Fax: +886-2-8792-7199.
1097-0215/asset/olbannerleft.jpg?v=1&s=45719cd7de57873027993264fcc568b335a8cd56)
1097-0215/asset/olbannerright.jpg?v=1&s=5e0fba63c1309b3036eb9215a0e1e83dd02efd19)
Early Detection and Diagnosis
You have full text access to this OnlineOpen article
Identification of novel DNA methylation markers in cervical cancer†
Article first published online: 8 APR 2008
DOI: 10.1002/ijc.23519
Copyright © 2008 Wiley-Liss, Inc.
Additional Information
How to Cite
Lai, H.-C., Lin, Y.-W., Huang, T. H., Yan, P., Huang, R.-L., Wang, H.-C., Liu, J., Chan, M. W., Chu, T.-Y., Sun, C.-A., Chang, C.-C. and Yu, M.-H. (2008), Identification of novel DNA methylation markers in cervical cancer. International Journal of Cancer, 123: 161–167. doi: 10.1002/ijc.23519
- †
The application for patents using these DNA sequences in cancer screening is in processing.
Publication History
- Issue published online: 24 APR 2008
- Article first published online: 8 APR 2008
- Manuscript Accepted: 23 JAN 2008
- Manuscript Received: 19 JAN 2008
Funded by
- Tri-Service General Hospital. Grant Number: TSGH-C96-2-S01∼4
- National Science Council, Republic of China. Grant Numbers: NSC95-2622-B-016-001, NSC95-2314-B-016-058-MY2
- C.Y. Chai Foundation for Advancement of Education, Sciences and Medicine
- Human Medical Research Foundation. Grant Number: B962114-1
Keywords:
- cervical cancer;
- epigenetics;
- hpv;
- methylation;
- microarray
Abstract
Testing for DNA methylation has potential in cancer screening. Most previous studies of DNA methylation in cervical cancer used a candidate gene approach. The aim our study was to identify novel genes that are methylated in cervical cancers and to test their potential in clinical applications. We did a differential methylation hybridization using a CpG island (CGI) microarray containing 8640 CGI tags to uncover methylated genes in squamous cell carcinomas (SCC) of the uterine cervix. Pooled DNA from cancer tissues and normal cervical swabs were used for comparison. Methylation-specific polymerase chain reaction, bisulfite sequencing and reverse transcription polymerase chain reaction were used to confirm the methylation status in cell lines, normal cervices (n = 45), low-grade lesions (n = 45), high-grade lesions (HSIL; n = 58) and invasive squamous cell carcinomas (SCC; n = 22 from swabs and n = 109 from tissues). Human papillomavirus (HPV) was detected using reverse line blots. We reported 6 genes (SOX1, PAX1, LMX1A, NKX6-1, WT1 and ONECUT1) more frequently methylated in SCC tissues (81.5, 94.4, 89.9, 80.4, 77.8 and 20.4%, respectively) than in their normal controls (2.2, 0, 6.7, 11.9, 11.1 and 0%, respectively; p < 0.0001). Parallel testing of HPV and PAX1 methylation in cervical swabs confers an improved sensitivity than HPV testing alone (80% vs. 66%) without compromising specificity (63% vs. 64%) for HSIL/SCC. Testing PAX1 methylation marker alone, the specificity for HSIL/SCC is 99%. The analysis of these novel DNA methylations may be a promising approach for the screening of cervical cancers. © 2008 Wiley-Liss, Inc.
