Amplification and overexpression of COPS3 in osteosarcomas ...

Oncogene (2003) 22, 5358–5361

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

SHORT REPORTS

Amplification and overexpression of COPS3 in osteosarcomas potentially target TP53 for proteasome-mediated degradation Jørn Henriksen1, Trude H Aagesen2, Gunhild M Maelandsmo1, Ragnhild A Lothe2, Ola Myklebost*,1 and Anne Forus1 1 Department of Tumour Biology, Institute of Cancer Research, The Norwegian Radium Hospital, 0310 Oslo, Norway; 2Department of Genetics, Institute of Cancer Research, The Norwegian Radium Hospital, 0310 Oslo, Norway

In sarcomas, the TP53 tumour suppressor pathway may be altered either by TP53 mutations or by amplification of MDM2, encoding a protein that inhibits TP53 and targets it for 26S-proteasome degradation. However, in the majority of the analysed clinical samples, neither of these types of aberrations are found, suggesting that additional mechanisms are involved. The present study shows that COPS3, located in 17p11 and encoding a component of the proteasome pathway, is more frequently amplified in osteosarcomas (OS) than is MDM2. We present detailed analysis of TP53 mutations and MDM2 and COPS3 expression levels in a set of 23 OS. Our results show that none of the tumours with COPS3 amplification had MDM2 amplification nor TP53 mutations, consistent with the hypothesis that one of the three aberrations is sufficient. The results suggest that inactivation of otherwise intact TP53 by aberrations in the proteasome pathway may contribute to the characteristic aneuploidy observed in OS. Oncogene (2003) 22, 5358–5361. doi:10.1038/sj.onc.1206671 Keywords: COP9 signalosome; chromosome 17; mdm2; sarcoma; CSN3; CGH

Osteosarcomas (OS) are highly malignant tumours of bone, derived from primitive bone-forming mesenchymal cells. They occur mostly during the adolescent growth period, suggesting a relation between rapid bone formation and tumour development (Link and Eilber, 1993). Cytogenetic studies have demonstrated highly aneuploid and extremely complex karyotypes with numerous abnormalities (Mertens et al., 1993; Tarkkanen et al., 1993; Fletcher et al., 1994; Sandberg and Bridge, 1994). The most common numerical changes are gain of chromosome 1, losses of chromosomes 9, 10, 13 and 17. Frequent structural changes are losses of chromosome bands 13q14 and 17p13 as well as partial or complete loss of 6q (Sandberg and Bridge, 1994; Bridge et al., 1997). Comparative genomic hybridization (CGH) studies reported high-level amplification of 1q21–23, 6p, 8q and 17p11–12 to be among the most common aberrations of *Correspondence: O Myklebost; E-mail: [email protected]

