CPSF4 Activates Telomerase Reverse Transcriptase and Predicts Poor Prognosis in Human Lung Adenocarcinomas

MOLECULAR ONCOLOGY XXX (2014) 1e13

available at www.sciencedirect.com ScienceDirect

www.elsevier.com/locate/molonc

CPSF4 activates telomerase reverse transcriptase and predicts poor prognosis in human lung adenocarcinomas

Wangbing Chena,1, Lijun Qinb,1, Shusen Wanga,*, Mei Lia, Dingbo Shia, Yun Tiana, Jingshu Wanga, Lingyi Fua, Zhenglin Lic, Wei Guoc, Wendan Yuc, Yuhui Yuanc, Tiebang Kanga, Wenlin Huanga,d,*, Wuguo Denga,d,* aSun Yat-Sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center of Cancer Medicine, Guangzhou, China bDepartment of Pediatrics, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, Guangzhou, China cInstitute of Cancer Stem Cell, Dalian Medical University Cancer Center, Dalian, China dState Key Laboratory of Targeted Drug for Tumors of Guangdong Province, Guangzhou Double Bioproduct Inc., Guangzhou, China

ARTICLE INFO ABSTRACT

Article history: The elevated expression and activation of human telomerase reverse transcriptase (hTERT) Received 9 October 2013 is associated with the unlimited proliferation of cancer cells. However, the excise mecha- Received in revised form nism of hTERT regulation during carcinogenesis is not well understood. In this study, we 6 January 2014 discovered cleavage and polyadenylation specific factor 4 (CPSF4) as a novel tumor- Accepted 5 February 2014 specific hTERT promoter-regulating protein in lung cancer cells and identified the roles Available online - of CPSF4 in regulating lung hTERT and lung cancer growth. The ectopic overexpression of CPSF4 upregulated the hTERT promoter-driven report gene expression and activated Keywords: the endogenous hTERT mRNA and protein expression and the telomerase activity in CPSF4 lung cancer cells and normal lung cells. In contrast, the knockdown of CPSF4 by siRNA Telomerase had the opposite effects. CPSF4 knockdown also significantly inhibited tumor cell growth hTERT in lung cancer cells in vitro and in a xenograft mouse model in vivo, and this inhibitory ef- Promoter fect was partially mediated by decreasing the expression of hTERT. High expression of both Lung cancer CPSF4 and hTERT proteins were detected in lung adenocarcinoma cells by comparison with the normal lung cells. Tissue microarray immunohistochemical analysis of lung adenocarcinomas also revealed a strong positive correlation between the expression of CPSF4 and hTERT proteins. Moreover, KaplaneMeier analysis showed that patients with high levels of CPSF4 and hTERT expression had a significantly shorter overall survival than those with low CPSF4 and hTERT expression levels. Collectively, these results demonstrate that CPSF4 plays a critical role in the regulation of hTERT expression and lung tumorigenesis and may be a new prognosis factor in lung adenocarcinomas. ª 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

* Corresponding authors. Sun Yat-Sen University Cancer Center, 651 Dongfeng East Road, Guangzhou 510060, China. Tel.: þ86 20 87342282; fax: þ86 20 87343170. E-mail addresses: [email protected] (S. Wang), [email protected] (W. Huang), [email protected] (W. Deng). 1 These authors contributed equally to this manuscript. 1574-7891/$ e see front matter ª 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.molonc.2014.02.001

Please cite this article in press as: Chen, W., et al., CPSF4 activates telomerase reverse transcriptase and predicts poor prognosis in human lung adenocarcinomas, Molecular Oncology (2014), http://dx.doi.org/10.1016/j.molonc.2014.02.001 2 MOLECULAR ONCOLOGY XXX (2014) 1e13

1. Introduction significantly over-represented in the hTERT promoter probe- protein complexes in nuclear proteins prepared from telome- Telomerase activation and the maintenance of telomeres are rase positive lung cancer cells compared to telomerase nega- critical steps in the unlimited proliferation of cancer cells. tive normal cells. Telomeres are composed of tandem repeat arrays of TTAGGG CPSF4 is a member of the cleavage and polyadenylation nucleotide DNA sequences and serve as protective “caps” at specificity factor (CPSF) complex, whose other components the ends of human chromosomes, protecting them from are CPSF160, CPSF100, CPSF73 and Fip1 (Kiefer et al., 2009). DNA degradation and unwanted repair (Blackburn, 2005; In addition to the evidence suggesting that CPSF4 functions Blackburn et al., 2006; Harley, 2008; Tian et al., 2010). In normal as a 30 mRNA processing factor that participates in the matu- somatic cells, telomeres are progressively lost at each cell di- ration of mRNA 30 ends (Barabino et al., 1997; de Vries et al., vision due to the inability of the linear DNA replication ma- 2000; Kaufmann et al., 2004; Nemeroff et al., 1998), the role chinery to replicate the telomere ends; this phenomenon of CPSF4 as a transcriptional coactivator has also been leads to the loss of approximately 50e200 bp of telomeric described (Rozenblatt-Rosen et al., 2009). We considered the DNA in each cell division cycle (Chiu and Harley, 1997; Liu, hypothesis that the differential expression of CPSF4 in cancer 1999). The loss of telomeres beyond a critical length can elicit cells and normal cells may be associated with the tumor- DNA-damage responses, activate cell cycle check points and specific activation of hTERT transcription. lead to cell senescence and apoptosis (Saretzki et al., 1999). In this study, we showed that the overexpression of CPSF4 Thus, the progressive loss of the terminal nucleotides of chro- activates the hTERT promoter, which in turn increases hTERT mosomes with each cell division limits the replicative capac- expression and activates telomerase. These results support ity of eukaryotic cells. However, cancer cells can activate the hypothesis that CPSF4 may be an important regulator of telomere maintenance mechanisms (TMMs) to overcome telomerase activity and cell growth in lung adenocarcinomas. this limitation and are thus able to proliferate indefinitely As hTERT expression is closely related to tumorigenesis and (Colgin and Reddel, 1999). Telomerase is a reverse transcrip- strictly controlled at the transcription level, our findings indi- tase that adds telomeric repeats to chromosomal ends; the cate the role of CPSF4 as a tumor-specific hTERT promoter addition of telomere repeats is the most thoroughly studied regulator to promote hTERT gene expression in human lung TMM (Bryan and Cech, 1999). Telomerase-independent mech- cancers and a potential novel therapeutic target for the treat- anisms, which are referred to as alternative lengthening of ment of lung cancers. telomeres, may also maintain the length of telomeres in cancer cells (Bryan et al., 1997). Telomerase is a ribonucleoprotein enzyme consisting of 2. Materials and methods human telomerase RNA (hTR), telomerase-associated protein 1 (TEP1) and human telomerase reverse transcriptase (hTERT), 2.1. Cell lines and cell culture the catalytic unit (Cohen et al., 2007; Feng et al., 1995; Nakamura et al., 1997; Weinrich et al., 1997). Telomerase activ- Telomerase positive three human adenocarcinoma cell lines ity is often directly correlated with the uncontrolled growth of (H1299, A549 and H322) were obtained from the American cells, which is an established hallmark of cancer (Harley, 2008; Type Culture Collection (ATCC, Manassas, VA) and cultured Shay and Keith, 2008). Telomerase (specifically its catalytic in RPMI-1640 medium (Invitrogen, Carlsbad, CA) supple- subunit hTERT) is overactive in 85e90% of cancers but is pre- mented with 10% fetal bovine serum. Human lung fibroblasts sent at very low or almost undetectable levels in normal cells WI-38 and normal human bronchial epithelial (HBE) cells, (Bisoffi et al., 2006; Ruden and Puri, 2013). hTERT is a major which express very low levels of hTERT because of promoter oncoprotein involved in aberrant cell proliferation, immortal- repression (Milyavsky et al., 2003) were cultured in Dulbecco’s ization, metastasis and the maintenance of stemness in the Modified Eagle Medium (Invitrogen, Carlsbad, CA) supple- majority of tumors (Lu et al., 2012). The overexpression of mented with 10% fetal bovine serum. All cells were main- hTERT in cancer cells can be achieved by gene amplification tained in a humidified atmosphere and 5% CO2 at 37 C. (Zhang et al., 2000), but most studies have focused on the transcriptional regulation of hTERT expression. The transcrip- 2.2. Streptavidin-agarose pulldown assay tional activity of the hTERT promoter is selectively up- regulated in tumors but silent in most normal cells (Ducrest The binding of transactivators to hTERT promoter DNA was et al., 2002; Takakura et al., 1999). This observation is largely assayed by streptavidin-agarose pulldown as described previ- attributed to several key transcription factors in tumor cells ously (Deng et al., 2006). Briefly, cell lines were grown to that can up-regulate hTERT transcription (Kyo et al., 2008). 80e90% confluence in 150-cm2 flasks, and nuclear extracts However, the factors that have been identified so far do not were prepared. A biotin-labeled double-stranded DNA probe completely account for the cancer-specific expression of corresponding to nucleotides 378 to þ60 of the hTERT pro- hTERT. To identify these potentially critical unknown factors, moter sequence was synthesized by Sigma (SigmaeAldrich, we have successfully established a screening system that St. Louis, MO). The binding assay was performed by mixing combines a streptavidin-agarose pulldown assay and high- 1 mg of nuclear protein extract, 10 mg of DNA probe, and throughput proteomics (Deng et al., 2007). One of the proteins 100 ml of streptavidin-agarose beads (SigmaeAldrich). The that we identified using this systematic approach is cleavage mixture was incubated at room temperature for 2 h with and polyadenylation specific factor 4 (CPSF4), which was agitation and then centrifuged at 500 g to pulldown the

