CIP2A facilitates the G1/S cell cycle transition via B‐Myb in human papillomavirus 16 oncoprotein E6‐expressing cells

Abstract Infection with high‐risk human papillomaviruses (HR‐HPVs, including HPV‐16, HPV‐18, HPV‐31) plays a central aetiologic role in the development of cervical carcinoma. The transforming properties of HR‐HPVs mainly reside in viral oncoproteins E6 and E7. E6 protein degrades the tumour suppressor p53 and abrogates cell cycle checkpoints. Cancerous inhibitor of protein phosphatase 2A (CIP2A) is an oncoprotein that is involved in the carcinogenesis of many human malignancies. Our previous data showed that CIP2A was overexpressed in cervical cancer. However, the regulation of CIP2A by HPV‐16E6 remains to be elucidated. In this study, we demonstrated that HPV‐16E6 significantly up‐regulated CIP2A mRNA and protein expression in a p53‐degradation‐dependent manner. Knockdown of CIP2A by siRNA inhibited viability and DNA synthesis and caused G1 cell cycle arrest of 16E6‐expressing cells. Knockdown of CIP2A resulted in a significant reduction in the expression of cyclin‐dependent kinase 1 (Cdk1) and Cdk2. Although CIP2A has been reported to stabilize c‐Myc by inhibiting PP2A‐mediated dephosphorylation of c‐Myc, we have presented evidence that the regulation of Cdk1 and Cdk2 by CIP2A is dependent on transcription factor B‐Myb rather than c‐Myc. Taken together, our study reveals the role of CIP2A in abrogating the G1 checkpoint in HPV‐16E6‐expressing cells and helps in understanding the molecular basis of HPV‐induced oncogenesis.

Viral oncogenes have provided significant insights into important biological activities. HPV oncogenes E6 and E7 are consistently expressed in HPV-positive cervical cancers, 5 and the sustained expression of these genes is essential for the maintenance of the transformed state of HPV-positive cells. 6 E6 and E7 proteins promote the degradation of the tumour suppressors p53 and retinoblastoma protein (pRb), respectively, thus modulating multiple biological functions including immortalization of primary cells, transformation of mouse fibroblasts, tumorigenesis in animals, abrogation of cell cycle checkpoints and chromosomal instability. [7][8][9] The ability of highrisk HPV E6 protein to degrade p53 is thought to be a primary mechanism in inducing cellular transformation.
Cancerous inhibitor of PP2A (CIP2A) is an oncoprotein that was first identified as an endogenous physiological inhibitor of tumour suppressor protein phosphatase 2A (PP2A), a serine/threonine phosphatase. 10 CIP2A is believed to execute its oncogenic functions mainly through stabilizing c-Myc by inhibiting PP2A dephosphorylation of c-Myc serine 62 (S62). 10 Various independent studies have found that CIP2A is overexpressed in many types of human carcinomas, including breast, lung, gastric and hepatocellular cancers. In addition to the role of CIP2A in promoting cellular transformation and cancer aggressiveness, CIP2A is also associated with a high tumour grade (for a review see Ref. 11 ). CIP2A is related to a poor patient prognosis and may be applied as a prognosis biomarker in evaluating the risk of tumour metastasis. In addition, it is overexpressed in 70% of most solid malignancies samples, while it is rarely expressed in normal tissues, which makes CIP2A a possible therapeutic target (for a review see Ref. 12 ). Although the oncogenic role of CIP2A in human malignancies has been well elucidated, how it modulates cell proliferation and cell cycle remains largely unknown.
We have recently demonstrated that CIP2A is overexpressed and positively associated with HPV-16E7 in cervical cancer tissues and cells, and the expression of CIP2A is correlated with tumour grade. 13 However, as another important oncoprotein encoded by HPV, how 16E6 protein regulates CIP2A and the role of CIP2A in 16E6-expressing cells remain unclear. In this report, we detected the mRNA and protein expression of CIP2A in 16E6-expressing primary human keratinocytes and explored the role of CIP2A in cell proliferation and G1 checkpoint regulation. We showed that HPV-16E6 protein up-regulated CIP2A by degrading p53 in 16E6-expressing cells and that CIP2A facilitated the G1/S transition by modulating Cdk1 and Cdk2 proteins in a B-Myb-dependent manner.

