BioTechniques

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/ibtn20

DNA Methylation: A Profile of Methods and Applications

Mario F. Fraga & Manel Esteller

To cite this article: Mario F. Fraga & Manel Esteller (2002) DNA Methylation: A Profile of
Methods and Applications, BioTechniques, 33:3, 632-649, DOI: 10.2144/02333rv01
To link to this article: https://doi.org/10.2144/02333rv01

© 2018 Author(s)

Published online: 24 Sep 2018.

Submit your article to this journal

Article views: 4074

View related articles

Citing articles: 12 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ibtn20
Review
DNA Methylation: A Profile of Methods
and Applications
BioTechniques 33:632-649 (September 2002)

Mario F. Fraga and INTRODUCTION sured by high-performance separation

Manel Esteller techniques or by enzymatic/chemical
Spanish National Cancer Methylation of cytosines in the 5′ po- means. The latter are never as sensitive
sition of the pyrimidinic ring is the most as the former, and sometimes their res-
Research Center (CNIO), important epigenetic alteration in eu- olution is restricted to endonuclease
Madrid, Spain karyotes. In animals, methylcytosine is cleavage sites. Despite the drawbacks,
mainly found in cytosine-guanine enzymatic/chemical approaches are
(CpG) dinucleotides (10), whereas in still commonly used because, unlike
plants it is more frequently located in separation techniques, they do not re-
cytosine-any base-guanine (CpNpG) quire expensive and complex equip-
trinucleotide sequences. The presence ment that is not always available. When
ABSTRACT of 5-methylcytosine in the promoter of separation devices are available, high-
specific genes alters the binding of tran- performance capillary electrophoresis
Ever since methylcytosine was found in scriptional factors and other proteins to (HPCE) may be the best choice since it
genomic DNA, this epigenetic alteration has DNA in plants (26) and animals (18). It is faster, cheaper, and more sensitive
become a center of scientific attraction, es- also attracts methyl-DNA-binding pro- than HPLC (33). Although in situ hy-
pecially because of its relation to gene si- teins and histone deacetylases that mod- bridization methods for studying cyto-
lencing in disease. There is currently a wide ify chromatin structure around the gene sine methylation status sometimes give
range of methods designed to yield quantita- transcription start site. Both mecha- accurate measures of the degree of total
tive and qualitative information on genomic nisms block transcription and cause DNA methylation, one of the most in-
DNA methylation. The earliest approaches gene silencing (10). Thus, methylation teresting aspects of these approaches is
were concentrated on the study of overall of C residues in genomic DNA plays a that they provide information on tissue-
levels of methylcytosine, but more recent ef- key role in the regulation of gene ex- specific methylation patterns. By
forts have focused on the study of the methy- pression (99). means of labeled anti-methylcytosine
lation status of specific DNA sequences. DNA methylation research can be antibodies, DNA methylation can be
Particularly, optimization of the methods approached from several standpoints monitored in metaphase chromosomes,
based on bisulfite modification of DNA per- because there is a wide range of tech- hetero/euchromatin, and, most impor-
mits the analysis of limited CpGs in restric- niques available for the study of the oc- tantly, on a cell-by-cell basis within the
tion enzyme sites (e.g., combined bisulfite currence and localization of methyl- same sample. The latter alternative,
restriction analyses and methylation-sensi- cytosine in the genome (45,82,83). which generally yields qualitative re-
tive single nucleotide primer extension) and Each technique has its own peculiari- sults, is of great interest in cancer re-
the overall characterization based on differ- ties, which implies that there is a suit- search because it can reveal methyla-
ential methylation states (e.g., methylation- able technique for each specific prob- tion differences between normal and
specific PCR, MethyLight, and methylation- lem. In any case, DNA methylation tumor tissues in the same sample.
sensitive single-stranded conformational status can only be established by the Two alternatives are currently used
polymorphism) and allows very specific pat- joint use of various methodologies. to study the degree of DNA methyla-
terns of methylation to be revealed (bisulfite Here we classify the available methods tion in particular DNA sequences: non-
DNA sequencing). In addition, novel meth- for studying the degree of DNA methy- bisulfite and bisulfite methods. The
ods designed to search for new methylcyto- lation, with respect to the type of infor- first relies on the use of methylation-
sine hot spots have yielded further data mation produced (Figure 1), including sensitive restriction endonucleases
without requiring prior knowledge of the the degree of whole genomic DNA combined with Southern blot analysis
DNA sequence. We hope this review will be methylation, the DNA methylation sta- or PCR detection, which sometimes
a valuable tool in selecting the best tech- tus of a specific sequence, and how to means that results are limited to cleav-
niques to address particular questions con- find new methylation hot spots. age sites. This problem can be avoided
cerning the cytosine methylation status of Levels of methylcytosine occur- by the bisulfite modification of the
genomic DNA. rence in the genomic DNA can be mea- DNA, which comprises a wide range of

