Các phương pháp phân tích DNA methylation ngoại bào hiện nay

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/316965450

Current methods of extracellular DNA methylation analysis

ArticleinMolecular Biology · March 2017

DOI: 10.1134/S0026893317010071

CITATIONS

READS

1,993

2 authors:

Olga E Bryzgunova

Institute of Chemical Biology and Fundamental Medicine, Russian Academy of Sci…

66 PUBLICATIONS1,762 CITATIONS

SEE PROFILE

Pavel Petrovich Laktionov

Russian Academy of Sciences, SD, Novosibirsk

278 PUBLICATIONS5,591 CITATIONS

SEE PROFILE

All content following this page was uploaded by Olga E Bryzgunova on 21 March 2018.

The user has requested enhancement of the downloaded file.

167

ISSN 0026-8933, Molecular Biology, 2017, Vol. 51, No. 2, pp. 167–183. © Pleiades Publishing, Inc., 2017.

Original Russian Text © O.E. Bryzgunova, P.P. Laktionov, 2017, published in Molekulyarnaya Biologiya, 2017, Vol. 51, No. 2, pp. 195–214.

Current Methods of Extracellular DNA Methylation Analysis

O. E. Bryzgunovaa, b, * and P. P. Laktionova, b

aInstitute of Chemical Biology and Fundamental Medicine Siberian Division

Russian Academy of Sciences, Novosibirsk, 630090 Russia

bE.N. Meshalkin Siberian Federal Biomedical Research Center, Ministry of Health Care of Russian Federation,

630055, Novosibirsk, Russia

*e-mail: olga.bryzgunova@niboch.nsc.ru

Received February 27, 2016; in final form, March 21, 2016

Abstract⎯The discovery of the enormous role methylated cytosine plays in regulating gene expression has led

to the development of a variety of techniques for detecting cytosine modification. A majority of these tech-

niques are geared towards analyzing genomic DNA, which is typically available in large quantities. The con-

centration of cell-free DNAs (cfDNA) extracted from biological fluids including plasma, saliva, tears, or

urine is relatively low and their degree of the fragmentation is high. Moreover, for noninvasive diagnostics of

cancer, methylation patterns must be studied in minor cancer-specific fractions of DNA molecules substan-

tially diluted by excess unmethylated molecules. The above limitations complicate the application of tradi-

tional techniques for cfDNA methylation analysis. In this manuscript, we review the state-of-art analysis of

cfDNA methylation, hydroxymethylation, and noncanonical methylation (outside of CpG islands). The

review covers methodological approaches to studying individual CpGs and genomic loci, as well as tech-

niques for the large-scale analysis of methylation.

Keywords: methylation, hydroxymethylation, non-CpG methylation, cell-free DNA, restriction endonucle-

ase, emulsion PCR, microarray

DOI: 10.1134/S0026893317010071

INTRODUCTION

Methylation of cytosines in the 5'-position of the

pyrimidine ring is a physiological process that is active

in all normal cells of the body. The proportion of

5-methylcytosine in the human genome is about 1% of

all bases and varies slightly in different types of tissues

with 75% of CpG dinucleotides being methylated [1].

It is known that at least 10% of expressed genes are

regulated by methylation and up to 60% of genes have

local CpG islands [2]. Changing the methylation sta-

tus of the promoter regions of genes involved in onco-

genesis including tumor suppressor genes may lead to

the inactivation (methylation) or activation (demeth-

ylation) of their expression and may promote the

emergence and proliferation of tumor cells. Increased

methylation status that is not detected in the norm is

revealed by the PCR analysis of genes in blood plasma

or serum in pancreatic cancer (Р16, 24.6% of cases)

[3], prostate cancer (RASSF1, RAR

2, GSTP1; 24, 12,

and 13%, respectively) [4], breast cancer (АРС and

GSTP1, 17 and 26%, respectively) [5], hepatocellular car-

cinoma (P16, 19–81%) [6, 7], ovarian cancer (in the

simultaneous analysis of six genes, i.e., BRCA1, RASSF1A,

APC, P14arf, P16ink4a, and DAPK, with 100% specific-

ity and 82% sensitivity [8]), and many others. Some

epigenetic changes (gene hypermethylation) is known

to occur at early stages (IA or B) of cancer [9, 10].