DNA-protein complex. The bound proteins were eluted with 2.7. Promoter reporters and dual-luciferase assay cold phosphate-buffered saline (PBS) for further analysis. Cells (2 105 cells/well) were seeded into six-well plates, 2.3. Identification of hTERT promoter-binding proteins cultured overnight, and transfected with the hTERT promoter- luciferase plasmids or the GFP reporter vector (driven by a The potential transactivators of hTERT promoter DNA were CMV or an hTERT promoter) (1 mg per well of plasmid) with Lip- analyzed using a mass spectrometry. Briefly, the bound pro- ofectamine 2000 (Invitrogen, Carlsbad, CA). Meanwhile, cells teins were separated by 10% SDS-PAGE and visualized by coo- were co-transfected with either a CPSF4 overexpression vector massie blue staining. The protein bands of interest were cut (pcDNA3.1-CPSF4 or the control vector pcDNA3.1) or a CPSF4- out and digested with sequencing-grade trypsin solution. specific siRNA (CPSF4-specific siRNA or a nonspecific siRNA The identification of digested samples was performed using control). The amount of co-transfected CPSF4 overexpression a mass spectrometry. The identities of the proteins of interest vector or CPSF4-specific siRNA was as indicated in the figures. were verified via available databases and software. For dual-luciferase assay, the transfection efficiency was normalized by cotransfection with Renilla luciferase reporter. 2.4. Chromatin immunoprecipitation assay (ChIP) For GFP reporter assay, the transfection efficiency was normalized by cotransfection with red fluorescent protein (RFP) The ChIP assay was performed using the ChIP IT Express kit expressing plasmid vector. The transfection efficiency was (Active Motif, Rixensart, Belgium) according to the manufac- approximately 50e65% in these cell lines. Forty-eight hours af- turer’s instructions. Briefly, the cells were fixed with 1% form- ter transfection, the cells were assayed for luciferase activity aldehyde and sonicated on ice to shear the DNA into 200 to using a Dual-Luciferase Reporter assay system (Promega 500 bp fragments. One third of the total cell lysate was used Corp., Madison, WI). The expression of GFP was examined un- as the DNA input control. The remaining two thirds of the der a fluorescence microscopy. The cells were assayed for the lysate was subjected to immunoprecipitation with anti- percentage of the GFP-positive cell population and the fluores- CPSF4 antibodies or non-immune rabbit IgG (Proteintech cence intensity of the GFP protein using flow cytometry. Group, Inc., Chicago, USA). A 438-bp region (378 to þ60 bp) of the hTERT promoter was amplified by PCR using the primers 2.8. Telomerase activity assays (50- TGGCCCCTCCCTCGGGTTAC-30 and 50- CCAGGGCTTCC- CACGTGCGC-30). The PCR products were resolved electropho- Telomerase activity was analyzed by a telomerase PCR retically on a 2% agarose gel and visualized by ethidium enzyme-linked immunosorbent assay kit (Roche Applied bromide staining. Science).

2.5. Plasmid vectors 2.9. Cell viability assay

A fragment of the hTERT promoter (400 to þ60) was ampliﬁed Cell viability was determined using an MTT assay kit (Roche by PCR and inserted into the SacI and SmaI sites of the lucif- Diagnosis, Indianapolis, IN) according to the manufacturer’s erase reporter vector pGL3-Basic (Promega Corp., Madison, protocol. Brieﬂy, A549 and H1299 cells plated in 96-well plates WI) to generate the hTERT promoter luciferase plasmid (2000 cells/well) were treated with CPSF4 siRNA or control pGL3-hTERT-400. A GFP reporter vector (driven by a CMV or siRNA (50 nmol/L). At 48 h hours after treatment, cells were an hTERT promoter) was constructed as previously described transfected with hTERT overexpression vector (pcDNA3.1- (Deng et al., 2007). The CPSF4 and hTERT overexpression vec- hTERT). Forty-eight hours after pcDNA3.1-hTERT transfection, tors pcDNA3.1-CPSF4 and pcDNA3.1-hTERT and the control the viability of the cells was determined. vector pcDNA3.1 were designed and synthesized by Cyagen (Cyagen Biosciences Inc., United States). 2.10. Western blot analysis

2.6. Transient transfection of lung cancer cells Western blot analyses of the cell lysates were performed with antibodies against CPSF4 (Proteintech Group, Inc., Chicago, To overexpress CPSF4, WI-38 and HBE and H322 cells were USA), hTERT (Epitomics), general transcription factor IIB (TFIIB) transfected with a pcDNA3.1-CPSF4 vector or a pcDNA3.1 con- and GAPDH (Cell Signaling Technology, Beverly, MA). Immuno- trol vector using Lipofectamine 2000 (Invitrogen, Carlsbad, reactive protein bands were detected using an enhanced chem- CA). To inhibit CPSF4 expression, A549 and H1299 cells were iluminescence kit (Amersham Pharmacia Biotech, Piscataway, transfected with CPSF4-specific siRNA oligonucleotide NJ) according to the manufacturer’s instructions. (50 nmol/L) with the sequences of the human CPSF4-specific siRNA were 50-CAU GCA CCC UCG AUU UGA ATT-30 (siRNA-1) 2.11. RT-PCR and 50-GGU CAC CUG UUA CAA GUG UTT-30 (siRNA-2); the sequence of the nonspecific siRNA was 50-UUC UCC GAA CGU Total cellular RNA extraction and first strand cDNA synthesis GUC ACG UTT-30 (50 nmol/L). CPSF4 siRNA and nonspecific were performed as described previously (Deng et al., 2007). siRNA were purchased from Shanghai GenePharma Co. Reverse transcriptionepolymerase chain reaction (RTePCR) (Shanghai China). Forty-eight hours after transfection, RNA was performed using EF-Taq polymerase (SolGent, Korea) to and protein were isolated, and RT-PCR, telomerase activity amplify hTERT. The primers for hTERT were as follows: 50- and western blot analyses were carried out as described below. GTCGAGCTGCTCAGGTCTT-30 and 50-