632 BioTechniques Vol. 33, No. 3 (2002)

Review
techniques that allow the quantitative methylation questions. To do this, we HPLC-Based Methods
and accurate determination of the categorized the most important meth-
methylation status of the allele, even at ods currently used by the type of infor- Relative methylcytosine contents of
the cell population level. All bisulfite- mation they provide. We also describe genomic DNA can be analyzed by
associated methods require PCR ampli- all the outstanding approaches and pro- chemical hydrolysis to obtain the total
fication of the bisulfite-modified DNA, vide more detailed information about base composition of the genome and the
and differences in methylcytosine pat- the techniques we consider most signif- subsequent fractionation and quantifica-
terns are displayed by methylation-de- icant, discussing their advantages, dis- tion of hydrolysis products using HPLC
pendent primer design (e.g., methyla- advantages, and possible artifacts. technologies. DNA hydrolysis can be
tion-specific PCR), in conjunction with carried out by incubation with formic
methylation-sensitive restriction en- acid at high temperature (27). However,
donucleases [e.g., combined bisulfite QUANTIFICATION OF Catania et al. (14) suggested the use of
restriction analyses (COBRA)], ge- OVERALL METHYLCYTOSINE hydrofluoric acid for chemical hydroly-
nomic sequencing, and other approach- IN GENOMIC DNA sis of DNA to prevent deamination of
es. Some of these approaches even pro- cytosine and methylcytosine, which of-
vide quantitative data about the average Global DNA methylation has been ten occurs with formic acid.
proportion of methylated and unmethy- proposed as a molecular marker for a In any case, enzymatic hydrolysis of
lated alleles in a population. variety of biological processes such as the DNA is reported to be a better alter-
Thus, because of the large assort- cancer (32,76) and ontogenetic devel- native for quantifying the degree of
ment of methods available for evaluat- opment in both plants (34) and animals DNA methylation (58). The 2′-deoxy-
ing the degree of genomic DNA methy- (55). Methylcytosine concentration in mononucleosides are obtained using
lation and the wide range of data types genomic DNA is currently quantified Nuclease P1, DNase I, and snake ven-
they yield, the goal of this paper is to by high-performance separation means om phosphodiesterase (25) or micro-
help researchers to choose the best (HPCE and HPLC) or by enzymatic/ coccal nuclease and phosphodiesterase
method for addressing specific DNA chemical approaches. II (61), followed by hydrolysis with al-
kaline phosphatase. Resulting deoxyri-
bonucleosides are subsequently separat-
ed by HPLC, and the methylcytosine
levels are quantified by comparing the
relative absorbance of cytosine and
methylcytosine at 254 nm in the sample
with external standards of known bases.
The degree of DNA methylation of sev-
eral animal (38) and plant tissues (98)
has been quantified by this method, but
at least 2.5 µg DNA are generally re-
quired to quantify 5-methylcytosine
with a low standard deviation for repli-
cate samples. Moreover, this method
suffers from occasional problems
whereby elution buffers precipitate into
the column (unpublished observation).
The sensitivity of the system can be
increased with mass spectrometry de-
tection, which has a detection limit 106
times the limit of absorption spec-
troscopy detectors. Annan et al. (3)
used HPLC-mass spectrometry coupled
via a moving-belt interface to analyze
electrophoretic derivatives of methyl-
cytosine and 5-hydroxymethyluracil
nucleobases. As little as 9.9 pg (signal-
to-noise ratio, 5) and 180 fg (signal-to-
noise ratio, 10) of the respective nucle-
Figure 1. Main methods available for the study of genomic DNA methylation status. MSP, methyl- obases were detected in the negative
ation-specific PCR; MS-SnuPE, methylation-sensitive single-nucleotide primer extension; MS-SSCP, electron-capture chemical ionization
methylation-sensitive single-stranded conformational polymorphism; MS-REs, methylation-sensitive re-
striction endonucleases; LM-PCR, ligation-mediated PCR; MS-ISH, methylation-specific in situ hy- mode, and linear responses were ob-
bridization; IPEM, incomplete primer extension mixture; CPBS, competitive primer binding sites; SP-PE, served over a moderate dynamic range.
solid-phase primer extension; and ERMA, enzymatic regional methylation assay. Alternatively, Gowher et al. (44) have