| Cell culture and plasmids
Primary human keratinocytes (PHKs) were derived from neonatal human foreskins collected from the University of Massachusetts Hospital. PHKs were cultured on mitomycin-treated J2-3T3 mouse fibroblast feeder cells in F-medium plus 5% foetal bovine serum (FBS) with all supplements as described previously. 14

| RNA extraction, RT-PCR and real-time quantitative RT-PCR (qRT-PCR)
Total cellular RNA extraction was performed with the RNeasy mini kit (Qiagen). RNA was then reverse-transcribed into cDNA with an iScript cDNA synthesis kit (Bio-Rad). The cDNA samples were subjected to PCR amplification with the primers described previously 16  b-Actin was used as a control for RNA loading and reverse transcription efficiency.
Quantitative RT-PCR of CIP2A and cell cycle-related proteins were performed as described previously. 13 Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) oligos were used as a control. Primer pairs used for qRT-PCR were as follows:  The relative expression was defined by the comparative Ct value. The grey value of the protein of interest was divided by the grey value of the corresponding loading control to get the relative value of each band. The relative value of the first band of the protein of interest in each image was set as "1," and the relative values of the other bands were divided by the relative value of the first band to obtain a ratio, and the ratio was the "final relative value" of the band. When Western blot experiments were repeated 3 times, we obtained 3 ratios. These ratios were used for the calculation of mean and SD values.

| Small interfering RNA and cell transfection
Chemically modified small interfering RNA (siRNA) oligonucleotides against CIP2A have been described previously, 13  HPV-16E6 siRNA 209, 5 0 -UCCAUAUGCUGUAUGUGAU-3 0 , described in Ref. 18 A scrambled siRNA was synthesized for use as the control.
Retinal pigment epithelium or SiHa cells were seeded in 6-cm dishes to ensure 30%-50% confluence at the time of transfection and cultured in medium without antibiotics for 24 hours. siRNAs were transfected into cells at a final concentration of 20 nmol/L using Lipofectamine RNAiMAX (Invitrogen). For knockdown analysis, cells were harvested 48 hours after transfection, and protein levels were analysed by Western blot. CIP2A was knocked down approximately 70% after siRNA transfection, which was close to the level expressed in the Babe vector cells. For cell cycle analysis, 36 hours after transfection, cells were treated with bleomycin, incubated for an additional 24 hours and harvested.

| Cell viability assay
The cell viability assay was assessed by use of the Cell Counting Kit 8 (CCK-8, Dojindo, Japan) following the manufacturer's protocol.
To evaluate whether B-Myb regulates the transcription of Cdk1 and Cdk2, RPE1 cells were cotransfected with the Cdk1 or Cdk2 promoter-luciferase constructs and renilla luciferase control plasmid together with the B-Myb siRNA plasmid. Cells were harvested after 48 hours, and luciferase activity was measured with the dualluciferase reporter assay system following the manufacturer's recommendations.

| Statistical analysis
All data are expressed as the means and standard deviations (SD).
Statistical analyses were based on nonparametric tests and performed with SPSS for Windows software, version 12.0. In all analyses, P ≤ .05 was deemed significant.

| High-risk HPV E6 up-regulates CIP2A in a p53-degradation-dependent manner
To explore the mechanism by which E6 regulates CIP2A, we utilized retrovirus-mediated successive infection to establish PHKs expressing wild-type HPV-16E6. The ability to promote the degradation of p53 protein has been suggested as a mechanism by which E6 protein from the high-risk HPV types exhibits oncogenic activities, so a 16E6 mutant-F2V (Phe-2 to Val mutation) that does not degrade p53 yet retains 16E6 activities for immortalization of HMECs was constructed as we previously described. 19 The expression of HPV- We then examined CIP2A protein and mRNA levels in PHKs expressing HPV-16E6. We showed that the CIP2A protein level was significantly up-regulated (approximately 2.7-fold; P < .01) by HPV-16E6 ( Figure 1C,D). In addition, the CIP2A mRNA level was elevated approximately 6.7-fold (P < .001) in PHKs expressing 16E6 (Figure 1E). However, there were no significant differences in CIP2A protein or mRNA expression between F2V-expressing PHKs and Babe vector cells, which indicates that the up-regulation of CIP2A by HPV-16E6 is p53-degradation-dependent.

DNA replication in 16E6-expressing cells
It has been reported that CIP2A promotes cell proliferation (for a review see Ref. 12 ). To explore the role of CIP2A in cellular proliferation of 16E6-expressing cells, small interfering RNA (siRNA) specific to CIP2A that was previously shown to effectively down-regulate CIP2A expression was used. 13 We have constructed HPV-16E6expressing RPE1 cells and confirmed the expression of 16E6 in these cells previously. 20

