Oncogene (2006) 25, 5436–5445 & 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc ORIGINAL ARTICLE Aberrant promoter methylation and silencing of the POU2F3 in cervical cancer

Z Zhang1, PC Huettner2, L Nguyen1, M Bidder1, MC Funk1,JLi1 and JS Rader1

1Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, Washington University School of Medicine, St Louis, MO, USA and 2Department of Pathology and Immunology, Washington University School of Medicine, St Louis, MO, USA

POU2F3 (OCT11, Skn-1a)is a keratinocyte-specific POU2F3, also known as Skn-1a and OCT11,isa POU transcription factor whose expression is tied to keratinocyte-specific POU transcription factor squamous epithelial stratification. It is also a candidate expressed in stratified squamous epithelia, including tumor suppressor gene in cervical cancer (CC)because it the epidermis, cervix and foreskin (Andersen et al., 1993; lies in a critical loss of heterozygosity region on11q23.3 in Goldsborough et al., 1993; Yukawa et al., 1993). Its that cancer, and its expression is lost in more than 50% of expression pattern parallels those of the E6 and E7 CC tumors and cell lines. We now report that the loss of oncoproteins of papillomavirus (HPV), showing POU2F3 expression is tied to the hypermethylation of primary expression in differentiating suprabasal cells CpG islands in the POU2F3 promoter. Bisulfite sequen- and very low expression in proliferating basal keratino- cing analysis revealed that methylation of specific CpG cytes (Yukawa et al., 1996). Skn-1a activates sites (À287 to À70 bp)correlated with POU2F3 expres- encoding cytokeratin 10 and SPRR2A, two major sion, which could be reactivated with a demethylating markers for differentiating keratinocytes agent. Combined bisulfite restriction analysis revealed (Andersen et al., 1993; Fischer et al., 1996). hSkn-1a aberrant methylation of the POU2F3 promoter in 18 of 46 contributes to epidermal stratification by promoting (39%)cervical tumors but never in normal epithelium. keratinocyte proliferation and enhancing subsequent POU2F3 expression was downregulated and inversely keratinocyte differentiation (Hildesheim et al., 2001). In correlated with promoter hypermethylation in 10 out of 11 addition, hSkn-1a can stimulate HPV transcription by CC cell lines. Immunohistochemical analysis on a cervical activating E6/E7 promoters (Yukawa et al., 1996; tissue microarray detected POU2F3 protein in the Andersen et al., 1997; Kukimoto and Kanda, 2001). epithelium above the basal layer. As the disease pro- POU2F3 lies in 11q23.3, a region that gressed, expression also decreased, especially in invasive often displays loss of heterozygosity in invasive cervical squamous cell cancer (70% loss). Thus, aberrant DNA cancer (CC) (Zhang et al., 2005). We previously methylation of the CpG island in POU2F3 promoter screened the entire coding region of POU2F3 in CCs appears to play a key role in silencing this gene expression with 11q23.3 deletions and found no mutations (Zhang in human CC. The results suggested that POU2F3 might et al., 2005). However, we found that POU2F3 was be one of the CC-related tumor suppressor genes, which expressed in normal cervix but absent in 50% of CCs are disrupted by both epigenetic and genetic mechanisms. and CC cell lines. Oncogene (2006) 25, 5436–5445. doi:10.1038/sj.onc.1209530; To determine if the downregulation of POU2F3 published online 10 April 2006 results from epigenetic silencing, we evaluated the methylation status of the CpG island in the gene’s Keywords: POU2F3; promoter; hypermethylation; cer- promoter and first exon. We also treated CC cell lines vical cancer with 5-aza-20-deoxycytidine (5-aza-dC) to determine whether the demethylating agent could reactivate POU2F3 expression. To evaluate the clinical relevance of POU2F3 at the various stages of CC, we developed Introduction and used a tissue microarray to examine the expression of POU2F3 protein by immunohistochemical analysis in POU transcription factors help in regulating viral normal cervical tissues, precancerous cervical intrae- transcription, keratinocyte differentiation and other pithelial neoplasias (CIN) and invasive CCs. cellular events (Andersen and Rosenfeld, 2001).

