p15PAF Is an Rb/E2F-Regulated S-Phase Protein Essential for DNA Synthesis and Cell Cycle Progression

The p15PAF/KIAA0101 protein is a proliferating cell nuclear antigen (PCNA)-associated protein overexpressed in multiple types of cancer. Attenuation of p15PAF expression leads to modifications in the DNA repair process, rendering cells more sensitive to ultraviolet-induced cell death. In this study, we identified that p15PAF expression peaks during the S phase of the cell cycle. We observed that p15PAF knockdown markedly inhibited cell proliferation, S-phase progression, and DNA synthesis. Depletion of p15PAF resulted in p21 upregulation, especially chromatin-bound p21. We further identified that the p15PAF promoter contains 3 E2F-binding motifs. Loss of Rb-mediated transcriptional repression resulted in upregulated p15PAF expression. Binding of E2F4 and E2F6 to the p15PAF promoter caused transcriptional repression. Overall, these results indicate that p15PAF is tightly regulated by the Rb/E2F complex. Loss of Rb/E2F-mediated repression during the G1/S transition phase leads to p15PAF upregulation, which facilitates DNA synthesis and S-phase progression.


Introduction
The p15 PAF protein, also known as KIAA0101 or OEATC-1, is a 15-kDa nuclear protein initially identified using a yeast two hybrid screen for proteins that bind to proliferating cell nuclear antigen (PCNA) [1]. p15 PAF binds to PCNA through the conserved PCNA-interacting protein motif (PIP box, Qxx(L/I/ M]xx[F/Y][F/Y]) at amino acids 62-69 (62-QKGIGEFF-69) [1]. No other functional domains or motifs have been identified using bioinformatic methods. p15 PAF is overexpressed in multiple types of human cancer, including hepatocellular carcinoma [2], lung cancer [3], breast cancer [4], and pancreatic cancer [5], and its overexpression is associated with poor patient outcome [2][3][4]. Overexpression of p15 PAF promotes cancer cell growth, whereas attenuation of expression by small interfering RNA (siRNA) leads to reduced cell proliferation [5]. These results indicate that p15 PAF has a growth-promoting role; however, the molecular mechanisms underlying its effects have yet to be identified. p15 PAF was reported to play a role in DNA repair. In a study by Simpson et al., p15 PAF expression was upregulated in response to ultraviolet (UV) irradiation. Besides, the associations of p15 PAF and p33 ING1b with PCNA were enhanced after UV irradiation [6]. Similar to p15 PAF , p33 ING1b is a PCNA-interacting protein, and is an isoform encoded by the ING1 tumor suppressor locus. Overexpression of p33 ING1b confers increased efficiency of repair of UV-damaged DNA in melanoma cells, and p53 is required for the repair process [7]. Overexpression of p15 PAF also protects cells from UV-induced cell death [6]. p15 PAF is a direct transcriptional target of ATF3. ATF3 and p15 PAF expression are sufficient to trigger the DNA repair machinery against UV damage [8]. p15 PAF was also reported to interact with BRCA1 and regulate the centrosome number [9]; however, it is not yet known if p15 PAF is a component of the double-strand break repair pathway.
Several PCNA-interacting proteins are regulators of cell cycle progression or components of DNA synthesis machinery [10]. In this study, we evaluated the role of p15 PAF in cell cycle progression and DNA synthesis, and showed that p15 PAF is a direct transcriptional target of the Rb/E2F pathway.

Immunohistochemical stain
The tissue distribution of p15 PAF protein was evaluated using a tissue array containing adult tissues of major organs, obtained from normal regions of surgically resected patient specimens. A 2 mm tissue core was removed for each organ using a manual tissue array device (Beecher Instruments, Silver Spring, MD, USA) and inserted into a recipient paraffin block. The arrayed tissues were cut into 4 mm slices and placed on positively charged slides. Tissue sections were dewaxed and rehydrated. Antigen retrieval was performed by incubating slides in 0.01 M citric acid buffer (pH 6.0) at 100 uC for 10 min. After blocking with 3% hydrogen peroxide and 5% fetal bovine serum (FBS), slides were incubated with a monoclonal antibody against p15 PAF (Abnova, Taipei, Taiwan) at 1:500 dilution at 4 uC overnight. Slides were then incubated with polymer-horseradish peroxidase (HRP) reagent (BioGenex, San Ramon, CA, USA). Peroxidase activity was visualized using diaminobenzidine tetrahydroxychloride solution (BioGenex). Sections were counterstained with hematoxylin. Dark brown nuclear staining was defined as positive, and no staining was defined as negative. For negative controls, the primary antibody was replaced with 5% FBS. This study was approved by the Research Ethical Committee of National Taiwan University Hospital. Written informed consent was obtained from all patients.