AGTGCTGTCTGATTCCAATGCTT-30. The PCR products were buffer (0.1 mol/L, pH 6.0) for 90 min (for hTERT detection) visualized under ultraviolet light and the band density was and with Target Retrieval Solution (pH 9; DakoCytomation) measured by Quantity One software (Bio-Rad). for 15 min (for CPSF4 detection) using a pressure cooker. The slides were then immersed in methanol containing 3% 2.12. In vivo tumor model and tissue processing hydrogen peroxide for 20 min to block endogenous peroxidase activity. After pre-incubation in 2.5% blocking serum to reduce Animal experiments were carried out in accordance with the Na- nonspecific binding, the sections were incubated overnight tional Institute of Health Guide for the Care and Use of Laboratory with a primary antibody, either anti-CPSF4 (1:50 dilution), or Animals, with the approval of the Animal Research Committee anti-hTERT antibody (1:50 dilution), in a humidified container of Sun Yat-sen University Cancer Center, Guangzhou. at 4 C. The TMA slides were processed with horseradish Cholesterol-conjugated CPSF4 siRNA and negative control siRNA peroxidase immunochemistry according to the manufac- for in vivo delivery were obtained from Shanghai GenePharma turer’s recommendations (DakoCytomation). As a negative Co. (Shanghai China). The knockdown efficiency of these siRNAs control, the staining procedure was performed in parallel was validated in vitro. To investigate the effect of CPSF4 inhibition with the primary antibody replaced by a normal rabbit IgG. on lung cancer cell growth in vivo, A549 cells (2 106)wereinoc- CPSF4 and hTERT proteins were mainly detected in the ulated subcutaneously into the flank of the nude mice. Treat- nuclei of the cancer cells. Nuclear staining intensity was ment began 2 weeks after the injection of the tumor cells. Mice graded as follows: absent staining as 0, weak as 1, moderate were randomly divided into 2 groups (5 mice per group): (a) as 2, and strong as 3. The percentage of stained cells was CPSF4 siRNA and (b) control siRNA. For delivery of cholesterol- graded as follows: 0 (no positive cells), 1 (<25% positive cells), conjugated siRNA, 10 nmol RNA in 0.1 ml saline buffer was 2 (25%e50% positive cells), 3 (50%e75% positive cells), and 4 injected intratumorally twice a week for 3 weeks (Hou et al., (>75% positive cells). The score for each tissue was calculated 2011). The tumor volume was calculated as V ¼ by multiplying the intensity and the percentage value (the (width2 length)/2 using digital calipers. Three weeks later, the range of this calculation was therefore 0e12). The receiver mice were sacrificed, and the tumor weight was recorded. Tumor operating characteristic (ROC) curve analysis was employed specimens were fixed in formalin and embedded in paraffin for to determine cutoff score for tumor “high expression” by using CPSF4 and hTERT protein expression analysis. The immunohis- the 0, 1-criterion. Briefly, the sensitivity and specificity for the tochemical staining was performed as described below. patient outcome being studied at each score was plotted to To confirm the effect of CPSF4 knockdown on tumor cell generate a ROC curve. The score was selected as the cutoff growth in a xenograft mouse model in vivo, the A549 cell lines value, which was closest to the point with both maximum stably expressing CPSF4 short-hairpin RNA (shRNA) or the sensitivity and specificity. Tumors designated as “low expres- scrambled non-target control shRNA were established and sion” for CPSF4 and hTERT were those with scores below or used. The lentivirus for shRNAs were obtained from Santa equal to the cutoff value, while “high expression” tumors Cruz Biotechnology Inc. To rescue hTERT expression, an were those with scores above the value. To facilitate ROC hTERT-expressing lentivirus was used to co-infect A549 cells curve analysis, the clinicopathologic features were dichoto- stably expressing CPSF4 shRNA. Four stable A549 cell lines mized: survival status (death due to lung adenocarcinomas (2 106) were respectively inoculated subcutaneously into or censored). the flank of the nude mice (5 mice per group): (1) CPSF4 shRNA; (2) non-target control shRNA; (3) CPSF4 shRNA þ hTERT; (4) 2.15. Statistical analysis CPSF4 shRNA þ Control empty vector (EV). Once palpable tumors were observed, tumor volume measurements were Student’s t-tests were used to compare two independent taken every 3 days using calipers. groups of data. ROC curve analysis was utilized to define the cutoff score for high expression of CPSF4 and hTERT. Pear- 2.13. Human lung adenocarcinoma specimens son’s correlation test was applied to analyze the association between the abundance of CPSF4 and hTERT. Survival curves The human lung adenocarcinoma tissue microarray used for were constructed using the KaplaneMeier method and were immunostaining analysis of CPSF4 and hTERT protein compared using the log-rank test. Statistical analyses were expression was purchased from Shanghai Outdo Biotech performed using the SPSS 16.0 software. The results were re- (Shanghai, China) and contains 171 lung adenocarcinomas ported as the mean SE. Values of P < 0.05 were considered and their corresponding adjacent non-malignant lung tis- to be statistically significant. sues. These tissue samples had been obtained before anticancer treatment and with prior written consent from patients. The overall survival (OS) for the corresponding pa- 3. Results tients was calculated from the day of surgery to the day of death or to the last follow-up. 3.1. Identification of proteins with differential hTERT promoter binding in telomerase positive lung 2.14. Immunohistochemistry (IHC) staining adenocarcinoma cells and telomerase negative normal lung cells The tissue microarray (TMA) slides were deparaffinized in xylene and rehydrated through graded alcohol. Antigen Previously, we reported that the streptavidin-agarose bead retrieval was performed by incubating samples with citrate pulldown assay is a useful and feasible approach to detect