In addition to genetic changes, epigenetic alterations such as DNA methylation and histone modifications can result in heritable gene silencing without changes to genetic sequences and are recognized as important causes of cancer.1–3 DNA methylation mostly occurs at the 5′ cytosine in the palindromic sequence, 5′-CpG-3′. CpG islands are CpG-rich areas of ∼1 kb that are usually located in the vicinity of genes, often near the promoters of widely expressed genes.4, 5 Methylation of CpG sites in the human genome is catalyzed by a family of DNA methyltransferases (DNMTs). DNMT1 is a maintenance methyltransferase with a preference for hemimethylated DNA whereas DNMT3a and DNMT3b are de novo methyltransferases with approximately equal preferences for methylated and unmethylated DNA.2, 6 The addition of methyl groups by DNMTs recruits complexes with transcription repressors that modify histones and thus silence genes. Global DNA hypomethylation and site-specific hypermethylation result in genomic instability and transcriptional gene inactivation, respectively, both of which are associated with cancer.7, 8 As epigenetic silencing of tumor suppressor genes by promoter hypermethylation is commonly observed in human cancers, DNA methylation could serve as a marker for early diagnosis of cancer and as a means of assessing the prognosis of cancer patients.3
Cervical cancer is one of the main causes of death of women worldwide.9 Infection with oncogenic human papillomavirus (HPV) is the most significant risk factor in the etiology of cervical cancer and is present in nearly all the cases of cervical cancer.10 Although HPV infection is necessary for the development of cervical cancer, its presence alone is insufficient to cause cervical cancer. The molecular mechanism responsible for the inefficiency of HPV-initiated cervical carcinogenesis remains elusive.11 In addition to genetic changes, epigenetic changes may play a role in the development of cervical cancer,12–16 which indicates that DNA methylation may be useful as a marker for cervical cancer screening.12, 17
Most previous studies of DNA methylation in cervical cancer used candidate gene approaches involving classical tumor suppressor genes or genes that are known to be methylated in other cancers. There is a need for a genome-wide approach to identify hypermethylated genes in invasive cervical cancer. A technique involving inhibition of DNA methyltransferases and histone deacetylases coupled with the use of expression microarrays was recently developed to study cervical cancer cell lines.12 Yet to-date, there are no data on genome-wide analysis of methylation in cervical cancer samples. Elucidation of methylation changes could identify new tumor suppressor genes and biomarkers, which could further our understanding of cervical carcinogenesis and would be useful for cervical cancer screening. We conducted a genome-wide differential analysis of the methylation status of cervical cancer using a CpG island microarray to discover novel genes methylation-silenced in cervical cancer. A panel of cervical scrapings and tissues from patients with pathologically proven cervical lesions was used to test the diagnostic utility of the methylation status of these genes in combination with HPV testing.
Material and Methods
Patients
Patients with normal uterine cervixes (n = 45) and patients with LSIL (n = 45) HSIL (n = 58), and invasive squamous cell carcinoma (SCC; n = 109 with tissue DNA and n = 22 with cervical swab DNA) of the uterine cervix participated in our study. The patients were diagnosed, treated, and tissue banked at theTri-Service General Hospital, Taipei, Taiwan, since 1993, as described previously.18 Cytological, histological and clinical data for all patients were reviewed by a panel of colposcopists, cytologists and pathologists. Patients with LSIL by cytology were followed up without treatment. Patients with HSIL by cytology underwent colposcopic cervical biopsy with subsequent conization or major surgery. The diagnosis of LSIL in the final analysis was done based on cytology. The diagnosis of HSIL in the final analysis includes cervical intraepithelial neoplaisa (CIN) II, III and carcinoma in situ (CIS) according to histological reports. All SCCs were confirmed by histology. Controls were recruited from healthy women who underwent routine Pap screening during the same period. Informed consent was obtained from all patients and control subjects. Exclusion criteria included pregnancy, chronic or acute systemic viral infection, a history of cervical neoplasia, skin or genital warts, an immune-compromised state, the presence of other cancers and past surgery of the uterine cervix. The study was approved by the Institutional Review Board of the Tri-Service General Hospital.
Clinical specimens and preparation of genomic DNA
Genomic DNA was extracted from specimens collected from subjects using an established protocol for tissue banking.18 The concentration of DNA was determined using the PicoGreen fluorescence absorption method, and DNA quality was verified using agarose gel electrophoresis. DNAs from cervical swabs of patients with normal cervixes, LSILs, HSILs and from invasive cervical cancer tumors were used. An independent set of DNA (n = 22) from cervical swabs of SCC patients was also tested.
