[Epigenetics 4:4, 221-230; 16 May 2009]; 2009 Landes Bioscience

Brief Report

Quantitative analysis of DNA methylation after whole bisulfitome amplification of a minute amount of DNA from body fluids
Thomas Vaissire,1 Cyrille Cuenin,1 Anupam Paliwal,1 Paolo Vineis,2 the Genair-EPIC Investigators,3 Pierre Hainaut1 and Zdenko Herceg1,*
1International Agency for Research on Cancer (IARC); Lyon, France; 2Imperial College London; St. Marys Campus; London, UK; 3The Genair-EPIC collaborators (35): Hoek G, Krzyzanowski M, Airoldi L, Dunning A, Garte S, Hainaut P, Malaveille C, Overvad K, Clavel-Chapelon F, Linseisen J, Boeing H, Trichopoulou A, Trichopoulos D, Kaladidi A, Palli D, Krogh V, Tumino R, Panico S, Bueno-De-Mesquita HB, Peeters PH, Kumle M, Gonzalez CA, Martinez C, Dorronsoro M, Barricarte A, Navarro C, Quiros JR, Berglund G, Janzon L, Jarvholm B, Day NE, Key TJ, Saracci R, Kaaks R, Riboli E

Abbreviations: qMAMBA, quantitative methylation analysis of minute DNA amounts after whole bisulfitome amplification; MSP, methylation-specific PCR; Q-MSP, quantitative methylation specific PCR; MEP, methylation enrichment pyrosequencing; PMA, pyrosequencing methylation assay; Gen-AIR, the study on genetic susceptibility and environmental factors; EPIC, the european prospective investigation into cancer and nutrition; UADT, upper aerodigestive tract; RASSF1A, Ras-association domain family 1 gene; MTHFR, methylenetetrahydrofolate reductase gene; MLH1, mutL homolog 1 gene (colon cancer, nonpolyposis type 2 gene); LINE1, long interspersed nuclear element-1 repetitive sequence Key words: DNA methylation, quantitative detection, plasma DNA, biomarkers, cancer, epigenetics

Cell-free circulating DNA isolated from the plasma of individuals with cancer has been shown to harbor cancer-associated changes in DNA methylation, and thus it represents an attractive target for biomarker discovery. However, the reliable detection of DNA methylation changes in body fluids has proven to be technically challenging. Here we describe a novel combination of methods that allows quantitative and sensitive detection of DNA methylation in minute amounts of DNA present in body fluids (quantitative Methylation Analysis of Minute DNA amounts after whole Bisulfitome Amplification, qMAMBA). This method involves genome-wide amplification of bisulphitemodified DNA template followed by quantitative methylation detection using pyrosequencing and allows analysis of multiple genes from a small amount of starting DNA. To validate our method, we used qMAMBA assays for four genes and LINE1 repetitive sequences combined with plasma DNA samples as a model system. qMAMBA offered high efficacy in the analysis of methylation levels and patterns in plasma samples with extremely small amounts of DNA and low concentrations of methylated alleles. Therefore, qMAMBA will facilitate methylation studies aiming to discover epigenetic biomarkers, and should prove particularly valuable in profiling a large sample series of body

fluids from molecular epidemiology studies as well as in tracking disease in early diagnostics.

Introduction
Tumor-derived cell-free circulating DNA isolated from the plasma and serum of individuals with cancer has been shown to contain cancer-associated alterations.1,2 While the origin and possible function of this free circulating DNA is not fully understood, it represents an attractive target for biomarker discovery. In addition to genetic changes (mutations, microsatellite alterations), plasma DNA from individuals with tumors was shown to harbor epigenetic changes, namely alterations in DNA methylation at CpG sites in the promoter regions of a wide range of tumor suppressor genes and other cancer-associated genes. DNA methylation changes are tumor specific and thus have the potential to serve as highly specific biomarkers. In addition, DNA methylation changes appear early in tumor development and can be found in virtually every type of human cancer, thus they can provide particularly attractive markers with broad application in diagnostics and risk assessment.3-5 The development of epigenetic markers for cancer-bearing individuals could similarly enhance the management of their disease. However, reliable detection of DNA methylation changes in body fluids has proven to be technically challenging. In particular, the amount of circulating DNA in body fluids is rather limited and the DNA harboring CpG methylation changes is likely to represent only a tiny fraction of the total circulating DNA. Many efforts have been made to develop a reliable technique that would allow sensitive detection of DNA methylation in minute amounts of nucleic acids derived from body fluids. Methylationspecific PCR (MSP) and quantitative methylation specific PCR (Q-MSP) are relatively inexpensive and highly sensitive techniques
221