and discover promoter-binding proteins, such as transactiva- and A549 cells had higher CPSF4 expression than H322 cells, tors, general transcription factors and coactivators (Deng while the expression of CPSF4 proteins in normal lung cell et al., 2007, 2006). In this study, to screen and identify novel lines WI-38 and HBE were almost undetectable. regulators of the hTERT promoter in lung adenocarcinomas, a50-biotinylated 438-bp DNA probe corresponding to nucleo- 3.3. CPSF4 protein activates the hTERT promoter in tides 378 to þ60 of the hTERT promoter sequence was gener- lung cancer cell lines ated, and a streptavidin-agarose bead pulldown method was used. Three human adenocarcinoma cell lines (H1299, A549 To test this hypothesis, we next determined the effect of and H322) are telomerase positive (Deng et al., 2007). WI-38 CPSF4 on hTERT promoter activity. We performed the overex- is an excellent telomerase negative control cell line since it expression experiments in H322 and HBE cells, which have lower press very low levels of hTERT because of promoter repression expression of CPSF4, and the knocking down experiments in (Milyavsky et al., 2003). First, nuclear protein extracts har- A549 and H1299 cells, which have high expression of CPSF4. vested from human lung adenocarcinoma cells (H1299, A549 A GFP reporter and luciferase reporter assay were first per- and H322) and normal lung cell lines (WI-38) were incubated formed. We co-transfected H322, HBE, A549 and H1299 cells with the hTERT promoter probe and streptavidineagarose with a GFP reporter driven by the hTERT or CMV promoter. beads. Next, the candidate regulators that specifically bound As shown in Figure 2A and B, overexpression of CPSF4 upregu- to the hTERT promoter probe were pulled down, separated lated hTERT promoter-mediated GFP expression in H322 and by SDS-PAGE, and visualized by coomassie blue staining. As HBE cells co-transfected with an expression vector encoding shown in Figure 1A (arrow), one of the protein bands (at CPSF4 (pcDNA3.1-CPSF4) and hTERT-GFP plasmids by compar- approximately 30 kDa) was significantly elevated in the lung ison with those cells co-transfected with a control vector lack- cancer cells in comparison with the normal lung cells. ing CPSF4 (pcDNA3.1) and hTERT-GFP plasmids. To identify the candidate tumor cell-elevated hTERT Overexpression of CPSF4 protein significantly increased the promoter-binding protein, a proteomics approach using GFP-positive cell population and the fluorescence intensity MALDI-TOF/TOF MS was utilized. The candidate hTERT of the GFP protein in H322 and HBE cells (Figure 2A and B). In promoter-binding protein (Figure 1A, arrow) was predicted contrast, exogenous expression of CPSF4 did not alter CMV to be the cleavage and polyadenylation specific factor 4 promoter-driven GFP expression in H322 and HBE cells (CPSF4). (Figure 2A and B). Inhibition of CPSF4 expression by CPSF4- specific siRNA (siCPSF4) attenuated hTERT promoter-driven 3.2. Validation of the interaction between CPSF4 protein GFP expression and dramatically decreased the GFP-positive and the hTERT promoter cell population and the fluorescence intensity of the GFP protein in A549 and H1299 cells (Figure 2C and D). Similarly, To verify that CPSF4 is a novel hTERT promoter-binding pro- knockdown of CPSF4 expression did not affect CMV tein in lung cancer cells, we pulled down the CPSF4 in nuclear promoter-driven GFP expression in A549 and H1299 cells protein extracts using a 50-biotinylated hTERT promoter (Figure 2C and D). probe and a streptavidin-agarose bead, and then detected Next, to further confirm the role of CPSF4 in regulating the the CPSF4 proteins on the complexes consisting of nuclear activity of the hTERT promoter, H322 and HBE cells were co- proteins, hTERT promoter and streptavidin-agarose beads transfected with the hTERT promoter-luciferase construct by Western blot using anti-CPSF4 antibody. As shown in pGL3-hTERT-400 and increasing concentrations of Figure 1B, the high levels of CPSF4 proteins which bound to pcDNA3.1-CPSF4. Using the pTK-RL vector as a control to the hTERT promoter probe were detected in lung cancer normalize for transfection efficiency, we showed that lucif- cell lines H1299, A549 and H322. However, the CPSF4 proteins erase activity driven by the hTERT promoter was activated by binding to the hTERT promoter probe were very low in CPSF4 in a dose-dependent manner in H322 and HBE cell lines normal lung cell lines WI-38 and HBE. To further confirm (Figure 2E). Conversely, to assess the effect of decreased CPSF4 that CPSF4 is a novel hTERT promoter-binding protein, a expression on the activity of the hTERT promoter, we knocked ChIP assay was performed in lung cancer cell lines and down CPSF4 expression by co-transfecting A549 and normal lung cell lines. As shown in Figure 1C, CPSF4 protein H1299 cells with siCPSF4 and the hTERT promoter-luciferase bound to the endogenous hTERT promoter in all cell lines. construct. The knockdown of CPSF4 by CPSF4-specific siRNA More importantly, a very weak CPSF4 binding was observed decreased the luciferase activity driven by the hTERT promoter in normal lung cell lines HBE and WI-38, but a strong CPSF4 in a dose-dependent manner in A549 and H1299 cells binding on hTERT promoter was shown in all three lung can- (Figure 2F). Taken together, our results suggest that CPSF4 cer cell lines H1299, A549 and H322 (Figure 1C). These results upregulates hTERT promoter activity. suggest that CPSF4 is recruited to the endogenous hTERT promoter during transcription and that more CPSF4 protein is 3.4. CPSF4 protein promotes the expression of hTERT bound to hTERT promoter in lung cancer cells. Therefore, and telomerase activity in lung cancer cell lines we hypothesized that CPSF4, the tumor cell-elevated hTERT promoter binding protein, drives the transcription of hTERT To verify the role of CPSF4 in regulating the transcription of in lung cancer cells. hTERT, we evaluated the effect of CPSF4 on endogenous We also detected the expression of CPSF4 in the nuclear ex- hTERT mRNA expression in lung cancer cell lines. We found tracts of lung cancer cell lines and norma lung cells by West- that the ectopic expression of CPSF4 using a CPSF4 expression ern blot analysis. As shown in Figure 1D, lung cancer H1299 plasmid significantly increased the level of hTERT mRNA in

Figure 1 e Detection and identiﬁcation of hTERT promoter-binding proteins that differ between telomerase positive lung adenocarcinoma cells and telomerase negative normal lung cells. (A) The potential hTERT promoter-binding proteins were pulled down, separated by the SDS-PAGE, and coomassie blue stained. A representative SDS-PAGE image is shown, and the arrow indicates the candidate hTERT promoter-binding protein. (B) Binding of CPSF4 to a biotinylated hTERT promoter probe (L378 to D60) or a nonspeciﬁc probe (NSP). CPSF4 proteins in the nuclear protein-hTERT probe-streptavidin bead complexes was detected by Western blot using anti-CPSF4 antibody. (C) Chromatin immunoprecipitation assays were carried out using the hTERT promoter from normal lung cells and various lung adenocarcinoma cells. PCR products were separated on 2% agarose gels. (D) The expression of CPSF4 proteins in cell nuclear extracts were analyzed by Western blot. The general transcription factor IIB (TFIIB) was used as loading controls.

WI-38, HBE and H322 cells (Figure 3A). By contrast, the inhibi- role in the regulation of hTERT expression and telomerase tion of CPSF4 by CPSF4 siRNA signiﬁcantly decreased the level activity. of hTERT mRNA in A549 and H1299 cells (Figure 3C). These results indicate that CPSF4 is involved in hTERT transcription in 3.5. The silencing of CPSF4 suppresses tumor growth by lung cancer cells. downregulating hTERT expression in lung adenocarcinoma The expression of the hTERT gene and the activation of in vitro and in vivo telomerase are primarily regulated at the transcriptional level (Kyo et al., 2008). We then further illustrated the biological hTERT promotes the survival and proliferation of cancer cells importance of CPSF4 in the regulation of hTERT expression and has become a very promising target for anticancer ther- by analyzing hTERT protein expression and hTERT activity. apy. Because we observed that hTERT is directly regulated Our results showed that the ectopic expression of CPSF4 up- by CPSF4 (Figure 1e3), we reasoned that CPSF4 inhibition regulated hTERT protein expression in WI-38, HBE and may be effective as a potential therapeutic strategy in lung H322 cells (Figure 3A). Furthermore, hTERT activity, as indi- adenocarcinoma treatment. To test this hypothesis and to cated by telomerase activity, was markedly increased examine the biologic effects of the CPSF4 knockdown on (Figure 3B). In contrast, the blockade of CPSF4 expression by lung adenocarcinoma growth, we performed an in vitro prolif- CPSF4 siRNA suppressed the expression of hTERT protein eration assay and a tumorigenicity assay using a xenograft and telomerase activity in A549 and H1299 cells (Figure 3C mouse model. As shown in Figure 4AeD, the knockdown of and D). These ﬁndings suggest that CPSF4 plays a positive CPSF4 dramatically suppressed the growth of lung cancer

