- Research article
- Open Access
- Open Peer Review
Discovery of DNA methylation markers in cervical cancer using relaxation ranking
- Maté Ongenaert†1,
- G Bea A Wisman†2Email author,
- Haukeline H Volders2,
- Alice J Koning2,
- Ate GJ van der Zee2,
- Wim van Criekinge1, 3Email author and
- Ed Schuuring4
© Ongenaert et al; licensee BioMed Central Ltd. 2008
- Received: 03 June 2008
- Accepted: 24 November 2008
- Published: 24 November 2008
To discover cancer specific DNA methylation markers, large-scale screening methods are widely used. The pharmacological unmasking expression microarray approach is an elegant method to enrich for genes that are silenced and re-expressed during functional reversal of DNA methylation upon treatment with demethylation agents. However, such experiments are performed in in vitro (cancer) cell lines, mostly with poor relevance when extrapolating to primary cancers. To overcome this problem, we incorporated data from primary cancer samples in the experimental design. A strategy to combine and rank data from these different data sources is essential to minimize the experimental work in the validation steps.
To apply a new relaxation ranking algorithm to enrich DNA methylation markers in cervical cancer.
The application of a new sorting methodology allowed us to sort high-throughput microarray data from both cervical cancer cell lines and primary cervical cancer samples. The performance of the sorting was analyzed in silico. Pathway and gene ontology analysis was performed on the top-selection and gives a strong indication that the ranking methodology is able to enrich towards genes that might be methylated. Terms like regulation of progression through cell cycle, positive regulation of programmed cell death as well as organ development and embryonic development are overrepresented. Combined with the highly enriched number of imprinted and X-chromosome located genes, and increased prevalence of known methylation markers selected from cervical (the highest-ranking known gene is CCNA1) as well as from other cancer types, the use of the ranking algorithm seems to be powerful in enriching towards methylated genes.
Verification of the DNA methylation state of the 10 highest-ranking genes revealed that 7/9 (78%) gene promoters showed DNA methylation in cervical carcinomas. Of these 7 genes, 3 (SST, HTRA3 and NPTX1) are not methylated in normal cervix tissue.
The application of this new relaxation ranking methodology allowed us to significantly enrich towards methylation genes in cancer. This enrichment is both shown in silico and by experimental validation, and revealed novel methylation markers as proof-of-concept that might be useful in early cancer detection in cervical scrapings.
- Cervical Cancer
- Imprint Gene
- Methylation Marker
- Cervical Cancer Cell Line
- Normal Cervix
DNA methylation represents a modification of DNA by addition of a methyl group to a cytosine, also referred to as the fifth base . This epigenetic change does not alter the primary DNA sequence and might contribute to overall genetic stability and maintenance of chromosomal integrity. Consequently, it facilitates the organization of the genome into active and inactive regions with respect to gene transcription . Genes with CpG islands in the promoter region are generally unmethylated in normal tissues. Upon DNA hypermethylation, transcription of the affected genes may be blocked, resulting in gene silencing. In neoplasia, hypermethylation is now considered as one of the important mechanisms resulting in silencing expression of tumour suppressor genes, i.e. genes responsible for control of normal cell differentiation and/or inhibition of cell growth . In many cancers, various markers have been reported to be hypermethylated . The detection of DNA hypermethylation was revolutionized by two discoveries. Bisulfite treatment results in the conversion of cytosine residues into uracil, except the protected methylcytosine residues . Based on the sequence differences after bisulfite treatment, methylated DNA can be distinguished from unmethylated DNA, using methylation specific PCR (MSP) .
In the last few years, hypermethylated biomarkers have been used in cancer research and diagnostics [7–9]. Presently, DNA hypermethylation of only few markers is of clinical relevance . Two classical examples are hypermethylation of MGMT in the prediction of treatment response to temozolomide in glioblastoma  and DNA hypermethylation of GSTP1 in the early detection of prostate cancer . The search for markers that are hypermethylated in specific cancer types resulted in a large list of genes but more recent evidence revealed that many of these markers are methylated in normal tissues as well [12, 13].
To discover novel markers that are specific for certain stages of cancer with a high specificity and sensitivity, large-scale screening methods were developed such as Restriction Landmark Genomic Scanning , Differential Methylation Hybridization [15–17], Illumina GoldenGate® Methylation, microarray-based Integrated Analysis of Methylation by Isoschizomers (MIAMI)  and MeDIP  in combination with methylation-specific oligonucleotide microarray . These approaches demonstrated that large-scale screening techniques have a large potential to find novel methylation targets in a whole range of cancers. To identify cancer related hypermethylated genes, also pharmacological unmasking expression microarray approaches were suited [21–23]. In this approach, the re-activation of gene expression using microarray analysis was studied during functional reversal of DNA methylation and histone acetylation in cancer cell lines using demethylating agents and histone deacetylase inhibitors. This methodology generally results in a list of several hundreds of candidate genes. Although the analysis of the promoter (e.g. screening for dense CpG islands) is used to narrow down the number of candidate genes, the number list is still too large. This methodology has proven relevant as its application resulted in the identification of new potential methylated genes [24, 25].
However, the initial large scale screening approach will also detect many genes that are not directly methylation targets themselves but become re-activated due to the re-expression of for instance transcription factors . Furthermore, in most studies only re-expression data after demethylation in cell lines were used. Smiraglia and co-workers  calculated that more than 57% of the loci methylated in cell lines were never methylated in 114 primary cancers of different malignancy types. The small number of cell lines used to identify methylated genes does not allow to draw conclusions on the relevance of such cancer-specific genes without testing a large series of primary tumours, which is not done in most studies.
Finally, the completion of the sequence of the human genome provided information on genes, promoter gene structure, CG-content and chromosomal localization. These data are useful to define criteria for the candidate genes to act as appropriate targets for DNA methylation.