*Correspondence to: Zdenko Herceg; Epigenetics Group; International Agency for Research on Cancer (IARC); 150 Cours Albert-Thomas; Lyon F-69008 France; Tel.: +33.4.72.73.83.98; Fax: +33.4.72.73.83.29; Email: herceg@iarc.fr Submitted: 03/03/09; Accepted: 04/24/09 Previously published online as an Epigenetics E-publication: http://www.landesbioscience.com/journals/epigenetics/article/8833

www.landesbioscience.com

Epigenetics

Quantitative analysis of methylation in plasma DNA

that have been widely used to characterize DNA methylation changes in tumor tissues and body fluids including plasma, sputum and urine.6-10 However, these techniques lack in-built measures of completeness of bisulphite conversion (conversion control) and are prone to mispriming, both of which can be a frequent source of false positives. These problems are further aggravated when these techniques are applied on low concentrations of starting DNA templates (typical for DNA samples isolated from body fluids) and when high numbers of PCR cycles are used. Although problems with mis-priming can be remedied by subsequent confirmation steps such as re-analysis using bisulphite sequencing11 or methylation sensitive restriction enzymes,12 these additional validation steps represent significant logistical disadvantages by reducing both time- and cost-effectiveness. To overcome the problems associated with false positives Liloglou and colleagues developed the Methylation Enrichment Pyrosequencing (MEP) method,13 which benefits from high sensitivity of MSP and specificity of the pyrosequencing methylation assay (PMA) that serves as a confirmatory step.13 While the MEP technique has important advantages over MSP and Q-MSP, its intrinsic weakness lies in the fact that it cannot be used to monitor quantitative levels of the methylated alleles in samples with minute amounts of DNA. Because the primers of the first step (MSP) are designed to specifically amplify methylated DNA, and exclude unmethylated DNA, this technique cannot provide a quantitative measure of different amounts of methylated and unmethylated DNA in samples. Since the fraction of circulating DNA containing DNA methylation changes varies between samples and typically represents only a tiny fraction of total circulating DNA, an ideal method should be able to provide a quantitative measure of methylated alleles in total circulating DNA. Furthermore, such a method would allow methylation analysis of a large number of genes in Figure 1. General outline of the quantitative Methylation Analysis of Minute Amount samples with low amounts of DNA, the requirement that of DNA (qMAMBA) method. The DNA isolated from body fluids (plasma) [1] was treated with sodium bisulphite [2] and subjected to genome-wide amplification [3]. none of the available techniques is able to fulfill. Amplified DNA was then used for pyrosequencing [4] using specific amplification In the present study we describe a new combina- and pyrosequencing primers designed to analyze a series of CpG dinucleotides in tion of techniques that allows quantitative and sensitive the promoter region of the genes of interest or LINE1 repetitive sequence. detection of DNA methylation in minute amounts of DNA present in body fluids (quantitative Methylation Analysis methylation in samples with low concentration of DNA, we have of Minute DNA amounts after whole Bisulfitome Amplification, sought to develop a new method that would satisfy this key criteqMAMBA). This method involves genome-wide amplification of rion. Here we describe the qMAMBA method, a new combination bisulphite-modified DNA template and quantitative methylation of techniques that allows multiple quantitative methylation assays to detection using pyrosequencing and offers significant advantages be performed on a small amount of starting DNA present in body over previously described methods. Its main strength is that it fluids. The general outline of the qMAMBA method is depicted provides both quantitative accuracy and sensitivity of detection in Figure 1. The DNA samples isolated from body fluids (e.g., of DNA methylation patterns as well as compatibility with high human plasma) were subjected to bisulphite treatment followed throughput settings. by genome-wide amplification and pyrosequencing. Genome-wide amplification of bisulphite-converted DNA provides a sufficient Results and Discussion DNA template for analysis of DNA methylation for a large number Because existing techniques have not been shown to satisfy the of genes. This method also takes advantage of pyrosequencing, a main criteria for their applicability in quantitative analysis of DNA highly reliable and quantitative sequencing-by-synthesis method for
222 Epigenetics 2009; Vol. 4 Issue 4