That is why the study of the methylation status of

the genes in cell-free DNA (cfDNA) is an important

task in modern biology. Every year, a growing number

of studies are devoted to analyzing the aberrant meth-

ylation of cytosines in cfDNA of various biological

fluids because this DNA is a convenient source of

diagnostic material for developing systems for early

noninvasive diagnostics and monitoring anticancer

therapy. Not all methodological approaches to study-

ing the methylation of genomic DNA can be applied to

analyzing aberrant methylation in cfDNA. Their use is

limited by the low content of cfDNA (up to 53 ng/mL

and 10–2160 ng/mL in the blood plasma of healthy

Abbreviations: cfDNA, cell-free DNA; MPS, massively parallel

sequencing; BLM-PCR blunt-ended ligation-mediated PCR;

HPLC, high-performance liquid chromatography; MALDI-

TOF MS, matrix-assisted laser desorption/ionization, time-of-

flight mass spectrometry; MeDIP, methylated DNA immuno-

precipitation; MethDet56, microarray-mediated methylation

analysis of 56 fragments in each sample; Methyl-BEAMing,

beads, emulsion, amplification, and magnetics; MIRA, methyl-

ated CpG island recovery assay using MBD2 proteins; MSP,

methylation specific PCR; Ms-SNuPE, methylation-sensitive

single nucleotide primer extension; qMAMBA, quantitative

methylation analysis of minute DNA amounts after whole bisul-

fitome amplification); SMRT, single molecule real-time

sequencing); tSMS, true single molecule sequencing

REVIEWS

UDC 57.088.1

168

MOLECULAR BIOLOGY Vol. 51 No. 2 2017

BRYZGUNOVA, LAKTIONOV

donors [11–15] and in cancer patients [13, 16–18],

respectively, and from 2 ng/mL to 96 ng/mL in the

urine of healthy donors and cancer patients [19–21].

Moreover, as a rule, cfDNA is low molecular DNA;

the size of cfDNA fragments in the blood plasma and

serum of healthy and sick people is most often 100–

200 bp, although there is high molecular cfDNA with

sizes of up to 10000 bp; the sizes of cfDNA fragments

in the urine of men and women are 150–250 bp [16–

20, 22–28]. It should be noted that the content of

tumor DNA in the blood of cancer patients is 1–2% of

the total pool of cfDNA [22]; there is little information

that the content of tumor DNA can achieve 93% [16].

The small amount of cfDNA complicates its analysis

and requires special methodology.

Thus, the methods of studying the methylation of

cytosines in cfDNA must meet the following require-

ments: they should make it possible to work with a

minimum amount of mostly short DNA fragments

and to analyze methylation of minor pool of DNA

molecules (DNA from cancer cells) in the presence of

the excess of ballast DNA (from normal tissues).

Obviously, this task cannot be solved without amplify-

ing initial DNA. For this purpose, both the PCR of

specific predefined loci and the whole-genome ampli-

fication of cfDNA samples may be used. The applica-

tion of the amplification methods for analyzing

cfDNA is determined by the methodology of the sub-

sequent analysis of DNA. These methods will be

described in the corresponding chapters.

All approaches to studying the methylation of both

genomic and cell-free DNA may be divided into three

groups as follows:

(1) chemical methods, which are based on the

selective modification of methylated and nonmethyl-

ated cytosines;

(2) enzymatic methods, which are based on the use

of restriction endonucleases sensitive to m5С in the

recognition site;

(3) mixed methods, which are the combination of

the above approaches.

All methods are used to solve two classes of prob-

lems, i.e., the study of the methylation of selected loci

or methylation on the level of the whole genome. Two

chapters of this review describe the methodological

approaches that are used for solving the tasks of the

first type; the separate chapter describes the methods

of the large-scale study of methylation of extracellular

DNA.

