Methylenetetrahydrofolate Reductase CpG Islands: Epigenotyping

Abstract

Background

Determination of the differential DNA methylation patterns of methylenetetrahydrofolate reductase (MTHFR) that are associated with differential MTHFR activity is important to understand the pathogenesis of ischemic stroke. However, to date, no data are available on the differential DNA methylation profiles of Kelantanese Malays. Therefore, we developed a rapid and efficient serial pyrosequencing assay to determine differential DNA methylation profiles of MTHFR, which help to further our understanding of the pathogenesis of ischemic stroke. The developed assay also served as the validation platform for our previous computational epigenetic research on MTHFR.

Methods

Polymerase chain reaction primers were designed and validated to specifically amplify the cytosine that is followed by guanine residues (CpGs) A and B regions. Prior epigenotyping on 110 Kelantanese Malays, the serial pyrosequencing assays for the CpGs A and B regions were validated using five validation controls. The mean values of the DNA methylation profiles of CpGs A and B were calculated.

Results

The mean DNA methylation levels for CpGs A and B were 0.984 ± 0.582 and 2.456 ± 1.406, respectively. The CpGs 8 and 20 showed the highest (5.581 ± 4.497) and the lowest (0.414 ± 2.814) levels of DNA methylation at a single‐base resolution.

Conclusion

We have successfully developed and validated a pyrosequencing assay that is fast and can yield high‐quality pyrograms for DNA methylation analysis and is therefore applicable to high throughput study. Using this newly developed pyrosequencing assay, the MTHFR DNA methylation profiles of 110 Kelantanese Malays were successfully determined. It also validated our computational epigenetic research on MTHFR.

INTRODUCTION

Methylenetetrahydrofolate reductase (MTHFR) is a flavin adenine dinucleotide‐dependent enzyme that can either reduces 5, 10‐methylenetetrahydrofolate to 5‐methyltetrahydrofolate or oxidises 5, 10‐methylenetetrahydrofolate to 10‐formyltetrahydrofolate 1. The 5‐methyltetrahydrofolate is required for the biogenesis of the universal methyl donor S‐adenosylmethionine, which can donate a methyl group to the 5′ position of a cytosine that is followed by guanine residues (CpG) during DNA methylation in eukaryotes 2.

DNA methylation is important for the regulation of mammalian genes and the maintenance of chromatin structure and imprinting, not only for normal development, but also in the development of pathological diseases 3. Aberrant DNA methylation profiles are associated with ischemic stroke and atherosclerosis, especially in relation to MTHFR gene. Our recent computational epigenetics study 1 confirmed that DNA methylation of MTHFR is indeed associated with ischemic stroke. The result is concordance to Castro et al. 4, who demonstrated that the level of hyperhomocysteinemia, which is higher than 75 μmol/l, is associated with a 35% reduction in DNA methylation and the risk for atherosclerosis. This observation indicates that increased risk of ischemic stroke may be explained, in part, by the effects of differential DNA methylation profiles on brain vulnerability in response to ischemic injury 5.

With the advancement of DNA sequencing technologies, bisulfite conversion followed by sequencing or pyrosequencing has become increasingly popular for investigating the epigenetic profiles of the human genome. Pyrosequencing is based on the concept of sequencing through synthesis, facilitating both qualitative and quantitative DNA methylation analyses that detect the intensity of luminescence proportional to the amount of nucleotide present. The intensity of pyrophosphate (PPi) is dependent on the incorporation of nucleotides into the complementary DNA strand 6. The CpG methylation status is read as a C/T single nucleotide polymorphism or is represented as Y in the pyrogram, where the percentage of methylation is calculated from the ratio of the intensity of the methylated Cs to the combination of the intensities of the Cs and Ts at each CpG, as the incorporation of Cs and Ts indicate methylated and unmethylated states, respectively 6.