To identify genes that are downregulated due to promoter hypermethylation and to enrich for those genes that are most frequently involved in cervical cancer, we performed the following experiments:
Affymetrix expression microarray analysis on a panel of frozen tissue samples from 39 human primary cervical cancers to identify cancer specific down-regulated genes
To select those genes that are hypermethylated in cervical cancer, Affymetrix expression microarray analysis on a panel of 4 different cervical cancer cell lines in which the expression of (hyper)methylated genes was re-activated upon treatment with 5-aza-2'deoxycytidine (DAC) (blocking DNA methylation), and/or trichostatin A (TSA) (inhibiting histone deacetylase (HDACi))
Data from both approaches were combined, and a novel non-parametrical ranking and selection method was applied to identify and rank candidate genes. Using in silico promoter analysis we restricted the analysis to those candidate genes that carry CpG islands
To validate whether our new approach resulted in a significant enrichment of hypermethylated genes, we compared the first 3000 high-ranking candidate probes with lists of imprinted genes, X-chromosome located genes and known methylation markers. In addition, to investigate whether the promoters of these selected gene probes are hypermethylated and this methylation is present in cancer and not in normal tissue, we determined the hypermethylation status of the 10 highest ranking candidate genes in both cervical cancers and normal cervices using COBRA (COmbined Bisulfite Restriction Analysis). These data revealed a highly significant enrichment of methylated genes.
Primary cervical tissue samples
For the expression microarray analysis, tissues from 39 early stage frozen cervical cancer samples were used from a collection of primary tumours surgically removed between 1993 and 2003. All patients were asked to participate in our study during their initial visit to the outpatient clinic of the University Medical Centre Groningen (UMCG, Groningen, The Netherlands). Gynaecological examination under general anaesthesia was performed in all cervical cancer patients for staging in accordance with the International Federation of Gynaecology and Obstetrics (FIGO) criteria . Tumour samples were collected after surgery and stored at -80°C. The stage of cervical cancer patients included 33 FIGO stage IB (85%) and 6 FIGO stage IIA (15%). The median age of the cervical cancer patients was 46 years (IQ range 35 – 52 yr.).
For COBRA and BSP (Bisulfite Sequencing PCR), 10 (of the 39) primary cervical cancers and 5 controls (normal cervix) were used. The age-matched normal cervical controls were women without a history of abnormal Pap smears or any form of cancer and planned to undergo a hysterectomy for benign reasons during the same period. Normal cervices were collected after surgery and histologically confirmed.
Informed consent was obtained from all patients participating in this study. The study was approved by the ethics committee of the UMCG.
Cervical cancer cell lines
Four cervical carcinoma cell lines were used: HeLa (cervical adenocarcinoma, HPV18), SiHa (cervical squamous cell carcinoma, HPV16), CSCC-7 (nonkeratinizing large cell cervical squamous cell carcinoma, HPV16) and CC-8 (cervical adenosquamous carcinoma, HPV45). HeLa and SiHa were obtained from the American Tissue Type Collection. CSCC-7 and CC-8  were a kind gift of Prof. GJ Fleuren (Leiden University Medical Centre, Leiden, the Netherlands). All cell lines were cultured in DMEM/Ham's F12 supplemented with 10% fetal calf serum.
Cell lines were treated for 3 days with low to high dose (200 nM, 1 μM or 5 μM) 5-aza-2'deoxycytidine (DAC), 200 nM DAC with 300 nM trichostatin A (TSA) after 48 hours, or left untreated as described previously [21, 23, 30]. Cells were split to low density 24 hours before treatment. Every 24 hours DAC was refreshed. After 72 hours cells were collected for RNA isolation.
RNA and DNA isolation
From the frozen biopsies, four 10-μm-thick sections were cut and used for standard RNA and DNA isolation. After cutting, a 3-μm-thick section was stained with haematoxylin/eosin for histological examination and only tissues with >80% tumour cells were included. Macrodissection was performed to enrich for epithelial cells in all normal cervices.
For DNA isolation, cells and tissue sections were dissolved in lysis buffer and incubated overnight at 55°C. DNA was extracted using standard salt-chloroform extraction and ethanol precipitation for high molecular DNA and dissolved in 250 μl TE-4 buffer (10 mM Tris; 1 mM EDTA (pH 8.0)). For quality control, genomic DNA was amplified in a multiplex PCR containing a control gene primer set resulting in products of 100, 200, 300, 400 and 600 bp according to the BIOMED-2 protocol .
RNA was isolated with TRIzol reagent (Invitrogen, Breda, the Netherlands) according to manufacturer's protocol. RNA was treated with DNAse and purified using the RNeasy mini-kit (Qiagen, Westburg, Leusden, the Netherlands). The quality and quantity of the RNA was determined by Agilent Lab-on-Chip analysis (ServiceXS, Leiden, the Netherlands, http://www.serviceXS.com.
Gene expression for 39 primary cancers and 20 cell line samples (4 cell lines treated with various combinations of DAC/TSA) was performed using the Affymetrix HGU 133 Plus 2.0 array with 54,675 probes for analysis of over 47,000 human transcripts. The labelling of the RNA, the quality control, the microarray hybridization and scanning were performed by ServiceXS according to Affymetrix standards. For labelling, 10 μg of total RNA was amplified by in vitro transcription using T7 RNA polymerase.
Quality of the microarray data was checked using histograms, box plots and a RNA degradation plot. One cell line sample was omitted because of poor quality. Using BioConductor  present (P), absent (A) or marginal (M) calls were determined with the MAS5 algorithm (see Additional file 1 for data). MAS5 uses a non-parametric statistical test (Wilcoxon signed rank test) that assesses whether significantly more perfect matches show more hybridization signal than their corresponding mismatches to produce the detection call for each probe set . The relaxation ranking approach only relied on P-calls. Some samples were analyzed in duplicate, and the profile of P-calls is highly similar (93–95% of the probesets have an identical P/M/A call).
Relaxation ranking algorithm
In order to identify the most promising markers that are methylated in cervical cancer, we assumed that such markers should be silenced in cancer cells and upregulated upon re-activation after DAC/TSA treatment, Therefore, the ideal methylation markers will be genes represented by probes with:
no expression in primary cervical cancers: P-calls = 0 out of 39 cancers
no expression in (untreated) cervical cancer cell lines: P-calls = 0 out of 4 cell lines
expression in cervical cancer cell lines treated with DAC (or DAC in combination with TSA): P-calls = 15 out of 15 treated cell lines
To select for those gene probes that would be the best candidate hypermethylated genes in cervical cancer, we present the relaxation ranking algorithm. Probesets were ranked, not primarily based on the number of P-calls and thus explicitly setting thresholds, but primarily driven by the number of probesets that would be picked up, based on selection criteria (the number of P-calls in primary cancers, untreated and treated cell lines). The stricter (e.g. P-calls: 0-0-15) these selection criteria, the lower the number of probes that meet with these criteria; while if the conditions become more and more relaxed (higher number of P-calls in primary cancers and untreated cell lines, and lower number of P-calls in treated cell lines), the more probes will comply. In the end, using P-calls: 39-4-0 as criteria, all probe sets were returned. This way, there was no need to define a 'prior' threshold for the number of P-calls.