IDENTIFICATION OF METHYLATED

CYTOSINES WITHOUT USING

METHYLATION-SENSITIVE RESTRICTION

ENDONUCLEASES

The method for analyzing the cytosine methylation

status that was developed over 20 years ago is based on

the modification of cytosine with bisulfite [29]. The

studied sample (cfDNA in our case) is treated with

sodium bisulfite, which leads to the complete sulfona-

tion of cytosines without the m5C modification. Sub-

sequent desulfonation under alkaline conditions leads

to the deamination of cytosines and their conversion to

uracils. It is clear that after the conversion and amplifi-

cation, m5C and nonmethylated cytosine are deter-

mined as cytosine [30, 31] and thymine, respectively.

At the next stage, one or both DNA strands are

amplified using primers (containing or not containing

CpG dinucleotides) complementary to converted

DNA (after the deamination of nonmethylated cyto-

sines, DNA loses its ability to complementary interac-

tions). After bisulfite conversion, it is possible to

increase the amount of cfDNA by the whole-genome

amplification as was performed, e.g., in the qMAMBA

method (quantitative methylation analysis of minute

DNA amounts after whole bisulfitome amplification

[32]). The authors believe that this approach makes it

possible to study methylation starting from 100 pg of

DNA. The resulting PCR products were sequenced [31],

pyrosequenced [33], cloned in vector, and the inser-

tions are sequenced [34] or the sequencing was per-

formed by methods of mass parallel sequencing

(MPS) [35]. The use of methylation independent

primers, which are complementary to the regions con-

taining no CpG dinucleotides, makes it possible to

determine the composition and concentration of

DNA. In the case of low amounts of DNA, whole

genome amplification methods are when direct

amplification is impossible or when DNA methylation

is studied by methods such as microarray analysis or

MPS. The advantages of the bisulfite conversion

method include its high sensitivity (from 2 pg of DNA

[36]), the accuracy due to the simultaneous analysis of

initial complementary strands, and the possibility to

study the methylation status of individual DNA mole-

cules. The drawbacks of this method are partial DNA

degradation during modification/deamination [37],

incomplete DNA denaturation in the regions of

extended repeats or complicated secondary structure,

and the incomplete modification of cytosines or

incomplete deamination. These drawbacks may be

minimized by optimizing conditions for the bisulfite

conversion. It should be noted that modern commer-

cial kits for the bisulfite conversion of cfDNA, e.g.,

Zymo Research (United States) and Qiagen (Ger-

many) allow one to efficiently perform this reaction in

a high yield of target DNA [33].

One of the examples is bisulfite sequencing using

primers containing no CpG dinucleotides. This

method allows one to analyze the aberrant methyla-

tion of individual cytosines in cfDNA from biological

fluids [31]. Isolated cfDNA is initially treated with

sodium bisulfite; PCR is then carried out with primers

containing no CpG dinucleotides; the resulting PCR

products are purified by electrophoresis in polyacryl-

amide gel to sequence the desired PCR product. The

MOLECULAR BIOLOGY Vol. 51 No. 2 2017

CURRENT METHODS OF EXTRACELLULAR DNA 169

purified PCR products are sequenced by the Sanger

method using forward and reverse primers. The

advantage of analyzing DNA methylation by Sanger

sequencing is the possibility of studying extended

sequences, while a disadvantage of this method is the

inability to quantify the results.

After bisulfite conversion, cfDNA is analyzed by

the pyrosequencing method, which is based on detect-

ing pyrophosphate released during the synthesis of the

second DNA strand on the studied single-stranded

DNA template. To enhance the sensitivity of the

method, PCR is often used in two stages (nested

PCR) using primers containing no CpG dinucleo-

tides. At the second stage, one biotinylated primer is

used for the isolation of the sequenced strand. The sin-

gle-stranded biotinylated DNA is isolated using sep-

harose beads covered with streptavidin, followed by

pyrosequencing on a PSQ 96MA pyrosequencer using

sequencing primer and the PyroGold SQA reagent kit

(Pyrosequencing AB, Sweden) [33]. On a model sys-

tem, the authors showed that amplification does not

influence the ratio of methylated and unmethylated

DNA. It should be noted that the analysis of DNA by

pyrosequencing can be carried out on 96-, 48-, and

24-well platforms (www.qiagen.com).

