human papilloma virus (hpv) e7-mediated attenuation of retinoblastoma (rb) induces hpygopus2...

2013;11:19-30. Published OnlineFirst January 2, 2013.Mol Cancer Res

Youlian R. Tzenov, Phillip G. Andrews, Kim Voisey, et al.

Cervical CancerRetinoblastoma (Rb) Induces hPygopus2 Expression via Elf-1 in Human Papilloma Virus (HPV) E7-Mediated Attenuation of

Updated version

10.1158/1541-7786.MCR-12-0510doi:Access the most recent version of this article at:

MaterialSupplementary

http://mcr.aacrjournals.org/content/suppl/2013/06/06/1541-7786.MCR-12-0510.DC1.htmlAccess the most recent supplemental material at:

Cited Articles

http://mcr.aacrjournals.org/content/11/1/19.full.html#ref-list-1This article cites by 49 articles, 18 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

SubscriptionsReprints and

[email protected] order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

[email protected] request permission to re-use all or part of this article, contact the AACR Publications Department at

on August 11, 2014. 2013 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst January 2, 2013; DOI: 10.1158/1541-7786.MCR-12-0510



Cell Cycle and Senescence

Human Papilloma Virus (HPV) E7-Mediated Attenuation ofRetinoblastoma (Rb) Induces hPygopus2Expression via Elf-1in Cervical Cancer

Youlian R. Tzenov1, Phillip G. Andrews1, Kim Voisey3, Paul Popadiuk4, Jieying Xiong1, Catherine Popadiuk2,and Kenneth R. Kao1,3

AbstractThe human papillomavirus (HPV) is the etiologic agent of cervical cancer. In this study, we provide evidence for

the human Pygopus (hPygo)2 gene as a cellular biomarker for HPV-related disease. In a tumor microarray of cervicalcancer progression, hPygo2 levels were greater in high-grade lesions and squamous cell carcinomas than in normalepithelia. Similarly, hPygo2 mRNA and protein levels were greater in HPV-positive cervical cancer cells relative touninfected primary cells. RNA interference (RNAi)-mediated depletion of HPV-E7 increased whereas E74-likefactor (Elf)-1 RNAi decreased association of Retinoblastoma (Rb) tumor suppressor with the hPygo2 promoter incervical cancer cell lines. Transfection of dominant-active Rb inhibited Elf-1-dependent activation of hPygo2,whereas Elf-1 itself increased hPygo2 expression. Chromatin immunoprecipitation assays showed that Rb repressedhPygo2 by inhibiting Elf-1 at the Ets-binding site in the hPygo2 promoter. These results suggested that abrogation ofRb byE7 resulted in derepression of Elf-1, which in turn stimulated expression of hPygo2. Thus, initiation of hPygo2expression by Elf-1 was required for proliferation of cervical cancer cells and its expression therefore may act as asurrogate marker for dysplasia. Mol Cancer Res; 11(1); 1930. 2012 AACR.

IntroductionThe Wnt/b-catenin transcription complex component,

Pygopus, was originally identied as a protein required forWnt-dependent transcriptional activation during embryon-ic development (1). Pygopus is also overexpressed in and isrequired for the growth of a number of cancer cell lines ofdiverse origin (26). Because of its use by malignant cells,Pygo is a candidate biomarker both for diagnostic andpotentially therapeutic benet, but our understanding ofthe mechanisms that augment its expression in disease islimited.An important factor required for hPygo2 expression

in breast and cervical carcinoma cell lines is the E74-likefactor (Elf-1; ref. 7), an E26 transformationspecic(ETS) family member. ETS factors have well-established

roles in carcinogenesis, regulating oncogenes and tumorsuppressors involved in apoptosis, angiogenesis, invasion,and metastasis (8). Elf-1 itself activates genes involved intumorigenesis (912), is overexpressed in several cancers(1316), and correlates with poor prognosis (1720). Innormal resting T lymphocytes, the pocket region of theRetinoblastoma (Rb) tumor suppressor interacts with theN-terminal LXCXE motif of promoter bound Elf-1 andblocks its transactivation (21).Paradigmatic of pathogenically mediated Rb deregulation