these tumours (Forus et al., 1995a; Tarkkanen et al., 1995). Amplification of these regions was frequently observed also in other sarcoma subtypes. In particular, amplification of 1q21–q22 is a common event in liposarcomas (Forus et al., 1995b; Szymanska et al., 1996; Pedeutour et al., 1999) and malignant fibrous histiocytomas (Forus et al., 1995b; Simons et al., 2000), whereas amplification of 17p occurred in 24% of the leiomyosarcomas studied (El-Rifai et al., 1998). So far, no candidate genes have been associated with the amplicon at 17p11–12, but some markers that were occasionally amplified in tumours with gain of 17p11–12 have been identified (Hulsebos et al., 1997; Wolf et al., 1999). In a recent transcriptome study of OS using cDNA microarrays (Forus et al., unpublished), we found particularly high relative expression levels of COPS3 in a sample with amplification of 17p11–12. The map position of COPS3 is 17p11.2 (Elsea et al., 1999), and the gene encodes a subunit of the COP9 signalosome (Potocki et al., 1999), which is implicated in several regulatory processes, many involving ubiquitinmediated proteolysis (Bech-Otschir et al., 2002, and references therein). Recently, it was shown that COP9 signalosome-specific phosphorylation targets TP53 to MDM2-mediated ubiquitination and subsequent degradation by the 26S-proteasome (Bech-Otschir et al., 2001). Interestingly, MDM2, located in 12q13–15 (Oliner et al., 1992), may also be amplified in OS, and such amplification has been associated with metastases (Ladanyi et al., 1993; Nakayama et al., 1995; Berner et al., 1996). The link to TP53 and MDM2 makes COPS3 an interesting candidate gene for the 17p amplifications. We therefore analysed amplification and expression levels of COPS3 in 23 OS samples, including five with known 17p amplification or gain of chromosome 17, and compared this to MDM2 amplification and expression levels and mutation analyses of TP53. The analysed material included 11 patient samples (osteosarcomas), 10 human OS grown subcutaneously as xenografts in nude mice and two OS cell lines. The 11 samples were from primary tumours and 10 were from metastatic samples (Table 1). Two sample pairs were included: patient sample OS3/xenograft OS3x and sample OS11/xenograft OS11x. The tissues were collected immediately after surgery, snap-frozen in liquid nitrogen and stored at 801C. All the tumours

Amplification and overexpression of COPS3 J Henriksen et al

5359 Table 1 Copy number and expression levels of MDM2 and COPS3 in relation to status of TP53 in OS samples

Tumor OS3 OS3x OS4x OS5x OS6x OS7x OS8x OS9x OS11 OS11x OS12x OS13x OS14 OS15 OS17 OS18 OS20 OS21 OS24 OS34 OS41 OSA cl U2-OS cl

Sample information Stage Grade CGH1 met 4 NA prim 4 NA prim 4 17p11-12 prim 4 17p prim 4 12q14 prim 4 met 4 17p11-12 prim 4 17 met 2 NA prim 2 12q14 met met 4 prim 4 NA met 3 NA met 4 NA met 3 NA prim 3-4 NA prim 4 17p11-12 prim 3 NA met NA prim NA NA NA

MDM2 COPS3 DNA mRNA DNA mRNA N NA N NA N 1 N 3 2-3 2 N 0 N 0 N 2 0 N 1 N3 N 0 N 0 3 N 0 N 0 N 1 N 0 N 1 2-3 0 N 3 N 0 N 1 N 0 N 1 3 N 0 N 0 N 1 N 0 N 1 N NA N NA N 0 2 N NA NA N 0 2-3 2 N NA N NA N NA N NA 3 N 2 N 2 N 2

Mutations no no no 856 G → A no no no 493 C → T no no no 843 C → A no no no no NA no no 612 G → T 581-610 del no no

TP532 Codon alteration no no no E286K no no no Q165stop no no no D281E no no no no NA no no E204stop in-frame deletion no no

1 Only amplification of 12q13-15 or 17p is listed. “-“ = analysed, but no amplification of these regions were detected (Forus et al., 1995a). 2 Screening for mutations in exons 5, 7 and 8 of TP53 gene were published previously (Flørenes et al., 1994) for fourteen of the samples. All of these samples were reanalysed for mutations within exon 6 and the rest of the samples were screened for mutation residing in exon 5-8 using temporal temperature gel electrophoresis and sequence analysis as previously described (Lothe et al., 2001). 3 This sample has CDK4 amplified.