Figure 2 e CPSF4 protein activates the hTERT promoter in lung cancer cells. A, B, activation of hTERT promoter-driven GFP gene expression by CPSF4 overexpression. H322 and HBE cells were co-transfected with pCDNA3.1-CPSF4 and hTERT or CMV promoter-driven GFP-expressing plasmids for 48 h, and the expression of GFP was analyzed by fluorescence microscopy (403 magnification). pCDNA3.1 empty vector was used as a negative control. The percentage of the GFP-positive cell population (A) and the fluorescence intensity (B) of the GFP protein derived from 5000 cells were determined by flow cytometry analysis (*, P < 0.05，**, P < 0.01). C, D, inhibition of hTERT promoter-driven GFP gene expression by CPSF4_ knockdown. A549 and H1299 cells were co-transfected with CPSF4-specific siRNA (siCPSF4) and hTERT or CMV promoter-driven GFP-expressing plasmids for 48 h, and the expression of GFP was analyzed by fluorescence microscopy (403 magnification). Nonspecific control small interfering RNA (siCtrl) was used as a negative control. The percentage of the GFP-positive cell population (C) and the fluorescence intensity (D) of the GFP protein derived from 5000 cells were determined by flow cytometry analysis (*, P < 0.05，**, P < 0.01). (E) H322 and HBE were co-transfected with 1 mg of hTERT-400-luciferase and the indicated doses of pCDNA3.1 empty vector or pCDNA3.1- CPSF4 for 48 h. Luciferase activity was measured using a dual-luciferase assay. The activation of luciferase was calculated relative to cells transfected with empty vector. All of the measurements represent the means ± SE of three independent experiments (*, P < 0.05). (F) A549 and H1299 were co-transfected with 1 mg of hTERT-400-luciferase and the indicated doses of CPSF4 siRNA or a nonspecific control siRNA for 48 h. Luciferase activity was measured using a dual-luciferase assay. The inhibition of luciferase was calculated as a percentage relative to cells transfected with control siRNA. All of the measurements represent the means ± SE of three independent experiments (*, P < 0.05).

cells. To further assess whether hTERT expression is involved growth partially through the downregulation of hTERT in this tumor growth suppression by siRNA against CPSF4, we expression. rescued hTERT expression using an hTERT-expressing plasmid after CPSF4-specific siRNA treatment. The CPSF4- 3.6. A positive correlation between the protein levels of specific siRNA-induced tumor inhibition was partially rescued CPSF4 and hTERT in lung adenocarcinoma tissues by the ectopic expression of hTERT (Figure 4A). Moreover, as shown in Figure 4E, the knockdown of CPSF4 did decrease To further determine the biological relevance of CPSF4- the hTERT protein levels in xenografts. mediated expression of hTERT, the expression of CPSF4 and To confirm the inhibitory effect of CPSF4 knockdown on hTERT was analyzed in 171 human lung adenocarcinoma tis- tumorcellgrowthinaxenograftmousemodelin vivo,we sues by IHC staining (Figure 5A). We quantified the immuno- established the A549 cell lines stably expressing CPSF4 histochemical staining of CPSF4 and hTERT in the human short-hairpin RNA (shRNA) or a scrambled non-target con- lung adenocarcinoma specimens on the 0 to 12 scale and trol shRNA. To see whether the hTERT can rescue the tu- then analyzed the scores. We found that CPSF4 expression mor volume change induced by CPSF4 knockdown in levels correlated positively with hTERT expression levels in mice, the A549 cells stably expressing CPSF4 shRNA were lung adenocarcinoma samples (Pearson’s correlation test co-infected with a hTERT-expressing lentivirus. The results r ¼ 0.55; P < 0.001) (Figure 5B). showed that the stable expression of CPSF4 shRNA consid- We also detected the expression of CPSF4 and hTERT in the erably inhibited tumor volume by comparison with the whole cell lysates of lung cancer cell lines and norma lung control shRNA. However, the overexpression of hTERT in cells by Western blot analysis. The results showed that the A549 cells stably expressing CPSF4 shRNA effectively lung cancer lines (H1299, A549 and H322) highly expressed rescued the tumor volume change induced by CPSF4 knock- CPSF4 and hTERT proteins, while the expression of CPSF4 down (Figure 4F). These results suggest that CPSF4 knock- and hTERT proteins in normal lung cell lines (WI-38 and down exerts its inhibitory effect on tumor cell HBE) could not be detected (Figure 5C). These results suggest

Figure 3 e CPSF4 protein promotes hTERT expression and telomerase activity in lung cancer cell lines. A, B, the up-regulation of hTERT mRNA and protein expression and telomerase activity by CPSF4 overexpression. WI-38 and HBE and H322 cells were transfected with pCDNA3.1 empty vector or pCDNA3.1-CPSF4 for 48 h, and the expression of hTERT mRNA and protein was analyzed by RT-PCR and Western blot (A). Telomerase activity was measured in WI-38 and HBE and H322 cells using a telomeric repeat ampliﬁcation protocol assay- based TeloTAGGG telomerase PCR enzyme-linked immunosorbent assay (**, P < 0.01) (B). C, D, down-regulation of hTERT mRNA and protein expression and telomerase activity by CPSF4 knockdown. A549 and H1299 cells were transfected with 50 nmol/L of CPSF4 siRNA or nonspeciﬁc control siRNA for 48 h, and the expression of hTERT mRNA and protein was analyzed by RT-PCR and Western blot (C). Telomerase activity was measured in A549 and H1299 cells by telomerase PCR enzyme-linked immunosorbent assay (**, P < 0.01) (D).

that CPSF4 expression is correlated with hTERT expression in highly specific to malignant cells using a streptavidin- human lung cancer cells. agarose pulldown assay and high-throughput proteomics. We observed that human lung adenocarcinoma cells 3.7. High CPSF4 and hTERT protein expression are expressed higher levels of CPSF4 proteins than normal lung associated with a poor clinical outcome in patients with lung cells. Moreover, we found that the promoter activity of hTERT adenocarcinoma was dependent on CPSF4. CPSF4 enhanced the expression of the hTERT promoter-driven GFP reporter gene without As shown in Figure 6A, CPSF4 and hTERT protein were local- affecting the expression of the CMV promoter-driven GFP ized to the nucleus of cancer cells. CPSF4 and hTERT protein gene. The forced ectopic expression of CPSF4 by transfection are highly expressed in tumor tissues compared to adjacent with a CPSF4 expression vector significantly up-regulated non-malignant lung tissues (Figure 6A, D). The ROC curves hTERT expression and telomerase activity of lung adenocarci- for CPSF4 and hTERT (Figure 6B, C) clearly show the point nomas cell, whereas the inhibition of CPSF4 by transfection on the curve closest to (0.0, 1.0) which maximizes both with CPSF4 siRNA did the opposite. Furthermore, our in vitro sensitivity and specificity for the OS. The CPSF4 and hTERT and in vivo results revealed that CPSF4 knockdown inhibited IHC cut-off scores for OS were 5.0 (Figure 6B, C). Thus, the lung cancer cell growth by decreasing hTERT expression. expression of CPSF4 and hTERT in each sample was subse- Importantly, our results also show that the levels of CPSF4 quently classified as either high level (score > 5) or low level expression are strongly correlated with the levels of hTERT (score 5). Moreover, the lung adenocarcinoma patients expression in lung adenocarcinoma specimens. Thus, we with high CPSF4 and hTERT expression had a significantly have identified the overexpression of CPSF4 as a cause of shorter OS than those with low CPSF4 and hTERT expres- deregulated hTERT expression in some lung sion (Figure 6E). Therefore, our results suggest that adenocarcinomas. high CPSF4 and hTERT protein expression are associated We have previously demonstrated that the streptavidin- with a poor clinical outcome in patients with lung agarose pulldown assay is an effective screening system to adenocarcinoma. analyze the tumor-specific transactivators and coactivators that regulate the promoters of carcinogenic genes COX-2 and hTERT (Deng et al., 2007; Deng et al., 2006, 2004; Deng 4. Discussion and Wu, 2003). These tumor-specific transcriptional regulation factors may become valuable diagnostic and therapeutic Our study sought to discover and identify several potential target molecules for cancer. One example of these factors is hTERT promoter-regulating proteins that are expected to be AP-2b, which exhibits tumor-specific binding to the hTERT