is the action of human papillomavirus (HPV) E7 protein, akey etiologic agent in the initiation of cervical cancer (22). E7is expressed following infection of proliferation-competentcervical cells with certain high-riskHPV subtypes. Similar toother viral proteins such as papovavirus T antigen andadenovirus E1A (23), E7 is a 98 amino acid nuclear phos-pho-oncoprotein, which cooperates with the HPV E6 pro-tein to cause immortalization of epithelial cells (24). E7induces degradation of the Rb protein, resulting in activationof several transcriptional regulators such as E2F and Elf-1,which promote cell-cycle progression (25).We are interested in understanding the mechanism of

expression and requirement of hPygo2 in cancer. In theinitiation of cervical carcinoma, HPV genome integration inabnormally proliferating cervical cells raises the probabilitythat they will progress to malignancy because the HPV E2genomic region is spliced out and the HPV E2 regulatoryprotein is no longer produced (26). The loss of E2 causes theconstitutive expression of E6 and E7. E6 recruits the cellularE3 ubiquitin ligase E6-associated protein and forms a

Authors' Afliations: 1Divisionsof Biomedical Science, 2Women'sHealth,Memorial University; 3Immunohistochemistry Lab, Laboratory Medicine,Eastern Health, St. John's, Newfoundland; and 4Dalla Lana School ofPublic Health, University of Toronto, Toronto, Ontario, Canada

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

This research was conducted in partial fulllment of the degree of PhD forY.R. Tzenov.

Corresponding Author:Kenneth R. Kao, Division of Biomedical Sciences,Faculty of Medicine, Memorial University of Newfoundland, 300 PrincePhilip Drive, St. John's, NL, Canada, A1B 3V6. Phone: 709-777-6860; Fax:709-777-7010; E-mail: [email protected]

doi: 10.1158/1541-7786.MCR-12-0510

2012 American Association for Cancer Research.

MolecularCancer

Research

www.aacrjournals.org 19



trimeric complex with p53 (27) leading to proteasomaldegradation (28), thereby preventing cell-cycle arrest and/or inducing apoptosis. The E7 protein, through the LXCXEmotif, binds to the pocket region of Rb and drives itscontinuous proteolytic degradation (29), permitting Elf-1to initiate expression of target genes. We hypothesized thatthe increase in hPygo2 expression required for growth oftransformed cervical cells may be the result of Elf-1 activa-tion caused by HPV E7-mediated attenuation of Rb.In this study, we provide evidence that hPygo2 is over-

expressed in cervical cancer and is required for growth oftransformed HPV-infected human endo- and ectocervicalcells. Our data indicated that hPygo2 overexpression incervical cancer is augmented by abrogation of Rb functionby E7 leading to derepression of Elf-1, suggesting a mech-anistic link between HPV and the cellular response of theoncogenic hPygo2 transcription factor.

Materials and MethodsTissue microarray and immunostainingCervical tumor microarrays (Cybridi, Lot No. CC10-11-

004 (192-194) and US Biomax, Inc., (Cat No. CIN481;Supplementary Table S1) were stained using a VentanaBenchmark Ultra automated clinical immunostainer opti-mized for each of the antibodies as provided in Supplemen-tary Table S2 (staining protocols as per manufacturer'sinstruction or as indicated in Supplementary Table S2).Cells were processed and stained for immunouorescence aspreviously described (3).ImageJ (Rasband, W.S., ImageJ, U.S. NIH, Bethesda,

MD; http://imagej.nih.gov/ij/, 19972012) counting soft-ware was used to quantify staining intensity in immunohis-tochemical images. Only cells in which nuclei were visiblewere stained and were categorized into negative, weak, andstrong staining (Supplementary Fig. S1). The percentages ofnegative, weak, and strong staining cells were calculated foreach core representing 5 diagnoses of disease progression[normal, CIN 1, CIN 2, CIN 3, and squamous cell carci-noma (SCC)] for 3 observers.

