIR2 were identi®ed (e.g., Genbank accession nos. AA156251, AA075294 and W39392). Further analysis of CIR2 was not pursued. On the other hand, the CIR1 sequence was found to perfectly match the 3'-end sequences of two GeneBank entries previously identi®ed as the CROC1 gene (CROC-1A, accession no. U39360, and CROC-1B, accession (Rothofsky and Lin, 1997). no. U39361) Expression of CIR1/CROC1 in human tissues and tumor-derived cell lines The CROC1 genes were originally isolated on the basis of their ability to mediate transcriptional activation of Figure 1 Representative banding patterns on mRNA dierential display gels showing up-regulated PCR fragments in immortal cells. CIR1 (a) and CIR2 (b) are indicated by arrowheads. Lanes 1 and 2, RNA from pre-immortal A33 and HA1 cells; lanes 3 and 4, RNA from immortal HA1 and A33 cells. Genetically matched pairs of pre-immortal and immortal HA1 cells were established from SV40-transformed human embryonic kidney (HEK) cells and cultured as described previously (Stewart and Bacchetti, 1991; Counter et al., 1992). Three HA1 subclones, A33, A34, and C12 were obtained by limiting dilution of HA1 cells at a very early stage after SV40 transformation (S Bacchetti, unpublished data). Total cellular RNA was isolated using the TRIzol reagent according to the manufacturer's guidelines (Gibco/BRL Life Technologies) and was treated with RNase-free DNase I (Pharmacia) to remove contaminating DNA (Liang et al., 1993). RNA was quantitated by UV absorbance. Dierential display was performed essentially as described (Liang and Pardee, 1992; Liang et al., 1993). Brie¯y, an aliquot of DNA-free total RNA (0.2 mg in 20 ml) was reverse transcribed with 100 units of Mo-MLV reverse transcriptase for 1 h at 378C, in the presence of 20 mM deoxynucleotide triphosphates and 0.2 mM H-T11M (5'-AAGCTTTTTTTTTTTM-3', where M is G, C, or A). Two ml of the resultant cDNA was ampli®ed by PCR in a 20 ml reaction volume, with 2.0 mM deoxynucleotide triphosphates, 0.2 mM downstream H-T11M primer, 0.2 mM upstream arbitrary primers (H-AP1-8 and H-AP41-48; RNAimage kits No. 1 and 6; GeneHunter, Nashville, TN), in the presence of [a-33P]dATP (2000 Ci/mmol, DuPont/NEN) and 1 unit of Taq DNA polymerase (Boehringer Mannheim). The 5' arbitrary primer that gave CIR1 and CIR2 products was H-AP42 (5'-AAGCTTTGCACCG-3') and 3' anchored primers were H-T11A and H-T11C, respectively. PCR conditions were 948C for 5 min, followed by 40 cycles at 948C for 30 s, 408C for 2 min and 728C for 30 s. A ®nal extension was carried out at 728C for 5 min. An aliquot of each PCR reaction (5 ml) was analysed on a 6% polyacrylamide DNA sequencing gel. Bands on the display gels that showed dierential intensity in pre-immortal and immortal cells were visualized by exposing the dried gels to X-ray ®lms, excised from the gel, and soaked in water to elute the DNA, which was then reampli®ed by PCR using the same primer set (Liang et al., 1993) 1323 CIR1/CROC1 over-expression in immortal cells L Ma et al 1324 Figure 2 Up-regulation of CIR1 and CIR2 in immortal cells. The blot was sequentially hybridized to reampli®ed PCR fragments corresponding to CIR1 (upper) and CIR2 (middle) and to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA. Lanes 1, 3, 5 and 7, RNA from pre-immortal HA1, A33, A34 and C12 cells; lanes 2, 4, 6 and 8, RNA from immortal HA1, A33, A34 and C12 cells. Fifteen mg of total RNA was size fractionated on a 1.2% agarose-formaldehyde gel and transferred onto ZetaProbe GT membranes (BioRad) by standard capillary blotting techniques in 106SSC. Speci®c cDNA probes were labeled with [a-32P]dCTP using a random priming DNA labeling kit (Boehringer Mannheim). The ®lters were hybridized