After bisulfite conversion, aberrant methylation of

cytosines in cfDNA in the presence of 97–99% of

unmethylated DNA may be detected by one of the

PCR variants. Methylation-specific PCR (MSP) uses

two pairs of primers, one of which is specific to bisul-

fite-converted methylated DNA target sequences, and

the other is specific to unmethylated counterpart

sequences of bisulfite-treated DNA. Thus, the former

primers are only attached to regions containing methyl-

ated cytosines, while the latter primers are only attached

to unmethylated cytosine-containing regions. The

other variant of PCR is Combined Bisulfite Restriction

Analysis (COBRA). Endonucleases, such as BstUI, are

used for restriction. The results of MSP and COBRA

are analyzed by electrophoresis in a polyacrylamide

gel (Fig. 1) [38–41]. Another variant of the analysis of

methylated cfDNA is quantitative TaqMan or

SybrGreen I PCR using primers/probes that are com-

plementary to one of the converted and initially meth-

ylated strand, i.e., a strand containing CpG dinucleo-

tides, and unmethylated DNA strand [42].

The Methyl-BEAMing (beads, emulsion, amplifi-

cation, and magnetics) PCR method allows one to

detect methylated cfDNA in the presence of 99.9% of

the unmethylated form (Fig. 2) [43]. In this case,

cfDNA is initially hydrolyzed by nucleases, diluted to

the concentration of several nanograms in a milliliter,

and subjected to bisulfite conversion. The resulting

individual cfDNA molecules are amplified in an

emulsion of so-called water nanoparticles in a contin-

uous oil phase. Each water nanoparticle contains

DNA polymerase, the necessary cofactors, dNTP,

one DNA template molecule, and a DNA-binding

magnetic bead. Thus, each bead binds thousands of

identical copies of one DNA template molecule within

its own water nanoparticle. This process is similar to

the cloning of individual DNA fragments into a plas-

mid vector with the formation of the bacterial colony.

After PCR completion, the oil emulsion is destroyed

and the beads are collected. The individual status of

each DNA molecule is estimated by hybridization of

resultant PCR products with fluorescent probes that

specifically bind to initially methylated or unmethyl-

ated DNA sequences. The results are analyzed by flow

cytometry.

The Methyl-BEAMing method is similar to ana-

lyzing methylated cfDNA by digital PCR when the

reaction mixture is divided into a plurality of microre-

actors and one or several DNA molecules are ampli-

fied in each of them. This amplification can be carried

out using the commercial QX100TM Droplet DigitalTM

PCR system (Bio-Rad), which makes it possible to

work with small amounts of cfDNA (0.5 mL of blood

plasma) without molecular standards and ethalon

[44]. The analysis of the methylation status of cyto-

sines in cfDNA by digital PCR may be performed

using the commercial BioMark System (Fluidigm) on

platform 12.765 Digital Arrays (Fluidigm) containing

12 panels of 765 reaction chambers each. This system is

used to study fetal cfDNA in the maternal blood [45].

Another method suitable for identifying aberrantly

methylated cytosines in cfDNA after chemical conver-

sion is MethyLight PCR (Fig. 3а) [46]. Probes con-

taining different fluorescent labels specifically bind to

methylated or unmethylated DNA regions. This

method makes it possible to identify methylated DNA

in the presence of a 10000-fold excess of unmethylated

DNA [47].

The variant of HeavyMethyl PCR is also applicable

to analyzing the cytosine methylation status in cfDNA

(Fig. 3b). This method uses methylation-specific oli-

gonucleotide blockers and probes for the methylation-

specific amplification and detection of DNA [48]. The

amount of 30 pg of methylated DNA is sufficient for

analysis in the presence of 50 ng of unmethylated

DNA. The principle of the HeavyMethyl method is as

follows. Blocker oligonucleotides cannot bind to

methylated DNA, and this region is free to bind to

primers (gray arrows), thus making amplification of

DNA possible in this case. The fluorescent methyla-

tion-specific probe provides visualization of the sig-

nal. In the case of unmethylated DNA (Fig. 3b),

blocker oligonucleotides bind to DNA and prevent the

access of primers, thus inhibiting amplification.