DMH using CpG island microarrays
Differential methylation hybridization (DMH) was performed according to Yan et al.19 Pooled DNA from 30 cancer tissues and 10 normal cervical swabs were used for comparison. DNA was digested using MseI, ligated to linkers, and sequentially digested with methylation-sensitive restriction enzymes (HpaII and BstUI). The digested linker-ligated DNA was used as a template for polymerase chain reaction (PCR) amplification (20 cycles) and coupled to fluorescence dyes (Cy3: pooled normal cervical sample; Cy5: pooled cervical cancer sample) before hybridizing to the CpG island microarray containing 8640 CpG island tags (University of Toronto). The identity of selected CpG islands (CGIs) was obtained from the CGI database (http://data.microarrays.ca/cpg/). The microarray data were analyzed using the circular-features mode of GenePix 6.0 software. Spots representing repetitive clones were flagged and unacceptable features were removed by filtering. Loci with ratios >2.0 were accepted as hypermethylated in the pooled cervical cancer sample.
Bisulfite modification, methylation-specific PCR (MSP), and bisulfite sequencing
A DNA modification kit (Chemicon, Ternecula, CA) was used according to the manufacturer's recommendations. MSP was done in a total volume of 25 μl containing 1 μl of modified template DNA, 1.5 pmol of each primer (Supplementary 1), 0.2 mmol/L deoxynucleotide triphosphates, and 1 unit of Gold Taq DNA polymerase (Applied Biosystems, Foster City, CA). MSP reactions were subjected to an initial incubation at 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, and annealing at the appropriate temperature for 30 s and at 72°C for 30 s. The final extension was done at 72°C for 5 min. Normal DNA from human peripheral blood was modified with sodium bisulfite and used as an unmethylation control. Normal human DNA was treated with SssI methyltransferase (New England Biolabs, Beverly, MA) to generate a positive control for methylated alleles. Amplification products were visualized on 2.5% agarose gel. All MSP data were derived from at least 2 independent modifications of DNA. The absence of signal in duplicate experiments was scored as negative methylation. Bisulfite-treated genomic DNA was amplified using primers (Supplementary 2) and amplified PCR product was purified and cloned into pCR4-TOPO vectors (Invitrogen, Carlsbad, CA). Bisulfite sequencing was performed on at least 5 individual clones using the 377 automatic sequencer (Applied Biosystems, Foster City, CA).
Reexpression of methylated genes by 5′-aza-2′-deoxycytidine treatment in cancer cell lines
The methylation status of candidate genes was tested in HeLa and CaSki cervical cancer cell lines using MSP. Reexpression of methylated genes in cervical cancer cell lines after treatment with 10 μM of 5′-aza-2′-deoxycytidine (AZC) (Sigma Chemical) for 4 days was assessed by RT-PCR. Total RNA was extracted using a Qiagen RNeasy kit (Qiagen, Valencia, CA). An additional DNase I digestion procedure (Qiagen) was included in the isolation of RNA to remove DNA contamination. One microgram of total RNA from each sample was subject to cDNA synthesis using Superscript II reverse transcriptase and random hexamer (Invitrogen). The cDNA generated was used for PCR amplification with the reagents in the PCR master mix reagents kit (Applied Biosystems) as recommended by the manufacturer. The reactions were carried out in a thermal cycler (GeneAmp 2400 PE, Applied Biosystems). The primers and conditions for the PCR are listed in Supplementary 3.
HPV detection
The presence of HPV DNA in SCC was detected by L1 consensus PCR followed by a reverse line blot previously described.20, 21 In brief, extracted DNA (100 ng) was PCR amplified with biotin-labeled primers that hybridized with HPV (PGMY primers) or β2-microglobulin. An aliquot of the amplified product was visualized by ethidium-bromide staining after agarose (1.5%) gel electrophoresis. The integrity of the extracted DNA and the HPV genotype were confirmed by hybridization with a strip containing probes for 27 HPV types and for the β2-microglobulin control, and visualized with streptavidin and alkaline phosphatase staining. Only high-risk types of HPV were included in the analysis.