Quantitative analysis of methylation in plasma DNA

the analysis of DNA methylation at multiple CpG sites with built-in internal controls for completeness of bisulphite treatment.14-16 We first tested whether qMAMBA could provide quantitative results across different DNA methylation levels by analyzing the dilution matrix simulating very low concentrations of DNA and varying degree of DNA methylation. For this, we selected two lung tumor samples exhibiting low (5%) and high (60%) levels of DNA methylation in the RASSF1A promoter, as previously verified by pyrosequencing (data not shown), and a series of dilutions (1:10, 1:100, 1:1,000 and 1:10,000). The samples from the dilution matrix were prepared prior to bisulphite treatment and genome-wide amplification and processed as independent samples throughout the entire qMAMBA protocol. As a control, dilution matrix DNA samples were subjected to the identical protocol as for qMAMBA with the exception of genome-wide amplification. Modified DNA samples (amplified and non-amplified) were subjected to PCR using the amplification primers specific for the RASSF1A gene promoter. The quality of PCR products was verified by agarose gel electrophoresis and ethidium bromide staining. As shown in Figure 2A, DNA samples subjected to genomewide amplification yielded PCR products of expected size even when diluted 10,000-fold (at concentration of 2 pg/ml), whereas samples without genome-wide amplification generated detectable PCR bands only up to dilution of 1:10. These gel results showed that genome-wide amplification of bisulphite-treated DNA samples allows a significant increase (1001,000-fold) in sensitivity of detection irrespective of the presence of different amounts of methylated alleles. PCR products were next subjected to pyrosequencing in order to verify their identity as well as to obtain a quantitative measure of methylated alleles. As shown in Figure 2B, pyrosequencing analysis confirmed the identity of the PCR products obtained from all matrix dilutions, with the exception of one sample at 1:1,000 dilution, demonstrating the absence of mis-priming even at high dilutions of matrix. Note that at 1:10,000 dilution pyrosequencing reaction was successful (Fig. 2B) despite the fact that the PCR band is hardly visible on agarose gel (Fig. 2A). This can be explained by the fact that only a smaller part (10 ml) of the PCR reaction was run on agarose gel for verification and that a larger part of the reaction (40 ml) was used for pyrosequencing reaction. Alternatively, this may reflect a higher sensitivity of pyrosequencing reaction. Comparison of qMAMBA results obtained on samples with different dilutions revealed a high degree of correspondence between qMAMBAderived methylation levels and expected methylation levels in the presence of either high or low amounts of methylated alleles (Fig. 2B), indicating that whole genome amplification of sodium bisulphite-treated DNA does not bias DNA methylation results.17 The mean methylation levels and methylation levels across six CpG sites of RASSF1A were similar in any given sample (Fig. 2B, and data not shown). To further validate our method on different genes, we have performed qMAMBA assays for the MTHFR and LINE1 sequences using serial dilution of genomic DNA and found consistent pyrosequancing results of both MTHFR and LINE1 when diluted up to 10,000-fold (Fig. 2C and D). Together these results demonstrate that qMAMBA offers a high sensitivity
www.landesbioscience.com

and specificity for quantitative detection of DNA methylation in samples with minute amounts of DNA templates. To test the use of the qMAMBA assay across different genes in body fluids, we chose to analyze DNA methylation patterns in the CpG island present in the promoter region of four genes (RASSF1A, CDKN2A, MLH1 and MTHFR) and the LINE1 repetitive sequence (human retrotransposon sequence commonly used as a surrogate for global methylation levels). As a proof of principle, we tested whether qMAMBA could be applied to the routine analysis of extremely diluted DNA templates, using plasma samples obtained from a case-control study.18 We analyzed DNA methylation levels in 21 plasma DNA samples by qMAMBA and, as controls, corresponding blood lymphocyte DNA samples were also analyzed by PMA. For qMAMBA analysis, DNA extracted from plasma samples was treated with sodium bisulphite and subjected to genome-wide amplification followed by PCR amplification. All of the 21 plasma DNA samples were subjected to PCR amplification using primers for five independent assays (four genes and LINE1 sequences), representing a total of 105 PCR reactions. The quality of PCR products was verified on agarose gel (Fig. 3A, and data not shown). Of the 105 plasma DNA samples (subjected to PCR amplifications) 93 samples (88%) produced PCR bands of the expected size on agarose gel (Table 3). The identity of the PCR products was confirmed by pyrosequencing. Typical pyrograms for four genes and LINE1 sequences analyzed are shown in Figure 3B. These results show that genome-wide amplification of bisulphite-treated low amounts of DNA can be reliably amplified, and that methylation levels in these amplified samples can be accurately monitored by pyrosequencing. The results of quantitative analysis of methylation status by qMAMBA across CpG sites for each gene and LINE1 sequences analyzed are shown in Figure 4 (left). Our analysis showed a diverse spectrum of DNA methylation levels in plasma DNA samples, ranging from 0% of methylation for RASSF1A and MLH1 in several samples to relatively high levels of DNA methylation for MTHFR and LINE1 in most of the plasma DNA samples analyzed (Fig. 4, left). Methylation levels across CpG sites were similar in a given sample, although some samples exhibited significant differences between different CpG sites analyzed. For comparison purposes, we used the pyrosequencing methylation assay (PMA) to analyze DNA methylation states of the same genes and LINE1 sequences in DNA isolated from corresponding blood lymphocytes. Relatively high concentrations of DNA in lymphocyte samples allowed methylation analysis using the PMA assay without prior genome-wide amplification of DNA. The results of PMA analysis of lymphocyte DNA samples are shown in Figure 4 (right). In contrast to the methylation levels in plasma DNA, methylation levels and patterns for all four genes and LINE1 sequences analyzed were almost identical among all lymphocyte DNA samples, consistent with the absence of interindividual variation in methylation levels in a given healthy tissue sample.19 These results demonstrate that qMAMBA can be used for quantitative detection of methylation levels in multiple genes and repetitive sequences in DNA isolated from plasma with sufficient sensitivity and specificity.
223

