Alterations of anaphase-promoting complex genes in human ... - Nature

Oncogene (2003) 22, 1486–1490

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

Alterations of anaphase-promoting complex genes in human colon cancer cells Qing Wang1,5, Caroline Moyret-Lalle1,2,5, Florence Couzon1,5, Christine Surbiguet-Clippe1, Jean-Christophe Saurin3, Thierry Lorca4, Claudine Navarro1 and Alain Puisieux1,2 1 Centre d’Oncologie Geńe´tique, INSERM U453, Centre León Be´rard, 28 rue Lae¨nnec, 69008 Lyon, Cedex 08, France; 2Faculte´ de Pharmacie, 8 avenue Rockefeller, 69373 Lyon, Cedex 08, France; 3Fe´de´ration de Spećialitie´s Digestives, Hoˆpital Edouard Herriot, Place d’Arsonval, 69437 Lyon, Cedex 03, France; 4Centre de Recherches de Biochimie Macromolećulaire, CNRS UPR 1086, 1919 route de Mende, 34293 Montpellier, Cedex 5, France

Ubiquitin-mediated proteolysis of cell cycle regulators is a major element of the cell cycle control. The anaphasepromoting complex (APC/C) is a large multisubunit ubiquitin-protein ligase required for the ubiquitination and degradation of G1 and mitotic checkpoint regulators. APC/C-dependent proteolysis regulates cyclin levels in G1, and triggers the separation of sister chromatids at the metaphase–anaphase transition and the destruction of mitotic cyclins at the end of mitosis. Furthermore, it was recently shown that APC/C regulates the degradation of crucial regulators of signal transduction pathways. We report here gene alterations in several components of this complex in human colon cancer cells, including APC6/CDC16 and APC8/CDC23 which are known to be key function elements. The experimental expression of a truncation mutant of APC8/CDC23 subunit (CDC23DTPR) leads to abnormal levels of APC/C targets such as cyclin B1 and disturbs the cell cycle progression of colon epithelial cells through mitosis. Overall, these data support the hypothesis of a deleterious role of these mutations during colorectal carcinogenesis. Oncogene (2003) 22, 1486–1490. doi:10.1038/sj.onc.1206224 Keywords: APC/C; cyclosome; ubiquitination; colon cancer; cyclin B1

the chromosomes along the metaphase plate (spindle checkpoint). Several lines of evidence suggest that the alteration of regulators specific of these checkpoints leads to cell transformation (Cahill et al., 1998). In yeast, a key regulator of both the G1 checkpoint and the spindle-assembly checkpoint is the anaphase-promoting complex or cyclosome (APC/C). APC/C is an ubiquitincyclin ligase required for degrading mitotic cyclins and other cell cycle regulators (Peters, 2002). Mitotic cyclin degradation is an essential mechanism for mitotic exit and the initiation of a new cell cycle (Wasch and Cross, 2002). In the yeast cell cycle, the induction of APC/C proteolytic activity is involved in at least three phenomena: at the metaphase–anaphase transition, it is needed for separating sister chromatids; at the end of anaphase, it is required for promoting exit from mitosis; during G1, it is necessary for preventing inappropriate entry into S phase by regulating cyclin levels (Irniger and Nasmyth, 1997). Subsequent studies demonstrated that this system is highly conserved in mammalian cells. In mammals, APC/C is a large multimeric complex composed of at least 11 subunits (Gieffers et al., 2001). Considering the prominent role of APC/C in cell-cycle regulation, it was tempting to speculate that its dysregulation triggered a major perturbation of cell-cycle progression and contributed to cell transformation.