Cell culture and treatment
HeLa cells, MCF7 cells, and viral package 293T cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% FBS in an incubator at 37 uC, with a humidified atmosphere of 95% air and 5% carbon dioxide (CO 2 ). For synchronization of the cell cycle, HeLa cells were treated with 2 mM thymidine (Sigma-Aldrich, St. Louis, MO, USA) for 14 h and then washed 3 times with phosphate buffered saline (PBS). They were then incubated in normal growth medium for 10 h prior to treatment with aphidicolin (5 mg/mL; Sigma-Aldrich) or thymidine (2 mM) for an additional 14 h to arrest cells at the G1/ S boundary.

Cell proliferation assay
The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazoliumbromide (MTT) assay was used to evaluate the rate of cell proliferation. This colorimetric assay measures the activity of cellular enzymes that reduce MTT dye to insoluble MTT formazan, giving a purple color. One-thousand living cells were seeded into 96-well plates and incubated at 37 uC in a humidified atmosphere with 5% CO 2 . After an appropriate time interval, MTT was added and incubated for 4 h. The resulting color reaction product was extracted using dimethyl sulfoxide and absorbance was measured at 570 nm.

Flow cytometry
The harvested cells were resuspended in 100 mL PBS at a density of 2610 6 cells/mL. The cells were added dropwise to 1 mL ice-cold 70% alcohol for fixation at 220uC overnight. The cells were then washed twice with PBS and incubated with 1 mL of PBS containing 10 mg/mL RNase A and 1 mg/mL propidium iodide (Sigma-Aldrich) for 30 min to degrade RNA and stain DNA. DNA ploidy was analyzed using a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA).

BrdU incorporation assay
Incorporation of BrdU during the S phase was evaluated using the BD BrdU FITC Assay Kit (BD Biosciences) according to the manufacturer's instructions.
For detection of incorporated BrdU, cells were pulsed with BrdU (10 mM, Roche, Basel, Switzerland) for 30 min at 37 uC and harvested at the indicated time points prior to fixation. Cells were fixed with cold methanol for 20 min at 220uC. DNA was denatured by treating with 4 M hydrogen chloride for 15 min and washed 3 times with PBS. Immunofluorescence was performed as described using FITC-anti-BrdU antibody (1:100, AbCam, Cambridge, MA, USA).

Reporter assay
The p15 PAF promoter fragment 2208 to +94 bp was obtained using PCR. The primers used were p15 PAF /SacI-208: 59-AGAGCTCATTTCTGTAGTTCAGAAAC-3 and p15 PAF /Hin-dIII+94: TAAGCTTCCCAGCCGAGGGTGTTTC. The PCR p15 PAF in Cell Cycle Progression PLOS ONE | www.plosone.org fragments were cloned into a promoterless pGL3-basic vector. Mutations were introduced by site-directed mutagenesis in the putative E2F-binding sites using the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA). The AURKA and MCL1 promoters, E2F4 and E2F6 expression plasmids, were gifts from Dr. Ju-Ming Wang (Institute of Bioinformatics and Biosignal Transduction, National Cheng Kung University). The Rb expression vector pSLX-CMV-Rb was kindly provided by Dr. Ming-Fu Chang (Institute of Biochemistry and Molecular Biology, National Taiwan University). Transient transfection was performed using the Turbofect reagent (Thermo Fisher, Waltham, MA, USA). Reporters were cotransfected with expression vectors of Rb, E2F4, E2F6 or the control plasmid, and the Renilla luciferase plasmid TK-Renilla, into MCF7 cells, which were grown in 12-well plates at 70% confluence. Twenty-four h after transfection, cell extracts were prepared, and luciferase activities were quantified using the Dual-Glo Luciferase Assay System (Promega, Madison, WI, USA) in an Orion II luminometer (Berthold Detection Systems, Pforzheim, Germany). All experiments were performed in triplicate.