Epigenetics

Quantitative analysis of methylation in plasma DNA

Figure 2. For figure legend, see page 225.

We have described qMAMBA method for quantitative and sensitive detection of DNA methylation in minute amounts of DNA present in body fluids. Its main advantage is that it allows quantitative detection of the amount of methylation in small amount of starting DNA found in plasma and other body fluids (Table 4). qMAMBA is a rapid and relatively inexpensive technique with high sensitivity, and specificity and it allows methylation profiling of multiple genes and simultaneous analysis of a large
224

series of body fluids collected during routine clinical sampling or as a part of large molecular epidemiology studies. We used four genes and LINE1 repetitive sequences and plasma DNA samples as a model system to test our method, and demonstrated the suitability of qMAMBA to accurately quantify methylation levels. Five qMAMBA assays (RASSF1A, MLH1, MTHFR, CDKN2A and LINE1) were initially developed and tested on plasma DNA samples and all five have passed the
2009; Vol. 4 Issue 4

Epigenetics

Quantitative analysis of methylation in plasma DNA Figure 2. Analysis of DNA methylation at different dilutions of DNA templates before and after genome-wide amplification. (A) Samples exhibiting low (5%, SL) and high (60%, SH) levels of DNA methylation in the RASSF1A gene promoter were serially diluted as indicated. The dilutions were treated with sodium bisulphite then subjected to either PCR amplification followed by pyrosequencing (PMA, left) or to genome-wide amplification followed by PCR amplification and pyrosequencing (qMAMBA, right). PCR reactions were carried out using the identical amplification primers specific for the CpG island in the RASSF1A gene promoter. The quality of PCR products was verified by running one part of the PCR reaction on agarose gel and staining with ethidium bromide. (B) The methylation levels of RASSF1A in samples with different dilutions of template DNA. The remaining parts of the PCR reactions verified in (A) were subjected to pyrosequencing using specific sequencing primers. Quantitative results of DNA methylation in the RASSF1A gene obtained by PMA (left) and qMAMBA (right) are expressed as means of all six CpG sites analyzed. (C and D) Analysis of DNA methylation for MTHFR and at different dilutions of DNA templates by qMAMBA. DNA samples were serially diluted as indicated and the dilutions were treated with sodium bisulphite and subjected to genome-wide amplification followed by PCR amplification and pyrosequencing. The quality of PCR products was verified on agarose gel (C) and the methylation levels were determined by pyrosequencing using specific sequencing primers (D). Quantitative results of DNA methylation are expressed as means of all six (LINE1) or four (MTHFR) CpG sites analyzed. Asterisk (*) indicates the sample for which pyrosequencing analysis failed.