Statistical analysis
Data analysis was carried out using statistical package SAS version 9.1. Associations between the methylation of genes and clinical parameters, including HPV status, were analyzed using a χ2 test and Fisher's exact test, wherever appropriate. Age and HPV infection were adjusted using a logistic regression model. The alpha level of statistical significance was set at p = 0.05. The sensitivities and specificities using HPV and methylation markers alone and in parallel or sequential combinations for the diagnosis of cervical lesions were calculated. The sensitivities for SCC and HSIL/SCC were defined as (the number of positive testing results in SCC/the total number of SCC) and (the number of positive testing results in HSIL and SCC/ the total number of HSIL and SCC), respectively. The specificities for SCC and HISL/SCC was defined as (the number of negative testing results in Normal + LSIL + HSIL/the total number of Normal + LSIL + HSIL) and (the number of negative testing results in Normal + LSIL/the total number of Normal + LSIL), respectively.
Results
Identification of methylated genes in invasive squamous cell carcinoma of the cervix
A total of 216 spots shown to be differential methylation between the pooled tumor and pooled normal samples were identified. Clones with repetitive sequences were excluded and reduced the list to 54 unique tags representing 26 promoter CGIs (supplementary 4). MSP primers were generated against these promoters and tested in 2 sets of pooled DNA (30 patients with SCC and 10 patients with normal cervixes; Fig. 1a). Genes that were methylated in pooled SCC samples but not in pooled normal cervix samples were subjected to further testing in individual SCC tumor samples as shown in Figure 1b and these genes validated include: SOX1, PAX1, LMX1A, NKX6-1, WT1 and ONECUT1. Detailed information regarding these 6 genes can be found in Table I.
Figure 1. Validation of genes in clinical samples and cell lines. (a) Pooled DNA samples from 30 squamous cell carcinomas (SCC) and 10 normal cervixes (N) were tested by methylation-specific polymerase chain reaction (MSP) amplification. Peripheral blood lymphocyte (PBL) DNA was used as a negative control (NC) and PBL DNA treated with SssI methylase was used as a positive control (PC). (MSP) products are shown. (b) The MSPs were further validated in a subset of individuals with SCC (T) and normal cervixes (N). Representative MSP results are shown. Key: M, methylation-specific PCR; U, unmethylation-specific PCR. (c) The effect of 5′-aza-2′-deoxycytidine (AZC) treatment was tested on gene expression in the HeLa cervical cancer cell line. The untreated cells were fully methylated (left). In the presence of AZC, demethylation (middle) of the promoter was detected by MSP and the reexpression (right) levels of genes were confirmed by reverse-transcription PCR.
| Symbol | UniGene | Chromosomal location | Name | Known molecular function |
|---|---|---|---|---|
| SOX1 | NM_005986 | 13q34 | Sex determining region Y-box 1 | DNA binding. Transcription factor activity |
| PAX1 | NM_006192 | 20p11.2 | Paired box gene 1 | DNA binding |
| LMX1A | NM_177398 | 1q22-q23 | LIM homeobox transcription factor 1 alpha | Transcription factor activity. Zinc ion binding |
| NKX6-1 | NM_006168 | 4q21.2-q22 | NK6 transcription factor related locus 1 | Transcription factor activity |
| ONECUT1 | NM_004498 | 15q21.1-q21.2 | One cut domain family member 1 | Transcription factor activity Transcriptional activator activity |
| WT1 | NM_024426 | 11p13 | Wilm's tumor 1 | Transcription factor activity. Zinc ion binding |
Association of DNA methylation and gene expression in cervical cancer cell lines
It is important to show that the genes of interest are indeed regulated by DNA methylation. The methylation status of the 6 genes was determined in 2 cervical cancer cell lines. Our data indicated that all 6 genes were methylated in HeLa cells. Four of the 6 genes were methylated in CaSki. As such HeLa cells are the logical choice for our demethylation experiment. We treated the cells with 10 μM AZC for 4 days and examined promoter demethylation and mRNA reexpression of these 6 genes. Indeed this dosage and time course treatment results in promoter demethylation patterns and gene reexpression as evident by MSP, RT-PCR (Fig. 1c) and bisulfite sequencing (BS) results (Figs. 2a and 2b). The patterns of hypermethylation were also confirmed in clinical tissues (Fig. 2c).
