RNA Polymerase II Transcription Is Required for Human Papillomavirus Type 16 E7- and Hydroxyurea-Induced Centriole Overduplication
Total Page:16
File Type:pdf, Size:1020Kb
Oncogene (2007) 26, 215–223 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc ORIGINAL ARTICLE RNA polymerase II transcription is required for human papillomavirus type 16 E7- and hydroxyurea-induced centriole overduplication A Duensing1,2, Y Liu2, N Spardy2,3, K Bartoli3, M Tseng2, J-A Kwon2, XTeng 2 and S Duensing2,4 1Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA; 2Molecular Virology Program, University of Pittsburgh Cancer Institute, Pittsburgh, PA, USA; 3Biochemistry and Molecular Genetics Graduate Program, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA and 4Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA Aberrant centrosome numbers are detected in virtually all interphase and mitosis in most animal and human human cancers where they can contribute to chromosomal cells (Bornens, 2002). A normal centrosome consists instability by promoting mitotic spindle abnormalities. of a pair of centrioles, short microtubule cylinders, Despite their widespread occurrence, the molecular embedded in pericentriolar material (Urbani and mechanisms that underlie centrosome amplification are Stearns, 1999). The single centrosome duplicates only beginning to emerge. Here, we present evidence for a precisely once before mitosis in order to form the novel regulatory circuit involved in centrosome over- two spindle poles (Hinchcliffe and Sluder, 2001). duplication that centers on RNA polymerase II (pol II). Centrosome duplication is synchronized with the We found that human papillomavirus type 16 E7(HPV-16 cell division cycle and starts in late G1 phase with E7)- and hydroxyurea (HU)-induced centriole overdupli- the splitting of the two centrioles. During S phase, a cation are abrogated by a-amanitin, a potent and specific newly synthesized centriole forms adjacent to each of the RNA pol II inhibitor. In contrast, normal centriole older centrioles in order to generate two new centro- duplication proceeded undisturbed in a-amanitin-treated somes. The molecular mechanisms involved in the cells. Centriole overduplication was significantly reduced regulation of this process as well as the structural basis by siRNA-mediated knock down of CREB-binding of centriole reproduction, however, remain poorly protein (CBP), a transcriptional co-activator. We identi- characterized. fied cyclin A2 as a key transcriptional target of RNA pol Numerical and structural centrosome anomalies are II during HU-induced centriole overduplication. Collec- widespread in human cancers, where they can promote tively, our results show that ongoing RNA pol II chromosomal instability through an increased risk for transcription is required for centriole overduplication cell division errors (Salisbury et al., 1999; Doxsey, 2002; whereas it may be dispensable for normal centriole Nigg, 2002; Duensing, 2005; Fukasawa, 2005). In duplication. Given that many chemotherapeutic agents principal, aberrant centrosome numbers can develop function through inhibition of transcription, our results through two distinct mechanisms: centrosome accumu- may help to develop strategies to target centrosome- lation and centrosome overduplication (Duensing and mediated chromosomal instability for cancer therapy and Munger, 2001; Nigg, 2002). It is conceivable that an prevention. accumulation of centrosomes due to impaired cytokin- Oncogene (2007) 26, 215–223. doi:10.1038/sj.onc.1209782; esis accounts for a large number of centrosome published online 3 July 2006 aberrations in malignant tumors (Bunz et al., 1998). However, it has recently been shown that a bona fide Keywords: HPV-16 E7; hydroxyurea; centrioles; RNA centriole overduplication can be induced in cells arrested polymerase II; cyclin A in early S phase using the ribonucleotide reductase inhibitor hydroxyurea (HU) or following overexpression of the human papillomavirus type 16 (HPV-16) E7 oncoprotein (Duensing et al., 2000; Guarguaglini et al., Introduction 2005). HPV-16 E7 is a major transforming oncoprotein of high-risk HPVs and disrupts the G1/S cell cycle Centrosomes are small cytoplasmic organelles that checkpoint on multiple levels including binding and function as major microtubule organizing centers during degradation of the retinoblastoma tumor suppressor protein (pRB) and inactivation of the cyclin-dependent- Cip1 Kip1 Correspondence: Dr S Duensing, Molecular Virology Program, kinase (CDK) inhibitors p21 and p27 (Munger University of Pittsburgh Cancer Institute, Hillman Cancer Center, and Howley, 2002). Cells treated with HU or over- Research Pavilion Suite 1.8, 5117 Centre Avenue, Pittsburgh, PA expressing HPV-16 E7 show a phenotype in which 