634 BioTechniques Vol. 33, No. 3 (2002)

Review
developed a highly specific and sensi- modification of the previously described
tive assay for detecting trace amounts HPCE method (33). Approximately one
of methylcytosine in genomic DNA. methylcytosine in 200 cytosine residues
They degraded DNA to nucleosides, can be detected by this method using 1
purified methylcytosine by HPLC, and, µg genomic DNA (Figure 2). To in-
for detection by 1-D and 2-D thin-layer crease sensitivity, laser-induced fluores-
chromatography (TLC), radiolabeled cence and mass spectrometry detectors
using deoxynucleoside kinase and γ should be used. However, there are no
[32P]ATP. Using this assay, they references describing how to quantify
showed that methylcytosine occurs in methylcytosine. Nevertheless, capillary
the DNA of Drosophila melanogaster electrophoretic separation and online
at a level of 1 in 1000–2000 cytosine detection by monitoring laser-induced
residues in adult flies. This labeling fluorescence have already been applied
strategy has already been used to quan- successfully to quantify other modified
tify methylcytosine concentration (61), bases in DNA (100). Using laser-in-
but, in this case, it was employed in duced fluorescence monitoring systems,
conjunction with 1-D high-perfor- a concentration detection limit of ap-
mance TLC using alkylamino-modified proximately 3 × 10-10 M (91) or 1
silica plates. adduct/107 normal nucleotides/µg DNA
(85) can be achieved.
HPCE-Based Methods
Analyses of Genome-Wide
The development of CE techniques Methylation by Chemical or
has given rise to an approach to research Enzymatic Means
that has several advantages over current
methodologies used to quantify the ex- As stated earlier, quantifying the de-
tent of DNA methylation. From its be- gree of DNA methylation by HPLC or
ginnings, CE has proved to be extremely HPCE requires access to sophisticated
helpful in separating various DNA com- equipment that is not always available.
ponents, including a number of base The radioactive labeling of CpG sites
adducts (71), and recently Fraga et al. using the methyl-acceptor assay (101)
(33) have reported the quantification of has been developed to address this prob-
the relative methylcytosine content of lem, but, among the technique’s other
the genomic DNA of plants using a drawbacks, it can only monitor CpG
HPCE system to analyze acid-hy- methylation changes, thus CpNpG
drolyzed genomic DNA. In this context, methylation cannot be detected. This
the separation and quantification of cyto- method uses bacterial SssI DNA methyl-
sine and methylcytosine are only possi- transferase to transfer tritium-labeled
ble by the use of an SDS micelle system. methyl groups from S-adenosylmethio-
This method is faster than HPLC (taking nine (SAM) to unmethylated cytosines
less than 10 min/sample), reasonably in-
expensive because it does not require
continuous running buffers, and displays
a great potential for fractionation (theo-
retically up to 106 plates). Nevertheless,
almost no preparative analyses are possi-
ble with HPCE systems because of the
low injection volumes involved.
The release of bases from DNA by
chemical means involves the production
of a complicated mixture of molecules
that makes the detection and quantifica-
tion of methylcytosine difficult when
DNA is not sufficiently purified and/or
concentrated. This problem can be
avoided by enzymatic hydrolysis with Figure 2. Resolution of nucleosides obtained
from enzymatic hydrolysis of genomic DNA
Nuclease P1 and alkaline phosphatase to from human cancer cell line HT29 by HPCE.
produce 2′-deoxymononucleosides, mC, 5-methylcytidine; A, adenosine; C, cytidine;
which are then fractionated using a G, guanosine; and T, thymidine.