quality control and produced reliable results. This method exploits genome-wide amplification of bisulphite-treated DNA that allows both a significant (up to 1,000-fold) increase in sensitivity and a sufficient amount of DNA templates for methylation analysis of a large number of genes. qMAMBA also takes advantage of pyrosequencing, a direct sequencing by synthesis method, that allows analysis of multiple CpG sites in one reaction.14,15 Moreover, qMAMBA could be easily modified for use in conjunction with other quantitative methods for methylation analysis, such as MethyLight.20 In addition to measuring DNA hypermethylation of tumor suppressor genes, this method can be used to monitor methylation levels of repetitive sequences (status of global methylation). However, the ultimate validation of the method in biomarker discovery will be in testing whether DNA methylation patterns found in circulating DNA can reflect those of tumor DNA. To this end, it will be necessary to apply qMAMBA in conjunction with series of plasma/tumor pairs collected in the context of clinical and epidemiologic protocols. In summary, qMAMBA will facilitate methylation studies aiming to discover and validate new epigenetic biomarkers and should prove particularly valuable in profiling large sample series of body fluids from molecular epidemiology studies and clinical studies for early diagnostics and prognostic purposes as well as in monitoring the efficacy of epigenetics-based cancer therapies and preventive strategies.

Materials and Methods

Plasma and blood lymphocyte samples and DNA extraction. Plasma and lymphocyte DNA samples were obtained from the study on genetic susceptibility and environmental factors (GenAIR),18 a case-control study nested within the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort.21,22 Cases are subjects with bladder, lung or upper aerodigestive tract (UADT) cancers or leukemia, all newly diagnosed after recruitment. Lymphocyte and plasma DNA samples from leukemia cases were excluded from our analyses. Plasma DNA was extracted using affinity columns (Qiagen, Hilden, Germany) from 300 ml plasma and eluted in 200 ml buffer as described previously.23 Genomic DNA from lymphocytes was extracted using buffy coat separated from blood samples by automated equipment (Autopure LS by Gentra Systems) as described previously,24 and DNA concentrations were quantified with the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE)
www.landesbioscience.com

and with QuantiT PicoGreen dsDNA reagent (Molecular Probes) as described previously.24 Sample DNA concentrations were calculated based on a standard curve established with Lambda DNA (Table 1). DNA dilution matrix. To simulate low concentrations of DNA template containing different amounts of methylated alleles, we prepared serial dilutions (1:10, 1:100, 1:1,000 and 1:10,000) of two bisulphite-modified DNA samples obtained from lung cancer tissues (at concentration of 20 ng/ml) containing low (5%) and high (60%) levels of methylated alleles. The dilutions prepared from the samples were then subjected to either PCR amplification followed by pyrosequencing (PMA) or to genome-wide amplification followed by PCR amplification and pyrosequencing (qMAMBA). qMAMBA. Figure 1 illustrates a general outline of the quantitative Methylation Analysis of Minute Amount of Bisulphite-treated and Amplified DNA (qMAMBA) assay. Bisulphite treatment. DNAs from plasma samples and blood lymphocyte samples were modified by treatment with sodium bisulphite using a bisulphite conversion kit (EpiTect, Qiagen). Lymphocyte DNAs were subjected to bisulphite conversion as described previously.24 The DNAs extracted from plasma samples are modified by treatment with sodium bisulphite using a slightly modified protocol. Briefly, DNA samples (containing 100 pg to 4.5 ng DNA in 50 ml buffer) were mixed with 85 ml of bisulphite DNA conversion and 15 ml of DNA protect buffer provided with the kit. Bisulphite conversion was carried out in a thermal cycler at 60C for 4 hours and 45 min separated by three steps of denaturation at 99C for 5 min each. To this mix, 560 ml of Buffer BL containing 10 mg/ml carrier RNA was added and transferred into an EpiTect spin column and spun at 13,000 rpm for 1 min. The columns were washed with 500 ml of washing buffer BW (containing EtOH) and incubated with 500 ml of Buffer BD (desulfonation buffer) at room temperature for 15 min. The columns were centrifuged at maximum speed for 1 min, and the flow-through was discarded. The columns were then washed twice with Buffer BW and the purified DNA was eluted in 10 ml of Buffer EB. Samples were kept at +4C for short-term storage or frozen (-20C) for long-term storage. Genome-wide amplification of bisulphite converted DNA. Genome-wide amplification of plasma DNA was carried out using the EpiTect Whole Bisulfitome (WBA) system (Qiagen). 30 ml of Reaction Buffer (EpiTect WBA Reaction Buffer, 29 ml; REPLI-g
225