Chromatin immunoprecipitation
After culturing to 60%-70% confluency, formaldehyde (1% final concentration) was added for 10 min crosslinking, and then glycine (125 mM final concentration) was added to arrest crosslinking. DNA was sheared by sonication. The sheared chromatin fragments were immunoprecipitated with an antibody specific for E2F4 (Thermo Scientific), E2F6 (Abcam), or control IgG (Santa Cruz) at 4 uC for 16 h. The antibody-chromatin complexes were precipitated using protein-G-sepharose slurry (Sigma-Aldrich) at 4 uC for 1 h. After dissociation from the immunoprecipitated chromatin, DNA underwent PCR amplification using primers for GAPDH and the putative E2F-binding sites.

GST-tagged protein-protein interaction assay
Full-length open-reading frames of p15 PAF and p21 were cloned to pET-15b (Novogen, Madison, WI, USA). His-tagged proteins were expressed in E.Coli strain BL21 (DE3) (Stratagene) and purified using nickel ion chromatography. pGEX-4T-1 containing the full-length open-reading frame of PCNA was transformed into BL21, and PCNA-GST fusion protein expression was induced using 0.5 mM isopropyl-D-1-thiogalactopyranoside (IPTG) for 4 h at 18 uC. The GST fusion proteins were purified using glutathione-sepharose beads (GE Healthcare, Little Chalfont, UK). His-tagged p21 (25 ng) and different amounts of His-tagged p15 PAF were then incubated with PCNA-GST fusion protein in binding buffer (PBS containing 1% NP-40 and 2 mM DTT) for 3 h at 4 uC. The glutathione-sepharose beads, along with the PCNA-GST fusion protein and binding proteins, were collected by centrifugation (1 min, 16,000 g), washed extensively in binding buffer, and analyzed employing SDS-PAGE and western blotting using anti-p21 antibody.

p15 PAF is expressed in proliferative tissues
To evaluate p15 PAF tissue distribution, we constructed a tissue array containing nontumorous adult tissues of major organs. Immunohistochemical staining revealed p15 PAF nuclear staining in the epidermis, lymph node, endometrium, and small and large intestines, but not in the brain, liver, heart, pancreas, lung, and myometrium (Fig. 1). In the lymph node, p15 PAF was specifically expressed in germinal centers. In the small and large intestines, p15 PAF was predominantly expressed in the lower half of the epithelial crypts, where proliferative cells reside. In the epidermis, p15 PAF was expressed in the suprabasal cells only, which are the major proliferative epidermal cells. This pattern of tissue distribution indicated that p15 PAF expression is restricted to proliferative cells.

p15 PAF is an S-phase protein
To investigate the regulation of p15 PAF expression during the cell cycle, we synchronized HeLa cells at the G1/S boundary using the thymidine/aphidicolin double block method. We used immunofluorescence to determine the percentage of p15 PAFpositive cells. As shown in Figs. 2A and 2B, p15 PAF expression peaked 6 h after release from aphidicolin, and rapidly declined 9 h after release, indicating the predominant expression of p15 PAF during the S phase. Western blotting also showed high p15 PAF expression 3 h and 6 h after release from aphidicolin, and downregulated p15 PAF expression 9 h after release (Fig. 2C). To verify this result, we pulse-labeled HeLa cells using 5-bromo-29deoxy-uridine (BrdU). Immunofluorescence showed that nearly all cells with BrdU staining were also positive for p15 PAF and vice versa (90% concordant rate; Figs. 2D and 2E). These results indicated that p15 PAF is predominantly expressed during the S phase.