Cell linesNormal human endocervical (HEN; ref. 30) and ecto-

cervical (HEC; ref. 31) primary cells and their HPV16 or18 and cigarette smoke condensate transformed subclones,HEN 16T (32) and HEC 18T (33), were generous giftsfrom Dr. A. Pater. Primary cell lines were maintained inDulbecco's Modied Eagle's Media (DMEM) supplemen-ted with 2.1 mL/mL bovine pituitary extract (Gibco) and0.06 mL/mL EGF (Gibco BRL) and cancer cells in DMEMplus 10%heat-inactivated FBS.Detailed information on celllines is provided in Supplementary Table S4 and Supple-mentary Fig. S4.

RNA extraction, cDNA generation, and quantitativePCRCellular RNA isolation using the Nucleospin RNA II Kit

(Macherey-Nagel) and cDNA generation was previously

described (34). cDNA samples were subjected to real-timequantitative PCR using RT2 SYBR green master mix (SABiosciences; ref. 7) and analyzed by the relative quantitativecomparative threshold cycle (DDCt) method (3). Primersequences are listed in Supplementary Table S5.

Protein extraction, SDS-PAGE, and immunoblottingCellular protein extraction and separation on SDS-PAGE

were transferred and detected as previously described (2).Antibody information is provided in Supplementary TableS2.

RNA interference and rescue assaysFour siRNA oligonucleotides were synthesized by Dhar-

macon, including siPy2-X (which targets the 30-UTRsequence of hPygo2 of endogenous hPygo2 mRNA),siPy2-Z (which targets the coding region of hPygo2), siE7(34), and a nonspecic negative control, siNTC. We havepreviously described the Elf-1 targeting siRNA, siElf-1, (7).Oligonucleotide sequences are listed in SupplementaryTable S5.A total of 3 105 cells per well were seeded in 6-well plates

and forward transfected with siRNA oligos at nal concen-trations of 10 to 25 nmol/L using Lipofectamine RNAiMAX(Invitrogen) as per themanufacturer's instructions. In rescueassays, cells were additionally transfected with 1 mg/well ofpCS2 or pCS2 hPygo2 24 hours after the siRNAtransfection using Lipofectamine with Plus Reagent (Invi-trogen). Cells were harvested 72 hours after seeding.

Fluorescent-activated cell sortingTrypsinized and PBS washed cells were xed in 2%

paraformaldehyde and permeabilized in 90% methanol/1 PBS. After washing, cells were incubated in PBS contain-ing 1 mL of 10 mg/mL propidium iodide and 10 mL of 10mg/mL RNase at 37C for 20 minutes in the dark. Sampleswere analyzed using a FACSCalibur ow cytometer andgraphically displayed using ModFit analysis software.

Chromatin immunoprecipitationChromatin immunoprecipitation (ChIP) assays were con-

ducted as previously described (7). Cross-linked cells wereand sonicated to produce 500 bp genomic DNA fragments.About 300 mg of precleared chromatin were immunopreci-pitated with 2 mg of anti-Rb and anti-Elf-1 antibodies.Normal rabbit and mouse IgGs were used as negativeimmunoprecipitation controls. After washing, reversingcross-links and purifying DNA, hPygo2, CCNA, andHCCS1 promoters were amplied by qPCR (sequencesgiven in Supplementary Table S5). Promoter occupancywas calculated as a percentage relative to input chromatinlevels.

PlasmidspECE-PSM-Rb was generously provided by Dr. Brenda

Gallie (35). The pECE-DKPN vector control for pECE-PSM-Rb was generated by digesting pECE-PSM-Rb withKpnI and re-ligating to remove the transcriptional start site.

Tzenov et al.