One of the disadvantages of the MethyLight,

HeavyMethyl, and other PCR variants that use bisul-

fite-converted DNA is the inability to assess the meth-

ylation of individual cytosines in the studied DNA

region. However, this problem is solved by combining

the methods of HeavyMethyl PCR and methylation-

sensitive single nucleotide primer extension (Ms-

170

MOLECULAR BIOLOGY Vol. 51 No. 2 2017

BRYZGUNOVA, LAKTIONOV

SNuPE) [49]. This analysis requires 14 pg of DNA,

and aberrantly methylated sites can be identified in the

sample with a 2000-fold excess of normal DNA. The

principle of the Ms-SNuPE method is the amplifica-

tion of the studied region after bisulfite conversion of

DNA using primers specific to converted DNA [50].

The resulting PCR products are incubated with the

Ms-SNuPE primer(s), which are designed to be

hybridized just before cytosine that should be investi-

gated for methylation. Incubation is carried out in a

buffer containing [32Р]dСTP, [32Р]dТTP or fluores-

cently labeled dCTP/dTTP, and Taq polymerase. The

labeled products are visualized after electrophoresis in

denaturing polyacrylamide gel. The methylation status

is determined by the base that is incorporated in the

polymer chain. It was suggested to use ddCTP/ddTTP

or α-S-ddCTP/α-S-ddTTP instead of dСTP/dТTP and

to assess the incorporation of a nucleotide by high perfor-

mance liquid chromatography (HPLC) [51] or by matrix-

assisted laser desorption/ionization and time-of-flight

mass spectrometry (MALDI-TOF-MS) [52, 53].

The methylation status of individual cytosines in

bisulfite-converted cfDNA is analyzed by the combi-

nation of the HeavyMethyl and Ms-SNuPE methods,

with the former being followed by the latter. The

results are analyzed by HPLC [49].

It is clear from the above that the modern methods

allow for the successful identification of the cfDNA

methylation status in biological fluids, including in the

presence of a significant (up to 2000-fold) excess of

unmethylated DNA.

IDENTIFICATION OF METHYLATED

CYTOSINES IN DNA BY METHYLATION-

SENSITIVE RESTRICTION ENDONUCLEASES

Another approach that makes it possible to detect the

cytosine methylation status of specific DNA sequences is

fundamentally different from the above methods. This is

based on the use of restriction endonucleases that are

sensitive or insensitive to methylated cytosines. Both the

individual enzymes (e.g., Bsh1236I (BstUI) [54] and

Tsp509I [55]) and the pairs of endonucleases (e.g.,

SmaI/XmaI and HpaII/MspI [56, 57]) can be used for

this purpose. The most commonly used pairs are

MspI/HpaII, which recognizes the CCGG sequence

(HpaII is sensitive to methylation of the second cyto-

sine in the site) [58] and HpaII/HinP1I that recog-

Fig. 1. Concept of methods of MSP (methylation specific PCR) and COBRA (combined bisulfite restriction analysis) for ana-

lyzing the cytosine methylation status.

GG CCGG

CCGG

Bisulﬁte

treatment

GGTT

Unmethylated DNA (Unmet) Methylated DNA (Met)

Unmethylated

DNA

Methylated

DNA

PCR with primers

speciﬁc for methylated

or unmethylated DNA

MSP COBRA

Electrophoresis

Met

met

MetUnmet Unmet

PCR with primers

containing no CpG

dinucleotides + restriction

endonuclease BstUI

Met Unmet

CCGG

met

CG C

GGTT

Current methods of extracellular DNA methylation analysis

Bài viết tổng quan các phương pháp phân tích methylation cfDNA hiện đại (bisulfite, MSP, COBRA, pyrosequencing), ứng dụng trong chẩn đoán và theo dõi ung thư.

Chủ đề:

Tài liệu liên quan

Tài liêu mới

AI tóm tắt

Giới thiệu tài liệu

Đối tượng sử dụng

Từ khoá chính

Nội dung tóm tắt

Hỗ trợ

Phương thức thanh toán

Theo dõi chúng tôi