Correspondence: Professor JS Rader, Department of Obstetrics and Results Gynecology, Division of Gynecologic Oncology, Washington Uni- versity School of Medicine, 4911 Barnes-Jewish Hospital Plaza, POU2F3 mRNA expression in cervical cancer cell lines Campus Box 8064, St Louis, MO 63110, USA. E-mail: [email protected] andclinical samples Received 31 August 2005; revised 6 February 2006; accepted 7 February We initially screened a panel of CC cell lines and a series 2006; published online 10 April 2006 of clinical samples for POU2F3 mRNA expression by Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5437 reverse transcription–polymerase chain reaction Methylation status of the POU2F3 CpG islands in (RT–PCR). Expression of POU2F3 was observed in all cervical cancer cell lines normal epithelia, whereas the expression was only In silico promoter and CpG island prediction identified a detected in four of 13 (30.8%) CIN (three of six CIN putative promoter, which locates À1000 to þ 50bp 1, none of three CIN 2 and one of four CIN 3), five of relative to the transcription start site. A typical CpG 17 (29.4%) CCs and five of 11 (45.5%) CC cell lines. island with a CpG observed/expected ratio of >0.60 Figure 1 shows representative expression results of and a GC content of >50% was also identified. This RT–PCR for POU2F3 (Figure 1). island extends from À613 nucleotides to þ 319 nucleo- tides relative to the transcription start site, and it contained 67 CpG sites (Figure 2). Using bisulfite sequencing, we looked for methylation at all 67 CpG sites in the POU2F3 promoter, comparing two cell lines that do not express POU2F3 (Ca Ski and SW756) with two that do (SiHa and ME-180). The results are shown in Figure 2. DNAs from SiHa and ME180 displayed little or no methylation of the CpG sites. In contrast, DNAs from Ca Ski and SW756 were heavily methylated across the CpG sites. Methylation of the 20th to 49th CpG sites (À287 to À70 bp) correlated with POU2F3 expression in these cell lines. We previously provided evidence for allele loss in CC cell lines at 11q23.3, as determined by homozygosity mapping-of-deletions (Zhang et al., 2005). Therefore, we decided to correlate these results with POU2F3 gene expression and the methylation status of the CpG islands in the promoter region (Table 1). The data show a good correlation (Po0.05) of POU2F3 expression by RT–PCR with promoter methylation by combined bisulfite restriction analysis (COBRA) and loss of heterozygosity (LOH) (Zhang et al., 2005) in 10 out of 11 cell lines examined. The lack of expression of POU2F3 with promoter methylation and regional LOH was seen in cell line HeLa, C-4 I, Ca Ski and Figure 1 Analysis of POU2F3 expression by reverse transcrip- SW756. The expression of POU2F3 was seen in cell line tion–polymerase chain reaction (RT–PCR) in RNA samples from normal cervical biopsies and cervical cancer (CC) cell lines (a), DoTc24510, CCI, ME-180 and SiHa, which show from microdissected normal cervical epithelium (Nl Cx Ep) and CC neither promoter methylation nor regional LOH. (b) and from microdissected Nl Cx Ep, cervical intraepithelial neoplasia (CIN) and CC (c). FOU2F3 was expressed in all the normal cervical samples examined, whereas its expression was only Methylation status of the POU2F3 CpG islandin cervical seen in some of the cases from CIN to CC. Case numbers are tissues shown on top of each panel. þ , Positive expression; À,no detectable expression. SiHa serves as an expression-positive As methylation status of the 20th to 49th CpG sites control. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (À287 to À70 bp) correlated with POU2F3 expression in serves as an RNA/cDNA loading and integrity control. four CC cell lines. We investigated the methylation

Figure 2 Proximal POU2F3 gene CpG islands and methylation analysis of the CpG islands in cervical cancer cell lines. (a) CpG islands identified by the CpGplot program. The numbers on the top show the number of bases upstream and downstream from the transcription start site (TSS) (base 0, arrow). The line under the TSS shows the CpG island region and the vertical lines show the 67 CpG sites in the promoter. Restriction enzyme sites for BstUI and Hpy99I are shown as vertical lines above the TSS. (b) Methylation status of POU2F3 CpG islands in four cervical cancer (CC) cell lines. Dark filled circles represent methylated cytosine in a CpG dinucleotide; white open circles, unmethylated cytosine; and striped circles, a mixture of methylated and unmethylated CpG sites.

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5438 Table 1 Summary of POU2F3 expression, methylation and allele loss in cervical cancer cell lines Cell lines POU2F3 mRNA expressiona POU2F3 promoter methylationb Homozygosity at 11q23.3c

HeLa À ++ C-4 I À ++ Ca Ski À ++ SW756 À ++ Hs588d À +ND C-33Ad 7 + À HT-3d ÀÀÀ DoTc24510 + À ND CCI + ÀÀ ME-180 + ÀÀ SiHa + ÀÀ

aPOU2F3 mRNA expression was measured by RT–PCR: + ¼ expression positive; À¼expression negative; 7 ¼ reduced expression. bPOU2F3 promoter methylation detected by COBRA: + ¼ methylation-sensitive restriction enzyme digested positive; À¼methylation-sensitive restriction enzyme digested negative. cHomozygosity at 11q23.3: + ¼ show homozygous for at least three markers; À¼a retention of both alleles for at least three markers; ND ¼ not determined (Zhang et al., 2005). dThese cell lines are HPV negative.