Figure 2. Bisulfite sequencing (BS) analysis of 6 genes discovered using a CpG microarray in cell lines and human tissues. a: The CpG plot of each promoter region is shown at the top. “+1” indicates the transcription start site defined by UCSC Genome Browser (http://genome.ucsc.edu/cgi-bin/hgGateway). The black bars indicate sequences analyzed by MSP. The arrowed lines indicate the sequences analyzed by BS. b: BS results are summarized as filled circles representing methylated CpGs and open circles representing unmethylated CpGs. Each line is an independently sequenced clone. Key: Cell line-0, Day 0 after AZC treatment; Cell line-4, Day 4 after AZC treatment; Tumor, DNA from SCC; Normal, DNA from normal cervix. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
Methylation analysis of genes in clinical cervical samples
The mean ages of patients with normal cervix and with LSIL, HSIL and SCC were 51.0 ± 11.3, 39.7 ± 9.6, 46.4 ± 14.4 and 53.3 ± 10.9 years, respectively (p < 0.05). At first, the methylation of novel genes were tested in DNA from normal cervical swabs and cervical cancer tissues (Table II). The methylation rates of all 6 genes (SOX1, PAX1, LMX1A, NKX6-1, WT1 and ONECUT1) between normal and SCC were significantly different (p < 0.0001). Clinical stages and the status of lymph node metastasis of invasive cervical cancer had no effect on methylation frequency. Then, the HPV status and methylation of novel genes were tested in a panel of DNA from swabs of various cervical lesions (Table III and Fig. 3). As expected, the positive rate of carrying high-risk HPV DNA in normal, LSIL, HSIL and SCC was significantly different (p < 0.0001). The methylation rates of all 6 genes (SOX1, PAX1, LMX1A, NKX6-1 and WT1) in normal, LSIL, HSIL and SCC were significantly different (p < 0.0001). There was none in the normal group and only 1 in 44 patients (2.3%) with LSIL were positive of PAX1 methylation. However, the methylation rates of PAX1 in HSIL and SCC were 42.1 and 86.4%, respectively. The methylation rate of SOX1 was low in precancerous lesions, but increased substantially between HSIL and SCC. WT1 exhibited a severity-dependent increase in methylation frequency. The methylation of ONECUT1 could be detected in 3 of the 22 (13.6%) swabs, however, the difference was not statistically different (p > 0.05).
Figure 3. The distribution of methylation frequencies in cervical swabs of normal, LSIL, HSIL and SCC. The detection rates of high-risk HPV were also shown in accordance with disease severities.