Mol Cancer Res; 11(1) January 2013 Molecular Cancer Research20



pCS2 Elf-1 (7) contains the human Elf-1 cDNA. pCS2was used as the empty vector control for Elf-1 plasmids.hPygo2 promoter constructs: pGL3-1494 antisense, pGL3-1494, pGL3-48, and pGL3-basic were used previously (7).Site-directed mutagenesis to generate the dominant-active(DA) Rb bindingdecient Elf-1 mutant, pCS2DA-Elf-1, using sequences inWang and colleagues (21), and the Ets-binding site mutant, pGL3-48 mutEBS was conductedusing the QuickChange Kit (Stratagene; primers listed inSupplementary Table S5). Plasmids were veried bysequencing.

Transient transfectionsCells were seeded 24 hours before transfection using

Lipofectamine and Plus reagent in 6 well plates (3.0 105 cells/well) using 1 mg of DNA (per well) or 15-cm plates(7.5 106 cells) using 5 mg of DNA. Three types oftransfections were conducted: (i) 2 expression plasmids(pECE-PSM-Rb, pCS2 DA-Elf-1, pCS2 Elf-1) andtheir empty plasmid counterparts (0.5 mg of each per well);(ii) one luciferase reporter plasmid (pGL3-1494 antisense,pGL3-1494, pGL3-48, pGL3-48 mutEBS, or pGL3-basic)with a plasmid expressing b-galactosidase, pRSV-bgal (0.5mg of each per well or 2.5 mg of each per plate); and (iii) 2expression plasmids (0.25 mg of each per well), one luciferasereporter plasmid (0.25 mg per well) and pRSV-bgal (0.25 mgper well). Luciferase assays were conducted and normalizedto b-galactosidase activity as previously described (7).

Image acquisition and analysisFilms were scanned at a resolution of 600 dpi, set to

grayscale, cropped, and used to make gures. No softwareadjustments were made. Densitometry was conducted onscanned lms to quantify relative protein levels using ImageJ(NIH) software.

ResultshPygo2 protein is overexpressed in CIN3 and SCCThe relative expression of hPygo2 in HPV-infected cer-

vical tissues was assessed using a microarray of tissue coresamples representing various stages of disease progression,stained by immunohistochemistry, with antibodies againstHPV L1, hPygo2, PCNA, and Elf-1.HPVproteinwas detected in normal ectocervical as well as

dysplastic and malignant tissue. The staining intensity usingthis antibody was higher in nondiseased ectocervical but notin endocervical epithelia (Supplementary Fig. S2A; ref. 36).The proportion of HPV-positive cells did not varysignicantly.hPygo2 protein was detected in both normal and path-

ologic tissues, with the intensity of staining appreciablyhigher in cervical intraepithelial neoplasia grade 3 (CIN3)and SCC cores. The proportion of positively stained cellsincreased signicantly from 7% in normal tissues to 93% inseverely dysplastic (CIN3) and malignant tissues. The cel-lular localization of hPygo2 was either solely cytoplasmic orboth cytoplasmic and nuclear (Supplementary Fig. S2B).

Both the expression and staining intensity of the prolif-eration marker proliferating cell nuclear antigen (PCNA)(37) signicantly increased with severity of disease, increas-ing from approximately 20% in normal epithelia and 41.5%in mild dysplasia (CIN1) to 81% and 76% in CIN3 andSCCs, respectively. In positive cores, PCNA protein wasprimarily localized to the nucleus (Supplementary Fig. S2B).Elf-1 protein was detected in all tissues and tumors.While

staining intensity clearly decreased with disease severity, theproportion of positively stained cells remained high in allepithelia (above 40%) with the exception of normal endo-cervical cores (24%). As expected, Elf-1 was localized to boththe cytoplasm and nucleus (13).The relatively higher levels of hPygo2 in CIN3 and SCCs

prompted further examination.We compared the expressionof hPygo2 and 2 established diagnostic markers for cervicalneoplasia (p16 and Ki-67; ref. 38), in a separate cervicalintraepithelial neoplasia tissue array, by immunohistochem-istry. The numbers of negative, weak, and strongly stainingcells were determined (Fig. 1 and Supplementary Figs. S1and S3 and Supplementary Table S3).The expression of Ki-67 increased progressively with the