status of this region in cervical samples and other CC cell lines using COBRA. The resulting PCR fragment was digested with BstUI or Hpy99I, which cut the fragment only if the CGCG target sites (BstUI) or CGWCG target sites (Hpy99I) had been retained through bisulfite-mediated deamination. Enzymatic cleavage thus indicates methylation, whereas lack of digestion indicates the absence of methylation. As expected, COBRA confirmed the methylation status of the POU2F3 promoter in the CC cell lines. The PCR Figure 3 Methylation analysis of POU2F3 promoter by combined bisulfite restriction analysis (COBRA). Representative restriction amplicons generated from SW756 and Ca Ski cells were digestion patterns are shown. After the enzymes digestion, completely digested (Figure 3), which is consistent with unmethylated alleles reveal a 218-bp polymerase chain reaction dense methylation from bisulfite sequencing. The SiHa (PCR) fragment, which is indicated by an arrow, and methylated and ME-180 amplicons were totally undigested and alleles show bands of various sizes. The PCR fragment made from Ca Ski and SW756 cells were completely digested, indicating were therefore considered to be predominately un- complete methylation of CpG sites in this region. Partial digestion methylated. Partial digestions of the C-33A and C-4 I was observed with DNAs from cell line C-33A, C-4 I and tumor amplicons were observed, indicating partial methylation 289, 225, 252, 378, 404, 380, 384, 142 and 383, indicating partial in these cells. methylation. Tumor numbers and cell line names are on the top of The aberrant promoter hypermethylation was the panel. M: 100 bp DNA ladder (Seegene, Rockville, MD, USA). Top panel showed the results when digested with BstUI and detected in 39% (18 of 46) of the cancers tested bottom panel showed the results when digested with Hpy99I. by COBRA (Figure 3). The methylation status of POU2F3 did not correlate statistically with clinico- pathological features such as tumor histology, stage or HPV type (P>0.05). The POU2F3 promoter was not of this gene reactivation, we sequenced the 20th–49th methylated in normal epithelium, suggesting that CpG sites of the POU2F3 promoter in the cell lines methylation of the POU2F3 promoter was unique to SW756 and Ca Ski before and after treatment. Bisulfite CC cells (Table 2). DNA sequencing confirmed that 5-aza-dC alone could partially demethylate most of the CpG sites in SW756. The effect of demethylation of 5-aza-dC on the CpG Effect of demethylating agents and histone deacetylase sites in Ca Ski is less frequent than that in SW756. inhibitor on expression of POU2F3 However, in four of these sites (CpG 21st, 22nd, 32nd To determine whether methylation is responsible for the and 44th) demethylation was significant at all concen- decreased level of POU2F3 mRNA in CC cell lines and trations used. Furthermore, as we increased the con- to investigate the role of histone deacetylation in centration of 5-aza-dC, the demethylation of the CpG transcriptional silencing, we treated the cell lines sites did not increase. And combination use of TSA did SW756 and Ca Ski with the demethylating agent not significantly enhance the demethylation effect of 5-aza-dC alone or in combination with trichostatin A 5-aza-dC in both cell line (Figure 5). (TSA), a potent inhibitor of histone deacetylase. Both real-time PCR and RT–PCR analysis revealed that 5-aza-dC could restore the expression of POU2F3 Immunohistochemical pattern of POU2F3 expression transcript at all concentrations used, even as low as in cervical neoplasia 1 mM. The addition of TSA did not increase levels of We prepared a tissue microarray to examine the mRNA further (Figure 4). To examine the mechanism expression of POU2F3 protein in cervical carcinogenesis.

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5439 Table 2 Hypermethylation of the POU2F3 promoter in cervical cancer cell lines and cervical tissue Samples Tumor classification Methylation

Positive Negative Frequency of methylation (%)

Cervical cancers 18 28 18/46 (39%) Histologya Adenocarcinoma 1 4 Adenosquamous 4 5 Squamous 13 19

HPV typea HPV16 7 11 HPV18 4 9 Other 2 5 Multiple 53

Stagea I1625 II 2 3

Lymph node metastasisa Positive 3 6 Negative 10 17 ND 55

CC cell lines 6 5 6/11 (54.5%) Normal cervical epithelium 0 7 0/7 (0%)

ND: not determined. aThe methylation of POU2F3 CpG sites did not correlate statistically with clinicopathological features such as tumor histology, HPV type, stage and lymph node metastasis (P>0.05).

CIN 3 (n ¼ 15), squamous carcinomas (n ¼ 9) and adenocarcinomas (n ¼ 10). Each sample was evaluated for staining intensity, location of the cells that stained and percentage of cells stained. Two to three replicate sections were available for 59% of the cases. Three cases were not evaluated owing to missing epithelium. The intra-specimen correlation per case was excellent. No two specimens within a case differed by more than one intensity grade, and only eight of the cases differed in staining by more than 30% (30% in three cases, and 40– 50% in five cases). In normal epithelium, POU2F3 staining was intense throughout the cytoplasm of cells above the basal cells. In one case, 70% of the cells stained at an intensity of 1. In the remaining cases, 70–100% of the cells stained at Figure 4 Pattern of expression determined by reverse transcrip- an intensity of 2–3. No membrane staining was tion polymerase chain reaction (RT–PCR) of POU2F3 transcript identified. The stroma and basal epithelium were before and after 5-aza-dC or/and trichostatin A (TSA) treatment in negative in all cases. In the samples of koilocytotic SW756 and Ca Ski cervical cancer cells. The expression levels of atypia and CIN 1, 65% of the epithelium stained at an POU2F3 mRNA were normalized to the level of Glyceraldehyde-3- phosphate dehydrogenase mRNA. 5-aza-dC could restore the intensity of 2–3. There was also glandular epithelium on expression of POU2F3 mRNA in both cell lines. There is no these sections, although fewer than 20% of its cells obvious dose-effect relationship between expression level of showed cytoplasmic staining, which varied in intensity. POU2F3 and the amount of 5-aza-dC used. The combination use However, in four of the cases, 100% of the cervical of TSA did not show significant synergetic or antagonized effect on reserve cells, located beneath the glandular epithelium at the expression. H2O was used as a negative control for POU2F3 expression and SiHa as a POU2F3 expression-positive control. the transformation zone, stained at an intensity of 2. In Concentrations of agents are shown on the top of the cell line; (À) the CIN 2 and 3 samples, respectively, 76 and 60% of untreated cells. the epithelium stained at an intensity of 1–2, except for two cases, which had an intensity of 3. Four cases of invasive squamous cancer did not stain, three cases had less than 30% staining at an intensity of 1 and two cases The samples on the array came from 76 different had staining of 80–90% at an intensity of 1–2. The women, and represented normal cervix (n ¼ 15), koilo- quantity and intensity of staining varied widely in the cytotic atypia (n ¼ 7), CIN 1 (n ¼ 10), CIN 2 (n ¼ 10), adenocarcinoma samples. Loss of POU2F3 protein was