Pyrosequencing employs four enzymatic reactions and two substrates in a single‐closed tube system for real‐time DNA synthesis 7, using Escherichia coli DNA polymerase, Saccharomyces cerevisiae adenosine‐5′‐triphosphate (ATP) sulfurylase, Photinus pyralis luciferase, and Solanum tuberosum apyrase, and the two substrates adenosine 5′ phosphosulfate (APS) and luciferin 7. During pyrosequencing, the E. coli DNA polymerase incorporate the first nucleotide into the complementary strand based on the order of dispensation generated from the sequences analyzed. This event is accompanied by the release of PPi in an equimolar concentration to the amount of the incorporated nucleotide. The release of pyrophosphate (PPi) activates S. cerevisiae APS to quantitatively convert PPi to adenosine triphosphate (ATP). In general, 1 pmol of DNA yields 6 × 10¹¹ ATP molecules, generating more than 6 × 10⁹ photons 7, 8, 9.

The ATP produced is detected by the thermally unstable P. pyralis luciferase, which oxidizes luciferin to an excited state oxyluciferin. The subsequent decay of oxyluciferin is accompanied by the production of visible light (λ = 560 nm), which is detected using a charge‐coupled camera and represented as a peak in the Pyrogram®, where each peak is proportional to the amount of nucleotides incorporated. The nucleotide‐degrading enzyme S. tuberosum apyrase continuously degrades unincorporated nucleotides and ATP before the addition of other nucleotides. As the process continues, a complementary sequence is generated and the synthesized nucleotide sequence is determined from the Pyrogram® 7, 8, 9.

Regardless of detection methods, to our knowledge, differential DNA methylation has not been fully elucidated among Kelantanese Malays, which is pertinent for the further understanding of ishcemic stroke pathogenesis, especially among people in this ethnic group. Herein, we have designed a serial pyrosequencing assay to investigate the DNA methylation profiles of MTHFR (CpGs A and B) that can also be adapted for higher throughput studies. In addition, the developed assay can also serve as the validation platform for computational epigenetics analysis of MTHFR 1. The flow of the study is shown in Appendix.

MATERIALS AND METHODS

Study Design

This prospective study was conducted at the Hospital Universiti Sains Malaysia between October 2010 and March 2012, and the serial pyrosequencing analysis was performed in Griffith University, Australia. The study was conducted in accordance with the local institutional ethics committee and complies with the Declaration of Helsinki. All Kelantanese Malays were confirmed by three generations and were healthy subjects (n = 110) between 18 and 55 years of age. The subjects were randomly selected healthy volunteers with no previous personal or family history of stroke, brain tumors, or other forms of cancer. Subjects who had taken any form of vitamin B supplementation 6 months prior to the study were excluded before informed written consent was obtained.

Bisulfite Conversion

Twenty microliters of genomic DNA (50–100 ng) was bisulfite‐converted using the EZ DNA Methylation‐Gold® Kit (Zymo Research, Irvine, CA) that combines bisulfite conversion and DNA clean up with a spin protocol.

Bisulfite‐converted DNA was qualitatively investigated using agarose gel electrophoresis. A total volume of 4 μl of bisulfite‐converted DNA (10–50 ng) and 0.8 μl of 6× NEB® blue loading dye (containing 2.5% Ficoll 400, 11 mM ethylenediaminetetraacetic acid, 3.3 ml Tris‐HCl, 0.017% sodium dodecyl sulfate, and 0.015% bromophenol blue, pH 8.0; NEB®, Hitchin, Herts, England) were mixed before loading into a 2% ethidium bromide‐stained agarose gel (Bioline®, London, UK). DNA without bisulfite conversion was used as the positive control for each agarose gel electrophoresis run. A 100 bp DNA ladder (containing 10 mM Tris‐HCl, 1 mM EDTA, pH 8.0; NEB®) at 0.5 μg/lane was loaded, and the agarose gel was run at 75 V for 1 hr.

Primer Design and Validation

The methylation‐independent primers and sequencing primers for CpG A were designed according to the commercial PyroMark® CpG Assay (Qiagen, Hilden, Germany). The primers for CpG B were designed using PyroMark® Assay Design Software (version SW 2.0, Qiagen) to obtain a pair of methylation‐independent primers with an optimal melting temperature and high primer set score. The properties of the methylation‐independent primers are shown in Table 1. Together with the methylation‐independent primers, three independent sequencing primers for CpG B were codesigned. Because the forward polymerase chain reaction (PCR) primers for CpG B were biotinylated, the sequencing primers were designed on the complementary reverse strand (Table 2).