Figure 4 e The knockdown of CPSF4 inhibits tumor growth by downregulating hTERT expression. (A) A549 and H1299 cells were transfected with CPSF4 siRNA or nonspecific control siRNA. At 48 h after treatment, cells were transfected with an hTERT overexpression vector (pcDNA3.1-hTERT) or an empty vector. Forty-eight hours later, cell viability was measured using an MTT assay. The mean and SE obtained from three independent experiments are plotted. (B) A representative picture of nude mice comparing the sizes of tumor grafts in nude mice 21 days after intratumoral injection of nonspecific control siRNA or CPSF4-specific siRNA. (C) Tumor volumes ± SE in nude mice. N [ 5; ***, P < 0.001. (D) Mean tumor weights ± SE in nude mice. N [ 5; ***, P < 0.001. (E) Immunohistochemistry of CPSF4 and hTERT from tumor xenografts in nonspecific control siRNA- and CPSF4-specific siRNA-treated nude mice (4003 magnification). (F) The A549 stable cell lines were injected into the flanks of nude mice. The tumor volumes were measured and recorded every 3 days, and tumor growth curves were created for each group (n [ 5). Dots represent the mean, while bars indicate the SEM. (*, P < 0.05; **, P < 0.01).

promoter in non-small cell lung cancer (Deng et al., 2007), an oncogenic function in lung adenocarcinomas, and CPSF4 and, when overexpressed, predicts poor survival in patients is frequently overexpressed in lung adenocarcinoma samples with stage I non-small cell lung cancer (Kim et al., 2011). In and correlates with poor clinical outcome. Therefore, our re- this study, we detected CPSF4 as a candidate hTERT sults once again suggest that the streptavidin-agarose pull- promoter-binding protein using streptavidin-agarose pulldown assay will be a useful approach to identify potential down technology. We then veriﬁed that CPSF4 regulates the molecular targets for the diagnosis and/or treatment of promoter activity and expression of hTERT. CPSF4 may have cancers.

Figure 5 e There is a positive correlation between the protein levels of CPSF4 and hTERT in lung adenocarcinoma specimens. (A) Representative immunohistochemical staining examples of high or low CPSF4 and hTERT expression in the serial sections from the same tumor tissues are shown. Scale bar, 200 mm. Case 1 and 2 means two different patients. (B) The tissue sections were quantitatively scored according to the percentage of positive cells and staining intensity as described in Materials and Methods. The percentage and intensity scores were multiplied to obtain a total score (range, 0e12). CPSF4 expression levels correlated positively with hTERT expression levels in lung adenocarcinoma samples (Pearson’s correlation test, r [ 0.55; P < 0.001). (C) The expression of CPSF4 and hTERT proteins in the total cell lysates of lung normal cells and cancer cells were analyzed by Western blot.

hTERT overexpression is observed in approximately 90% of that various cellular factors directly or indirectly regulate human cancers, including lung cancer, but the level of hTERT the hTERT promoter in lung cancer, including cellular tran- in most normal tissues is almost undetectable (Daniel et al., scriptional activators (c-Myc, Sp1, HPV-16 E6 oncoprotein, 2012; Lantuejoul et al., 2004; Toomey et al., 2001; Zhu et al., etc.) and repressors (p53, p73, etc) (Beitzinger et al., 2006; 2006). However, little is known about how hTERT is reacti- Cheng et al., 2008; Fujiki et al., 2007). Nevertheless, previous vated during tumorigenesis. Studies have indicated that studies on the regulation of hTERT expression have revealed hTERT gene ampliﬁcation is one of the mechanisms for hTERT the diversity and complexity of hTERT transcriptional regula- overexpression in non-small-cell lung cancer (Zhu et al., 2006). tion (Daniel et al., 2012). Thus, our current study adds a new In this report, we provide both clinical and mechanistic evi- mechanism to the body of research on the transcriptional dence that an increase in the abundance of CPSF4 protein is regulation of hTERT expression by providing evidence that another mechanism that may explain the up-regulated hTERT hTERT is a direct transcriptional target of CPSF4 and that expression in some lung adenocarcinomas. Interestingly, CPSF4 critically controls hTERT expression in lung adenocarci- there were 29 tumors that contained high levels of hTERT nomas cells. and low levels of CPSF4 (Figure 6D), indicating that other fac- CPSF4 belongs to the CPSF complex which cooperates with tors, in addition to the CPSF4, may be involved in the regula- other 30 mRNA-processing factors participating in the maturation of hTERT expression in lung adenocarcinomas. For tion of the 30 ends of mRNA (Barabino et al., 1997; de Vries example, AP-2b activates the expression of the hTERT in et al., 2000; Kaufmann et al., 2004; Nemeroff et al., 1998). lung cancer, as we reported previously (Deng et al., 2007). How these primary transcript-processing factors regulate The transcriptional regulation of the hTERT gene is the major gene transcription is an interesting question. Recent studies mechanism for cancer-speciﬁc activation of hTERT (Daniel have suggested that the transcriptional, splicing and et al., 2012; Kyo et al., 2008). Several studies have indicated cleavage-polyadenylation factors assemble an mRNA ’factory’

Figure 6 e CPSF4 and hTERT protein are highly expressed in lung adenocarcinomas tissues compared to adjacent non-malignant lung tissues and higher expression of CPSF4 and hTERT indicates a poor prognosis. (A) Immunohistochemical staining of CPSF4 and hTERT was performed in a tumor tissue microarray of samples from patients with lung adenocarcinomas. Representative examples of CPSF4 and hTERT staining in the tumor tissues and adjacent non-malignant lung tissues are shown (2003 magniﬁcation). B, C, ROC curve analysis was used to determine the cutoff score for high expression of CPSF4 and hTERT protein in lung adenocarcinoma tissues. The sensitivity and speciﬁcity for OS were plotted: (B) CPSF4; p [ 0.001 (C) hTERT; p [ 0.008. (D) The protein level of CPSF4 correlates positively with the protein level of hTERT in lung adenocarcinoma tissues (P < 0.001, c2 tests). (E) KaplaneMeier analysis of overall survival with high or low CPSF4 and hTERT expression ( p < 0.001, log-rank test).