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5440 significant in squamous carcinoma compared with Discussion normal cervix (Po0.01). The percentage of cell staining of POU2F3 was significantly decreased in CIN and CC We have shown that expression of POU2F3 mRNA and than in normal cervix (Po0.01), and the staining protein is significantly reduced or absent in both CIN intensity significantly decreased as the disease grade and CC compared to normal cervical epithelium. Loss increased (Po0.01). The results are summarized in of expression correlates with the promoter methylation Table 3. The representative staining sections were shown as assessed by bisulfite sequencing and COBRA in Figure 6. techniques. Treating CC cell lines with micromolar

Figure 5 Effects of 5-aza-dC and trichostatin A (TSA) on POU2F3 promoter CpG island methylation. CpG sites 20th to 49th are shown along the top row. Wide range of CpG sites demethylation was observed after the treatment in SW756. Four CpG dinucleotides 21, 22, 32 and 44 revealed the most significant demethylation after the treatments in Ca Ski cells. Dark filled circles represent methylated CpG sites; white open circles, unmethylated CpG sites; dotted circles, hemimethylated CpG sites with unmethylated CpG dominating; striped circles, hemimethylated CpG sites with methylated CpG dominating.

Table 3 Immunochemistry staining of POU2F3 in cervical tissues Type of epithelium Evaluated Cytoplasm staining Percent cellular Comments cases Number of cases staining** (%)

No staining Intensity 1 Intensity 2–3*

Normal cervix 13 0 1 12 70–100 No staining epithelial membrane; stroma or basal cells

Koilocytotic atypia Squamous 0 0 7 70–100 100% of reserve cells stained and CIN 1 (low grade) (four cases) 11 0 1 50–70 0 3 20–50 Glandular 0 2 0 20–50 6 0 4 5–20

CIN 2 and 3 Squamous 0 6 3 70–100 (high grade) 21 3 4 50–70 2 1 20–50 0 2 5–20 Glandular 0 2 1 20–40 3

Squamous carcinoma*** 9 4 2 0 80–90 30o30

Adenocarcinoma 10 2 1 1 100 2 2 50–70 2 10–25

CIN ¼ cervical intraepithelial neoplasia. *Po0.01, staining intensity decreased as the CIN grade increased compared to normal cervix. **Po0.01, percentage of cellular staining in CIN is significantly decreased compared to normal cervix. ***Po0.01 overall, loss of POU2F3 in squamous carcinoma is significant compared to normal cervix.

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5441 activate upstream genes that are responsible for the activation of POU2F3, and the methylation status of the POU2F3 CpG island is not directly linked to the expression of this gene. The treatment of 5-aza-dC causes a variety of changes in cells, including the decondensation of chromatin and global genomic hypomethylation (Christman, 2002; Suzuki et al., 2002). Alternatively, the pattern of the DNA demethy- lation after 5-aza-dC treatment suggests that DNA methylation of only specific CpG is linked with expression. The reproducible demethylated CpG sites after the treatment in Ca Ski might be important in the transcriptional regulation of POU2F3 in this cell line. The distance between the CpG sites (21st, 22nd, 32nd and 44th) is 133–151 bp, which is around one turn of double-strand DNA (146 bp) wrapped around a nucleo- some. This raises a possibility that methylated CpGs cause the lack of capability for nucleosomal histones to bind to the specific DNA sequence. CpG methylation indeed affects histone–DNA interaction by reducing DNA backbone flexibility and dynamics (Pennings et al., 2005). The most positive evidence of CpG methylation repositioning a nucleosome was found on the promoter sequence of the chicken adult b-globin gene. A strong histone octamer positioning site was largely abolished by CpG methylation. The methyla- tion-sensitive nucleosome position was proposed to act as a switch between methylated and unmethylated overlapping CpG positions, with implications for access to promoter elements (Davey et al., 2004). We did not observe the dose-dependent demethyla- tion of CpG sites of 5-aza-dC treatment on the treated cell lines. This same observation was also reported by a Figure 6 Immunohistochemical staining of POU2F3 in cervical resent microassay study, which revealed that 5-aza-dC samples from the tissue microarray. (a) Strong staining of POU2F3 and TSA effects on gene expression patterns were in the cytoplasm of normal cervical epithelium above the basal cells independent of dose and duration of exposure (Gius ( Â 40). (b) Intermediate staining in cervical intraepithelial neopla- et al., 2004). sias (CIN) 1 ( Â 40). (c) Diminished staining in CIN 3 ( Â 40). (d) Interestingly, we previously identified a somatic Absent POU2F3 expression in invasive squamous cervical cancer (CC) ( Â 40). (e–h) POU2F3 staining in the reserve cells of normal mutation in this region of the promoter in one of 12 and low-grade lesions ( Â 40). CCs. The C>A mutation occurs À221 base pairs (bp) upstream of the transcription initiation site in a predicted Sp1 binding site. (Zhang et al., 2005). Whereas transcription factors that include CpG in their binding concentrations of 5-aza-dC for 5 days demethylated site, such as the CCCTC binding factor (CTCF) and the many of the CpG sites in the promoter and led to CpG-binging protein, are sensitive to the presence of the partially transcriptional reactivation of POU2F3. Add- 5-methyl group in the DNA major groove and are ing TSA to 5-aza-dC treatment did not enhance the prevented from binding to these sites when methylated. degree of demethylation. This is the case even for the ubiquitous transcription Although 5-aza-dC treatment could induce expression factor SP1, previously thought to be insensitive to the of POU2F3 mRNA, many CpG sites remained partially DNA methylation state of a promoter (Pennings et al., methylated, and the combination of TAS subsequent to 2005). This warrants further experiment to investigate 5-aza-dC treatment did not extend the degree of the relationship between SP1 binding site mutation in demethylation seen with 5-aza-dC alone. The possible the promoter region and POU2F3 expression in CC. explanation is that the CpG sites that were demethylated Cell cycle disruption and cellular transformation are following the treatment of 5-aza-dC are critical in known to occur when the HPV viral oncoproteins E6 maintaining the transcriptional activity of POU2F3. and E7 target tumor p53 and pRB, respectively Also, the reactivation of the gene could emanate from (Howley, 1991). However, chromosome complementa- only a small number of alleles, not detected by bisulfite tion experiments suggest that alteration of additional sequencing analysis, which become extensively demethy- tumor suppressor proteins likely contributes to cervical lated following 5-aza-dC treatment (Cameron et al., carcinogenesis. Microcell transfer of a partial copy of 1999). Another possibility is that 5-aza-dC might suppresses tumorigenicity in CC cell