highest intensity (P

expected, Rb tumor suppressor levels were lower in thetumorigenic cell lines. The protein levels of Elf-1 and PCNAclosely paralleled mRNA levels (Fig. 2B). The 80 kDa(cytoplasmic) and 98 kDa (nuclear) are the predominantforms of Elf-1. Henceforth, we focus on the nuclear isoformas it is the only isoform capable of binding to the hPygo2promoter. hPygo2 is detected as 2 differentially migratingbands on gel electrophoresis, possibly identifying 2 isoforms,which were observed using a variety of antibodies (data notshown). Both forms also paralleled the correspondingmRNA levels of expression. While not detectable in HENcells, hPygo2was highly localized to nuclei inHEN16T cellsusing immunouorescence (Fig. 2C and Supplementary Fig.S4). These results suggested that hPygo2mRNA and protein

are higher in HPV-transformed cervical cells relative toprimary cells and that hPygo2 expression may be regulatedat the transcriptional level.

hPygo2 and Elf-1 are required for cervical cancer cellproliferationThe higher levels of hPygo2 observed inHPV-infected cells

and tissues suggested that itmay, as in other cancers (26, 39),have an important function in cervical cancer. HEN16T cellswere therefore transfected with 2 siRNAs directed againsthPygo2, siPy2-X, targeting its 30-UTR and siPy2-Z targetingthe coding region to determine the requirement of hPygo2.Either siPy2-Xor siPy2-Z signicantly reduced both isoformsof hPygo2 as compared with the nontarget control siRNA(siNTC). When transfected with siNTC, the proportion ofcells that accumulated in G1 was 33% but was appreciablyhigher at 66% (P< 0.01) after siPy2-X treatment and 55% (P< 0.01) after siPy2-Z treatment (Fig. 3A). The increase of cellsin G1 was accompanied by a signicant decrease in thepercentage of cells in S-phase. HEC 18T cells redistributedin a similar manner after hPygo2 siRNA treatment with thelargest proportion of cells arresting in G1 (Fig. 3B).Associated with G1-dependent cell-cycle arrest is the

activation of the tumor suppressor protein 53 (p53; ref. 40).Indeed, higher levels of p53 and its target gene cyclin-dependent inhibitor 1 (p21) coincided with hPygo2 knock-down-mediated, G1 phase cell-cycle arrest in HEN 16T andHEC18T cells (Fig. 3C andD). Interestingly, the increase inp53 levels was reduced to normal levels by transiently co-transfecting cells with hPygo2, suggesting that elevated levelsof hPygo2 protein are a requirement for the proliferation ofHPV-transformed HEN 16T and HEC 18T cervical cells.Our previous work showed that the oncogenic v-Ets

family transcription factor E74-like factor 1 (Elf-1) boundto the hPygo2 promoter and could promote its expression inMCF-7 breast cancer cells (7). We therefore used RNAi todetermine whether Elf-1 was involved in hPygo2-dependentcervical cancer cell proliferation. siElf-1 increased the pro-portion of HEN 16T cells in G1 by 27% (P < 0.01) relativeto siNTC (Fig. 3E), which was accompanied by a signicantdecrease in the proportion of cells in S-phase. siElf-1 treat-ment of HEC 18T cells had a similar effect, increasing theproportion of cells in G1 by 34% (P < 0.01; Fig. 3F).In both HEN 16T and HEC 18T cells, the reduction of

nuclear Elf-1 protein levels by siElf-1 paralleled a signicantdecrease in hPygo2 protein expression (Fig. 3G and H).Similar to our results with the hPygo2 knockdown, there wasan increase in p53 and p21 expression upon depletion ofnuclear Elf-1. The siElf-1induced increase in p53 wasattenuated back to control levels by overexpression ofhPygo2 (Fig. 3G and H), suggesting that Elf-1mediatedactivation of hPygo2 expression was required for cell-cycleprogression in these cells.