The following sorting method was applied (R-scripts are presented in Additional file 2):
All possible conditions were generated and the number of probes that were picked up under these conditions was calculated:
the number of samples with expression (P) of a certain probe in
primary cervical cancer samples is called xsample
cervical cancer cell lines is called ysample
treated cervical cancer cell lines is called zsample
all combinations of x, y and z are made
x (the number of P-calls in primary cancers) varies from 0 to 39
y (the number of P-calls in untreated cell lines) from 0 to 4
z (the number of P-calls in treated cell lines) from 0 to 15
In total, 3200 combinations of x, y and z can be made
a probeset was found under each of these generated conditions x, y and z if:
xsample ≤ x (number of P-calls for probe in primary cancers smaller or equal compared to condition) AND
ysample ≤ y (number of P-calls for probe in untreated cell lines smaller or equal compared to condition) AND
zsample ≥ z (number of P-calls for probe in treated cell lines larger or equal compared to condition)
under very strict conditions (x = 0, y = 0, z = 15) no probes were found, while under the most relaxed conditions (x = 39, y = 4, z = 0) all probes were returned. For all combinations of x, y and z, the number of probes that complied (w), was stored
The data was sorted with w as primary criterion (ascending), followed by x (ascending), y (ascending) and z (descending)
This sorted dataset was analyzed row per row. In row i, the wi probes retrieved with criteria xi yi zi were compared with the list of probes, already picked up in rows 1 to i-1. If a probe did not occur in this list, it was added to the list
This process continued until there were m (user-defined) probes in the list
All in silico statistical enrichment tests are chi-square tests with Yates' correction, given p values are two-tailed.
DNA methylation analysis using COBRA and bisulfite sequencing (BSP)
To validate the (hyper)methylation status of candidate gene probes, DNA extracted from 10 cervical cancers and 5 normal cervices were analyzed using BSP and COBRA. Bisulfite modification of genomic DNA was performed using the EZ DNA methylation kit (Zymogen, BaseClear, Leiden, the Netherlands). The 5' promoter region of the tested gene was amplified using bisulfite treated DNA. PCR primers for amplification of specific targets sequences are listed in Additional file 3. COBRA was performed directly on the BSP products as described by Xiong et al. . using digestions with BstUI, Taq1 and/or HinfI according the manufacture's protocol (New England Biolabs Inc., Beverly, MA). For sequence analysis, the BSP products were purified (Qiagen, Westburg, Leusden, the Netherlands) and subjected to direct sequencing (BaseClear, Leiden, the Netherlands). Leukocyte DNA collected from anonymous healthy volunteers and in vitro CpG methylated DNA with SssI (CpG) methyltransferase (New England Biolabs Inc., Beverly, MA) were used as negative and positive control, respectively.
To identify novel markers that are methylated in cervical cancer, we applied a multistep approach that combines re-expression of silenced hypermethylated genes in cervical cancer cell lines (using DAC and DAC/TSA), downregulated expression in 39 cervical cancers expression, and selection of candidate markers using a relaxing ranking algorithm. The best profile of a candidate marker would be: no expression in any of the 39 cervical primary cancers and 4 untreated cancer cell lines, but re-activation of expression after demethylation and/or blocking of histone deacetylation in all 15 cell lines treated with various combinations of DAC/TSA (P-calls: 0-0-15). However, none of the probe sets showed this ideal profile. To generate a list of candidate genes, a relaxation ranking algorithm was applied.
The validation of the TOP3000 probe-list selected using relaxing high-ranking
To validate whether the TOP3000 contains potential hypermethylated genes, we determined the occurrence of various gene sets that are known to be hypermethylated such as imprinted genes, chromosome-X genes, cervical cancer related hypermethylated genes and genes reported to be methylated frequently in cancers, other than cervical cancer.
A. Enrichment for imprinted genes
Imprinting is a genetic mechanism by which genes are selectively expressed from the maternal or paternal homologue of a chromosome. As methylation is one of the regulatory mechanisms controlling the allele-specific expression of imprinted genes [36–40], it is expected that known imprinted genes are enriched in the TOP3000 selection. According to the Imprinted Gene Catalogue , this TOP3000 list contains 16 imprinted (or parent-specific expressed) genes (Additional file 5). On the whole Affymetrix array in total 74 imprinted genes could be assigned with a probe. Taking into account duplicate probes and probes that are not associated with a gene symbol, 8.76 imprinted genes could be expected in the first 3000 probes if the imprinted genes were randomly distributed indicating a 1.83-fold [16/8.76] enrichment in the TOP3000 (X2 = 5.904; p = 0.0151). The enrichment towards imprinted genes is even more significant in the TOP100 candidate genes (3 versus only 0.31 expected; X2 = 14.9; p < 0.0001).
B. Enrichment for genes on the X-chromosome
X-chromosome-inactivation in females is initiated from an inactivation centre that produces the Xist transcript, an RNA molecule that covers one copy of the X-chromosome and results in silencing of gene expression. This coating initiates a number of chromatin changes including stable DNA methylation . Of the entire list of 54675 probes on this Affymetrix array, 40683 could be associated with a chromosomal location, 1325 probes are located on the X-chromosome. In the TOP3000 list (with 2239 chromosomal locations known), 93 probes are located on chromosome X (data not shown) indicating a significant enrichment (1.28-fold) of X-chromosome-located probes in the TOP3000 (X2 = 5.8; p = 0.0165). This enrichment is even more significant within the TOP1000 (42/708 chromosomal locations; X2 = 12.567; p = 0.0004) and TOP100 probes (13/71 chromosomal regions known; X2 = 36.097; p < 0.0001).