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5442 lines (Saxon et al., 1986; Koi et al., 1989). POU2F3 lies POU transcription factors respond to a wide variety of within a significant region of LOH at 11q23.3 in CC, extracellular and intracellular signals, and they exercise and is therefore a candidate tumor suppressor gene that important regulatory control over the maintenance and could be inactivated by both chromosomal loss (Rader differentiation of cells via specific interactions with et al., 1996; Huettner et al., 1998; O’Sullivan et al., 2001; DNA and other proteins (Andersen and Rosenfeld, Zhang et al., 2005) and aberrant methylation of its 2001). Therefore, POU2F3 may have two distinct roles promoter. Our data in CC cell lines show a good in the cervix: differentiation-dependent regulation of correlation of lack of POU2F3 expression with promo- HPV transcription and the development of invasive ter methylation and LOH. It supported the hypothesis cancer by epigenetic mechanisms. Alternatively, regula- that abnormal promoter hypermethylation can have the tion of viral expression by POU2F3 may affect same effect as a coding-region mutation in one copy of transcriptional co-activators that indirectly affect the gene (the first hit); often, loss of the other copy serves methylation of POU2F3, as it does with E7 binding as the second hit (Herman and Baylin, 2003). Further when the acetylation activity of pCAF is inhibited evidence that POU2F3 functions as a CC tumor (Avvakumov et al., 2003). Blocking pCAF reduces its suppressor gene was provided by Enomoto et al. acetyltransferase activity and alters cellular gene expres- (2004), who showed that growth is repressed and sion such as the differentiation-specific function of pRb differentiation is induced in CC cell lines expressing (Nguyen et al., 2004). exogenous hSkn-1a. In summary, our findings demonstrate that CpG sites Epithelial cancers arise when the normal program of in the POU2F3 promoter are often aberrantly methy- differentiation is disrupted and the tightly regulated gene lated in CCs and that such hypermethylation of the expression program is modified. The innermost basal promoter associates with transcriptional silencing of the layer of the cervix is made of undifferentiated keratino- POU2F3 gene. Thus, epigenetic mechanisms may cytes that have a high proliferation potential. The cells mediate POU2F3 downregulation in CC cells. The committed to terminal differentiation migrate upwards tissue specificity of POU2F3 and its loss of function in where they undergo distinct morphological and structur- more than 40% of squamous CCs make it a target for al changes. In the cervix, most cancers begin at the transcriptional therapy aimed at restoring differentia- transformation zone, where the glandular epithelium tion as a component of cancer treatment. undergoes squamous metaplasia by ingrowth from adjacent squamous epithelium or by squamous differ- entiation of reserve cells. The product of the POU2F3 Materials and methods gene promotes keratinocyte proliferation, leading to stratification. It subsequently influences differentiation Identification of the POU2F3 promoter, CpG island and by upregulating the expression of terminal differentiation transcriptional binding sites genes (Hildesheim et al., 2001). The immunohistochem- We used the complete cDNA sequence of POU2F3 (GenBank ical staining pattern seen in our human tissue microarray Accession No. AF133895) to search the was consistent with the gene’s purported role in vitro. database (May 2004 Assembly) at http://genome.UCSC.edu. POU2F3 was highly expressed in the cytoplasm of the The genomic sequence was located from 119 616 256 to epithelium above the basal cells and reserve cells. We 119 695863 bp (a span of 79 608 bp) on chromosome 11. An also observed loss of POU2F3 expression as the grade of interval of 3 kb upstream and downstream from the transcrip- CIN increased confirming the finding of Hietala et al. tion initiation site was used for in silico promoter region and CpG island identification. The putative promoter region was (1997), who found 20% less cytoplasmic staining in CIN predicted by PromoSer software (http://biosulf.buledu/cgi-bin/ 3 samples than in normal epithelium. However, we zlab/getprom.pl). The CpG plot program (http://www.ebi.a- observed the most marked decrease in expression in c.uk/emboss/cpgplot) was used to identify the CpG island, invasive squamous CC, where cytoplasmic expression in using the following parameters: length >200 bp, C þ G% the epithelium was 70% less than in normal epithelium. >50.00 and observed/expected ratio >0.60. Transcriptional Two characteristic features of HPV infection are binding sites were identified by TESS: Transcription Element tissue specificity and a requirement for advanced Search Software at http://www.cbil.upenn.edu/tess. differentiation of epithelial tissues. Thus, the cell layers above the basal layer support increased levels of viral Cervical tumors and cervical cancer cell lines DNA replication and synthesis of capsid proteins. Skn- A total of 46 snap-frozen, 22-optimum cutting temperature 1a has been linked to the regulation of HPV transcrip- OCT embedded CC biopsies (five adenocarcinomas, nine tion via the activation of E6/E7 promoters (Yukawa adenosquamous carcinomas and 32 squamous cell carcinomas), et al., 1996; Andersen et al., 1997; Kukimoto and 13 normal cervical epithelia, 13 CINs and 11 CC cell lines were Kanda, 2001). However, cellular transformation does analysed for POU2F3 expression or POU2F3 methylation. not occur as a consequence of this productive infection Normal cervix was defined as stratified squamous epithelium because differentiated cells that express the oncoproteins histopathologically absent of cellular atypia. In this study, the normal cervix was from ectocervix. DNAs/RNAs for eight of have lost the ability to proliferate and are sloughed off. the cancers, all normal epithelia and all the CIN specimens In contrast, viral and host changes within basal cells were obtained by laser capture microdissection (LCM, PixCells lead to viral integration and a malignant phenotype. It is IIe LCM System, Arcturus Engineering, Mountain View, CA, unclear at this time whether methylation of the POU2F3 USA). All other tumor DNAs used for methylation analysis promoter relates to the gene’s role in viral regulation. only were sectioned from the OCT-embedded tumor biopsies.