Epigenetics

Quantitative analysis of methylation in plasma DNA

Figure 3. Quantitative analysis of amplified bisulphite converted DNA. (A) PCR reactions were carried out on plasma DNA samples (14) subjected to sodium bisulphite treatment and genome-wide amplification, and PCR products were resolved on agarose gel. M, molecular weight marker. Asterisk (*) indicates the sample for which pyrosequencing analysis was successful despite a DNA smear indicating certain degree of non-specific amplification. (B) Quantitative analysis of DNA methylation in plasma DNA samples by qMAMBA. Representative pyrograms of each of four genes and LINE1 sequences analyzed in plasma DNA samples are shown. 226 Epigenetics 2009; Vol. 4 Issue 4

Quantitative analysis of methylation in plasma DNA

Figure 4. Quantitative profiles of DNA methylation levels across multiple CpG sites of five different genomic sequences (LINE1, RASSF1A, CDKN2A, MLH1 and MTHFR) in plasma DNA samples (as analysed by qMAMBA, left) and corresponding blood lymphocyte samples (as analyzed by PMA, right) of different patients.

www.landesbioscience.com

Epigenetics

227

Quantitative analysis of methylation in plasma DNA

Midi DNA Polymerase, 1 ml) from the EpiTect Amplification Master Mix was added to the 10 ml of bisulphite converted DNA and incubated for 8 h at 28C, and Repli-g Midi DNA Polymerase was inactivated by heating the sample for 5 min at 95C. Primers. Specific amplification primers (biotinylated) and pyrosequencing primers were designed using the criteria described previously.24 These primers were designed to probe a series of four to eight CpG dinucleotides in the promoter region of the Ras-association domain family 1 (RASSF1A) gene, the cyclin-dependent kinase inhibitor 2A (CDKN2A), methylenetetrahydrofolate reductase (MTHFR), the mutL homolog 1 (colon cancer, nonpolyposis type 2 gene, MLH1), and for the LINE1 repetitive sequence given in Table 2. Pyrosequencing analysis. PCR amplification of genome-wide amplified DNA was carried out using Hot-Start Taq Master Mix (Qiagen). Pyrosequencing analysis was performed using specific primers on the PSQ96MA pyrosequencer (Biotage). Briefly, Hot-start PCR was performed with HotStarTaq Master Mix kit (Qiagen), and pyrosequencing was carried out in accordance with the manufacturers protocol (Biotage). The target CpGs were evaluated by converting the resulting pyrograms to numerical values for peak heights. The percentage of methylation was calculated as described previously.24 Pyrosequencing methylation analysis (PMA) of lymphocyte DNA samples. Pyrosequencing analysis of methylation status in DNA samples isolated from blood lymphocytes was carried out as described previously.24
Acknowledgements

Table 1 Plasma DNA samples and quantification of DNA concentrations Sample

1 2 3 4 5 6 7 8 23 24 25 26 27 28 29 30 31 32 33 34 35

Plasma DNA concentration (ng/l)

0.088 0.023 0.027 0.041 0.028 0.027 0.019 0.033 0.019 0.043 0.020 0.009 0.024 0.017 0.050 0.010 0.008 0.044 0.020 0.056 0.009

DNA concentration after GWA (ng/l)

62.75 35.75 56.25 140.5 83.75 56.50 21.75 0.60 ND ND 12.50 ND 9.00 ND 20.25 ND 2.68 ND ND ND 22.25

T. Vaissire is supported by a PhD fellowship from La Ligue Nationale (Franaise) Contre le Cancer. The work in the IARC Epigenetics Group is supported by grants from the National Institutes of Health/National Cancer Institute (NIH/NCI), United States; the Association pour la Recherche sur le Cancer (ARC), France; La Ligue Nationale (Franaise) Contre le Cancer, France;

GWA, genome-wide amplification; ND, not determined.

the European Network of Excellence Environmental Cancer Risk, Nutrition and Individual Susceptibility (ECNIS), and the Swiss Bridge Award (to Z.H.).

Table 2 List of genes, primers and sequences used in qMAMBA assays Gene PCR amplification primers Sequencing primers
CDKN2A (sense) CDKN2A (antisense) LINE (sense) LINE (antisense) MTHFR (sense) 5'-TTGAGGGTGGGAAGATGGT 5'-BIOTIN-TAGGGAGTGTTAGATAGTGG 5'-GGAGGGAGAGGAA 5'-AACTCCCTAACCCCTTAC 5'-BIOTIN-CCCRAACCTCCAAAATCTC 5'-AACTCCCTAACCCCTTAC 5'-BIOTIN-TTTTAATTTTT 5'-GGGTTTGGATTTTGAG GTTTGGAGGGTAGT 5'-BIOTIN-AAAAAAA CCACTTATCACCAAATTC 5'-TTTAGGAGTGAAGGAGGT 5'-BIOTIN-CCCTATACCTAATCTATC 5'-AGTTTGGATTTTGGGGGAGG 5'-GTTTTGAYGTAGAY GTTTTATTAGGGT 5'-GGGTTAGTTTTGTGGTTT