Prim = primary, met = metastatic, NA = not analysed, N = normal copy number Amplification = signal intensity relative to normal samples N 2-3 normal borderline low-level (2-3 fold) (3-5 fold)

moderate (5-10 fold)

high (>10 fold)

mRNA expression: 0 = no expression, 1 = low expression, 2 = moderate expression, 3 = high expression

were classified according to the WHO International Histological Classification of Tumours (Weiss, 1994). Among the 10 samples analysed by CGH, three (OS4x, OS8x and OS21) showed amplification of 17p11– 12, one (OS5x) revealed gain of the whole short arm (17p) and one (OS9x) showed gain of chromosome 17. Only two of the 10 samples (OS6x and OS11x) demonstrated amplification of 12q14 (Forus et al., 1995a). CDK4 was amplified in both of these samples (Berner et al., 1996), whereas the amplicon included MDM2 in one (Table 1). As shown in Table 1 and Figure 1, Southern blot analyses revealed amplification (42 copy number gain) of COPS3 in six of the 23 samples. Two of the samples showed borderline amplification (OS4x and OS24), one showed low-level amplification (OS20) and three samples revealed moderate amplification levels (OS8x, OS14 and OS21).

MDM2 was amplified in only two of the samples, at borderline level in OS11x and high level in the OSA cell line. Expression of COPS3 and MDM2 mRNA was analysed in the 19 samples from which RNA was available. Four of the samples showed high expression of COPS3 mRNA, two of these had moderate amplification of COPS3, whereas the other two had normal copy numbers. A moderate expression of COPS3 was observed in six samples, of which three showed normal and three had borderline or low-level amplification. Expression of MDM2 mRNA was detected in three samples, at high level only in the sample with high level of amplification (OSA). Moderate or low expression of MDM2 was seen in U2-OS and one of the tumour samples (OS3x), both having normal copy number of MDM2 (Table 1). Oncogene


5360

Figure 1 Representative Southern and Northern blot analysis showing amplification and elevated expression of MDM2 and COPS3 in OS samples. (a) DNA was extracted, digested, separated and blotted onto nylon membranes as described previously (Forus et al., 1993). Southern blots were sequentially hybridized to probes from COPS3, MDM2 and to a control probe from chromosome 2 (APOB). The net signals from specific bands were corrected for unequal sample loading by calibration relative to the signal obtained with the APOB probe. The signals were compared to signals from control samples with a normal karyotype. N indicates a normal copy number and amplified genes are designated by an A. The probes used were PCR products of the IMAGE clone 813983 (COPS3) and cDNA probes for MDM2 (Oliner et al., 1992) and APOB (Huang et al., 1985). (b) RNA was extracted, separated and blotted onto nylon membranes and hybridizations were performed as described (Forus et al., 1993). Probes were sequentially hybridized to the filter membranes and signals from an 18S rRNA oligonucleotide were used to correct for unequal sample loading

Mutation analyses of TP53 (exons 5–8) identified four tumours with single-base substitutions and an in-frame deletion in one sample (Table 1), all with normal copy numbers of MDM2 or COPS3. Two of the mutations resulted in premature stop codons (OS11 and OS34) and thus shortened proteins. Two mutations resulted in amino-acid substitutions, but the alteration in codon 281 (Asp-to-Glu) might not have any functional implication, since both Asp and Glu belongs to the acidic group of amino acids. The alteration of codon 286, resulting in Glu-to-Lys, would probably disrupt the structure of the core domain that binds specific DNA sequences (Cho et al., 1994). The deletion of 30 base pairs (581–610) does not change the frame, but removes 10 amino acids from the central portion of TP53 and thereby probably affects its DNA binding ability. In our panel of tumours, amplification of COPS3 was observed in 6/24 tumours, compared to 2/24 demonOncogene