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5443 Cervical cancer cell lines were developed either by JR (CCI) Sodium bisulfite treatment and bisulfite sequencing (Rader et al., 1990) or purchased from American Type Culture Genomic DNA was obtained by digesting cervical tissues with Collection (HeLa, SiHa, Ca Ski, C-4 I, HT-3, ME-180, C-33A, proteinase K. It was extracted with phenol/chloroform and SW756, Hs588.T and DoTc2 4510) (American Type Culture treated with sodium bisulfite to convert unmethylated cytosine Collection (ATCC), Manassas, VA, USA). Cells were grown in to uracil. the media as recommended by the ATCC. One microgram genomic DNA was treated with sodium bisulfite at 501C for 18–20 h, using the CpGENOMEt universal DNA modification kit (Chemicon International Analysis of POU2F3 expression Inc., Temecula, CA, USA) (Li et al., 2005). Bisulfite-treated RNAs were extracted from epithelia using the PicoPuret RNA DNA was amplified by nested PCR, using overlapping primer Isolation Kit (Arcturus Biosciences Inc., Mountain View, CA, sets to amplify both methylated and unmethylated DNA in the USA). RNA from CC cell lines was extracted using TRIZOLs 933 bp CpG island region that extends from the 50-end of the Reagent (Invitrogen, Carlsbad, CA, USA). Reverse transcrip- POU2F3 CpG island into the first exon (Table 4). The tion–PCR was performed with POU2F3-specific primers resulting PCR products include all 67 CpG dinucleotides in the (forward primer, 50-gagccaggaaatgatcgaaa-30; reverse primer, island. PCR was performed in a 10 ml volume containing 0 0 5 -ctggcatttagcccagacat-3 ), which were designed to cross an 1 Â PCR buffer, 1 Â GC-rich buffer, 2.0 mM of MgCl2, 200 nM intron to avoid false-positive results. Amplification of glycer- of dNTPs (Amersham Biosciences Corp., Piscataway, NJ, aldehyde-3-phosphate dehydrogenase (GAPDH) from the same USA), 0.2 mM of each primer and 0.4 U of FastStartTaq cDNA samples provided an internal control of RNA loading. (Roche Diagnostics Co., Indianapolis, IN, USA). The PCR

Table 4 POU2F3 promoter methylation primer sets Primer Primer sequence (50-30) Locationa CpG included Product size (bp)

M1_Outer Forward gggttaggtgagtttggttat –681 to À318 364 Reverse cgaaacacctacctactcca M1_inner Forward agggaggtgaagaattttgt À595 to À436 1st–7th 160 Reverse acccgaacttccttctaaac M2_inner Forward atttcgcgatttgtttagaa À467 to À341 7th–13th 127 Reverse caacctaacaacacctcctc