Introduction In eukaryotes, a series of control systems, termed checkpoints, ensure the maintenance of genomic integrity (Elledge, 1996). Checkpoints monitor the status of DNA replication to prevent damaged or incompletely duplicated genomes from undergoing mitosis. They also modulate all events required for chromosome segregation, assembly of the spindle and correct positioning of *Correspondence: Professor A Puisieux, Centre d’Oncologie Geńe´tique/Unite´ INSERM U453, Centre León Be´rard, 28 rue Lae¨nnec, 69008 Lyon, France; E-mail: [email protected] 5 QW, CM-L and FC contributed equally to this work. Received 11 June 2002; revised 1 November 2002; accepted 6 November 2002

PCR-HA analysis of APC/C subunits To test the potential implication of APC/C in human carcinogenesis, we determined the gene status of nine of these subunits by using RT–PCR-based heteroduplex screening (PCR-HA) in cancer cells from different human tissue origins. Three APC/C subunits, APC6/CDC16, APC8/ CDC23 and APC3/CDC27, are known to be key functional components (Zachariae and Nasmyth, 1996). Mutation screening of these genes was exhaustively performed in a total of 52 cancer cell lines: 10 from colon cancer, 10 from breast cancer, 15 from neuroblastoma, five from hepatocarcinoma, six from

Mutations in APC/C subunits in colon cancer Q Wang et al

1487

Figure 1 Sequence analysis of alterations in APC/C component genes. Following cell lines were screened for APC/C mutations: 10 breast cell lines (HS578T, T47D, MCF7, BT20, HBL100, MDA-MB-436, Cal 51, ZR75.1, MDA-MB-361, MDA-MB-231), 10 colon cancer cell lines (HCT116, SW48, LoVo, LS174T, Co 115, EB, Colo320, SW480, HT29, HCT15), 15 neuroblastoma cell lines (CLB-ES, CLB-GA, IGR-N91, CLB-CA, SKN-AS, GHNO-NO, CLB-Ba, CLB-Bac, CLB-Bar, CLB-Be, CLB-Ber, IMR32, LS, CLB-PE, SHSY65Y), five hepatocarcinoma cell lines (Huh7, MV, Focus, HepG2, Hep3B), six melanoma cell lines (Gı¨ l, Mel32, IC8, Mel4, A-375, NTP), two glioma cell lines (G152, G142), and four others (one endometrium AN3CA, one ovarian SK-OV-3, one prostate Du 145 and one choriocarcinoma BeWo). Cell lines were obtained either from the American Type Culture Collection (www.ATCC.org) or established in the laboratory (Combaret et al., 1995). Cell line IGR-N91 was kindly supplied by J Beńard (Institute Gustave Roussy, Villjuif, France) and SKN-AS, Cal51 by C Theillet (Centre Val d’Aurelle, Montpellier, France). RT–PCR-based heteroduplex screening was described elsewhere (Wang et al., 1999). Primers for PCR are available upon request. Sequencing was performed with the Bigdye PRISM pre-ready kit (Applied Biosystems) and an automated sequencing apparatus (ABI 377). Nucleotide sequences containing mutations are shown and compared to a wild-type homologue (control). The position of the mutated nucleotide is indicated by an arrow

melanoma, two from glioma, and four from endometrium, ovarian carcinoma, prostate carcinoma and choriocarcinoma, respectively (for the complete list of

cell lines, see legend of Figure 1). For other components (APC2, APC4, APC5, APC7, APC10, APC11), screening was carried out in at least 40 of the above tumor cell Oncogene