Rb-dependent regulation of hPygo2 expression via Elf-1is deactivated by HPV E7 proteinThe requirement of Elf-1 and hPygo2 in cervical cancer

suggested that they are activated following HPV integration.

100

80

60

40

20

0

Ki67

A

B

C

p16

hPygo2

Norm

alCIN

ICIN

IICIN

III SCC

Norm

alCIN

ICIN

IICIN

III SCC

Norm

alCIN

ICIN

IICIN

III SCC

Negative

Weak

Strong

Negative

Weak

Strong

Negative

Weak

Strong

100

80

60

40

20

0

120100806040200

Figure 1. Quantication of protein expression by immunohistochemistry.A cervical intraepithelial neoplasia tissue microarray was stained for (A)Ki-67, (B) p16, and (C) hPygo2. Proportions of negative, weak, andstrongly positive staining cells in each set of cores representing eachstage of cervical disease progression (horizontal axes) determined by 3observers. Shownare themeansandSDsof eachof thecalculatedvaluesfor each core analyzed. The asterisks over the strongly expressing datafor CIN3 and SCC indicate signicant difference between thesediagnoses and the CIN2 samples.

Tzenov et al.




A critical effector for HPV pathogenesis is the deregulatedexpression and activity of its E7 viral oncoprotein. As E7induces degradation of Rb, we predicted that upregulation ofRb by reducing E7 would be expected to decrease Elf-1activity, leading to hPygo2 downregulation.To assess whether E7 was required for increased Elf-1

activity and subsequently hPygo2 expression, we knockeddown E7 protein using RNAi. Treatment of HEN 16T cellswith an E7-specic siRNA siE7 effectively reduced E7protein levels compared with the nontarget control siRNA(Fig. 4A). The reduction in E7was associated with a decreasein hPygo2, but an increase in Rb protein levels, while notaffecting levels of Elf-1. It is possible, therefore, that Rbattenuated hPygo2 levels by regulating the activity but notthe expression or stability of Elf-1.Elf-1 is an Ets-related transcription factor that binds to a

number of target genes including the hPygo2 promoterwhose expression is augmented in cancer (2, 9, 41, 42). Invitro protein interaction assays showed that Rb represses Elf-1 by binding to its transactivation domain in nondividingcells, attenuating its ability to contribute to the oncogenicphenotype (21). As Rb is amajor degradation target ofHPV-E7, we determined whether knockdown of E7 affected itspresence, along with Elf-1, at the hPygo2 promoter usingchromatin immunoprecipitation (ChIP) assays. We previ-ously examined a 1,568 bp (1,494 to74) segment of thehPygo2 promoter and showed that Elf-1 binds to the48 to25 region which contains an Ets-binding site (7). After

siNTC treatment of HEN 16T cells, the presence of Elf-1and Rb was detected at the hPygo2 promoter and at thepromoters of a known Elf-1 target gene, hepatocellularcarcinoma suppressor 1 (HCCS1; ref. 9; Fig. 4B) and anRb target gene, cyclin A (CCNA; ref. 43; Fig. 4C). Treatingcells with siE7 did not affect Elf-1 binding to the hPygo2or HCCS1 promoters. There was, however, a signicantincrease in the presence of Rb at both the hPygo2 (56%, P0.038) and the CCNA promoters. Thus, E7 reduced theassociation of Rb at the hPygo2 promoter while not affectingElf-1. This nding suggested that activation of hPygo2 byHPV in cervical cancer cells was due to derepression of Elf-1activity by the repressive action of E7 on Rb.To further analyze the derepression of hPygo2 by HPV

E7, we used a gain-of-function mutant of Rb (PSM-Rb;ref. 35), a dominant acting Rb-independent mutant of Elf-1(DA-Elf-1; ref. 21) and wild-type Elf-1 and assessed theireffect on hPygo2 expression. The PSM-Rb construct has 8phosphorylation sites substituted by alanines (Supplemen-tary Fig. S5A), making it refractory to cyclin/CDK com-plexes and thus rendering it constitutively repressive (35).Protein products synthesized using this construct are stillable to bind Elf-1 (Supplementary Fig. S5B). The DA-Elf-1construct has the 3 crucial amino acids in its LXCXE, whichare required for binding to Rb, substituted with RXRXH(Supplementary Fig. S5C), rendering it unable to interactwith Rb (Supplementary Fig. S5B). We hypothesized thatoverexpression of PSM-Rb would reduce the positive effect