Vol. 33, No. 3 (2002)

in CpG targets. The data obtained from region of interest with SssI and 3H-la- cytosine and adenine bases to yield the
a scintillation counter are used to calcu- beled SAM incubation using dam fluorescent adducts ethenocytosine and
late the number of methyl groups incor- methyltransferase and 14C-labeled ethenoadenine, respectively, while ura-
porated in the DNA. The amount of SAM as an internal control [primers for cil, thymine, and guanine do not under-
tritium incorporated is inversely propor- PCR are designed to contain two dam go this reaction. Following DNA depu-
tional to the level of CpG methylation sites (CATG)]. Under this combination rination, the extent of fluorescence as
(23). This method has revealed changes of conditions, the relative 3H/14C signal quantified with a luminescence
in the DNA methylation patterns of pa- ratio is linearly correlated with the spectrometer is proportional to the
tients who suffer from a number of dis- methylation density of the region of in- amount of methylcytosine in the
eases, including cancer (40). terest. In general, methods based on the genome. This method has several ad-
Recently, a modification of the SssI acceptance of methyl groups in vitro vantages over the methyl-acceptor as-
assay was reported to allow the quan- can give inaccurate but reproducible re- say: a chemical, rather than an enzy-
tification of the degree of DNA methy- sults. If these methods are used, then matic reaction, used quite stable and
lation in specific DNA regions (37). they should be validated on a few sam- inexpensive reagents, and its measure-
This quantitative approach, termed en- ples with a direct quantitative analysis ments are not restricted to the methyl-
zymatic regional methylation assay for global 5-methylcytosine levels. cytosine located in GpC targets.
(ERMA), provides high-quality infor- Combining the sodium bisulfite re- Diaz-Sala et al. (22) described other
mation about the methylation density action with the ability of chloroac- approaches in plant samples. They
of a sequence. However, as in the previ- etaldehyde to label DNA fluorescently, were able to measure the relative de-
ously described case of the SssI assay, Oakeley et al. (74) described a method gree of DNA methylation accurately by
the acquisition of data is limited to spe- for analyzing DNA methylation in any the fluorescence quantification of the
cific endonuclease cleavage sites. The sequence by fluorescent labeling. They online slab of DNA digestion products
method merges bisulfite modification treated the DNA with chloroacetalde- that were obtained with methylation-
of DNA and PCR amplification of the hyde, which reacts quantitatively with sensitive endonucleases.
Review
In Situ Hybridization Methods for somes (4,70) and in chromatin using immunofluorescence analyses can be
Studying Total Cytosine Methylation monoclonal antibodies (MAbs) com- accomplished using an image analyzer
bined with fluorescence staining (68). and a charge-coupled device camera
Global DNA methylation can also be The incorporation of 5-bromodeoxyuri- (97). For this purpose, sample results
quantified by methylcytosine-specific dine in DNA with a thermal denatura- must be compared with results obtained
antibodies (67). The accessibility of tion step may increase the binding from cells with known methylation lev-
methylcytosine to specific antibodies in efficiency of anti-methylcytosine anti- els. The study of global DNA methyla-
dsDNA was first reported in the mid- bodies on metaphase chromosomes and tion of human female X chromosomes
1980s (2). An outstanding advantage of increase the efficiency of antibody bind- using this method revealed global hy-
this approach is that it may be carried out ing, as revealed by immunofluorescence pomethylation of the late-replicating X.