p15 PAF is essential for cell cycle progression and DNA synthesis
We obtained 5 different lentiviral constructs carrying p15 PAF shRNA from the National RNAi Core Facility (Academia Sinica, Taipei, Taiwan), and used them to transduce HeLa cells. Cells transduced with shRNAs #4 and #5 displayed marked reductions in p15 PAF mRNA and protein expression (Fig. 3A). We therefore used them in subsequent analyses. Using the MTT assay, we showed that stable knockdown of p15 PAF by shRNAs #4 and #5 led to the inhibition of HeLa cell proliferation (Fig. 3B). To evaluate the effects of p15 PAF knockdown on cell cycle progression, we synchronized HeLa cells at the G1/S boundary using a double  Fig. 3C, the control cells progressed normally through the S phase to the G2 phase, and approximately one-third of cells completed the cell cycle and returned to the G1 phase 9 h after release from thymidine. In contrast, cells with attenuated p15 PAF expression remained at the G1/S boundary for up to 12 h after release. The major event that occurs during the S phase is DNA synthesis; therefore, we investigated the effects of p15 PAF knockdown on DNA synthesis. Our results from the BrdU incorporation assay showed that p15 PAF knockdown markedly inhibited DNA synthesis in asynchronous HeLa cells (Fig. 3D).

p15 PAF knockdown upregulates p21 expression and increases its chromatin binding
The cyclin-dependent kinase inhibitor p21 WAF1/CIP1 (p21) plays important roles in the regulation of cell cycle progression. To elucidate the mechanism by which p15 PAF knockdown arrests the cell cycle, we evaluated p21 expression using western blotting. In unsynchronized HeLa cells, p15 PAF knockdown induced marked upregulation of the p21 protein (Fig. 4A) and mRNA expression (Fig. 4B). These results indicated that regulation of p21 expression is mainly at the transcriptional level. p21 is also a PCNAinteracting protein [11] and inhibits the ability of PCNA to activate DNA polymerase d, the principal replicative DNA polymerase, which inhibits DNA replication [11]. We speculated that p15 PAF might compete with p21 for binding to PCNA, and that p15 PAF attenuation leads to excess interaction between p21 and PCNA. Therefore, we assayed the chromatin-bound fraction of p21 and identified that p15 PAF attenuation resulted in p21 accumulation in chromatin (Fig. 4C). An in vitro GST-tagged protein-protein interaction assay further showed that p15 PAF inhibited the binding of p21 to GST-PCNA fusion protein in a dose-dependent manner (Fig. 4D). The Rb/E2F pathway is the major regulatory mechanism for genes that are required for S-phase entry, such as DNA polymerase subunits, cyclin A, and cyclin E [12]. Therefore, we hypothesized that p15 PAF is a direct transcriptional target of the Rb/E2F pathway. We analyzed the p15 PAF promoter sequence using the TFSEARCH website, identifying 3 putative E2F binding motifs located at 294 to 287, 229 to 222, and 17 to 25 bp of the transcriptional start site (Fig. 5A). HeLa cells have a defective Rb function caused by the binding of human papilloma virus protein E7 to Rb [13]; therefore, we used the breast cancer cell line MCF7 to evaluate Rb/E2F pathway-mediated p15 PAF regulation. As shown in Fig. 5B, following Rb attenuation by siRNA, p15 PAF expression was upregulated. This observation indicated that Rb is a negative regulator of p15 PAF expression. To validate the biological functionality of the 3 putative E2F motifs, we cloned the promoter sequence containing the 3 putative E2F-binding sites into a position upstream of the luciferase reporter gene in the plasmid pGL3-basic. Results from the luciferase assay showed the promoter activity was repressed by Rb (Fig. 5C). Rb also inhibited the promoters of 2 known Rb/E2F targets AURKA and MCL1 [14,15]. To determine which E2F motifs are functional, we mutated each site using site-directed mutagenesis. When analyzing the effects of each mutation, we found that mutations in all 3 sites resulted in increased luciferase activity (Fig. 5D), indicating that all 3 sites are functional and that the Rb/E2F pathway exerts inhibitory effects on the p15 PAF promoter. Rb interacts with several E2F family members. Of these, E2F1-3 are transcriptional activators, whereas E2F4-8 are transcriptional repressors [16]. Cotransfection of E2F4 or E2F6 with the p15 PAF promoter reporter resulted in repressed promoter activity (Fig. 5E). We also performed chromatin immunoprecipitation (ChIP) to confirm the binding of E2F4 and E2F6 with the p15 PAF promoter in vivo (Fig. 5F). Overall, these results support the hypothesis that p15 PAF is a target of the Rb/E2F pathway.