Table 1.

Methylation‐Independent Primers for CpG B

Methylation‐independent primers	Sequence (5′–3′)	Length (base)	Temperature (°C)	GC content (%)	Delta G (kcal/mol)	Score (%)
CpG B_F	GGGAGGTTGTTTGTAGGTTTTAGGATTTTT	30	62.8	36.7	1.60–2.59	74
CpG B_R	CCCCCCTCTCCAACAACCTAACACC	25	63.3	60.0	−0.96 to 0.00	74

The score was retrieved from the PyroMark® Assay Design Software (Qiagen). The CpG B_F primer was biotinylated.

Table 2.

Sequencing Primers for CpG B

Sequencing primers	Sequence (5′–3′)	Length (base)	Temperature (°C)	GC content (%)	Delta G (kcal/mol)
CpGB_1st SQN	ACCCAACCTTCTAAATCACCC	21	54.7	47.6	2.33–2.68
CpGB_2nd SQN	CCCTACAACCACCACAATAA	20	52.2	45.0	2.69–3.28
CpGB_3rd SQN	CCAACAACCTAACAC	15	46.7	43.3	−0.96 to 0.00

The properties of the sequencing primer were determined using an OligoAnalyzer version 3.1.

The designed CpG B primer (180 bp) was aligned against a genomic reference sequence using the Basic Local Alignment Search Tool (BLAST; http://blast.ncbi.nlm.nih.gov/Blast.cgi) (21) to detect the presence of single nucleotide polymorphisms in the primer annealing sites or within the amplicon. The designed primer was further evaluated using a standard primer design parameter to assess the thermodynamic properties and potential for the formation of secondary structures, using OligoAnalyzer version 3.1 (http://www.idtdna.com/analyser/Applications/OligoAnalyzer/) (22). In addition, the predicted CpG B amplified product was aligned against the bisulfite‐modified DNA database, methBLAST (http://medgen.ugent.be/methBLAST/) (23), to confirm the specificity of the product.

Amplification of CpG Regions

The optimized PCR reaction for CpG A was performed in a total volume of 15 μl, containing 6.89 μl of highly purified water, 1X colorless GoTaq® Flexi Buffer (Promega, Madison, WI), 1.6 mM magnesium chloride solution (Promega), 2.0 units (U) of GoTaq® Hot Start Polymerase (Promega), 200 μM deoxynucleotide solution mix (NEB®), 1X PyroMark® CpG Assay PCR Primers (Qiagen), and 20 ng of bisulfite‐converted DNA template. A similar concentration of each reagent component was used for the PCR amplification of CpG B except that the concentration of each forward and reverse primer (IDT, Coralville, IA) was 5 pmol in a total volume of 15 μl.

Three negative controls containing highly purified water instead of genomic DNA were included in each PCR reaction. The reaction mixture was gently tapped and spun down before placing into the Applied Biosystems Veriti® 96‐well Thermal Cycler (Life Technologies, Mulgrave, Victoria, Australia). The PCR cycling conditions for CpG A were initiated at 95°C for 15 min to facilitate unbinding of the antibody to initiate the hot start PCR. Cycling was continued with 45 cycles at 95°C for 30 sec, 57.4°C for 30 sec, and 72°C for 30 sec. The final elongation step was performed at 72°C for 7 min. The initial denaturation and final elongation steps for CpG B amplification were similar to that of CpG A. However, the cycling conditions for CpG B were slightly modified, where 45 amplification cycles were performed at 95°C for 30 sec, 57.4°C for 45 sec, and 72°C for 45 sec.

In addition, all of the primer sets were tested using a nonbisulfite‐converted DNA template under similar conditions as described above. The absence of amplified products in agarose gel electrophoresis indicated that the PCR conditions did not amplify unconverted DNA templates 10.

Pyrosequencing and Validation

The CpG assay design and setup for the analysis of multiple consecutive CpG positions using a PyroMark Q24 browser (Qiagen) was conducted using the default values. Assay design using these targeted settings was applied for both CpGs A and B (Table 3). The CpG assay was analyzed using the samples from Kelantanese Malays and the assay was validated using five validation controls, based on the manufacturer's recommendations.

