Current methylation screening methods

DNA methylation is the covalent addition of a methyl group to the cytosines within a 5´-CpG-3´ dinucleotide. More than 25 years ago, the methylation of the cytosines of the promoter sequences was shown to play an important role in the regulation of transcriptional activity of the genes [1]. Subsequently, it was suggested that methylation might contribute to tumorigenesis [2]. Today, methylation-dependent gene-expression aberrations have been demonstrated to contribute to all of the typical hallmarks of cancer [3]. At the same time, increasing evidence shows that aberrant methylation of cancer-related genes could be used as a new target for effective early diagnosis and treatment of not only cancer but also many other diseases.

Aberrant methylation of the cancer genes occurs in the early stages of carcinogenesis, and tumor DNA has long been shown to circulate in blood flow and other body fluids. That opens the possibility of early detection of the neoplastic process in simple, for example, blood-based, tests, once powerful and tumor-specific biomarkers have been discovered [4]. Furthermore, methylation changes, due to their nature, are potentially reversible, and therefore new compounds targeting the methylation changes can be used in the treatment of neoplastic disease [5].

The fundamental discoveries within both the fields of DNA methylation and genetics have occurred at the same time. However, for years, the field of methylation studies has lagged behind the developments within genetics due to the technological limitations and difficulties of methods for studies of methylation changes.

The first technologies allowing for the investigation of methylation phenomenon relied on the endonucleases digesting the DNA, depending on the methylation status of the restriction site. The enzymatic digest was usually followed by Southern blotting. However, the fact that these methodologies require large amounts of DNA has been limiting their use, especially in cancer studies. Along with enzymatic approaches, at the beginning, field studies of total content of methylated cytosines in genomes were based on high-performance liquid chromatography and high-performance capillary chromatography techniques [6,7]. Those studies revealed the phenomena of the overall significant hypomethylation of cancer genomes. Currently, the protocols based on endonucleases and Southern blots are rarely used. However, methylation-sensitive endonucleases digestion has been combined with real time (RT)-PCR technologies and integrated into genome-wide methylation study protocols (discussed below). The quantitative analysis of DNA methylation using RT-PCR (qAMP) technique is an example of the combination of methylation-sensitive endonuclease digestion and RT-PCR, where RT-PCR reveals a presence or absence of a digestion within the locus of interest [8].

During in vitro amplification of genomic DNA, all methylation information is lost. Therefore, in order to study methylation marks, the methylation pattern has to be preserved prior to the use of PCR. Revolutionary for the field was the use of sodium bisulfite, which allows to preserve methylation marks during PCR amplification. In principle, sodium bisulfite deaminates cytosines to uracil, while methylated cytosines remain unchanged. During the subsequent PCR amplification, retained 5-methylcytosines are amplified as cytosines, whereas thymine is incorporated at unmethylated cytosines sites.

Current PCR-based methods for investigation of the methylation status of CpG sites within a locus if interest can be divided into two groups. One group of the technologies utilizes PCR for a specific amplification of methylated (or unmethylated) templates. The second group of protocols is based on simultaneous PCR amplification of templates originating from both methylated and unmethylated alleles, and investigation of the methylation status of the locus of interest by a secondary post-PCR method.

One of the first methods combining bisulfite modification and locus-specific PCR amplification was methylation-specific PCR (MSP) [9]. In this protocol, the methylation status of the locus was revealed by the presence or absence of the PCR product on standard agarose gel. The method requires a separate PCR with primers targeting unmethylated template to confirm unmethylated status of the locus. The MSP principle was adopted in the MethyLight™ technique, where primers specific for the methylated template flank a fluorescent probe [10]. The real-time nature of the MethylLight assay allows for the quantification of the levels of fully methylated template. An innovative variant of the above technology is the HeavyMethyl^® assay, in which the methylation-specific blockers are used along with methylation-specific primers and probe [11]. The blockers prevent mispriming and amplification of the unmethylated template and, hence, increase the sensitivity of the assay in detection of the methylated template. All the above methodologies interrogate only CpG sites within the primer/probe binding site and are based upon the assumption of a perfect match primer binding that allows only for the detection of a fully methylated variant of the template.

A separate group of the methodologies is based upon the principle of simultaneous amplification of methylated and unmethylated templates using one primer pair, and the subsequent investigation of the origins of the PCR product via secondary methods. Many secondary methods have been adopted over the years to distinguish (different in base composition) PCR products originating from methylated and unmethylated templates. One of the first methodologies utilizing the above principle was bisulfite sequencing, where bisulfite-treated template is subjected to dideoxynucleotide-based sequencing reaction and the methylation status of each CpG site of the amplicon is interrogated [12]. The nonquantitative drawback of this method can be overcome by combining it with cloning of the PCR product and sequencing of the subset of the clones; however, that makes this protocol very time- and labor-consuming. Recently, a new sequencing technology named pyrosequencing has been developed. Pyrosequencing is based on the luminometric detection of pyrophosphate release, which follows the nucleotide incorporation [13] and can potentially reveal in a quantitative manner the methylation status of a single CpG site. The quantitative data on each of the CpG sites within the PCR product can also be obtained by applying base-specific and matrix-assisted laser desorption/ionization time-to-flight spectrometry (MALDI-TOF MS) [14]. However prior to the MALDI-TOF MS analyses, the locus of interest has to be PCR amplified from bisulfite-modified DNA, in vitro RNA transcribed and RNaseA-base specifically cleaved.