C. Enrichment for cervical cancer specific methylation markers
Reported DNA methylation markers in cervical cancer present in the TOP3000
D. Enrichment for known hypermethylation markers in cancers other than cervical cancer
Listing of cancer associated hypermethylation markers that have been reported previously within the 250 highest ranking genes, as found by literature search (trough NCBI E-fetch, using GeneCards to search aliases)
Pediatric acute leukemia
Different cancer cell lines (Leukemic, Hepatic, Prostate, Breast, Colon)
Head-and neck cancer
Prostate and bladder cancer
Interestingly, this analysis revealed that in the TOP-250 known methylation markers seem to be significantly enriched and highly-ranked. This also showed that the known cervical cancer-specific markers are not enriched to the same extend (CCNA1 is highest at position 234), implying the existence of better hypermethylated markers, involved in cervical cancer, in the TOP3000 list.
In summary, as the top-list contains a relatively large number of imprinted, chromosome-X and known methylation markers, our analysis revealed that the ranking strategy was able to enrich the candidate gene list with possible new (hyper)methylated genes.
E. Gene Ontology
Associated with the development and progression of cancer, silencing by hypermethylation often affects genes in important pathways . Therefore, we investigated whether our selected TOP3000 candidate genes are associated with specific pathways or harbour related functions. Multiple GO-terms using Gene Ontology (GO) by GOstat  (Additional file 6) and specific pathways using Ingenuity Pathway Analysis (IPA) (Additional file 7), were significantly over-represented within the TOP3000 list when compared to all annotated human genes.
These terms include regulation of transcription – DNA-dependent, transcription from RNA polymerase II promoter, regulation of progression through cell cycle, positive regulation of programmed cell death as well as organ development and embryonic development (all p-values < 10-6 ; corrected for multiple hypothesis testing ).
Genes in these processes were reported to be often methylated during cancer progression . Genes responsible for development and differentiation are mainly silenced by methylation in normal tissues. On the other hand, in cancer tissues, genes responsible for cell cycle control and induction of apoptosis are often aberrantly expressed as many of these genes have tumour suppressor activity. DNA hypermethylation is one mechanism to regulate expression of tumour suppressor genes .
The GO-analysis provided additional strong indication that our highest ranking genes in the top-list are significantly enriched for methylated genes involved in cervical cancer tissue or cell lines.
Validation of the 10 highest-ranking candidate genes using COBRA
Methylation status using COBRA of the 10 highest-ranking gene promoters.
Methylation in cancer
Methylation in normal
represent a known gene (i.e. gene symbol)
contain a CpG island surrounding the TSS
are located on any chromosome except chromosome X
are expressed (present) in less than 15 carcinomas, in order to identify markers to be methylated in ≥60% of cervical cancers (no P-call)
Table 3 summarizes the methylation status of the 10 highest ranking genes in 10 cervical cancer and 5 normal cervices using COBRA. One gene (ADARB1 at rank 2) could not be analyzed for methylation as no specific BSP products could be amplified using several combinations of primer pairs. Interestingly, using the BSP products of the other 9 listed genes, 7 (78%) showed methylation in carcinomas (Table 3).
Enrichment of cervical cancer specific methylation markers
A cancer specific cervix hypermethylation marker is only of relevance for the diagnosis of (pre-) malignant disease in case normal cervical epithelium is not methylated. COBRA analysis of 5 normal cervices for all 9 genes revealed that 4 genes (DAZL,SYCP3,ZFP42 and NNAT) are hypermethylated in all 5 samples (Table 3).
On the other hand, of the 7 genes hypermethylated in cervical cancer specimens, 3 genes (SST, HTRA3 and NPTX1) did not show DNA methylation in any of the normal cervices of 5 independent individuals. We observed the same methylation profile for CCNA1 that was reported previously as a cervical cancer specific gene  with hypermethylation in only 6 of 10 tumours but none of the 5 normals (Table 3). This analysis revealed that the relaxation ranking algorithm not only resulted in a very significant enrichment for genes with a positive methylation status, but also for hypermethylated genes that are specifically methylated in cancers and not in the normal cervices.
In this study, we optimized the identification of methylation markers after pharmacological unmasking microarray approach combined with microarray expression data of primary cancer samples. For the integration of data from both cell lines and primary cancers, we developed a novel ranking strategy, which combines re-activation in cell lines and no expression in primary cancer tissue. The relaxation ranking algorithm uses a non-parametrical method of sorting. No threshold on expression level or P-calls has to be set and no overlap between different cell lines has to be chosen. The only parameter needed is the number of probes/genes that should be included in the top list. Using this algorithm, genes can still be selected for further analysis, even if it is not in (almost) all cell lines re-expressed or not silenced in most primary tumour samples.
In this study, we showed that the experimental design in combination with the ranking strategy is able to enrich a list of probes for methylated genes. Imprinted genes and genes on the X-chromosome are significantly enriched in the high-ranking TOP3000 probes. Pathway and gene ontology analysis illustrates that the high-ranking genes are involved in tumour development and progression. Enrichment of similar pathways or ontologies when selecting abnormal expressed genes is commonly reported in various cancer types [49, 50]. More importantly, methylation markers reported to be involved in various cancers (including cervical cancer) are significantly enriched in the top-lists as well. Interestingly, the highest ranking cervical cancer specific gene is CCNA1 (position 234 in Additional file 8; position 47 in Table 3). Apart from cervical cancer, CCNA1 was reported to be hypermethylated in colorectal, oral and head and neck cancer [30, 51, 52]. In good agreement with the reported data, we show that CCNA1 is hypermethylated in 6 of 10 cervical carcinomas and none of the normal cervices using COBRA and BSP sequencing (Table 3 and Figure 3).