strating MDM2 amplification. On the other hand, two of the samples with known dosage alterations of chromosome 17 but no localized amplification, OS5x and OS9x, showed normal copy numbers of COPS3. It seems likely that gain of the whole chromosome 17 or 17p, as observed in these cases, is selected by mechanisms different from that of local amplification of 17p11–12. Interestingly, amplification of MDM2 and COPS3 was never observed in the same tumour sample, consistent with previous CGH studies of OS where amplification of 12q13–15 and 17p11–12 was never observed in the same tumour (Forus et al., 1995a; Stock et al., 2000). Therefore, amplification of MDM2 or COPS3, or TP53 mutation, may be complementary events. Furthermore, we only found mutations in TP53 in tumour samples without COPS3 or MDM2 amplification. High expression of COPS3 was found in the samples with high amplification of the gene and in two additional samples with normal copy number. The high expression of COPS3 in the two samples with normal copy number shows that other mechanisms than amplification may induce overexpression of COPS3. One sample showed low copy number gain of MDM2, but the expression of MDM2 mRNA was below the level of detection. This alteration might be due to gain of a larger chromosomal area and may in fact reflect other gene targets such as CDK4, localized in the vicinity (Berner et al., 1996). The rather frequent amplification COPS3 suggests an important role in development or progression in OS. Furthermore, our results are consistent with the hypothesis that amplification of MDM2 and COPS3 are mutually exclusive events both causing the degradation of TP53 by the 26S-proteasome. In normal cells, this pathway must be tightly regulated. In tumour cells with a normal TP53, we suggest that amplification and overexpression of either MDM2 or COPS3 may shift this balance towards increased degradation of the TP53 protein, resulting in a phenotype equivalent to inactivation by mutation. This would explain why amplification of these genes is not observed in cells with mutated TP53. Recently, Hansen et al. (2002) proposed a model in which TP53 loss of function through 26S-proteasome degradation results in genomic instability and polyploidy due to amplification of the centrosome (Hansen et al., 2002, and references therein). Since COPS3 is part of the same pathway, one might speculate that amplification and overexpression of this gene may also affect centrosome instability, and that the aberrations affecting the proteasome pathway may be responsible for the complex karyotypes observed in OS. Clearly, further experiments are needed to confirm these hypotheses.

Acknowledgements The work was supported by the Norwegian Cancer Society. We like to thank Anine B Dahlberg and Marianne Berg for excellent assistance to this work.


5361 References Bech-Otschir D, Kraft R, Huang X, Henklein P, Kapelari B, Pollmann C and Dubiel W. (2001). EMBO J., 20, 1630–1639. Bech-Otschir D, Seeger M and Dubiel W. (2002). J. Cell Sci., 115, 467–473. Berner J-M, Forus A, Elkahloun A, Meltzer PS, Fodstad Ø and Myklebost O. (1996). Genes Chromosome Cancer, 17, 245–259. Bridge JA, Nelson M, McComb E, McGuire MH, Rosenthal H, Vergara G, Maale GE, Spanier S and Neff JR. (1997). Cancer Genet. Cytogenet., 95, 74–87. Cho Y, Gorina S, Jeffrey PD and Pavletich NP. (1994). Science, 267, 346–355. El-Rifai W, Sarlomo-Rikala M, Knuutila S and Miettinen M. (1998). Am. J. Pathol., 153, 985–990. Elsea SH, Mykytyn K, Ferrell K, Coulter KL, Das P, Dubiel W, Patel PI and Metherall JE. (1999). Am. J. Med. Genet., 87, 342–348. Fletcher JA, Gebhardt MC and Kozakewich HP. (1994). Cancer Genet. Cytogenet., 77, 81–88. Flrenes VA, Maelandsmo GM, Forus A, Andreassen A˚, Myklebost O and Fodstad Ø. (1994). J. Natl. Cancer Inst., 86, 1297–1302. Forus A, Flrenes VA, Maelandsmo GM, Meltzer PS, Fodstad Ø and Myklebost O. (1993). Cell Growth Differ., 4, 1065–1070. Forus A, Olde Weghuis D, Smeets D, Fodstad Ø, Myklebost O and Geurts van Kessel A. (1995a). Genes Chromosome Cancer, 14, 15–21. Forus A, Olde Weghuis D, Smeets D, Fodstad Ø, Myklebost O and Geurts van Kessel A. (1995b). Genes Chromosome Cancer, 14, 8–14. Hansen DV, Hsu JY, Kaiser BK, Jackson PK and Eldridge AG. (2002). Oncogene, 21, 6209–6221. Huang LS, Bock SC, Feinstein SI and Breslow JL. (1985). Proc. Natl. Acad. Sci. USA, 82, 6825–6829. Hulsebos TJ, Bijleveld EH, Oskam NT, Westerveld A, Leenstra S, Hogendoorn PC and Bras J. (1997). Genes Chromosomes Cancer, 18, 279–285. Ladanyi M, Cha C, Lewis R, Jhanwar SC, Huvos AG and Healey JH. (1993). Cancer Res., 53, 16–18. Link MP and Eilber F. (1993). Cancer: Principles & Practice of Oncology DeVita VT, Hellman S. and Rosenberg SA (eds). Lippincott: Philadelphia, pp. 1534–1551.