that carries out the coupled transcription, splicing and cleava- regulates the transcription of target genes. Due to the lack of geepolyadenylation of mRNA precursors. This ’factory’ exists DNA-binding domains in CPSF4 that are found in general tran- in place at the promoter and produces mature transcripts scription factors, we speculated that CPSF4 may execute its (Calvo and Manley, 2003; McCracken et al., 1997). Rozenblatt- co-activation effect on hTERT by recruiting the general tran- Rosen et al. (2009) found that CPSF4 interacted with transcrip- scription factors to assemble the hTERT transcriptional com- tional factors to form a complex that binds to promoters and plex in the nucleus. However, we cannot rule out the

possibility that CPSF4 may affect hTERT mRNA maturation, Bisoffi, M., Heaphy, C.M., Griffith, J.K., 2006. Telomeres: prognostic such as cleavage and polyadenylation of hTERT mRNA 30 markers for solid tumors. Int. J. Cancer 119, 2255e2260. Journal ends. Further detailed analyses are necessary to determine International du Cancer. Blackburn, E.H., 2005. Telomerase and Cancer: Kirk A. the effect of CPSF4 on hTERT mRNA maturation. e 0 Landon AACR prize for basic cancer research lecture. Mol. Recently, studies have described the role of some mRNA 3 cancer Res. : MCR 3, 477e482. end-processing factors in cancer, including FIP1L1, CSTF50, Blackburn, E.H., Greider, C.W., Szostak, J.W., 2006. Telomeres and CSTF2 and Neo-PAP (Aragaki et al., 2011; Cools et al., 2004; telomerase: the path from maize, Tetrahymena and yeast to Gotlib et al., 2004; Kleiman and Manley, 2001; Topalian et al., human cancer and aging. Nat. Med. 12, 1133e1138. 2001). For example, Aragaki and colleagues found that CSTF2 Bryan, T.M., Cech, T.R., 1999. Telomerase and the maintenance of e overexpression is an independent poor prognostic factor in chromosome ends. Curr. Opin. Cell Biol. 11, 318 324. Bryan, T.M., Englezou, A., Dalla-Pozza, L., Dunham, M.A., non-small-cell lung cancer patients. In addition, the suppres- Reddel, R.R., 1997. Evidence for an alternative mechanism for sion of CSTF2 expression inhibited the growth of lung cancer maintaining telomere length in human tumors and tumor- cells, whereas the exogenous expression of CSTF2 promoted derived cell lines. Nat. Med. 3, 1271e1274. the growth and invasion of lung cancer cells (Aragaki et al., Calvo, O., Manley, J.L., 2003. Strange bedfellows: polyadenylation 2011). The results from our in vitro and in vivo experiments factors at the promoter. Genes Dev. 17, 1321e1327. indicated that CPSF4 exerted its oncogenic function by regu- Cheng, Y.W., Wu, T.C., Chen, C.Y., Chou, M.C., Ko, J.L., Lee, H., 2008. Human telomerase reverse transcriptase activated by E6 lating hTERT expression in lung adenocarcinoma cells, oncoprotein is required for human papillomavirus-16/18- although the exact molecular mechanism responsible for infected lung tumorigenesis. Clin. cancer Res. : An Official J. CPSF4 overexpression in lung adenocarcinoma cells is still Am. Assoc. Cancer Res. 14, 7173e7179. unknown. Chiu, C.P., Harley, C.B., 1997. Replicative senescence and cell In summary, CPSF4 upregulates hTERT promoter activity immortality: the role of telomeres and telomerase. Proc. Soc. and thus transcriptionally activates the expression of hTERT Exp. Biol. Med. 214, 99e106. in lung cancer cells. Lung adenocarcinoma patients with Cohen, S.B., Graham, M.E., Lovrecz, G.O., Bache, N., Robinson, P.J., Reddel, R.R., 2007. Protein composition of catalytically active high expression levels of CPSF4 and hTERT protein had human telomerase from immortal cells. Science 315, shorter survival periods. CPSF4 knockdown inhibited tumor 1850e1853. growth in vitro and in vivo. CPSF4 may be a new therapeutic Colgin, L.M., Reddel, R.R., 1999. Telomere maintenance target in lung adenocarcinomas. mechanisms and cellular immortalization. Curr. Opin. Gen. Dev. 9, 97e103. Cools, J., Stover, E.H., Wlodarska, I., Marynen, P., Gilliland, D.G., 2004. The FIP1L1-PDGFRalpha kinase in hypereosinophilic 5. Disclosure of conflict of interest syndrome and chronic eosinophilic leukemia. Curr. Opin. Hematol. 11, 51e57. The authors declare no conflicts of interest. Daniel, M., Peek, G.W., Tollefsbol, T.O., 2012. Regulation of the human catalytic subunit of telomerase (hTERT). Gene 498, 135e146. de Vries, H., Ruegsegger, U., Hubner, W., Friedlein, A., Langen, H., Keller, W., 2000. Human pre-mRNA cleavage factor II(m) Acknowledgments contains homologs of yeast proteins and bridges two other cleavage factors. EMBO J. 19, 5895e5904. This work was supported by the funds from the National Nat- Deng, W.G., Jayachandran, G., Wu, G., Xu, K., Roth, J.A., Ji, L., 2007. ural Science Foundation of China (81272896, 81272195, Tumor-specific activation of human telomerase reverses 81071687, 81372133), the State “863 Program” of China transcriptase promoter activity by activating enhancer- (SS2012AA020403), the State “973 Program” of China binding protein-2beta in human lung cancer cells. J. Biol. Chem. 282, 26460e26470. (2014CB542005), the Doctoral Programs Foundation of Ministry Deng, W.G., Tang, S.T., Tseng, H.P., Wu, K.K., 2006. Melatonin of Education of China (20110171110077), and the State Key suppresses macrophage cyclooxygenase-2 and inducible nitric Laboratory of Oncology in South China (W Deng). oxide synthase expression by inhibiting p52 acetylation and binding. Blood 108, 518e524. REFERENCES Deng, W.G., Wu, K.K., 2003. Regulation of inducible nitric oxide synthase expression by p300 and p50 acetylation. J. Immunol. 171, 6581e6588. Deng, W.G., Zhu, Y., Wu, K.K., 2004. Role of p300 and PCAF in Aragaki, M., Takahashi, K., Akiyama, H., Tsuchiya, E., Kondo, S., regulating cyclooxygenase-2 promoter activation by Nakamura, Y., Daigo, Y., 2011. Characterization of a cleavage inflammatory mediators. Blood 103, 2135e2142. stimulation factor, 3’ pre-RNA, subunit 2, 64 kDa (CSTF2) as a Ducrest, A.L., Szutorisz, H., Lingner, J., Nabholz, M., 2002. therapeutic target for lung cancer. Clin. Cancer Res. An Official Regulation of the human telomerase reverse transcriptase J. Am. Assoc. Cancer Res. 17, 5889e5900. gene. Oncogene 21, 541e552. Barabino, S.M., Hubner, W., Jenny, A., Minvielle-Sebastia, L., Feng, J., Funk, W.D., Wang, S.S., Weinrich, S.L., Avilion, A.A., Keller, W., 1997. The 30-kD subunit of mammalian cleavage Chiu, C.P., Adams, R.R., Chang, E., Allsopp, R.C., Yu, J., et al., and polyadenylation specificity factor and its yeast homolog 1995. The RNA component of human telomerase. Science 269, are RNA-binding zinc finger proteins. Genes Dev. 11, 1236e1241. 1703e1716. Fujiki, T., Miura, T., Maura, M., Shiraishi, H., Nishimura, S., Beitzinger, M., Oswald, C., Beinoraviciute-Kellner, R., Stiewe, T., Imada, Y., Uehara, N., Tashiro, K., Shirahata, S., 2006. Regulation of telomerase activity by the p53 family Katakura, Y., 2007. TAK1 represses transcription of the member p73. Oncogene 25, 813e826.