Table 3.

Sequence for CpGs A and B Analyses

Assay	CpG A
Sequence to analyze	GGTTATTGAGTTATYGATGGGGGYGAGGATAYGGGT
Dispensation order	TGTCATGATCGATATCGAGTGGTCGAGATATCG
Prebisulfite converted DNA sequence	GGTCACTGAGTCACCGATGGGGGCGAGGAYACGGGC
Assay	CpG B_1st SQN
Sequence to analyze	RAAACTAAAAAATAAATAACRACRAAACCRAAATCRCCRAAAAAAAAATCCTA
Dispensation order	CAAGACTAATGATATCAGATCAAGATCAAGATACGATCAAGAATCT
Prebisulfite converted DNA sequence	CGGCTCCAGGATCTCCCTCCCGGCGACCCCGGCCTCGCCGTCACTCACCCTCC
Assay	CpG B_2nd SQN
Sequence to analyze	TACRACCRACRACCRAACRTTCT
Dispensation order	ATGATCAGATCAGATCAGATCAGATCGATC
Prebisulfite converted DNA sequence	CAGAACGCTCGGCCGCCGGCCGCA
Assay	CpG B_3rd SQN
Sequence to analyze	CTACRCCRCRCCCCTTCACTACRTTCCCCRCCCCTACAACRACCACAATAATAC
Dispensation order	ACTGATCGATCGATCGACCTCACTATCGATCCGACCTACATCAGACAC
Prebisulfite converted DNA sequence	CGCTGCAGGGGCGGGGAACGCAGTGAAGGGGCGCGGCG

The variable position of 5mC was represented with Y and R in the sequence to be analyzed; as G or C in both the dispensation order and in prebisulfite converted DNA sequence. The Y residue was the IUPAC for C or T residue in the forward strand, while the R residue was the IUPAC for A or G residue in the reverse complementary strand. The targeted internal control was represented with blue bolded A, T, or C residues.

Fifteen microliters of PCR product was used as the pyrosequencing template and was prepared according to the manufacturer's instruction, with some slight modifications. Serial pyrosequencing was performed to ensure coverage of all CpG positions within the 180 bp amplicon in the CpG B. A total volume of 15 μl of binding buffer containing sepharose beads was resuspended in the completed sequencing reaction, and the template preparation step was repeated for subsequent pyrosequencing.

Statistical Analysis

The DNA methylation profile of 110 Kelantanese Malays was analyzed using SPSS software version 2.0 (Armonk, CA). The mean DNA methylation of CpGs A and B, as well as DNA methylation at single base resolution, were determined.

RESULTS

The designed serial pyrosequencing assay was successfully epigenotyped using the 110 Kelantanese Malays samples. The pyrograms of CpGs A (Fig. 1) and B (Fig. 2) confirmed the feasibility and reproducibility of this method. Furthermore, the validation controls did not show any significant peaks in the pyrogram (Fig. 3), indicating that the designed sequencing and methylation‐independent primers were accurate and that the PCR amplification reactions were successful.

Figure 1.

Pyrogram of CpG A. The internal control is highlighted with a light blue background at Cs. The highlighted blue background indicates the variables position of Y (C/T), where the percentage indicates the degree of sample methylation.

Figure 2.

Pyrogram of CpG B. (A–C) The pyrograms of three sequencing primers covering the entire CpG B. The highlighted yellow background represents the internal control within the analyzed sequence.

Figure 3.

Pyrogram of validation controls. The figure represents five validation controls of the CpG A assay. The validation controls showed no significant peaks in the pyrogram.

Prior to performing pyrosequencing, the quality of the bisulfite‐converted DNA was investigated using agarose gel electrophoresis. An agarose gel with a clear band located above 1571 bp indicated the presence of bisulfite‐converted DNA (Fig. 4A). There was no smearing or fragmentations observed within the lane, indicating that bisulfite was successfully converted. Subsequent PCR amplifications were qualitatively investigated by agarose gel electrophoresis (3%) using 4 μl of the amplified product, where each band corresponded to 232 bp and 180 bp for CpGs A and B, respectively (Fig. 4B).