Discussion
p15 PAF is overexpressed in many types of solid tumor [2][3][4][5][6]. Previous studies have identified several functions for p15 PAF , including DNA repair, cell proliferation, and tumor invasion [5][6][7]17]. Using immunohistochemical staining, we found that p15 PAF is specifically expressed in proliferative cells in normal adult organs. Besides, p15 PAF is a PCNA-binding protein, so we propose that it is a component of cell proliferation machinery. In this study, we found that p15 PAF is expressed predominantly in the S phase, and is essential for S-phase progression and DNA synthesis.
The hypothesis that p15 PAF is involved in cell proliferation is not totally new. Consistent with our results, Mizutani et al. showed that p15 PAF knockdown markedly inhibited the growth of anaplastic thyroid cancer cells [18]. However, the distribution of p15 PAF during the cell cycle and its functional role in cell cycle progression remain controversial. Emanuele et al. concluded that p15 PAF expression peaks in the G2/M phase of the cell cycle and declines rapidly at the mitotic exit [19]. In contrast, we observed the predominant expression of p15 PAF during the S phase. We consider the discrepancies to be caused by differences in the interpretation of the results. In the figure 1C of their report, p15 PAF expression remained at high levels from 0 h to 7 h after release from the double thymidine block, and declined 8 h after release, coinciding with the appearance of the G2/M marker phospho-serine 10 on histone 3. This expression pattern is more consistent with that of an S-phase protein than that of a G2/M protein. In our study, we used immunofluorescence and western blotting to evaluate the cell cycle distribution. We identified that p15 PAF expression is highly concordant with BrdU incorporation, indicating that p15 PAF is predominantly expressed during the S phase. Besides, in the study by Emanuele et al., p15 PAF attenuation led to a decreased cell number during the S phase [19]. This observation was consistent with our result that p15 PAF knockdown inhibited S-phase progression. Therefore, although we cannot exclude p15 PAF function during other cell cycle phases, our results indicate that its main function is in promoting S-phase progression.
In this study, we observed that p15 PAF knockdown upregulated p21 expression. p21 protein has 2 major mechanisms for cell cycle arrest. It binds to and inhibits the activity of cyclin-dependent kinases CDK1, 2, and 4, to induce G1 arrest and block entry into the S phase [20]. The p21 is also a PCNA-interacting protein and inhibits the activation of DNA polymerase d by PCNA [11].We observed that p15 PAF knockdown upregulated p21 expression and increased the levels of p21 in the chromatin-bound fraction, indicating that p15 PAF is able to compete with p21 for binding to PCNA. Consistent with our results, Yu et al. observed that p21 overexpression reduced p15 PAF binding to PCNA, and that p15 PAF upregulation inhibited the binding of p21 to PCNA [21]. Therefore, equilibrium between p21 and p15 PAF expression might be an important mechanism in the determination of S-phase entry.
Our study results from luciferase assays, ChIP experiments, and transient transfection assays clearly show that E2F4 and E2F6 repressed transcription from the p15 PAF promoter by binding to the proximal promoter region. These observations supported our hypothesis that p15 PAF expression varies during cell cycle progression. Many E2F target genes, such as p15 PAF , have complex promoter structures that include 2 or more E2F consensus sites [22,23]. In some E2F target genes, the activating and repressing E2Fs bind to the same or different E2F-binding motifs to regulate gene expression [24]. In p15 PAF , we observed that all 3 E2F-binding motifs repressed p15 PAF expression. However, whether each site has a distinct function in the control of p15 PAF expression remains unknown.
In summary, our results indicate that p15 PAF is an S-phase protein tightly regulated by the Rb/E2F complex, and that loss of Rb/E2F-mediated inhibition during the G1/S transition leads to upregulated p15 PAF expression. p15 PAF then competes with p21 for binding to PCNA. Therefore, the presence of p15 PAF is essential for DNA synthesis and S-phase progression.