MTHFR (antisense) MLH1 (sense) MLH1 (antisense) RASSF1A (sense)

RASSF1A (antisense)

5'-BIOTIN-CAACTCAAT AAACTCAAACTCCCC

Modified sequence/corresponding unmodified sequence analysed* 5'-YGYGGGTTTTGAGTYGTTYGYGYGYGYG 5'-CGCGGGCCCTGAGCCGCCCGCGCGCGCG 5'-RCCCTACTTCRACTCRCRCACRATACR 5'-GCCCTGCTTCGGCTCGCGCACGGTGCG 5'-YGGTATGAGAGATTTYG GGAGAAGATGAGGYGGYG 5'-CGGCATGAGAGACTCCG GGAGAAGATGAGGCGGCG 5'-YGYGYGTTYGTYGTTYGTTATATATYGTTYG 5'-CGCGCGCTCGCCGTCCGCCACATACCGCTCG 5'-YGTTYGGTTYGYGTTTGTTAGYG TTTAAAGTTAGYG 5'-CGCCCGGCCCGCG CTTGCTAGCGCCCAAAGCCAGCG

*Analysed CpG sites are highlighted. 228 Epigenetics 2009; Vol. 4 Issue 4

Quantitative analysis of methylation in plasma DNA

Table 3 DNA methylation levels in 4 genes and LINE1 in plasma DNA samples Gene
RASSF1A MLH1 MTHFR CDKN2A LINE1 Total

PCR product detected on gel/ samples analyzed

21/21 14/21 16/21 21/21 21/21 93/105

Methylation levels in plasma DNA [mean (minmax)]

2.6 (0.010.2) 3.7 (0.016.7) 47.9 (9.990.3) 2.5 (0.014.0) 73.2 (57.093.0)

% of hypermethylated samples*
11% 64% 25% 14% 73%**

*DNA hypermethylation frequency for RASSF1A, MLH1, MTHFR and CDKN2A was calculated as the percentage of plasma DNA samples with methylation levels above 95th percentile levels in blood samples. **the number indicated corresponds to the percentage of plasma DNA samples exhibiting hypomethylation of LINE1, calculated as the percentage of tumor samples with methylation levels below 95th percentile levels in blood samples.

Table 4 Comparison of qMAMBA with other methods available for methylation analysis in minute amounts of DNA Method Major steps involved
MSP QMSP PMA MEP qMAMBA Bisulphite-conversion; PCR using methylated CpG specific primers Bisulphite-conversion; quantitative-PCR using methylated CpG specific primers Bisulphite-conversion; PCR using primers specific for modified DNA; pyrosequencing Bisulphite-conversion; PCR using methylated CpG specific primer; pyrosequencing Bisulphite-conversion; genome-wide DNA amplification; PCR with primers specific for modified DNA; pyrosequencing