M3_Outer Forward ggcgagtaggattaggagtt À440 to À63 378 Reverse gaaaccaaaatctccttcct M3_inner Forward tgttaggttgggttggagta À350 to À228 14th–19th 123 Reverse ccaaacccatacaaaacaat M4_inner Forward attgttttgtatgggtttgg À287 to À186 20th–29th 101 Reverse attaccacactccccaaata

M5_Outer Forward attgttttgtatgggtttgg À287 to +106 394 Reverse cgttatccctctccttcac M5_inner Forward gtgtttatttggggagtgtg À210 to À70 30th–49th 141 Reverse aaatctccttcctccaattc M6_inner Forward gaggaaggagattttggttt À83 to +76 50th–58th 160 Reverse cgcaataacacctatcaaaa

M7_Outer Forward gaggaaggagattttggttt À83 to +309 393 Reverse tacaactaccactccccaaa M7_inner Forward taggatggtgaatttggagt +4 to +109 56th–58th 106 Reverse aaacgttatccctctccttc M8_inner Forward ttttgataggtgttattgcg +57 to +210 58th–63th 154 Reverse ccctttaatccccaaatact

M9_Outer Forward cggttagtatttggggatta +186 to +469 284 Reverse cctccatttcacacatatacc M9_inner Forward tttgggataagaaattttgg +217 to +373 64–67th 157 Reverse ataaataaaccacccatccc aThe first nucleotide in exon 1 is numbered as +1, and primer location is based on the position of this nucleotide.

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5444 conditions were: 951C for 3 min, 941C for 30 s, 601C for 30 s CIN 1, 10 CIN 2, 15CIN 3, nine squamous cancers and 10 and 721C for 1 min with annealing temperature decreasing adenocarcinomas. 0.51C every cycle for 14 cycles, and 941C for 30 s, 531C for 30 s, 721C for 1 min for 26 cycles for both reactions. Amplified products were purified using AMPure (Agencourt Bioscience Immunohistochemical localization and grading of POU2F3 Corp., Beverly, MA, USA), and were sequenced directly. DNA Sections (5mm thick) were cut from the tissue microarray. sequencing reactions were performed using BigDyes Termi- Initial sections were stained for H&E to verify histology. nator v3.1 Sequencing Reagents according to the manufac- Owing to the small size of the lesions used in array, not all turer’s instructions and an AB 3100 sequencer (Applied samples were adequate for analysis; 73 samples proved Biosystems, Foster, CA, USA) was employed to examine the adequate with 43 of these available in duplicate or triplicate. DNA fragment. Avidin–biotin staining was used as described previously (Hietala et al., 1997). Combined bisulfite restriction analysis Briefly, POU2F3 antigen was retrieved with Antigen The promoter region encompassing 30 CpG islands (À287 to Retrieval Citra Solution (BioGenex, San Ramon, CA, USA). À70 bp) was amplified and purified as described above. Affinity-purified rabbit polyclonal antibody raised against a Purified PCR products were digested with methylation- peptide mapping to the carboxyl-terminus of rat Skn-1a (Skn- sensitive restriction enzymes (BstUI and Hpy99I; New 1a/i (C-20): sc-330; Santa Cruz Biotechnology Inc., Santa England BioLabs Inc., Beverly, MA, USA). Digested PCR Cruz, CA, USA) was diluted to 1:500. The antibody was products were identified by ethidium bromide staining on 3% incubated with the array section overnight in a humidified 1 agarose gels. The restriction enzyme BstUI cleaves (CGCG) chamber at 4 C. After they had reacted with the primary sequences, although failing to cut (TGTG) sites that result antibody, the sections were treated with a biotinylated anti- from bisulfite conversion of unmethylated CpG sites. Hpy99I rabbit IgG (H þ L) (Vector laboratories, Burlingame, CA, USA) for 30 min. The sections were then developed with three, cleaves (CGWCG) sequences, although failing to cut 0 (TGWTG) sites (W ¼ A or T). Variation in methylation is 3 -diaminobenzidine (DAKO, Carpinteria, CA, USA), coun- indicated by the creation of new or the retention of pre-existing terstained with hematoxylin, dehydrated in graded alcohols restriction sites. and placed under coverslips. Three investigators (PCH, ZZ, MB) independently assessed the patterns and percentages of stained cells in each section of the tissue microarray. Staining Treatment of cell lines with demethylating agents and histone intensity was graded as: 3 ¼ strongly staining, 2 ¼ moderately deacetylase inhibitor staining and 1 ¼ weakly staining. Ca Ski and SW756 cell lines in which the POU2F3 promoter had been methylated were seeded at a density of 1 Â 105 cells/ 100 mm dish and treated with 1, 3, 5or 10 mM 5-aza-dC Statistical methods (dissolved in dimethyl sulfoxide (DMSO)) (Sigma, St Louis, Statistical analyses of tumor characteristics and promoter MO, USA) for 5days. For the combination study, the cells methylation/POU2F3 protein expression were performed were treated for 4 days with 5or 10 mM 5-aza-dC, and 100 nM using the w2 test. Fisher’s exact test was used when the TSA (dissolved in ethanol) (Wako Chemicals USA Inc., expected frequencies were too low for the w2 test to be used Richmond, VA, USA) was added for the last 24 h. The media reliably. and drug additives were replaced daily. DNA and RNA were isolated after 5days of treatment. The control cultures were treated with equal amount of vehicle DMSO and/or ethanol. Acknowledgements