overnight at 428C in 50% formamide, 56SSC, 0.5 M sodium phosphate buer and 1% SDS, washed twice in 26SSC/0.1% SDS for 30 min followed by 0.16SSC/0.1% SDS for 15 ± 30 min at 558C. Quantitation of mRNA levels was done by PhosphorImager (Molecular Dynamics). Total counts of the CIR1 or CIR2 band (minus background) were divided by the total count of the GAPDH band (minus background) to normalize for dierences in lane loading Figure 3 Nucleotide sequences of CIR1 and CIR2 cDNAs. Primers originally used for dierential display are depicted in lowercase for CIR1 (a) and CIR2 (b) (Genbank accession no. AFO 49671 and AFO 49672, respectively). Putative polyadenylation signal sequences are in bold. The reampli®ed CIR1 and CIR2 fragments were cloned into pBluescript at the EcoRV site after the addition of T overhangs. Four independent clones of each were sequenced and all four contained identical CIR1 or CIR2 inserts the human c-fos promoter from its direct repeat enhancer sequences (Rothofsky and Lin, 1997). CROC-1A and CROC-1B are predicted to encode two proteins of 170 and 221 amino acids, respectively, that are identical in their carboxy-terminal 140 residues (Rothofsky and Lin, 1997). An isoform of CROC1, called UEV-1, was recently identi®ed as a transcript that was down-regulated upon dierentiation of a human colon carcinoma cell line (Sancho et al., 1998). CROC1/UEV-1 represents a class of newly de®ned ubiquitin-conjugating E2 enzyme variant proteins, which share signi®cant sequence similarity and predicted sedondary and tertiary structure with the ubiquitin-conjugating enzymes, but which lack ubiquitin-conjugating activity (Sancho et al., 1998). We have also identi®ed a Saccharomyces cerevisiae gene, MMS2 (orf YGL087c), that is closely related to CIR1/CROC1 (48% identity in a 136 amino-acid overlap) (Sancho et al., 1998; Broom®eld et al., 1998; LM and S Bacchetti, unpublished data) and found that haploid yeast strains carrying deletions of MMS2 were viable, suggesting that this gene is not essential for yeast cell survival. We have examined the levels of CIR1/CROC1 mRNA in human tissues and tumor-derived cell lines. Northern analysis of human multiple tissue blots revealed two major mRNA species of about 2.0 ± 2.1 kb (Figure 4), which represent the two forms of CROC1 (Rothofsky and Lin, 1997). CIR1/CROC1 mRNA was present in all normal tissues but levels of expression were highly variable (Figure 4a,b,d). We examined CIR1/CROC1 expression in eight dierent tumor cell lines (Figure 4c). All were positive for CIR1/CROC1 expression and levels of mRNAs were consistently higher than in most normal human tissues, and in particular substantially higher than in the corresponding normal tissues (e.g., compare A549 versus lung, SW480 versus colon, and three leukemia cell lines HL-60, K562 and MOLT-4 versus peripheral blood leukocytes and bone marrow). The CROC1/UEV1-1 gene is located on chromosome 20q13.2 (Sancho et al., 1998), a region frequently ampli®ed in many types of human cancer (Tanner et al., 1994). Others have reported that SW480 cells contain aberrant chromosomes harboring large portions of 20q (Brinkmann et al., 1996), a ®nding that may account, at least in part, for the high level of CIR1/CROC1 mRNA present in this cell line. In conclusion, by means of dierential display of mRNAs we have identi®ed two species, CIR1 and CIR1/CROC1 over-expression in immortal cells L Ma et al 1325 Figure 4 Expression of CIR1/CROC1 in human tissues and tumor cell lines. Northern blots containing 2 mg of poly(A)-selected mRNA from multiple human tissues and cancer cell lines were purchased from Clontech (Palo Alto, CA) and are as follows: (a) Human MTN blot, (b) Human III MTN blot, (c) Human