Cdk2 proteins
Cell cycle progression is monitored by several checkpoints, among which the G1 checkpoint is very important, as it determines whether cells enter S phase and initiate DNA replication. To explore how CIP2A modulates the G1 checkpoint, we examined the expression of cell cycle-related proteins including Cdk4, Cdk6, Cdk1, and Cdk2 and their associated cyclin partners-cyclin D1, cyclin A2, cyclin B1 and cyclin E1. In addition to Cdk2, which was proposed to be a major regulator in S phase entry, the protein level of the mitotic Cdk-Cdk1 was also markedly reduced with CIP2A knockdown (P < .05) (Figure 4A,B). The cyclin partner of Cdk1-cyclin B1 decreased significantly as well (P < .05). In contrast, no significant alterations were observed in the expression of structurally and functionally similar Cdk4 and Cdk6 proteins as well as their main partner cyclin D1 after CIP2A silencing. Meanwhile, no decrease in cyclin E1 and cyclin A2 was observed in CIP2A knockdown cells ( Figure 4A,B). We then We performed qRT-PCR to detect the mRNA levels of all cell cycle-related genes. Interestingly, knockdown of CIP2A did not cause significant changes in mRNA levels of these cell cycle regulators (Figure 4C), which suggests that CIP2A does not affect the transcription of these genes and that the regulation of Cdk1 and Cdk2 by CIP2A may be post-transcriptional.

| Regulation of Cdk1 and Cdk2 by CIP2A depends on B-Myb, instead of c-Myc
The transcription factor c-Myc is a basic helix-loop-helix leucine zipper protein that is involved in cell proliferation, cell cycle progression, metabolism and malignant transformation through targeting of downstream genes. 22 We speculated that cell cycle regulation by   To further verify the regulation and function of CIP2A in cervical cancer cells, we introduced HPV-16-positive SiHa cells. HPV-16E6specific siRNA was reported to be effective and specific when transfected into SiHa cells. 18 The decreased 16E6 protein and elevated p53 protein indicated that HPV-16E6 was efficiently knocked down ( Figure 6A). Our data showed that CIP2A was reduced after 16E6 knockdown. However, we noticed that the decrease in CIP2A was not dramatic, and we propose that this is mainly because HPV E7 also has the ability to up-regulate CIP2A. 13 Furthermore, we showed that B-Myb might be a downstream target of CIP2A as CIP2A silencing greatly inhibited the expression of B-Myb in SiHa cells ( Figure 6B). As expected, the levels of Cdk1 and Cdk2 proteins decreased correspondingly. Flow cytometric analysis showed that CIP2A silencing caused more cells to arrest in G1 phase (71.9 AE 5.0% vs 56.9 AE 5.6%, P < .05) and fewer cells to enter S phase (9.46% vs 13.0%) ( Figure 6C,D). Moreover, CIP2A silencing enhanced bleomycin-induced G1 arrest, which showed a more accumulated G1 phase (65.8 AE 8.1% vs 49.7 AE 3.1%, P < .05) in SiHa cells. The data further hint that CIP2A may be involved in the G1 checkpoint by regulating Cdk1 and Cdk2 in a B-Myb-dependent manner in cervical cancer cells.