Analysis of the methylation status of the highest ranking genes revealed that seven out of nine selected genes (78%) are methylated in cervical cancers, whereas 4 of these 7 genes (DAZL, SYCP3, ZFP42 and NNAT) were also hypermethylated in all 5 normal cervices (Table 3). Although hypermethylation of NNAT has been implicated in paediatric acute leukaemia , the hypermethylation status in both cancer and normal tissues suggests that NNAT acts as an imprinted gene (Additional file 5) rather than a cancer specific methylation marker in cervical cancer. The other three genes (SST, HTRA3 and NPTX1) might be cancer specific because these genes are, similar as CCNA1, hypermethylated in the cancers and not in the normal controls (Table 3). Of these genes, two were previously described as cancer specific genes: SST in colon carcinoma  and NPTX1 in pancreatic cancer . However, all 3 genes have not been described previously in cervical cancer. The exact involvement in cervical cancer development of these 3 genes has to be explored in the future, but the application of the relaxation ranking algorithm illustrates the power of enrichment for new hypermethylated genes that can discriminate between cervical cancer and normal cervical epithelium.
The combination of the initial setup and the analysis is unique. In most other studies either few genes are investigated for their methylation status in primary cancer samples or a large-screening approach is applied on cell lines only. Generally, only genes which are re-expressed in most cell lines can be retained for further investigation, as several hundreds of genes are upregulated in one or more cell lines after treatment with DAC/TSA. Most studies used additional filtering (such as pathway analysis, known mutated genes), but the list of candidate genes that need experimental validation to determine their methylation status is long. These markers need to go through a pipeline of DNA methylation detection in cell lines and cancer samples, in order to find only a few cancer specific markers with different sensitivity and specificity [21, 23, 53]. However, the success rate is relatively low, as many promoter regions do not show (differential) methylation. In addition, CpG arrays can be used to identify putative methylation markers, as recently described for cervical cancer . Again, this method requires the analysis of many markers to end up with only few cervical cancer specific methylation markers.
In the last few years it became apparent that many markers that are methylated in cancer have been shown to be methylated in normal tissues as well [12, 13, 56, 57]. Our present analysis once more illustrates that many more genes, preceded by a CpG island in the promoter region, are methylated in normal tissue as well than was previously anticipated. To be able to further increase the enrichment for these cancer specific methylated markers, the inclusion of expression microarray data from normal tissue in the relaxation ranking algorithm analysis might be helpful. To validate this, we performed global gene expression microarray analysis using the Affymetrix HGU 133 Plus 2.0 array with 54,675 probes (in parallel with the samples described in this study) on 5 independent age-matched normal cervices from healthy women. We assume that cancer specific methylated markers should be expressed in all normal cervices resulting in a positive P-call (most optimal P-call = 5). Including the P-call for normal expression on our 10 highest ranked methylated genes and CCNA1 (as listed in Table 3), revealed that all the four cervical cancer specific methylated genes (SST, HTRA3, NPTX1 and CCNA1) would not have been selected as none of the normal cervices showed a P-call for these probes (data not shown). It is generally accepted that tumor suppressor genes (including cancer specific methylated genes) are characterized by the fact that their expression can be downregulated as the result of methylation, mutations and/or deletions, but is still present in its normal counterpart tissue. However, the expression levels in normal tissue are relatively low for most of these genes when compared to those cancer tissues that do not show downregulation as was reported for p16INK4a . Thus, our data suggest that the addition of expression data of normal cervices would not enrich for cervical cancer specific methylated genes.
Other possibilities to further refine the selection of cancer relevant hypermethylated genes are to restrict the ranking to gene promoters that are likely to be methylated because of defined CG-content or the presence of conserved motifs (or similar sequence attributes) related to hypermethylated promoters [59–61]. Recently, we identified novel methylation markers, based on a genome-wide promoter alignment . Promoters, closely related in the alignment with known methylation markers show to have a high chance to be methylated as well.
In conclusion, the application of this new relaxation ranking methodology allowed us to significantly enrich towards methylation genes in cancer. This enrichment is both shown in silico and by experimental validation, and revealed novel methylation markers as proof-of-concept that might be useful in early cancer detection in cervical scrapings.
This study was supported by OncoMethylome Sciences SA, Liège, Belgium and by the Dutch Cancer Society (KWF-NKB-RUG 2004-3161). Prof AGJ van der Zee is a paid consultant for OncoMethylome Sciences SA, Liège, Belgium.
- Doerfler W, Toth M, Kochanek S, Achten S, Freisemrabien U, Behnkrappa A, et al: Eukaryotic Dna Methylation – Facts and Problems. Febs Letters. 1990, 268: 329-333.View ArticlePubMedGoogle Scholar