human telomerase reverse transcriptase gene. Oncogene 26, Nakamura, T.M., Morin, G.B., Chapman, K.B., Weinrich, S.L., 5258e5266. Andrews, W.H., Lingner, J., Harley, C.B., Cech, T.R., 1997. Gotlib, J., Cools, J., Malone 3rd, J.M., Schrier, S.L., Gilliland, D.G., Telomerase catalytic subunit homologs from fission yeast and Coutre, S.E., 2004. The FIP1L1-PDGFRalpha fusion tyrosine human. Science 277, 955e959. kinase in hypereosinophilic syndrome and chronic Nemeroff, M.E., Barabino, S.M., Li, Y., Keller, W., Krug, R.M., 1998. eosinophilic leukemia: implications for diagnosis, Influenza virus NS1 protein interacts with the cellular 30 kDa classification, and management. Blood 103, 2879e2891. subunit of CPSF and inhibits 3’end formation of cellular pre- Harley, C.B., 2008. Telomerase and cancer therapeutics. Nat. Rev. mRNAs. Mol. Cell 1, 991e1000. Cancer 8, 167e179. Rozenblatt-Rosen, O., Nagaike, T., Francis, J.M., Kaneko, S., Hou, J., Lin, L., Zhou, W., Wang, Z., Ding, G., Dong, Q., Qin, L., Glatt, K.A., Hughes, C.M., LaFramboise, T., Manley, J.L., Wu, X., Zheng, Y., Yang, Y., Tian, W., Zhang, Q., Wang, C., Meyerson, M., 2009. The tumor suppressor Cdc73 functionally Zhuang, S.M., Zheng, L., Liang, A., Tao, W., Cao, X., 2011. associates with CPSF and CstF 3’ mRNA processing factors. Identification of miRNomes in human liver and hepatocellular Proc. Natl. Acad. Sci. U S A 106, 755e760. carcinoma reveals miR-199a/b-3p as therapeutic target for Ruden, M., Puri, N., 2013. Novel anticancer therapeutics targeting hepatocellular carcinoma. Cancer Cell 19, 232e243. telomerase. Cancer Treat. Rev. 39, 444e456. Kaufmann, I., Martin, G., Friedlein, A., Langen, H., Keller, W., 2004. Saretzki, G., Sitte, N., Merkel, U., Wurm, R.E., von Zglinicki, T., Human Fip1 is a subunit of CPSF that binds to U-rich RNA 1999. Telomere shortening triggers a p53-dependent cell cycle elements and stimulates poly(A) polymerase. EMBO J. 23, arrest via accumulation of G-rich single stranded DNA 616e626. fragments. Oncogene 18, 5148e5158. Kiefer, H., Mizutani, A., Iemura, S., Natsume, T., Ando, H., Shay, J.W., Keith, W.N., 2008. Targeting telomerase for cancer Kuroda, Y., Mikoshiba, K., 2009. Inositol 1,4,5-triphosphate therapeutics. Br. J. Cancer 98, 677e683. receptor-binding protein released with inositol 1,4,5- Takakura, M., Kyo, S., Kanaya, T., Hirano, H., Takeda, J., triphosphate (IRBIT) associates with components of the mRNA Yutsudo, M., Inoue, M., 1999. Cloning of human telomerase 3’ processing machinery in a phosphorylation-dependent catalytic subunit (hTERT) gene promoter and identification of manner and inhibits polyadenylation. J. Biol. Chem. 284, proximal core promoter sequences essential for 10694e10705. transcriptional activation in immortalized and cancer cells. Kim, M.P., Chen, Y., Bekele, B.N., Lopez, A., Khanna, A., Chen, J.Q., Cancer Res. 59, 551e557. Spitz, M.R., Behrens, C., Solis, L., Wismach, M., Ji, L., Tian, X., Chen, B., Liu, X., 2010. Telomere and telomerase as Wistuba, I.I., Roth, J.A., Katz, R.L., 2011. Activating enhancer- targets for cancer therapy. Appl. Biochem. Biotechnol. 160, binding protein-2beta nucleolar localization predicts poor 1460e1472. survival after stage I non-small cell lung cancer resection. Toomey, D., Smyth, G., Condron, C., Kay, E., Conroy, R., Foley, D., Ann. Thorac Surg. 92, 1044e1050. Hong, C., Hogan, B., Toner, S., McCormick, P., Broe, P., Kelly, C., Kleiman, F.E., Manley, J.L., 2001. The BARD1-CstF-50 interaction Bouchier-Hayes, D., 2001. Immune function, telomerase, and links mRNA 3’ end formation to DNA damage and tumor angiogenesis in patients with primary, operable nonsmall cell suppression. Cell 104, 743e753. lung carcinoma: tumor size and lymph node status remain the Kyo, S., Takakura, M., Fujiwara, T., Inoue, M., 2008. most important prognostic features. Cancer 92, 2648e2657. Understanding and exploiting hTERT promoter regulation for Topalian, S.L., Kaneko, S., Gonzales, M.I., Bond, G.L., Ward, Y., diagnosis and treatment of human cancers. Cancer Sci. 99, Manley, J.L., 2001. Identification and functional 1528e1538. characterization of neo-poly(A) polymerase, an RNA Lantuejoul, S., Soria, J.C., Moro-Sibilot, D., Morat, L., Veyrenc, S., processing enzyme overexpressed in human tumors. Mol. Lorimier, P., Brichon, P.Y., Sabatier, L., Brambilla, C., Cell. Biol. 21, 5614e5623. Brambilla, E., 2004. Differential expression of telomerase Weinrich, S.L., Pruzan, R., Ma, L., Ouellette, M., Tesmer, V.M., reverse transcriptase (hTERT) in lung tumours. Br. J. Cancer 90, Holt, S.E., Bodnar, A.G., Lichtsteiner, S., Kim, N.W., 1222e1229. Trager, J.B., Taylor, R.D., Carlos, R., Andrews, W.H., Liu, J.P., 1999. Studies of the molecular mechanisms in the Wright, W.E., Shay, J.W., Harley, C.B., Morin, G.B., 1997. regulation of telomerase activity. FASEB J. Official Pub. Fed. Reconstitution of human telomerase with the template RNA Am. Soc. Exp. Biol. 13, 2091e2104. component hTR and the catalytic protein subunit hTRT. Nat. Lu, M.H., Liao, Z.L., Zhao, X.Y., Fan, Y.H., Lin, X.L., Fang, D.C., Gen. 17, 498e502. Guo, H., Yang, S.M., 2012. hTERT-based therapy: a universal Zhang, A., Zheng, C., Lindvall, C., Hou, M., Ekedahl, J., anticancer approach (Review). Oncol. Rep. 28, 1945e1952. Lewensohn, R., Yan, Z., Yang, X., Henriksson, M., Blennow, E., McCracken, S., Fong, N., Yankulov, K., Ballantyne, S., Pan, G., Nordenskjold, M., Zetterberg, A., Bjorkholm, M., Gruber, A., Greenblatt, J., Patterson, S.D., Wickens, M., Bentley, D.L., 1997. Xu, D., 2000. Frequent amplification of the telomerase reverse The C-terminal domain of RNA polymerase II couples mRNA transcriptase gene in human tumors. Cancer Res. 60, processing to transcription. Nature 385, 357e361. 6230e6235. Milyavsky, M., Shats, I., Erez, N., Tang, X., Senderovich, S., Zhu, C.Q., Cutz, J.C., Liu, N., Lau, D., Shepherd, F.A., Squire, J.A., Meerson, A., Tabach, Y., Goldfinger, N., Ginsberg, D., Tsao, M.S., 2006. Amplification of telomerase (hTERT) gene is a Harris, C.C., Rotter, V., 2003. Prolonged culture of telomerase- poor prognostic marker in non-small-cell lung cancer. Br. J. immortalized human fibroblasts leads to a premalignant Cancer 94, 1452e1459. phenotype. Cancer Res. 63, 7147e7157.