Figure 4.

Qualitative evaluation of bisulfite‐converted DNA and PCR amplifications. (A) Lane 1 contains the NEB® DNA ladder (100 bp), lanes 2 and 3 contain bisulfite‐converted DNA, lane 4 contains nonbisulfite‐converted genomic DNA as a positive control for bisulfite conversion, and lane 5 contains the negative control. (B and C) Lane 1 contains NEB® DNA ladder (100 bp), three negative controls were included in each electrophoresis run. With respect to (B) and (C), lanes 2 onward contain the PCR products for CpGs A and B, respectively.

The developed assay validated the computational epigenetics analysis of MTHFR 1 and suggested that CpGs A and B were showing differential DNA methylation profiles among ischemic stroke patients (Table 4). In particular, the mean DNA methylation levels at CpGs A and B were 0.984 ± 0.582 and 2.456 ± 1.406, respectively. The DNA methylation levels of 3 and 17 sites of CpGs A and B at the single base resolution were shown in Table 4. Of the CpG sites, CpGs 8 and 20 showed the highest (5.581 ± 4.497) and the lowest (0.414 ± 2.814) levels of DNA methylation, respectively (Table 4).

Table 4.

Mean Methylation Levels at CpGs A and B

CpG islets	CG dinucleotides	Kelantanese Malays (n = 110)
CpG A	CpG1	1.023 ± 0.695
	CpG2	0.735 ± 0.512
	CpG3	1.195 ± 0.795
CpG B	CpG4	1.354 ± 2.279
	CpG5	0.738 ± 2.729
	CpG6	1.364 ± 2.560
	CpG7	1.555 ± 2.474
	CpG8	5.581 ± 4.497
	CpG9	4.697 ± 0.770
	CpG10	1.010 ± 1.530
	CpG11	2.408 ± 2.798
	CpG12	3.202 ± 4.795
	CpG13	3.276 ± 4.087
	CpG14	3.149 ± 3.913
	CpG15	2.559 ± 1.929
	CpG16	0.630 ± 0.741
	CpG17	2.952 ± 1.669
	CpG18	2.104 ± 2.511
	CpG19	2.614 ± 3.607
	CpG20	0.414 ± 2.814

DISSCUSSION

We have successfully developed a serial pyrosequencing assay for the simultaneous detection of DNA methylation profiles of CpG islands located in MTHFR. The pyrosequencing assay was successfully applied to determine the DNA methylation profiles within MTHFR gene in 110 Kelantanese Malays. In addition, the developed assay also validated the computational epigenetic research on MTHFR 1.

Bisulfite conversion could pose several drawbacks in relation to the quantity and quality of the genomic DNA. Therefore, it is pertinent to evaluate the quantity and quality of the bisulfite‐converted DNA prior to DNA methylation. Traditional qualitative evaluations based on agarose gel electrophoresis require large amounts (2 μg) of converted DNA, which is not feasible for most research where only a limited amount of DNA can be obtained from human samples 11. However, our assay requires as little as 10–50 ng of bisulfite‐converted DNA for qualitative evaluation of bisulfite conversion. Nonetheless, the quantitative evaluation of bisulfite‐converted DNA via ultraviolet or fluorescence detection provides adequate quantification, but generates no information on DNA fragmentation and can therefore lead to mispairing of PCR primers or incorrect amplicon size 11.

Therefore, we opted for qualitative evaluation using agarose gel electrophoresis followed by downstream pyrosequencing. Each pyrosequencing run is coupled with an internal control for the confirmation of successful bisulfite conversion. In fact, incomplete double‐stranded DNA denaturation before conversion might lead to the overrepresentation of methylation in the final output because unmethylated Cs are typically deaminated to Us in single‐stranded DNA. Because deaminated Cs cannot base pair with unmodified guanines in the complementary strands, bisulfite‐treated DNA is not complementary and could therefore act as a starting template for PCR 8.