1488 Table 1

Alterations of APC/C genes in colorectal cancer cells

Cell line

Gene

Base change

Amino-acid change

Co115 LS174T

APC4 APC6/CDC16

811 T>C 1662-1667delA

HCT15

APC6/CDC16

HT29 Primary colon cancer

APC8/CDC23 APC8/CDC23

899 C >A 1361 G>A 715 G>T 127 C>A

C271R Frameshift (stop codon at 569) S300Y C454Y E239X L43I

lines, including all cell lines derived from colon cancer, breast cancer and neuroblastoma. Alterations affecting three genes: APC4, APC6/CDC16 and APC8/CDC23, were identified in four colon cancer cell lines (Table 1 and Figure 1). Two of these alterations, namely the deletion of an adenine in APC6/CDC16 in LS174 T cells and the nonsense mutation (E239X) of APC8/CDC23 in HT29 cells, were clear deleterious mutations resulting in premature stop codons. The APC8/CDC23 mutation leads to the loss of all six tetratrico peptide repeat (TPR) motifs downstream of the premature stop codon. These TPR motifs are essential functional domains of the protein predicted to mediate protein–protein interactions (Zachariae and Nasmyth, 1996). This observation raised the question of the frequency of APC8/CDC23 mutations in primary tumors. The mutation screening of APC8/CDC23 carried out in 22 available colorectal cancer tissues revealed a missense mutation at codon 43 (127C>A, L43I) in one colon cancer. This missense mutation was absent from normal tissues taken from this patient, indicating that it was not a polymorphism. A missense mutation in APC4 was found in Co 115 cell line (811T>C), leading to a replacement of a cysteine by an arginine at codon 271. In HCT15 cells, two missense mutations of the APC6/CDC16 gene were identified (899C>A and 1361G>A); both were located on the same allele, as determined by cloning. Although these variants were not found in more than 100 control chromosomes, further studies will be required to determine whether they functionally alter the gene products. All alterations were heterozygous and the second allele was wildtype (Figure 1), suggesting that mutants displayed a dominant-negative effect. Finally, an abnormal transcript of APC3/CDC27 was detected in the SW480 colon cancer cells. Sequencing of this transcript showed numerous nucleotide mismatches scattered along the RT–PCR products. Preliminary experiments suggested that this transcript was derived from a processed pseudogene (data not shown). Sustained levels of APC/C substrate proteins and delayed mitotic exit after nocodazole release in CDC23DTPR mutant cells Some have suggested that the degradation of B-cyclin type protein by inactivating cdk is the essential function of APC/C for mitotic exit and origin resetting during Oncogene

mitosis (Noton and Diffley, 2000). In order to explore the consequences of the presence of a mutated APC/C subunit on APC/C target protein levels and cell cycle distribution, HCT116 and HT29 cells were treated with nocodazole in order to activate the mitotic checkpoint (Shah and Cleveland, 2000). HCT116 cells exhibit wildtype APC8/CDC23 protein (CDC23 +/+), whereas HT29 cells express both wild-type and truncated mutant APC8/CDC23 proteins (CDC23 DTPR/+). Nocodazole is a chemical agent that inhibits spindle assembly by disrupting microtubules, and synchronizes cells in G2–M phase. After 24 h of nocodazole treatment, cells arrested at prometaphase were collected by shake-off. Mitotic cells were then released from arrest by adding fresh medium, then fractions were taken at hourly intervals. Asynchronous cells were included as controls. As shown in Figure 2a, HCT116 nocodazole-treated cells re-entered the cell cycle when released into fresh medium, whereas HT29 cells remained permanently arrested in G2–M. As shown by Western blot analysis (Figure 2b), levels of cyclin B1 in HCT116 cells increased significantly as cells entered mitosis (Noc, lane 2), then dropped as cells entered the subsequent cell cycle (2–4 h, lanes 4–6) and rose again following S phase (lane 7). In contrast, levels of cyclin B1 in HT29 cells slowly decreased after release (3–10 h, lanes 14–18). The variation of protein levels of another APC/C substrate, AIM-1 (Aurora and Ipll-like midbody-associated protein-1, also called auroral/B) (Kawasaki et al., 2001), highly mirrored that of cyclin B1 in both HCT116 and HT29 cell lines (Figure 2b). In order to directly implicate the truncated form of APC8/CDC23 subunit (CDC23DTPR) in mitotic exit delay, we cloned APC8/CDC23 truncation mutant cDNA from HT29 cells and transiently expressed this mutant in HCT116 cells. An experimental, transient overexpression of the CDC23DTPR mutant significantly increased the number of cells in G2–M phase (43.5% 6 h after release) when cells were synchronized by nocodazole then released into fresh medium, compared with mock-transfected cells (27% 6 h after release). This delay for entering a new cell cycle was correlated with abnormal cyclin B1 and AIM-1 protein levels (data not shown). Although this should be confirmed by stable overexpression experiments, our preliminary data suggest that the presence of the APC8/CDC23 mutant form may delay mitotic exit by altering the degradation schedule of mitotic key regulators. Overall, this is the first demonstration of APC/C mutations in human cancer cells. Further analyses, using additional APC/C components and different types of primary tumors, should be performed to determine the exact prevalence of these alterations. However, several observations may be pointed out. First, four of five abnormalities affected APC6/CDC16, APC8/CDC23 or APC3/CDC27, three key APC/C components that are essential for the ubiquitination of B-type cyclins in Saccharomyces cerevisiae, either in vitro or in vivo (Zachariae and Nasmyth, 1996). Their experimental inactivation provokes the premature appearance of mitotic cyclins and an unscheduled initiation of DNA