| DISCUSSION
By retrovirus-mediated overexpressing of HPV-16E6 and E6 mutant-F2V, on the one hand, we showed that 16E6 significantly increased CIP2A mRNA and protein expression in a p53-degradation-dependent manner. On the other hand, knockdown of CIP2A caused a reduction in the expression of p53 in HPV-16E6-expressing cells.
Although the interaction between p53 and CIP2A needs to be further investigated, our finding is in accordance with that of previous studies. One study demonstrated that in human breast cancer tissues, CIP2A mRNA expression was positively correlated with p53 mutations; in addition, in breast cancer mouse models with p53 depletion, CIP2A protein expression was induced. 24 Another study showed that after treatment with doxorubicin, CIP2A expression was increased in MDA-MB231 (mutant p53 breast cancer cell line) cells and decreased in MCF-7 (wild-type p53 breast cancer cell line) cells, which demonstrated that cells with mutant p53 have higher CIP2A expression. 25 Notably, treatment with p53 activator molecules resulted in a reduction in CIP2A expression in breast cancer cells. 26 Taken together, our data and results from these investigations strongly suggest a negative association between p53 and CIP2A proteins, and further research will focus on the underlying molecular basis.
Genomic instability is an important prerequisite and hallmark of carcinogenesis. Cell cycle progression is monitored at several cell cycle checkpoints by regulating the expression and activity of cyclins and Cdks, whose defects contribute to genomic instability. 27 The G1 checkpoint is the first to determine whether cells enter the S phase and initiate DNA replication. The G1 checkpoint is mainly regulated by Cdk2/cyclin E-cyclin A complexes by phosphorylation of pRb. 28 Upon DNA damage caused by exposure to genotoxic agents such as UV or chemicals, p53 is activated via multiple mechanisms and inactivates the Cdk2/cyclin E-cyclin A complexes by activating transcription of the Cdk inhibitor p21, resulting in pRb hypophosphorylation and G1 cell cycle arrest. 29 Because of the central role p53 plays in G1 checkpoint control and the efficient degradation of p53 by highrisk HPV E6 protein, it is reasonable that HPV-16E6 abrogates the G1 cell cycle checkpoint. 30 However, whether CIP2A is involved in the abrogation of the G1 checkpoint in HPV-16E6-expressing cells is unclear. Our data showed that knockdown of CIP2A by siRNA abrogated the ability of 16E6 to bypass the G1 checkpoint, thus proving the function of CIP2A in G1 checkpoint regulation.
Cdk2 was believed to be a major regulator and was essential in S phase entry. However, an increasing number of studies have shown that Cdk1 plays a vital role in the maintenance of cell cycle progression. A previous report showed that Cdk1 can substitute for Cdk2 during the G1/S transition. 31 Moreover, a study of Cdk2/Cdk3/ Cdk4/Cdk6-knockout mice demonstrated that Cdk1 binds all cyclins and drives the whole cell cycle progression. 32 Our data showed that both Cdk2 and Cdk1 function in the G1/S transition and reduced Cdk2 and Cdk1 by CIP2A knockdown lead to G1 arrest. Our results reveal that CIP2A facilitates the G1/S transition by regulating Cdk1 and Cdk2.
c-Myc is a transcriptional factor implicated in control of cell proliferation as well as a variety of cellular regulatory processes. The regulation of c-Myc by HPV-16E6 is conflicting. HPV-16E6 has been reported to activate and up-regulate the expression of c-Myc protein in a p53-dependent manner. 33 However, one study showed that HPV-16E6 did not up-regulate c-Myc, and c-Myc was not required for the activation of telomerase. 34 Moreover, other report showed in vitro degradation of c-Myc protein by 16E6. 35 Therefore, the activation or degradation of c-Myc by 16E6 may be cell-specific or merely reflect the rate of cell proliferation under distinct environmental conditions. Our studies could not find a significant up-regulation of c-Myc in 16E6-expressing PHKs, which is consistent with the results found in HFK (human foreskin keratinocytes, also named PHK). 34 More importantly, CIP2A silencing did not repress either c-Myc or phospho-S62-Myc expression, which indicates that the regulation of the G1 checkpoint by CIP2A is c-Myc-independent.
B-Myb is a transcription factor with specific DNA-binding activity. 36 Previous studies have shown that B-Myb is overexpressed in cervical cancers. 37 Importantly, the expression of B-Myb mRNA peaks during the G1/S phase transition. 38 In addition, overexpression of B-Myb bypassed p53-induced G1 arrest. 39 Notably, B-Myb was reported to bind to the Cdk1 promoter and activate the expression of Cdk1. 23 It is well documented that B-Myb was up-regulated in HPV E7-expressing cells. 40  We demonstrated that CIP2A is up-regulated by HPV-16E6 in a p53-dependent manner. CIP2A is essential to abrogate the G1 cell cycle checkpoint by modulating Cdk1 and Cdk2 expression in HPV-16E6-expressing cells. Furthermore, a luciferase assay showed that the regulation of Cdks by CIP2A in 16E6-expressing cells depends on the transcription factor B-Myb rather than c-Myc. Our results reveal a specific role of CIP2A in HPV-16E6-mediated regulation of G1/S cell cycle progression. Previous study demonstrated that CIP2A protein was specifically expressed in cervical cancer tissues and was undetectable in normal adjacent cervical tissues. 42 Our laboratory found that CIP2A protein was overexpressed in cervical cancer and high grade of cervical intraepithelial neoplasia (CIN), CIN III, which was the key step from the precancerous to invasive cancer, but not in normal cervical, CINI or CINII tissues. 13 This expression pattern indicates that CIP2A has a clinical prominence in the development of cervical cancer and makes CIP2A to be a potential diagnostic biomarker. In addition, CIP2A mutant mouse models showed no obvious abnormality in mouse development and viability. 43 Therefore, the expression specificity and limited toxicity indicate that CIP2A inhibitors may be a hopeful therapeutic strategy for treatment of cervical cancer. However, much effort is needed for translational medicine and clinical application of targeting CIP2A for cervical cancer diagnosis and treatment.

ACKNOWLEDG EMENTS
This work was supported by the National Natural Science Foundation of China (81571986) and Natural Science Foundation of Shandong Province (ZR2015HM084) and supported in part by the National Natural Science Foundation of China (81472552) and Natural Science Foundation of Shandong Province (ZR2014HQ070).

CONF LICT OF I NTEREST
The authors confirm that there are no conflict of interests.