| Clinical status/genes | SOX1 | PAX1 | LMX1A | NKX6-1 | WT1 | ONECUT1 |
|---|---|---|---|---|---|---|
| ||||||
| Methylation rate | ||||||
| Normal cervix | 1/45 (2.2) | 0/41 (0) | 3/45 (6.7) | 5/42 (11.9) | 5/45 (11.1) | 0/45 (0) |
| SCC | 88/108 (81.5) | 101/107 (94.4) | 98/109 (89.9) | 86/107 (80.4) | 84/108 (77.8) | 22/108 (20.4) |
| Probability | p < 0.0001 | p < 0.0001 | p < 0.0001 | p < 0.0001 | p < 0.0001 | p < 0.0001 |
| FIGO stage | ||||||
| Stage I (n = 63) | 49/62 (79.0) | 59/62 (95.2) | 58/63 (92.1) | 50/62 (80.7) | 50/62 (80.7) | 14/62 (22.6) |
| Stage II (n = 35) | 28/35 (80.0) | 31/34 (91.2) | 29/35 (82.9) | 28/34 (20.0) | 27/35 (77.1) | 6/35 (17.1) |
| Stage III (n = 11) | 11/11 (100) | 11/11 (100) | 11/11 (100) | 8/11 (72.7) | 7/11 (63.6) | 2/11 (18.2) |
| Probability | NS | NS | NS | NS | NS | NS |
| LN metastasis | ||||||
| Negative (n = 46) | 34/45 (75.6) | 44/46 (95.7) | 41/46 (89.1) | 36/46 (78.3) | 34/45 (75.6) | 8/45 (17.8) |
| Positive (n = 33) | 27/33 (81.8) | 30/32 (93.8) | 28/33 (84.9) | 26/32 (81.3) | 28/33 (84.9) | 6/33 (18.2) |
| Probability | NS | NS | NS | NS | NS | NS |
| Normal (n = 45) | LSIL (n = 45) | HSIL (n = 58) | SCC (n = 22) | p-value | |
|---|---|---|---|---|---|
| SOX1 | 1/45 (2.2) | 2/45 (4.4) | 5/54 (9.3) | 15/22 (68.2) | p < 0.0001 |
| PAX1 | 0/41 (0) | 1/44 (2.3) | 24/57 (42.1) | 19/22 (86.4) | p < 0.0001 |
| LMX1A | 3/45 (6.7) | 6/45 (13.3) | 8/50 (16.0) | 8/22 (36.4) | p < 0.0001 |
| NKX6-1 | 5/42 (11.9) | 24/45 (53.3) | 27/49 (55.1) | 14/22 (63.6) | p < 0.0001 |
| WT1 | 5/45 (11.1) | 9/45 (20.0) | 24/57 (42.1) | 17/22 (77.3) | p < 0.0001 |
| ONECUT1 | 0/45 (0) | 3/45 (6.7) | 4/54 (7.4) | 3/22 (13.6) | p > 0.05 |
| HPV | 10/43 (23.2) | 22/45 (48.9) | 35/58 (60.3) | 18/22 (81.8) | p < 0.0001 |
Diagnostic performance of DNA methylation markers
The sensitivities and specificities of HPV and DNA methylations, alone, in parallel or in sequential combinations, were determined to assess their usefulness as biomarkers for the diagnosis or screening of patients with HSIL and SCC (Table IV). The sensitivity and specificity for the diagnosis of SCC using HPV testing were 82 and 54%, respectively. The sensitivity and specificity for the diagnosis of HSIL/SCC using HPV testing were 66 and 64%, respectively. PAX1 conferred the best performance with sensitivities of 86% and specificities of 82% for SCC and with sensitivities of 54% and specificities of 99% for HSIL/SCC.
| SCC | HSIL/SCC | |||
|---|---|---|---|---|
| Sensitivity | Specificity | Sensitivity | Specificity | |
| ||||
| HPV | 82 | 54 | 66 | 64 |
| SOX1 | 68 | 94 | 26 | 97 |
| SOX/HPV1 | 95 | 53 | 71 | 61 |
| SOX/HPV2 | 55 | 96 | 21 | 99 |
| PAX1 | 86 | 82 | 54 | 99 |
| PAX1/HPV1 | 95 | 48 | 80 | 63 |
| PAX1/HPV2 | 73 | 88 | 41 | 99 |
| LMX1A | 36 | 88 | 22 | 90 |
| LMX1A/HPV1 | 91 | 47 | 76 | 57 |
| LMX1A/HPV2 | 27 | 94 | 15 | 97 |
| NKX6-1 | 64 | 59 | 58 | 67 |
| NKX6-1/HPV1 | 95 | 30 | 90 | 41 |
| NKX6-1/HPV2 | 50 | 81 | 38 | 90 |
| WT1 | 77 | 74 | 52 | 84 |
| WT1/HPV1 | 100 | 41 | 84 | 52 |
| WT1/HPV2 | 59 | 88 | 34 | 95 |
When combined parallel testing (CPT) was applied for HPV and each methylation marker, which means that either one being positive was counted as positive, the sensitivities and specificities for SCC were in the ranges of 91–100% and 30–53%, respectively. When combined sequential testing (CST) was applied for HPV and each methylation marker, which means testing for HPV first with methylation detection following for HPV (+) patients, the sensitivities and specificities for SCC were in the ranges of 27–73% and 81–96%, respectively. In the analysis for HSIL/SCC together, sensitivities and specificities were in the ranges of 71–90% and 41–63%, respectively in the CPT mode. Sensitivities and specificities were in the ranges of 15–41% and 90–99%, respectively in the CST mode. The CPT mode of PAX1/HPV testing confers a better sensitivity of detecting HSIL/SCC than HPV testing alone without compromising the specificity.