1489

Figure 2 Cell cycle and Western blot analysis of cyclin B1 and AIM-1 protein levels in HCT116 (CDC23+/+) and HT29 (CDC23 DTPR/+) cells after nocodazole release. HCT116 (CDC23+/+) and HT29 (CDC23 DTPR/+) were synchronized at prometaphase by nocodazole block at a final concentration of 0.4 mg/ml during 24 h. Mitotically arrested cells were collected by shake-off. Cells released into fresh medium were collected at indicated time points (hours). Asynchronous cells were used as controls. (a) DNA content profile of released cells was analysed by fluorescence-activated cell sorting. Briefly, cells were pulsed with 20 mm of BrdU for 30 min, washed and trypsinized, fixed with 70% ethanol, labeled with anti-BrdU antibody (monoclonal FITC, 1 : 500, Becton Dickinson) followed by treatment with 1 mg of RNase/ml and propidium iodide staining, then subjected to flow cytometry using a BectonDickinson FACScalibur flow cytometer. The percentages of G1, S and G2/M populations were calculated with the CellQuest analysis program. (b) Cells collected at indicated time points were lysed and proteins (50 mg) were analysed by Western blotting with mouse anti-cyclin B1 (1 : 1000, Upstate biotechnology) (upper panel) and anti-AIM-1 (1 : 500, Transduction Laboratories) (lower panel) monoclonal antibodies. AS, asynchronous (lane 1), Noc, nocodazole (lane 2)

replication (Zhao et al., 1998). We show here that the presence of a APC8/CDC23 truncation mutant in human colon epithelial cells is associated to abnormal levels of cyclin B1 and to a deregulation of the cell cycle progression through mitosis. It is noteworthy that similar observations were reported when human lung cancer cells A549 arrested in mitosis by nocodazole, were released back into fresh medium containing a proteasome inhibitor (Zhang and Lees, 2001), confirming the role of the proteasome activity in the control of the cell cycle. The second striking observation is that all identified mutations were heterozygous. This may be explained by the crucial cellular activity of APC/C, operating at multiple cell-cycle transition levels. During the cell cycle, APC/C activity is under the control of

combined regulatory events such as phosphorylation, functional sequestration by kinetochore-associated checkpoint proteins, and recruitment to the APC/C core of activating subunits such as CDC20/Slp l/Fizzy and Cdh1, two related but functionally distinct APC/Cactivating subunits (Visintin et al., 1997). We can speculate that a complete inhibition of APC/C activity would lead to cell death. By contrast, a quantitative or qualitative disruption of some of its subunits could be sufficient to cause a significant dysregulation of key cellcycle regulators, thus contributing to malignant transformation. It is noteworthy that several of the known substrates of APC/C have been shown to be overexpressed in human cancers, including colorectal cancers. These substrates include human securin, also called Oncogene


1490

pituitary tumor-transforming gene (PTTG), which is involved in sister chromatid separation, STK15/aurora2, required for separation and chromosome segregation, and cyclin B1 (Wang et al., 1997; Bischoff et al., 1998; Heaney et al., 2000; Soria et al., 2000; SarafanVasseur et al., 2002). The experimental overexpression of these genes is able to trigger the transformation of rodent fibroblasts (Bischoff et al., 1998; Yin et al., 2001), suggesting that a disruption of the APC/C complex may well deregulate cell division at multiple levels. Among these effects, we can speculate that the mutation or the deregulation of APC/C subunits might be one of the mechanisms that leads to the accumulation of cyclin B1 protein in colorectal tumors. Recently, the gene encoding hCdc4, an ubiquitin ligase that directs cyclin E to the degradation machinery, was found to be mutated in a breast cancer cell line