Feasibility with low amounts of DNA

YES YES YES

Analysis across multiple CpG sites

NO NO YES

Validation step
NO NO YES

Quantitative analysis of DNA methylation

NO YES YES

YES

YES

YES

NO

YES

YES

YES

YES

References
1. Chen XQ, Stroun M, Magnenat JL, Nicod LP, Kurt AM, Lyautey J, et al. Microsatellite alterations in plasma DNA of small cell lung cancer patients. Nat Med 1996; 2:1033-5. 2. Diehl F, Schmidt K, Choti MA, Romans K, Goodman S, Li M, et al. Circulating mutant DNA to assess tumor dynamics. Nat Med 2008; 14:985-90. 3. Herceg Z. Epigenetics and cancer: towards an evaluation of the impact of environmental and dietary factors. Mutagenesis 2007; 22:91-103. 4. Laird PW. The power and the promise of DNA methylation markers. Nat Rev Cancer 2003; 3:253-66. 5. Esteller M. Cancer epigenomics: DNA methylomes and histone-modification maps. Nat Rev Genet 2007; 8:286-98. 6. Sanchez-Cespedes M, Esteller M, Wu L, Nawroz-Danish H, Yoo GH, Koch WM, et al. Gene promoter hypermethylation in tumors and serum of head and neck cancer patients. Cancer Res 2000; 60:892-5. 7. Rosas SL, Koch W, da Costa Carvalho MG, Wu L, Califano J, Westra W, et al. Promoter hypermethylation patterns of p16, O6-methylguanine-DNA-methyltransferase, and death-associated protein kinase in tumors and saliva of head and neck cancer patients. Cancer Res 2001; 61:939-42. 8. Cairns P, Esteller M, Herman JG, Schoenberg M, Jeronimo C, Sanchez-Cespedes M, et al. Molecular detection of prostate cancer in urine by GSTP1 hypermethylation. Clin Cancer Res 2001; 7:2727-30. 9. Belinsky SA, Palmisano WA, Gilliland FD, Crooks LA, Divine KK, Winters SA, et al. Aberrant promoter methylation in bronchial epithelium and sputum from current and former smokers. Cancer Res 2002; 62:2370-7. 10. Bearzatto A, Conte D, Frattini M, Zaffaroni N, Andriani F, Balestra D, et al. p16(INK4A) Hypermethylation detected by fluorescent methylation-specific PCR in plasmas from non-small cell lung cancer. Clin Cancer Res 2002; 8:3782-7.

11. Kawamoto K, Enokida H, Gotanda T, Kubo H, Nishiyama K, Kawahara M, Nakagawa M. p16INK4a and p14ARF methylation as a potential biomarker for human bladder cancer. Biochem Biophys Res Commun 2006; 339:790-6. 12. Belinsky SA, Klinge DM, Dekker JD, Smith MW, Bocklage TJ, Gilliland FD, et al. Gene promoter methylation in plasma and sputum increases with lung cancer risk. Clin Cancer Res 2005; 11:6505-11. 13. Shaw RJ, Akufo-Tetteh EK, Risk JM, Field JK, Liloglou T. Methylation enrichment pyrosequencing: combining the specificity of MSP with validation by pyrosequencing. Nucleic Acids Res 2006; 34:78. 14. Tost J, Dunker J, Gut IG. Analysis and quantification of multiple methylation variable positions in CpG islands by Pyrosequencing. Biotechniques 2003; 35:152-6. 15. Colella S, Shen L, Baggerly KA, Issa JP, Krahe R. Sensitive and quantitative universal Pyrosequencing methylation analysis of CpG sites. Biotechniques 2003; 35:146-50. 16. Yang AS, Estecio MR, Doshi K, Kondo Y, Tajara EH, Issa JP. A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res 2004; 32:38. 17. Mill J, Yazdanpanah S, Guckel E, Ziegler S, Kaminsky Z, Petronis A. Whole genome amplification of sodium bisulfite-treated DNA allows the accurate estimate of methylated cytosine density in limited DNA resources. Biotechniques 2006; 41:603-7. 18. Peluso M, Hainaut P, Airoldi L, Autrup H, Dunning A, Garte S, et al. Methodology of laboratory measurements in prospective studies on gene-environment interactions: the experience of GenAir. Mutat Res 2005; 574:92-104. 19. Eckhardt F, Lewin J, Cortese R, Rakyan VK, Attwood J, Burger M, et al. DNA methylation profiling of human chromosomes 6, 20 and 22. Nat Genet 2006; 38:1378-85. 20. Eads CA, Danenberg KD, Kawakami K, Saltz LB, Blake C, Shibata D, et al. MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res 2000; 28:32. 21. Riboli E, Kaaks R. The EPIC Project: rationale and study design. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol 1997; 26:6-14.

www.landesbioscience.com

Epigenetics

229

Quantitative analysis of methylation in plasma DNA

22. Slimani N, Kaaks R, Ferrari P, Casagrande C, Clavel-Chapelon F, Lotze G, et al. European Prospective Investigation into Cancer and Nutrition (EPIC) calibration study: rationale, design and population characteristics. Public Health Nutr 2002; 5:1125-45. 23. Gormally E, Vineis P, Matullo G, Veglia F, Caboux E, Le Roux E, et al. TP53 and KRAS2 mutations in plasma DNA of healthy subjects and subsequent cancer occurrence: a prospective study. Cancer Res 2006; 66:6871-6. 24. Vaissiere T, Hung RJ, Zaridze D, Moukeria A, Cuenin C, Fasolo V, et al. Quantitative analysis of DNA methylation profiles in lung cancer identifies aberrant DNA methylation of specific genes and its association with gender and cancer risk factors. Cancer Res 2009; 69:243-52.

230

Epigenetics

2009; Vol. 4 Issue 4