Discussion
The cancer epigenome has potential in clinical diagnosis and therapeutic intervention. We used a CpG island microarray and DNA amplicons derived from clinical specimens to identify genes that are hypermethylated in cervical cancer. This approach is shown to be useful in identifying the differentially methylated genes.19, 22, 23 This report is the first to identify novel methylated genes in primary cervical tissues of various degrees of precancerous lesions and malignancy, and showed evidence that methylation markers may be useful in the screening of cervical cancer.
All of the 6 methylated genes are transcription factors important in the developmental process. SOX1, PAX1, LMX1A, NKX6-1 and WT1 are important for brain, roof plate, limb, pancreatic islet and urogenital development, respectively.24–29 ONECUT1 is important for liver and pancreas gene expression.30 Little is known about the roles of these transcription factors in cancer. It was recently reported that expression of SOX1 attenuates the tumorigenic potential of neuronal precursors after neural stem cell transplantation.31 Although there is no published evidence linking PAX1 to malignancies, it has been proposed that some members of the PAX family are oncogenic.29 Methylation of LMX1A was observed in a colon cancer cell line (HCT-116), and was demethylated in the DKO cell line genetically disrupted in DNMT1 and DNMT3b.32 A recent report using a CpG island microarray showed that NKX6-1 is methylated in some subtypes of lymphoma.23 WT1 was initially identified as a tumor suppressor gene involved in the development of Wilm's tumors and its oncogenic properties in various cancers have recently been reported.33 Down regulation of ONECUT1, also known as hepatocyte nuclear factor 6 alpha (HNF6a), was recently associated with the progression of hepatoma.34 These reports indicate that the transcription factors identified in our study have potential tumor suppressive effects. Elucidation of the roles of these transcription factors in the carcinogenesis process in cervical cancer as well as in other cancers may further the understanding of the association between embryonic development and cancer biology. However, every genome-wide approach has its complexity. The methylated promoters detected in our study may be affected by many factors such as the types of microarray, the number and types of restriction enzymes, the stringency of array conditions and the selection criteria. Many other genes may be discovered using similar approaches or different selection criteria. Our study did not identify other classical tumor suppressor genes.
From the clinical perspective, the introduction of the Pap smear in the 1940s has greatly reduced the mortality of invasive cervical cancer.35–37 However, the sensitivity of Pap smear is low, and varies substantially between areas with different screening infrastructures.38 The HPV test is used for cervical cancer screening because infection with oncogenic genotypes of HPV is necessary for cervical carcinogenesis.39 A recent comprehensive analysis of studies on HPV testing in primary cervical cancer screening showed that HPV testing is more sensitive than cytology but not as specific for the identification of women with CIN2 or worse lesions.38 The specificity of HPV testing varies substantially between populations. Indeed, the transient nature of most HPV infections may cause anxiety in patients who test positive for HPV infection and results in unnecessary referrals. As such, there leaves room for the development of other useful molecular biomarkers to triage HPV testing in the screening of cervical cancer.