The completeness of bisulfite conversion and the length of the fragmented genomic DNA template could significantly affect the final PCR result 8. Initial sodium hydroxide treatment on genomic DNA typically occurs simultaneously with the thermal denaturation step during bisulfite conversion to produce completely converted DNA 12. The exothermic sulfonation at high temperatures (90–98°C) can maintain single‐stranded DNA. However, in addition to a high concentration of bisulfite and long incubation time, high temperatures can also accelerate DNA fragmentation (Patterson et al., 2012) 13. Harsh bisulfite conversion can lead to the breakage of the glycosidic bonds joining the purine base and the deoxyribose moiety 14. The breakages randomly produce an apurinic site without disturbing the structural integrity of phosphodiester backbone, leading to DNA fragmentation 11, 14.

The fragmentation of DNA generates two main effects in subsequent PCR amplifications, as the available amount of template DNA not only determines the success of the PCR reaction, but is also pertinent for determining the precision of quantitative methylation ratio 11. Due to DNA fragmentation, the corresponding DNA fragments are unavailable for the targeted region and can lead to failed PCR amplifications. In addition, fragmented DNA occurring between the primer‐binding sites can drastically reduce the amount of template DNA for PCR, yielding a result similar to digital PCR 11.

DNA fragmentation during bisulfite conversion produces limited empirical sizes of PCR amplicon of approximately 400–500 bp 15, therefore, the amplicon was limited to 180 bp in the present study. Small amplicon size not only decreases background signals that may interfere with the pyrosequencing reaction, it also reduces the number of failed PCR amplifications. Ideally, the smaller the amplicon size, the better the performance, where 80–200 bp amplicons can improve the resolution of the pyrograms, as they consume only a small number of beads, and it has been reported that the efficiency of the biotinylated amplicon capture is inversely proportional with the amplicon size 8, 11, 16.

As for the primer design, the overlapping of the 5′ and 3′ ends of the primers with CpG is discouraged, as overlapping might lead to the preferential amplification of nucleotides and inaccurate methylation levels 8. Kristensen and Hansen 17 showed that the presence of more than two CpG sites in one of the primers lead to the failed amplification of the unmethylated DNA template. However, when at least five bases from the 3′ terminus do not overlap with the retained degenerate positions of CpG and single nucleotide polymorphism, primer slippage not only diminishes, but also results in the unbiased amplification of the region 8.

Therefore, we incorporated a CpG near the 5′ or the 3′ end of the forward primer sequence. This inclusion facilitated the selection of a larger pool of primers for CpG regions, which has an added advantage because they do not selectively amplify the methylated template located in the middle of primer sequence. Furthermore, Wojdacz and Dobrovic (2007) 18 suggested that the CpGs in the primer are necessary to avoid the significant underestimation of the degree of DNA methylation and increase the low analytical sensitivity of methylation‐independent PCR.

It has been reported that manipulating the annealing temperature during amplification neutralizes this bias, irrespective of the presence or absence of CpG on the primers 17. Wojdacz et al. 10 reported that the methylation‐independent primer favors the amplification of methylated primers, where the low annealing temperature and primer will predominantly bind to the methylated template with increased annealing temperature. An intermediate (53–60°C) annealing temperature value is ideal for both methylated and unmethylated templates where amplification biases are compensated 10.

The amplification annealing temperature can be enhanced through primers design, using 18–30 bases and a melting temperature ranging from 62°C to 68°C (Loo et al., 2012) 19. Ideally, primers should have a matched melting temperature value within ± 1°C 10. The melting temperature value not only affects the annealing temperature, but also determines the primer specificity and bisulfite conversion. Placing more GC‐rich sequences at the 5′ end of the primers rather than at the 3′ end, significantly increases the primer specificity to avoid palindromes within and between the complementary sequences and GTs at the 3′ end 8, Loo et al., 2012). Furthermore, the specificity of bisulfite conversion can be determined either by the inclusion of 25% CTs in the primer sequence or by the inclusion of Ts derived from non‐CpG Cs close to the 3′ end of the primer (10, Patterson et al., 2012).