(Strohmaier et al., 2001). We report here the second example of somatic mutations of the ubiquitin–proteasome system occurring in cancer cells. Proteolytic degradations of intracellular proteins by the ubiquitin– proteasome pathway are essential to processes like signal transduction and cell-cycle progression. It has now been made clear that malignancy often occurs when cells stop degrading or, contrarily, when they overdegrade a protein, thus triggering cancer-promoting defects in the cell cycle, DNA repair, apoptosis or signaling.

Acknowledegements This work was supported by grants from the Comite´ De´partemental du Rhoˆne and the Comite´ De´partemental de la Droˆme de la Ligue de Lutte contre le Cancer, and from the Association pour la Recherche sur le Cancer.

References Bischoff JR, Anderson L, Zhu Y, Mossie K, Ng L, Souza B, Schryver B, Flanagan P, Clairvoyant F, Ginther C, Chan CS, Novotny M, Slamon DJ and Plowman GD. (1998). EMBO J., 17, 3052–3065. Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, Markowitz SD, Kinzler KW and Vogelstein B. (1998). Nature, 392, 300–303. Combaret V, Turc-Carel C, Thiesse P, Rebillard AC, Frappaz D, Haus O, Philip T and Favrot MC. (1995). Int. J. Cancer, 61, 185–191. Elledge SJ. (1996). Science, 274, 1664–1672. Gieffers C, Dube P, Harris JR, Stark H and Peters JM. (2001). Mol. Cell, 7, 907–913. Heaney AP, Singson R, McCabe CJ, Nelson V, Nakashima M and Melmed S. (2000). Lancet, 355, 716–719. Irniger S and Nasmyth KJ. (1997). J. Cell. Sci., 110, 1523–1531. Kawasaki A, Matsumura I, Miyagawa J-I, Ezoe S, Tanaka H, Terada Y, Tatsuka M, Machii T, Miyazaki H, Furukawa Y and Kanakura Y. (2001). J. Cell Biol., 152, 275–287. Noton E and Diffley JF. (2000). Mol. Cell., 5, 85–95. Peters JM (2002). Mol Cell., 9, 931–943.

Oncogene

Sarafan-Vasseur N, Lamy A, Bourguignon J, Le Pessot F, Hieter P, Sesbouë´ R, Bastard C, Fre´bourg T and Flaman JM. (2002). Oncogene, 21, 2051–2057. Shah JV and Cleveland DW. (2000). Cell, 103, 997–1000. Soria JC, Jang SJ, Khuri FR, Hassan K, Liu D, Hong WK and Mao L. (2000). Cancer Res., 60, 4000–4004. Strohmaier H, Spruck CH, Kaiser P, Won KA, Sangfelt O and Reed SI. (2001). Nature, 413, 316–322. Visintin R, Prinz S and Amon A. (1997). Science, 278, 460–463. Wang A, Yoshimi N, Ino N, Tanaka T and Mori H. (1997). J. Cancer Res. Clin. Oncol., 123, 124–127. Wang Q, Lasset C, Desseigne F, Saurin JC, Maugard C, Navarro C, Ruano E, Descos L, Trillet-Lenoir V, Bosset JF and Puisieux A. (1999). Hum. Genet., 105, 79–85. Wasch R and Cross FR. (2002). Nature, 418, 556–562. Yin XY, Grove L, Datta NS, Katula K, Long MW and Prochownik EV. (2001). Cancer Res., 61, 6487–6493. Zachariae W and Nasmyth K. (1996). Mol. Biol. Cell, 7, 791–801. Zhao N, Lai F, Fernald AA, Eisenbart JD, Espinosa R, Wang PW and Le Beau MM. (1998). Genomics, 53, 184–190. Zhang Y and Lees E. (2001). Mol. Cell. Biol., 21, 5190–5199.