Alternative strategies for molecular screening such as the use of a higher cutoff value for HPV viral load, mRNA expression of E6/E7 oncogenes and the expression of p16 have been proposed and are being tested.40–42 The use of DNA methylation as a molecular marker for cervical cancer screening is being researched. Herein we identified transcription factors that are silenced by promoter methylation in invasive cervical cancer and high-grade lesions but are hardly if ever silenced in low-grade lesions and normal uterine cervixes. The methylation status of these genes may offer high sensitivities or specificities as a tool a country may take according to individual concerns. In a developed country or an area of low incidence of cervical cancer, a screening test may need higher sensitivity despite of a little compromise of specificity, suggesting that a combined parallel testing of HPV and methylation markers may be an alternative. However, in a developing country, a more specific screening method may be the choice to avoid the resource-demanding management of false positive patients, suggesting that testing methylation markers alone or a combined sequential testing of HPV and methylation markers may be a more logical approach. For example, in a country with with HSIL/SCC incidence of 50/100,000 and 30% HPV detection rate in female population, high-risk HPV testing in 100,000 women may result in 30,000 positive cases needing further investigation. Testing of a methylation marker such as PAX1 (the specificity is 99%) may narrow down the number of high risk women to around 1,000, which greatly reduce the demand of limited medical resources. A latest study reported the high concordance (85–89%) of methylation levels between PCR estimates generated from cervical swabs to that of cervical biopsies.13 In our study, the methylation rates of PAX1 and WT1 from cervical swabs were 89.6% (86.4/94.4) and 99.4% (77.3/77.8) of those from cancer tissues. The methylation rates of SOX1 and NKX6-1 from cervical swabs were 83.7% (68.2/81.5) and 79.1% (63.6/80.4) of those from cancer tissues. The methylation rates of LMX1A and ONECUT1 from cervical swabs were only 40.5% (36.4/89.9) and 66.7% (13.6/20.4) of those from cancer tissues. These results suggested that the concordance may be gene-specific. Reasons for the discordance warrant further investigation. The expected declining in the prevalence of HPV-positive lesions after HPV vaccination will substantially reduce the predictive power of cytology and HPV testing. HPV testing with a cytology triage or a triage of other molecular markers may be a better approach for future cervical cancer screening.43 In addition, attempts have been made to disseminate user-friendly self-sampling methods for collection of cervical material for HPV testing. This may increase the willingness of women to participate in cervical cancer screening, particularly in areas lacking in medical resources.44 The high concordance between the results of HPV tests of self-collected compared to physician-collected cervical-vaginal exfoliated cells implied that self-collected materials for HPV testing plus a methylation triage may be a feasible alternative especially for women living in rural areas. A standard assay for methylation of a specific gene is being discussed towards clinical application of methylated DNA sequences as cancer biomarkers.45 A user friendly method for the analysis of these methylations in combination with a standardized HPV testing may hold promise for the molecular screening of cervical cancer.
In summary, we identified new methylation markers that may hold promise for the molecular screening of cervical cancer. The utility of these methylation markers in conjunction with HPV testing warrants further investigation in different ethnic and geographic backgrounds. Limited by the retrospective hospital-based case-control design in our study, more comprehensive prospective population-based studies using a standardized methylation assay are needed before translating these methylated DNA sequences into practical cancer biomarkers. Many well-established population-based cohorts,38, 46–48 designed for the molecular screening of cervical cancer using HPV, would be suitable for testing the feasibility of this scheme. Identification of more methylation markers in cervical neoplasias, especially adenocarcinoma, may be of great value in achieving better performance.
References
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
- 28
- 29
- 30
- 31
- 32
- 33
- 34
- 35
- 36
- 37
- 38
- 39
- 40
- 41
- 42
- 43
- 44
- 45
- 46
- 47
- 48
Supporting Information
This article contains supplementary material available via the Internet at http://www.interscience.wiley.com/jpages/0020-7136/suppmat .
| Filename | Format | Size | Description |
|---|---|---|---|
| ijc23519-Supplementary_1.xls | 16K | Supporting Information file ijc23519-Supplementary_1.xls | |
| ijc23519-Supplementary_2.xls | 16K | Supporting Information file ijc23519-Supplementary_2.xls | |
| ijc23519-Supplementary_3.xls | 15K | Supporting Information file ijc23519-Supplementary_3.xls | |
| ijc23519-Supplementary_4.xls | 25K | Supporting Information file ijc23519-Supplementary_4.xls |
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.