The high-resolution melting (HRM) technology was recently adopted for cost- and labor-efficient methylation screening [15]. The method relies upon sequence-dependent melting properties of the PCR products in a denaturing gradient. The PCR products derived from methylated and unmethylated bisulfite-modified template have different base compositions, and therefore can be easily distinguished in a temperature gradient in the presence of DNA intercalating dye. Two other techniques developed prior to the methylation-sensitive HRM protocol, methylation-specific denaturing gradient gel electrophoresis (MS-DGGE) and methylation-specific denaturing high-performance liquid chromatography (MS-DHPLC), utilize similar principles to HRM analysis, but use different denaturing matrixes – polyacrylamide gels with a gradient of denaturing agents and reverse phase chromatography support under partly denaturing conditions, respectively [16,17]. One more commonly used technology based on nondiscriminatory PCR amplification of bisulfite-modified template is combined bisulfite restriction analysis (COBRA), where the PCR product is digested with endonucleases to the determine methylation levels of the specific loci [18].

The design of primer sequences specifically amplifying methylated/unmethylated sequence is relatively easy. The design of the primer pair that allows for the proportional and simultaneous amplification of methylated and unmethylated bisulfite-modified templates can be difficult for the unwary. Over the years, it has been noted on many occasions that the techniques utilizing one primer set were lacking sensitivity [19,20]. The lack of sensitivity was attributed to the PCR bias phenomena, which is described as the preferential amplification of certain DNA sequences. In methylation studies, PCR bias has been shown to favor the unmethylated (T-rich) variant of the bisulfite-modified template and cause under-amplification of the methylated allele [21–23]. As a consequence of the PCR bias, the post-PCR methods are not able to detect PCR product amplified from methylated template [23]. We have recently proposed a new strategy for primer design, which allows compensating for PCR bias, and the highly sensitive detection of sequences originating from methylated template [24]. The primers designed according to the new rules allow for the uniform detection of 0.1–1% methylation levels, which is similar to the detection levels of technologies based on locus-specific amplification (MSP) [23].

The last group of technologies relying on bisulfite modification are technologies that aim to investigate the methylation levels of single cytosines within CpG dinucleotides. Most of those techniques are adaptations of technologies previously used in the typing of SNPs, such as technology based on the principle of single nucleotide primer extension (SNuPE) [25]. One of the first applications of the above principle in interrogation of single CpG sites was MS–SNuPE [26]. Currently, many modifications of the initial protocol have been reported. In principle, all of them integrate different detection formats of the products of the SNuPE reaction [27–29].

One technology, which can potentially be adapted to screening either methylation status of single CpG sites or multiple sites within the locus of interest, is methylation-specific single-strand conformation analysis (MS–SSCA). The protocol consists of the amplification of the bisulfite-modified template, denaturation of the products and subsequent electrophoresis on nondenaturing polyacrylamide gels [30]. Since the technique was initially developed for the mutation screening in the genomic DNA, it has the power to effectively distinguish the methylation of single CpG sites. One further technique allowing for identification of methylation levels of single CpG sites is MethylQuant [31]. The technique relies on the comparison of the real-time PCR reactions, one of which amplifies target sequence irrespective of its methylation status, and the other amplifies only methylated sequence.

The above-described technologies focus on the investigation of the methylation status of a single locus of interest and, in most of the cases, are used when the location of the sequence of interest is known. Single-locus methods are not suitable for the high-throughput and genome-wide biomarker discovery applications. However, due to the significant technological progress in the field, methods are now available for genome-wide screening of methylation changes. Most of the genome-wide approaches focus on high-throughput microarray platforms. Three major techniques of methylation analysis have been combined with microarray platforms – bisulfite sequencing, methylation sensitive digestion and immunoprecipitation.

The hybridization of bisulfite-modified DNA to an array of nucleotides interrogating single CpG sites is an array-based technology with potentially a single CpG resolution. Two other techniques, methylation-sensitive digestion and immunoprecipitation, allow for the isolation/enrichment of differently methylated regions and the subsequent interrogation of those on microarray platforms following the principles of comparative genomic hybridization. Many variations of the methylation-sensitive enzymes and proteins binding to methylated sequences have been used for the discovery of new deferentially methylated regions. Nevertheless, all the findings from genome-wide screenings have to be validated via a single-locus method before the conclusion on the new biomarker discovery can be put forward.

The variety of technologies available for methylation studies allow researchers today to choose the one suitable for both the nature of the sample used in the experiments and the questions to be answered by the experiment. However, each of the techniques has to be treated with care and none of them is foolproof. Potential artifacts, such as PCR bias, can compromise the results of the experiments, and therefore each protocol has to be carefully optimized and validated before used in experiments and the conclusion from the results are drawn.

Acknowledgements

I would like to thank Dr Lise Lotte Hansen and Dr Alexander Dobrovic for the critical review of this editorial. At the same time, I would like to thank the Lundbeck, Toyota and Harboe Foundations and Roche Diagnostics for the support of my work.

Financial & competing interests disclosure

The author is listed as a co-inventor on a patent application regarding the aspects of MS-HRM technology. The patent application has been filed for by the University of Aarhus (Aarhus, Denmark) and Peter MacCallum Cancer Centre (VIC, Australia). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.