Constructing the tissue microarray This research was supported by National Institutes of Health Hematoxylin- and eosin (H&E)-stained slides from represen- (NIH) Grant CA94141-04. We thank Dr FT Kraus for his tative paraffin-embedded blocks of CC, CIN and normal generous advice on immunohistochemistry. Margo C Funk epithelium were reviewed and the location of the disease was supported by the Howard Hughes Medical Institute. We process was marked by two investigators (PCH and MF). The also thank the Alvin J Siteman Cancer Center at Washington Siteman Cancer Center Tissue Procurement Facility used a University School of Medicine and Barnes-Jewish Hospital in Beecher tissue puncher/array system (Beecher Instruments, St Louis, MO, for the use of the Tissue Procurement Core that Silver Spring, MD, USA) to assemble the array, as described provided the tissue array. The Siteman Cancer Center is previously (Kononen et al., 1998). It contained triplicate supported in part by an NCI Cancer Center Support Grant sections from 15normal cervixes, seven koilocytotic atypia, 10 No. P30 CA91842.

References

Andersen B, Hariri A, Pittelkow MR, Rosenfeld MG. (1997). Enomoto Y, Enomoto K, Kitamura T, Kanda T. (2004). J Biol Chem 272: 15905–15913. Oncogene 23: 5014–5022. Andersen B, Rosenfeld MG. (2001). Endocr Rev 22: 2–35. Fischer DF, Gibbs S, van De Putte P, Backendorf C. (1996). Andersen B, Schonemann MD, Flynn SE, Pearse II RV, Singh Mol Cell Biol 16: 5365–5374. H, Rosenfeld MG. (1993). Science 260: 78–82. Gius D, Cui H, Bradbury CM, Cook J, Smart DK, Zhao S Avvakumov N, Torchia J, Mymryk JS. (2003). Oncogene 22: et al. (2004). Cancer Cell 6: 361–371. 3833–3841. Goldsborough AS, Healy LE, Copeland NG, Gilbert DJ, Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin Jenkins NA, Willison KR et al. (1993). Nucleic Acids Res 21: SB. (1999). Nat Genet 21: 103–107. 127–134. Christman JK. (2002). Oncogene 21: 5483–5495. Herman JG, Baylin SB. (2003). N Engl J Med 349: 2042–2054. Davey CS, Pennings S, Reilly C, Meehan RR, Allan J. (2004). Hietala KA, Kosma VM, Syrjanen KJ, Syrjanen SM, Nucleic Acids Res 32: 4322–4331. Kellokoski JK. (1997). J Pathol 183: 305–310.

Oncogene Hypermethylation of POU2F3 in cervical cancer Z Zhang et al 5445 Hildesheim J, Kuhn U, Yee CL, Foster RA, Yancey KB, Pennings S, Allan J, Davey CS. (2005). Brief Funct Genomic Vogel JC. (2001). J Cell Sci 114: 1913–1923. Proteomic 3: 351–361. Howley PM. (1991). Cancer Res 51: 5019s–5022s. Rader JS, Golub TR, Hudson JB, Patel D, Bedell MA, Huettner PC, Gerhard DS, Li L, Gersell DJ, Dunnigan K, Laimins LA. (1990). Oncogene 5: 571–576. Kamarasova T et al. (1998). Hum Pathol 29: 364–370. Rader JS, Kamarasova T, Huettner PC, Li L, Li Y, Gerhard Koi M, Morita H, Yamada H, Satoh H, Barrett JC, Oshimura DS. (1996). Oncogene 13: 2737–2741. M. (1989). Mol Carcinogen 2: 12–21. Saxon PJ, Srivatsan ES, Stanbridge EJ. (1986). EMBO J 5: Kononen J, Bubendorf L, Kallioniemi A, Barlund M, Schraml 3461–3466. P, Leighton S et al. (1998). Nat Med 4: 844–847. Suzuki H, Gabrielson E, Chen W, Anbazhagan R, van Kukimoto I, Kanda T. (2001). J Virol 75: 9302–9311. Engeland M, Weijenberg MP et al. (2002). Nat Genet 31: Li J, Zhang Z, Bidder M, Funk MC, Nguyen L, Goodfellow 141–149 (E-pub ahead of print 2002 May 6). PJ et al. (2005). Gynecol Oncol 96: 150–158. Yukawa K, Butz K, Yasui T, Kikutani H, Hoppe-Seyler F. Nguyen DX, Baglia LA, Huang SM, Baker CM, McCance DJ. (1996). J Virol 70: 10–16. (2004). EMBO J 23: 1609–1618 (E-pub ahead of print, 2004 Yukawa K, Yasui T, Yamamoto A, Shiku H, Kishimoto T, Mar 25). Kikutani H. (1993). Gene 133: 163–169. O’Sullivan MJ, Rader JS, Gerhard DS, Li Y, Trinkaus KM, Zhang Z, Gerhard DS, Nguyen L, Li J, Traugott A, Huettner Gersell DJ et al. (2001). Hum Pathol 32: 475–478. PC et al. (2005). Genes Cancer 43: 95–103.

Oncogene