- Robertson KD: DNA methylation and chromatin – unraveling the tangled web. Oncogene. 2002, 21: 5361-5379.View ArticlePubMedGoogle Scholar
- Serman A, Vlahovic M, Serman L, Bulic-Jakus F: DNA methylation as a regulatory mechanism for gene expression in mammals. Coll Antropol. 2006, 30: 665-671.PubMedGoogle Scholar
- Paluszczak J, Baer-Dubowska W: Epigenetic diagnostics of cancer – the application of DNA methylation markers. J Appl Genet. 2006, 47: 365-375.View ArticlePubMedGoogle Scholar
- Hayatsu H: Bisulfite modification of nucleic acids and their constituents. Prog Nucleic Acid Res Mol Biol. 1976, 16: 75-124.View ArticlePubMedGoogle Scholar
- Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB: Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA. 1996, 93: 9821-9826.View ArticlePubMedPubMed CentralGoogle Scholar
- Esteller M: Cancer epigenetics: DNA methylation and chromatin alterations in human cancer. New Trends in Cancer for the 21St Century. 2003, 532: 39-49.View ArticleGoogle Scholar
- Herman JG: Epigenetic changes in cancer and preneoplasia. Cold Spring Harb Symp Quant Biol. 2005, 70: 329-333.View ArticlePubMedGoogle Scholar
- Esteller M: Cancer epigenomics: DNA methylomes and histone-modification maps. Nat Rev Genet. 2007, 8: 286-298.View ArticlePubMedGoogle Scholar
- Hegi ME, Diserens AC, Gorlia T, Hamou MF, de TN, Weller M, et al: MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med. 2005, 352: 997-1003.View ArticlePubMedGoogle Scholar
- Hoque MO, Topaloglu O, Begum S, Henrique R, Rosenbaum E, Van CW, et al: Quantitative methylation-specific polymerase chain reaction gene patterns in urine sediment distinguish prostate cancer patients from control subjects. J Clin Oncol. 2005, 23: 6569-6575.View ArticlePubMedGoogle Scholar
- Dammann R, Strunnikova M, Schagdarsurengin U, Rastetter M, Papritz M, Hattenhorst UE, et al: CpG island methylation and expression of tumour-associated genes in lung carcinoma. Eur J Cancer. 2005, 41: 1223-1236.View ArticlePubMedGoogle Scholar
- Wisman GBA, Nijhuis ER, Hoque MO, Reesink-Peters N, Koning AJ, Volders HH, et al: Assessment of gene promoter hypermethylation for detection of cervical neoplasia. Int J Cancer. 2006, 119: 1908-1914.View ArticlePubMedGoogle Scholar
- Costello JF, Fruhwald MC, Smiraglia DJ, Rush LJ, Robertson GP, Gao X, et al: Aberrant CpG-island methylation has non-random and tumour-type-specific patterns. Nat Genet. 2000, 24: 132-138.View ArticlePubMedGoogle Scholar
- Yan PS, Perry MR, Laux DE, Asare AL, Caldwell CW, Huang TH: CpG island arrays: an application toward deciphering epigenetic signatures of breast cancer. Clin Cancer Res. 2000, 6: 1432-1438.PubMedGoogle Scholar
- Strathdee G, Brown R: Abberant DNA methylation in cancer: potential clinical interventions. Expert Rev Mol Med. 2002, 4 (4): 1-17.View ArticlePubMedGoogle Scholar
- Huang TH, Perry MR, Laux DE: Methylation profiling of CpG islands in human breast cancer cells. Hum Mol Genet. 1999, 8: 459-470.View ArticlePubMedGoogle Scholar
- Hatada I, Morita S, Kimura M, Horii T, Yamashita R, Nakai K: Genome-wide demethylation during neural differentiation of P19 embryonal carcinoma cells. J Hum Genet. 2008, 53: 185-191.View ArticlePubMedGoogle Scholar
- Weber M, Davies JJ, Wittig D, Oakeley EJ, Haase M, Lam WL, et al: Chromosome-wide and promoter-specific analyses identify sites of differential DNA methylation in normal and transformed human cells. Nat Genet. 2005, 37: 853-862.View ArticlePubMedGoogle Scholar
- Shi H, Maier S, Nimmrich I, Yan PS, Caldwell CW, Olek A, et al: Oligonucleotide-based microarray for DNA methylation analysis: Principles and applications. J Cell Biochem. 2003, 88: 138-143.View ArticlePubMedGoogle Scholar
- Suzuki H, Gabrielson E, Chen W, Anbazhagan R, van Engeland M, Weijenberg MP, et al: A genomic screen for genes upregulated by demethylation and histone deacetylase inhibition in human colorectal cancer. Nat Genet. 2002, 31: 141-149.View ArticlePubMedGoogle Scholar
- Sova P, Feng Q, Geiss G, Wood T, Strauss R, Rudolf V, et al: Discovery of novel methylation biomarkers in cervical carcinoma by global demethylation and microarray analysis. Cancer Epidemiol Biomarkers Prev. 2006, 15: 114-123.View ArticlePubMedGoogle Scholar
- Yamashita K, Upadhyay S, Osada M, Hoque MO, Xiao Y, Mori M, et al: Pharmacologic unmasking of epigenetically silenced tumor suppressor genes in esophageal squamous cell carcinoma. Cancer Cell. 2002, 2: 485-495.View ArticlePubMedGoogle Scholar
- Guo M, Akiyama Y, House MG, Hooker CM, Heath E, Gabrielson E, et al: Hypermethylation of the GATA genes in lung cancer. Clin Cancer Res. 2004, 10: 7917-7924.View ArticlePubMedGoogle Scholar
- Yamashita K, Park HL, Kim MS, Osada M, Tokumaru Y, Inoue H, et al: PGP9.5 methylation in diffuse-type gastric cancer. Cancer Res. 2006, 66: 3921-3927.View ArticlePubMedGoogle Scholar
- Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB: Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet. 1999, 21: 103-107.View ArticlePubMedGoogle Scholar
- Smiraglia DJ, Rush LJ, Fruhwald MC, Dai ZY, Held WA, Costello JF, et al: Excessive CpG island hypermethylation in cancer cell lines versus primary human malignancies. Hum Mol Genet. 2001, 10: 1413-1419.View ArticlePubMedGoogle Scholar
- Finan MA, DeCesare S, Fiorica JV, Chambers R, Hoffman MS, Kline RC, et al: Radical hysterectomy for stage IB1 vs IB2 carcinoma of the cervix: does the new staging system predict morbidity and survival?. Gynecol Oncol. 1996, 62 (2): 139-147.View ArticlePubMedGoogle Scholar