Lothe RA, Smith-Srensen B, Hektoen M, Stenwig AE, Mandahl N, Sæter G and Mertens F. (2001). Genes Chromosome Cancer, 30, 202–206. Mertens F, Mandahl N, Orndal C, Baldetorp B, Bauer HC, Rydholm A, Wiebe T, Willen H, Akerman M and Heim S et al. (1993). Int. J. Cancer, 55, 44–50. Nakayama T, Toguchida J, Wadayama B, Kanoe H, Kotoura Y and Sasaki MS. (1995). Int. J. Cancer, 64, 342–346. Oliner JD, Kinzler KW, Meltzer PS, George DL and Vogelstein B. (1992). Nature, 358, 80–83. Pedeutour F, Forus A, Coindre JM, Berner JM, Nicolo G, Michiels JF, Terrier P, Ranchere-Vince D, Collin F, Myklebost O and Turc-Carel C. (1999). Genes Chromosomes Cancer, 24, 30–41. Potocki L, Chen KS and Lupski JR. (1999). Genomics, 57, 180–182. Sandberg AA and Bridge J. (1994). The cytogenetics of Bone and Soft Tissue Tumors. R.G. Landes Company: Austin. Simons A, Schepens M, Jeuken J, Sprenger S, van de Zande G, Bjerkehagen B, Forus A, Weibolt V, Molenaar I, van den Berg E, Myklebost O, Bridge J, van Kessel AG and Suijkerbuijk R. (2000). Cancer Genet. Cytogenet., 118, 89– 98. Stock C, Kager L, Fink FM, Gadner H and Ambros PF. (2000). Genes Chromosomes Cancer, 28, 329–336. Szymanska J, Tarkkanen M, Wiklund T, Virolainen M, Blomquist C, Asko-Seljavaara S, Tukiainen E, Elomaa I and Knuutila S. (1996). Genes Chromosom Cancer, 15, 89– 94. Tarkkanen M, Kaipainen A, Karaharju E, Bohling T, Szymanska J, Helio H, Kivioja A, Elomaa I and Knuutila S. (1993). Cancer Genet. Cytogenet., 68, 1–21. Tarkkanen M, Karhu R, Kallioniemi A, Elomaa I, Kivioja AH, Nevalainen J, Bohling T, Karaharju E, Hyytinen E, Knuutila S and Kallioniemi O-P. (1995). Cancer Res., 55, 1334–1338. Weiss SW. (1994). Histological Typing of Soft Tissue Tumours. Springer-Verlag: Berlin. Wolf M, Tarkkanen M, Hulsebos T, Larramendy F, Forus AOM, Aaltonen LA, Elomaa I and Knuutila S. (1999). Int. J. Cancer, 82, 329–333.

Note added in proof After submission of this work, we found that van Dartel, M and coworkers (Cancer Genet Cytogenet 139: 91–96, 2002) have described the complicated structure of this amplicon in a set of osteosarcomas. Interestingly they also find a peak of amplification around COP53, but also one around TOP3A, suggesting that more than one mechanism may be responsible for amplification in this region.

Oncogene