Cancer Line MTN blot, and (d) Human II MTN blot. Tissues and cell lines from which RNA was isolated are indicated. Top and bottom panels contain the results after hybridization with the 0.86 kb SalI ± BsaBI fragment of the CROC-1B cDNA (Rothofsky and Lin, 1997) and with a b-actin probe (Clontech), respectively. All four membranes were hybridized with the CROC1 probe and exposed to X-ray ®lm under identical conditions. Hybridization of Human III MTN blot (b) and Human Cancer Line MTN blot (c) with b-actin was performed together, whereas Human MTN blot (a) and Human II MTN (d) were hybridized on separate occasions CIR2, that are up-regulated early during cell immortalization. Our ®ndings extend the available information on transcriptional alterations that are associated with the immortal phenotype, although it remains to be established whether enhanced expression of CIR1 and CIR2 is causal to or merely a consequence of cell immortalization. Sequence analysis revealed that CIR1 is identical to the recently cloned CROC1/UEV-1 gene (Rothofsky and Lin, 1997; Sancho et al., 1998) which de®nes a novel subfamily of highly conserved, inactive variants of the ubiquitin-conjugating E2 enzymes (Sancho et al., 1998). Several observations suggest a potential involvement of CIR1/CROC1/UEV-1 in cell proliferation and dierentiation: (i) CROC1 is a nuclear protein that is capable of activating transcription of the c-fos promoter direct enhancer sequences, implicating the protein in the RAF-mediated signal transduction pathway (Rothofsky and Lin, 1997); (ii) constitutive expression of UEV-1 inhibits the capacity of HT-29-M6 cells to dierentiate upon reaching con¯uence and has profound inhibitory eects on the activity of the mitotic kinase cdk1, resulting in DNA endoreduplication and accumulation of cells in G2-M (Sancho et al., 1998). Inhibition of differentiation would be a prerequisite for the maintenance of proliferative capacity of immortalized cells; (iii) CROC1/UEV-1 maps to chromosome 20q13.2 (Sancho et al., 1998), a region which is moderately ampli®ed in many types of cancer (e.g. Kallioniemi et al., 1994; Muleris et al., 1994; Tanner et al., 1994, 1995; Brinkman et al., 1996) and highly ampli®ed in others, e.g. breast cancer, where it correlates with aggressiveness and poor prognosis (Tanner et al., 1995). There is also a strong association between 20q13.2 ampli®cation and genomic instability in HPVtransformed uroepithelial cells and between high frequency of 20q gain and cell immortalization (Savelieva et al., 1997; Yeager et al., 1998). Indeed, 20q gain is thought to re¯ect one genetic alteration required by tumor cells to overcome senescence (Savelieva et al., 1997; Yeager et al., 1998). Lastly, we have documented that expression of CIR1/CROC1 is up-regulated early in cell immortalization and is common in tumor-derived cell lines. Taken together, these ®ndings suggest that CIR1/CROC1 might be implicated in the immortalization process and consequently also in human carcinogenesis. Acknowledgements We thank Richard Austin for help in setting up dierential display, Brenda Andrews for help with the yeast experiments, BL Chow for technical assistance, Afra van der Beek and Ping Wang for helpful discussions and Afra van der Beek and Frank Graham for critical reading of the manuscript. This work was supported by grants from the Medical Research Council (MRC) and the National Cancer Institute of Canada (NCIC) to S Bacchetti and WX. LM is a postdoctoral fellow of the Dutch Cancer Society and was also supported by a MRC postdoctoral fellowship for six months during this work. WX is a Research Scientist of the NCIC and S Bacchetti is a Terry Fox Cancer Research Scientist of