- Koopman LA, Szuhai K, van Eendenburg JDH, Bezrookove V, Kenter GG, Schuuring E, et al: Recurrent integration of human papillomaviruses 16, 45, and 67 near translocation breakpoints in new cervical cancer cell lines. Cancer Res. 1999, 59: 5615-5624.PubMedGoogle Scholar
- Tokumaru Y, Yamashita K, Osada M, Nomoto S, Sun DI, Xiao Y, et al: Inverse correlation between cyclin A1 hypermethylation and p53 mutation in head and neck cancer identified by reversal of epigenetic silencing. Cancer Res. 2004, 64: 5982-5987.View ArticlePubMedGoogle Scholar
- van Dongen JJ, Langerak AW, Bruggemann M, Evans PA, Hummel M, Lavender FL, et al: Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia. 2003, 17: 2257-2317.View ArticlePubMedGoogle Scholar
- Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, et al: Bioconductor: open software development for computational biology and bioinformatics. Genome Biology. 2004, 5.Google Scholar
- Liu G, Loraine AE, Shigeta R, Cline M, Cheng J, Valmeekam V, et al: NetAffx: Affymetrix probesets and annotations. Nucleic Acids Res. 2003, 31 (1): 82-86.View ArticlePubMedPubMed CentralGoogle Scholar
- Xiong ZG, Laird PW: COBRA: A sensitive and quantitative DNA methylation assay. Nucleic Acids Res. 1997, 25 (12): 2532-2534.View ArticlePubMedPubMed CentralGoogle Scholar
- Rebhan M, ChalifaCaspi V, Prilusky J, Lancet D: GeneCards: Integrating information about genes, proteins and diseases. Trends in Genetics. 1997, 13: 163.View ArticlePubMedGoogle Scholar
- Holmes R, Soloway PD: Regulation of imprinted DNA methylation. Cytogenet Genome Res. 2006, 113: 122-129.View ArticlePubMedGoogle Scholar
- Feng W, Marquez RT, Lu Z, Liu J, Lu KH, Issa JP, et al: Imprinted tumor suppressor genes ARHI and PEG3 are the most frequently down-regulated in human ovarian cancers by loss of heterozygosity and promoter methylation. Cancer. 2008, 112: 1489-1502.View ArticlePubMedGoogle Scholar
- Zhao J, Dahle D, Zhou Y, Zhang X, Klibanski A: Hypermethylation of the promoter region is associated with the loss of MEG3 gene expression in human pituitary tumors. J Clin Endocrinol Metab. 2005, 90: 2179-2186.View ArticlePubMedGoogle Scholar
- Benetatos L, Dasoula A, Hatzimichael E, Georgiou I, Syrrou M, Bourantas KL: Promoter hypermethylation of the MEG3 (DLK1/MEG3) imprinted gene in multiple myeloma. Clin Lymphoma Myeloma. 2008, 8: 171-175.View ArticlePubMedGoogle Scholar
- Kuerbitz SJ, Pahys J, Wilson A, Compitello N, Gray TA: Hypermethylation of the imprinted NNAT locus occurs frequently in pediatric acute leukemia. Carcinogenesis. 2002, 23: 559-564.View ArticlePubMedGoogle Scholar
- Morison IM, Ramsay JP, Spencer HG: A census of mammalian imprinting. Trends in Genetics. 2005, 21: 457-465.View ArticlePubMedGoogle Scholar
- Heard E: Recent advances in X-chromosome inactivation. Current Opinion in Cell Biology. 2004, 16: 247-255.View ArticlePubMedGoogle Scholar
- Ongenaert M, Van Neste L, De Meyer T, Menschaert G, Bekaert S, Van Criekinge W: PubMeth: a cancer methylation database combining text-mining and expert annotation. Nucleic Acids Res. 2008, 36 (Database issue): D842-D846.PubMedGoogle Scholar
- Baylin SB, Ohm JE: Epigenetic gene silencing in cancer – a mechanism for early oncogenic pathway addiction?. Nat Rev Cancer. 2006, 6: 107-116.View ArticlePubMedGoogle Scholar
- Beissbarth T, Speed TP: GOstat: find statistically overrepresented Gene Ontologies within a group of genes. Bioinformatics. 2004, 20: 1464-1465.View ArticlePubMedGoogle Scholar
- Hochberg Y, Benjamini Y: More Powerful Procedures for Multiple Significance Testing. Stat Med. 1990, 9: 811-818.View ArticlePubMedGoogle Scholar
- Serman A, Vlahovic M, Serman L, Bulic-Jakus F: DNA methylation as a regulatory mechanism for gene expression in mammals. Coll Antropol. 2006, 30: 665-671.PubMedGoogle Scholar
- Kitkumthorn N, Yanatassaneejit P, Kiatpongsan S, Phokaew C, Trivijitsilp P, Termrungruanglert W, et al: Cyclin A1 promoter hypermethylation in human papillomavirus-associated cervical cancer. BMC Cancer. 2006, 6: 55.View ArticlePubMedPubMed CentralGoogle Scholar
- Kornberg LJ, Villaret D, Popp M, Lui L, McLaren R, Brown H, et al: Gene expression profiling in squamous cell carcinoma of the oral cavity shows abnormalities in several signaling pathways. Laryngoscope. 2005, 115: 690-698.View ArticlePubMedGoogle Scholar
- Wang Y, Yu Q, Cho AH, Rondeau G, Welsh J, Adamson E, et al: Survey of differentially methylated promoters in prostate cancer cell lines. Neoplasia. 2005, 7: 748-760.View ArticlePubMedPubMed CentralGoogle Scholar
- Xu XL, Yu J, Zhang HY, Sun MH, Gu J, Du X, et al: Methylation profile of the promoter CpG islands of 31 genes that may contribute to colorectal carcinogenesis. World J Gastroenterol. 2004, 10: 3441-3454.View ArticlePubMedPubMed CentralGoogle Scholar
- Shaw RJ, Liloglou T, Rogers SN, Brown JS, Vaughan ED, Lowe D, et al: Promoter methylation of P16, RARbeta, E-cadherin, cyclin A1 and cytoglobin in oral cancer: quantitative evaluation using pyrosequencing. Br J Cancer. 2006, 94: 561-568.View ArticlePubMedPubMed CentralGoogle Scholar
- Mori Y, Cai K, Cheng Y, Wang S, Paun B, Hamilton JP, et al: A genome-wide search identifies epigenetic silencing of somatostatin, tachykinin-1, and 5 other genes in colon cancer. Gastroenterology. 2006, 131: 797-808.View ArticlePubMedGoogle Scholar
- Hagihara A, Miyamoto K, Furuta J, Hiraoka N, Wakazono K, Seki S, et al: Identification of 27 5' CpG islands aberrantly methylated and 13 genes silenced in human pancreatic cancers. Oncogene. 2004, 23: 8705-8710.View ArticlePubMedGoogle Scholar
- Lai HC, Lin YW, Huang TH, Yan P, Huang RL, Wang HC, et al: Identification of novel DNA methylation markers in cervical cancer. Int J Cancer. 2008, 123: 161-167.View ArticlePubMedGoogle Scholar