It has been recommended that all designed primers contain delta G values of less than 10 kcal/mol, but higher than −9.0 kcal/mol, as low delta G values (<−9.0 kcal/mol) indicate a high tendency for hairpin structures and primer dimer formation (http://www.idtdna.com/analyzer/Applications/OligoAnalyzer/). The secondary structures might act as additional templates for PCR amplification, which reduces primer specificity (20, Loo et al., 2012). Due to the thermodynamic nature of PCR amplification, the specificity and efficiency of PCR are strongly influenced by the primer:template ratio (Loo et al., 2012). A minimum of 10 ng of bisulfite‐treated template is typically required for reproducible CpG methylation analysis because reduced reproducibility is highly dependent on reduced amounts of template 8. Nonetheless, excessive DNA template might enhance primer mispairing, while low amounts of template reduce the yield of the PCR product due to the difficulty of primers in securing template DNA. In contrast, increasing the number of amplification cycles will not only exhaust excess primer, which is pertinent to reducing or avoiding background signals in the pyrograms, but can also increases amplicon yield. It has been reported that background signals can be enhanced through excess unincorporated biotinylated primers captured on the beads, which act as additional sequencing primers 8.

The basic requirement of PCR primers for pyrosequencing is to biotinylate one of the PCR primers at the 5′ terminus to facilitate the immobilization of high performance streptavidin‐coated sepharose beads (GE Healthcare, Uppsala, Sweden). These primers should be of high performance liquid chromatography grade to avoid competition with free biotinylated PCR primers for bead‐binding sites with the biotinylated PCR products and to improve the signal strength in the pyrogram 20.

The pyrosequencing technology is applicable for the genotyping of single nucleotide polymorphisms, microbial typing via 16S rRNA, analyzing secondary hairpin structure, resequencing, and DNA methylation analysis. Pyrosequencing is useful for medical genetics, population genetics, pharmacogenomics, and forensic DNA analysis. Because pyrosequencing employs a bioluminometric, luciferase‐based enzymatic system for DNA sequencing, it is advisable to perform pyrosequencing at a temperature lower than 28°C to prevent the denaturation of the low thermostable luciferase and avoid background noise in the pyrogram 9. Therefore, single‐stranded DNA templates are utilized to increase the annealing specificity of the sequencing primer (Ronaghi, 2001 21; 9).

The sequencing primer was designed to attach to the opposite direction relative to the biotinylated PCR primer that includes at least one C that was not followed by a G, to act as the bisulfite conversion control 8. Usually, this C is unmethylated and should be fully converted to T during the bisulfite conversion, ideally with no peak observed within this variable position in the pyrogram. The C is then read as a T in the forward strand (or an A in the reverse strand), which was added either before or after the dispensation order at the beginning of the sequence for the analysis. The internal control will be highlighted with a light blue or yellow background in the pyrogram 7.

The signal intensity of a peak is proportional to the number of nucleotides incorporated; therefore, a single peak with at least 10 a.u. and a peak height with at least 40 relative light units is an accurate quantification for the DNA methylation status 7. The peaks in the pyrogram should not be accompanied by any background signals in the blank dispensation and variable positions 7. The sequence within the pyrogram should be consistent with the expected reference sequence retrieved from genomic DNA.

The main limitation of pyrosequencing technology in DNA methylation analysis is the read‐length. Although traditional direct bisulfite sequencing and cloning facilitates the analysis of CpG regions up to 600 bp in a single sequencing run, several sequencing primers are required in pyrosequencing for a region of comparable size 8. This drawback often persuades researchers to opt for direct bisulfite sequencing as a discovery tool 16. Nonetheless, pyrosequencing still outperforms direct bisulfite sequencing due to its rapid speed and high throughput. Therefore, the pyrosequencing method still remains a preferred choice for CpG methylation pattern analysis that requires the quantification of a large number of samples with limited targeted region sizes 8, 16.

CONCLUSION

We have successfully developed a bisulfite PCR with serial pyrosequencing method that significantly shortens the time required to obtain high‐quality pyrograms for DNA methylation analysis. In addition, by using this method, we have also successfully investigated the DNA methylation profiles of MTHFR in 110 Kelantanese Malays. The developed method also validated our previous research on computational epigenetics of MTHFR.

CONFLICTS OF INTEREST

None declared.