the NCIC. CIR1/CROC1 over-expression in immortal cells L Ma et al 1326 References Avilion AA, Piatyszek MA, Gupta J, Shay JW, Bacchetti S and Greider CW. (1996). Cancer Res., 56, 645 ± 650. Bacchetti S. (1996). Cancer Surv., 28, 197 ± 216. Brinkmann U, Gallo M, Polymeropoulos MH and Pastan I. (1996). Genome Res., 6, 187 ± 194. Broom®eld S, Chow BL and Xiao W. (1998). Proc. Natl. Acad. Sci. USA, in press. Counter CM, Avilion AA, LeFeuvre CE, Stewart NG, Greider CW, Harley CB and Bacchetti S. (1992). EMBO J., 11, 1921 ± 1929. Edwards PA. (1993). Cancer Surv., 16, 79 ± 96. Girardi AJ, Jensen FC and Koprowski H. (1965). J. Cell. Comp. Physiol., 65, 69 ± 84. Goldstein S. (1990). Science, 249, 1129 ± 1133. Harley CB and Sherwood SW. (1997). Cancer Surv., 29, 263 ± 284. Harrington EA, Fanidi A and Evan GI. (1994). Curr. Opin. Genet. Dev., 4, 120 ± 129. Hay¯ick L and Moorhead PS. (1961). Exp. Cell Res., 25, 585 ± 621. Imai S and Takano T. (1992). Biochem. Biophys. Res. Commun., 189, 148 ± 153. Kallioniemi A, Kallioniemi O-P, Piper J, Tanner MM, Stokke T, Chen L, Smith HS, Pinkel D, Gray JW and Waldman FM. (1994). Proc. Natl. Acad. Sci. USA, 91, 2156 ± 2160. Kim NW, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PL, Coviello GM, Wright WE, Weinrich SL and Shay JW. (1994). Science, 266, 2011 ± 2015. Liang P and Pardee AB. (1992). Science, 257, 967 ± 971. Liang P, Averboukh L and Pardee AB. (1993). Nucleic Acids Res., 21, 3269 ± 3275. Meyerson M, Counter CM, Eaton EN, Ellisen LW, Steiner P, Caddle SD, Ziaugra L, Beijersbergen RL, Davido MJ, Liu Q, Bacchetti S, Haber DA and Weinberg RA. (1997). Cell, 90, 785 ± 795. Muleris M, Almeida A, Gerbault-Seureau M, Malfoy B and Dutrillaux B. (1994). Genes Chromosomes Cancer, 10, 160 ± 170. Nakamura TM and Cech TR. (1998). Cell, 92, 587 ± 590. Ozer HL, Banga SS, Dasgupta T, Houghton J, Hubbard K, Jha KK, Kim SH, Lenahan M, Pang Z, Pardinas JR and Patsalis PC. (1996). Exp. Gerontol., 31, 303 ± 310. Pardinas J, Pang Z, Houghton J, Palejwala V, Donnelly RJ, Hubbard K, Small MB and Ozer HL. (1997). Cell. Physiol., 171, 325 ± 335. Rothofsky ML and Lin SL. (1997). Gene, 195, 141 ± 149. Sancho E, Vila MR, Sanchez-Pulido L, Lozano JJ, Paciucci R, Nadal M, Fox M, Harvey C, Bercovich B, Loukili N, Ciechanover A, Lin SL, Sanz F, Estivill X, Valencia A and Thomson TM. (1998). Mol. Cell. Biol., 18, 576 ± 589. Sasaki M, Honda T, Yamada H, Wake N, Barrett JC and Oshimura M. (1994). Cancer Res., 54, 6090 ± 6093. Satoh Y, Kashimura M, Kaneko S, Karasaki Y, Higashi K and Gotoh S. (1994). Mutat. Res., 316, 25 ± 36. Savelieva E, Belair CD, Newton MA, DeVries S, Gray JW, Waldman F and Rezniko CA. (1997). Oncogene, 14, 551 ± 560. Shay JW, Wright WE and Werbin H. (1991). Biochim. Biophys. Acta., 1072, 1 ± 7. Smith JR and Pereira-Smith OM. (1996). Science, 273, 63 ± 67. Stamps AC, Gusterson BA and O'Hare MJ. (1992). Eur. J. Cancer, 28A, 1495 ± 1500. Stewart N and Bacchetti S. (1991). Virology, 180, 49 ± 57. Tanner MM, Tirkkonen M, Kallioniemi A, Collins C, Stokke T, Karhu R, Kowbel D, Shadravan F, Hintz M, Kuo WL, Waldman FM, Isola JJ, Gray JW and Kallioniemi O-P. (1994). Cancer Res., 54, 4257 ± 4260. Tanner MM, Tirkkonen M, Kallioniemi A, Holli K, Collins C, Kowbel D, Gray JW, Kallioniemi O-P and Isola JJ. (1995). Clin. Cancer Res., 1, 1455 ± 1461. Weinrich SL, Pruzan R, Ma L, Oullette M, Tesmer VM, Holt SE, Bodnar AG, Lichtsteiner S, Kim NW, Trager JB, Taylor RD, Carlos R, Andrews WH, Wright WE, Shay JW, Harley CB and Morin GB. (1997). Natl. Genet., 17, 498 ± 502. Yeager TR, DeVries S, Jarrard DF, Kao C, Nakada SY, Moon TD, Bruskewitz R, Stadler WM, Meisner LF, Gilchrist KW, Newton MA, Waldman F and Rezniko CA. (1998). Genes & Dev., 12, 163 ± 174.