15213, USA. multiple immature daughter centrioles are frequently E-mail: [email protected] Received 4 November 2005; revised 2 May 2006; accepted 26 May 2006; found adjacent to a single mature centriole (Guargua- published online 3 July 2006 glini et al., 2005). RNA polymerase II transcription A Duensing et al 216 Both, ectopic expression of HPV-16 E7 or treatment as evidenced by BrdU incorporation (Figure 1c). Our with HU cause an abnormally prolonged S phase of the results indicate that HPV-16 E7 requires RNA pol II cell division cycle (Balczon et al., 1995; Martin et al., activity to induce abnormal centriole duplication 1998). S phase is governed by E2F-mediated gene and that this process can be effectively inhibited by transcription and previous studies have shown that a-amanitin. E2F activity is involved in the induction of aberrant To further dissect the role of RNA pol II in centriole centrosome duplication by both HPV-16 E7 or HU duplication, U2OS/centrin-GFP cells were treated with (Meraldi et al., 1999; Duensing et al., 2000). However, 1mM HU either alone or in combination with 0.5 mg/ml the role of the transcriptional machinery itself in or 1 mg/ml a-amanitin for 72 h (Figure 1d). No major centrosome duplication and differences in the require- changes of the normal centriole duplication status were ment for ongoing transcription between normal and detected between untreated cells and asynchronously aberrant centriole duplication have not been determined growing cells treated with a-amanitin alone (Figure 1e). in detail in human cells. Treatment with HU alone led to a significant increase We report here that both HPV-16 E7- and HU- of cells with overduplicated centrioles (more than four induced centriole overduplication require ongoing RNA per cell) from 0 to 43.1% (Pp0.0001). Simultaneous pol II-mediated gene transcription whereas normal treatment with HU and a-amanitin led to a significant centriole duplication was not sensitive to an inhibition reduction of cells with centriole overduplication from of RNA pol II activity. Ongoing transcription/transla- 43.1% in cells treated with HU alone to 20% in cells tion of cyclin A2 was indentified as a critical step in treated with HU and 0.5 mg/ml a-amanitin (2.2-fold; centriole overduplication induced by a prolonged Pp0.0005) and 3.9% in cells treated with HU and S phase, which is in line with previous studies showing 1 mg/ml a-amanitin (11.1-fold; Pp0.0005), respectively a requirement for CDK2 activity for centriole over- (Figure 1e). As HU-treated cells are cell cycle arrested in duplication (Duensing et al., 2006). Various chemo- S phase and the a-amanitin concentrations used did not therapeutic agents function through an inhibition of cause a cell cycle arrest (Figure 1c), this decrease of transcriptional activity and their effects on centriole numerical centriole abnormalities is unrelated to non- duplication errors as therapeutic principle or basis for specific cell cycle effects of a-amanitin. The finding that their use as cancer prevention agents deserves further not all cells treated with HU showed overduplication of investigation. centrioles is in line with previous reports (Balczon et al., 1995) and most likely based on the fact that a relaxation of an intrinsic block to reduplication (Wong and Stearns, 2003) occurs only in a proportion of cells Results (Matsumoto et al., 1999; Meraldi et al., 1999; D’Assoro et al., 2004). a-Amanitin inhibits HPV-16 E7- and HU-induced A significant increase of cells with normally dupli- centriole overduplication cated centrioles (3 or 4 per cell) was detected in cells We analysed whether a-amanitin, a potent and specific treated with HU alone (40.1%) when compared to inhibitor of RNA pol II derived from the green death untreated controls (19.8%; 2.0-fold; Pp0.001) indicat- cap mushroom Amanita phalloides (Gong et al., 2004), ing that normal centriole duplication without over- interferes with the effects of HPV-16 E7 and HU on duplication continues in a fraction of HU-arrested cells centriole duplication. The mode of action of a-amanitin (Figure 1e). Simultaneous treatment of cells with HU is not fully understood but there is evidence that it and 0.5 mg/ml a-amanitin or 1 mg/ml a-amanitin did not interacts with the largest RNA pol II subunit, Rpb1, lead to a decrease of cells displaying normally duplicated and thereby inhibits transcription initiation and elonga- centrioles (Figure 1e). These findings, together with the tion (Bushnell et al., 2002). a-Amanitin blocks nucleo- fact that a-amanitin alone does not reduce the number tide incorporation into nascent mRNA transcripts in the of cells with normally duplicated centrioles in asynchro- nucleus whereas nucleolar RNA transcription carried nously growing populations, suggest that RNA