Figure 2. Expression analysis of hPygo2 in cervical cancer cell lines. A, expression of hPygo2, PCNA, Elf-1, andRbmRNAwas analyzed by quantitative PCR innormal (HEN and HEC) and malignant cervical cell lines (HEN 16T and HEC 18T). mRNA levels were normalized to glyceraldehyde-3-phosphatedehydrogenase (GAPDH) and set to 2 in HEN cells, adjusted accordingly for the other cell lines and plotted on a logarithmic scale. B, expression levels ofhPygo2, PCNA, Elf-1, and Rb in normal andmalignant cervical cell lines weremeasured using immunoblots. b-Actin was used as a loading control. Molecularweight is expressed in kDa. C, normal HEN and cervical cancer (HEN 16T) cell lines were processed for immunocytochemical (left column) andimmunouorescent (middle and right columns) detection of hPygo2 (brown andgreen, respectively). Cells were additionally co-stainedwith 40,6-diamidino-2-phenylindole (DAPI) and b-catenin. Full-length blots are presented in Supplementary Fig. S6.

HPV E7 Induces hPygo2

www.aacrjournals.org Mol Cancer Res; 11(1) January 2013 23



of E7 on hPygo2 expression, and in contrast, the DA-Elf-1mutant would constitutively activate hPygo2 expression.Our previous work indicated that Elf-1 augmented hPygo2

gene expression by binding to the Ets-binding site (EBS)present within the rst 48 bp of the hPygo2 promoter (7). Toassess the effect of the Rb and Elf-1 constructs on hPygo2gene regulation, we overexpressed different combinations ofPSM-Rb, Elf-1, and DA-Elf-1 in conjunction with 2 report-er constructs, both of which included the EBS but differed inthe length of hPygo2 promoter DNA. These included theminimal hPygo2 promoter containing the rst 48 bp (pGL3

48), a large segment containing 1,494 bp (pGL3-1494), anda negative control with no promoter sequence (pGL3 basic;ref. 7).Overexpression of PSM-Rb in HEN 16T cells led to 2.9-

fold (P 0.037) and 7.1-fold (P 0.038) decreases in thepGL3-48 and pGL3-1494 samples, respectively, comparedwith the empty vector control (Fig. 5A). In contrast, trans-fection with either Elf-1 or DA-Elf-1 signicantly increasedhPygo2 reporter activity compared with the control. Elf-1overexpression caused a 17.6-fold increase (P 0.017) in theminimal promoter reporter and a 3.7-fold increase (P

Figure 3. Depletion of hPygo2 andElf-1 causes activation of p53 andG1 phase cell-cycle arrest. A andB,HEN 16T and HEC 18T cells weretreated with either siNTC orhPygo2 siRNAs. Fluorescent-activated cell sorting (FACS) wasconducted to show the percentageof cells in each phase of the cellcycle. C and D, HEN 16T and HEC18T cells were treated with eithersiNTC or hPygo2 siRNA incombination with either pCS2 orpCS2 hPygo2 expressionvectors. EandF,HEN16TandHEC18T cells were treated with eithersiNTC or Elf-1 siRNA (siElf-1) andsubsequently analyzed byFACS.Gand H, HEN 16T and HEC 18T cellswere treated with either siNTC orsiElf-1 in combination with eitherpCS2 or pCS2 hPygo2. Full-length blots are presented inSupplementary Fig. S6.

Tzenov et al.




0.047) in the large reporter. DA-Elf-1 overexpressionincreased the minimal reporter activity by 24.1-fold (P

human papilloma virus (hpv) e7-mediated attenuation of retinoblastoma (rb) induces hpygopus2...

Documents