- Cheong J, Yamada Y, Yamashita R, Irie T, Kanai A, Wakaguri H, et al: Diverse DNA methylation statuses at alternative promoters of human genes in various tissues. DNA Res. 2006, 13: 155-167.View ArticlePubMedGoogle Scholar
- Hoque MO, Kim MS, Ostrow KL, Liu J, Wisman GB, Park HL, et al: Genome-wide promoter analysis uncovers portions of the cancer methylome. Cancer Res. 2008, 68: 2661-2670.View ArticlePubMedPubMed CentralGoogle Scholar
- Kang S, Kim J, Kim HB, Shim JW, Nam E, Kim SH, et al: Methylation of p16INK4a is a non-rare event in cervical intraepithelial neoplasia. Diagn Mol Pathol. 2006, 15: 74-82.View ArticlePubMedGoogle Scholar
- Bock C, Paulsen M, Tierling S, Mikeska T, Lengauer T, Walter J: CpG island methylation in human lymphocytes is highly correlated with DNA sequence, repeats, and predicted DNA structure. PLoS Genet. 2006, 2: e26.View ArticlePubMedPubMed CentralGoogle Scholar
- Das R, Dimitrova N, Xuan Z, Rollins RA, Haghighi F, Edwards JR, et al: Computational prediction of methylation status in human genomic sequences. Proc Natl Acad Sci USA. 2006, 103: 10713-10716.View ArticlePubMedPubMed CentralGoogle Scholar
- Feltus FA, Lee EK, Costello JF, Plass C, Vertino PM: Predicting aberrant CpG island methylation. Proc Natl Acad Sci USA. 2003, 100: 12253-12258.View ArticlePubMedPubMed CentralGoogle Scholar
- Ivanova T, Vinokurova S, Petrenko A, Eshilev E, Solovyova N, Kisseljov F, et al: Frequent hypermethylation of 5 ' flanking region of TIMP-2 gene in cervical cancer. Int J Cancer. 2004, 108: 882-886.View ArticlePubMedGoogle Scholar
- Dowdy SC, Gostout BS, Shridhar V, Wu XS, Smith DI, Podratz KC, et al: Biallelic methylation and silencing of paternally expressed gene 3 (PEG3) in gynecologic cancer cell lines. Gynecol Oncol. 2005, 99: 126-134.View ArticlePubMedGoogle Scholar
- Kim TY, Lee HJ, Hwang KS, Lee M, Kim JW, Bang YJ, et al: Methylation of RUNX3 in various types of human cancers and premalignant stages of gastric carcinoma. Lab Invest. 2004, 84: 479-484.View ArticlePubMedGoogle Scholar
- Steenbergen RDM, Kramer D, Braakhuis BJM, Stem PL, Verheijen RHM, Meijer CJLM, et al: TSLC1 gene silencing in cervical cancer cell lines and cervical neoplasia. J Natl Cancer Inst. 2004, 96: 294-305.View ArticlePubMedGoogle Scholar
- Cheung TH, Lo KWK, Yim SF, Chan LKY, Heung MS, Chan CS, et al: Epigenetic and genetic alternation of PTEN in cervical neoplasm. Gynecol Oncol. 2004, 93: 621-627.View ArticlePubMedGoogle Scholar
- Shivapurkar N, Toyooka S, Toyooka KO, Reddy J, Miyajima K, Suzuki M, et al: Aberrant methylation of trail decoy receptor genes is frequent in multiple tumor types. Int J Cancer. 2004, 109: 786-792.View ArticlePubMedGoogle Scholar
- Mihara M, Yoshida Y, Tsukamoto T, Inada K, Nakanishi Y, Yagi Y, et al: Methylation of multiple genes in gastric glands with intestinal metaplasia – A disorder with polyclonal origins. Am J Pathol. 2006, 169: 1643-1651.View ArticlePubMedPubMed CentralGoogle Scholar
- Fradet Y, Picard V, Bergeron A, Larue H: Cancer-testis antigen expression in bladder cancer. Progres en Urologie. 2006, 16: 421-428.PubMedGoogle Scholar
- Manton KJ, Douglas ML, Netzel-Arnett S, Fitzpatrick DR, Nicol DL, Boyd AW, et al: Hypermethylation of the 5 ' CpG island of the gene encoding the serine protease Testisin promotes its loss in testicular tumorigenesis. Br J Cancer. 2005, 92: 760-769.View ArticlePubMedPubMed CentralGoogle Scholar
- Kempkensteffen C, Christoph F, Weikert S, Krause H, Kollermann J, Schostak M, et al: Epigenetic silencing of the putative tumor suppressor gene testisin in testicular germ cell tumors. J Cancer Res Clin Oncol. 2006, 132: 765-770.View ArticlePubMedGoogle Scholar
- Okino ST, Pookot D, Li LC, Zhao H, Urakatni S, Shiina H, et al: Epigenetic inactivation of the dioxin-responsive cytochrome P4501A1 gene in human prostate cancer. Cancer Res. 2006, 66: 7420-7428.View ArticlePubMedGoogle Scholar
- Wischnewski F, Pantel K, Schwarzenbach H: Promoter demethylation and histone acetylation mediate gene expression of MAGE-A1,-A2,-A3, and-A12 in human cancer cells. Mol Cancer Res. 2006, 4: 339-349.View ArticlePubMedGoogle Scholar
- Qiu G, Fan JC, He YS: 5' CpG island methylation analysis identifies the MAGE-A1 and MAGE-A3 genes as potential markers of HCC. Clin Biochem. 2006, 39: 259-266.View ArticlePubMedGoogle Scholar
- Sigalotti L, Coral S, Nardi G, Spessotto A, Cortini E, Cattarossi I, et al: Promoter methylation controls the expression of MAGE2, 3 and 4 genes in human cutaneous melanoma. J Immunother. 2002, 25: 16-26.View ArticlePubMedGoogle Scholar
- Rauch T, Li H, Wu X, Pfeifer GP: MIRA-assisted microarray analysis, a new technology for the determination of DNA methylation patterns, identifies frequent methylation of homeodomain-containing genes in lung cancer cells. Cancer Res. 2006, 66: 7939-7947.View ArticlePubMedGoogle Scholar
- Hong CB, Bollen AW, Costello JF: The contribution of genetic and epigenetic mechanisms to gene silencing in oligodendrogliomas. Cancer Res. 2003, 63: 7600-7605.PubMedGoogle Scholar
- Hoffmann MJ, Muller M, Engers R, Schulz WA: Epigenetic control of CTCFL/BORIS and OCT4 expression in urogenital malignancies. Biochem Pharmacol. 2006, 72: 1577-1588.View ArticlePubMedGoogle Scholar
- Blanchard F, Tracy E, Smith J, Chattopadhyay S, Wang YP, Held WA, et al: DNA methylation controls the responsiveness of hepatoma cells to leukemia inhibitory factor. Hepatology. 2003, 38: 1516-1528.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1755-8794/1/57/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.