on a cell-by-cell basis rather than in a staining (70). As a result of differential In addition, there is a method that allows
heterogeneous population. Every in situ immunofluorescence staining, this not only the mapping of structural
hybridization-based technique for the method allows the semi-quantitative (banding) but also functional (methyla-
study of DNA methylation must be pre- analysis of the antibody binding sites in tion status) features of different chromo-
viously validated with a control for the metaphase chromosomes. Furthermore, some domains (6). This strategy uses
accessibility of the antibody to the DNA. in situ hybridization with fluorescence MAbs on mildly denatured chromo-
Classical immunoassay detection ap- detection makes it possible to distin- somes. An alternative to fluorescence
proaches involve quantifying the guish different methylcytosine-rich detection is to connect a colored en-
retention of radiolabeled DNA by poly- DNA sites (4). For example, secondary zyme-dependent reaction. For instance,
clonal antibodies on nitrocellulose fil- constrictions, juxtacentromeric regions, peroxidase-tagged second antibodies,
ters, immunoprecipitation, gel filtration, and T-bands emit strong fluorescence, with 4-chloro-1-naphthol used as sub-
and visualization under electron mi- while the short arms of acrocentrics strate with anti-5-methylcytosine MAbs,
croscopy (2). Cytosine methylation can produce strong polymorphic signals. have been used to quantify the effects of
also be detected in metaphase chromo- Quantitative in situ DNA methylation the demethylating agent 5-azacytidine
on the constitutive heterochromatin of (69). This MAb displays excellent schizomer pair is able to cut the DNA
human chromosomes (21). specificity because it reacts with only when its target is unmethylated,
When comparing the results of glob- methylcytidine and methylcytosine but whereas the other is not sensitive to
al DNA methylation evaluated by the not with other nucleosides. When suffi- methylated cytosines. The most com-
radiolabeled methyl incorporation as- cient specificity is achieved and no mon isoschizomers used are the HpaII/
say (an alternative method for the quan- cross-reactions are detected, in situ MspI pair. In general, both cleave the
tification of global DNA methylation DNA methylation immunofluorescence DNA at the CCGG target, but HpaII is
levels) (19) with immunohistochemical analyses can be used to monitor signifi- not able to cut when the second cytosine
staining of the same tissue sections cant growth processes in animals [e.g., is methylated (CmCGG). Moreover, it
with MAbs against methylcytosine, embryo development (11)] and plants was reported that cleavage by MspI of
similar contents are generally observed. [e.g., pollen nuclei maturation (73)]. GGCCGG sequences is also inhibited
In addition, in situ hybridization can by the methylation of the C next to the
also reveal differences in the methyla- CpG (12). EcoRII/BstNI isoschizomers
tion status of cancer and normal adja- NON-BISULFITE METHODS FOR are more frequently used in plants,
cent tissues by analyzing the same tis- MAPPING METHYLCYTOSINES whereas most methylcytosine is located
sue section. IN SPECIFIC DNA SEQUENCES at CCNGG sequences (where N is any
Because of the high sensitivity and base other than G) (46). EcoRII recog-
specificity required, antibody design The most widely used methods for nizes CCNGG targets, but it only
and production must be carefully con- studying DNA methylation patterns of cleaves when unmethylated (9), where-
sidered, even though these topics de- specific regions of DNA with no base as BstNI is insensitive to cytosine meth-
pend on the detection method. One of modifications are based on the use of ylation. Although these pairs of en-
the highest sensitivities (in the pmol methylation-sensitive and insensitive zymes can cleave hemimethylated
range) can be achieved by the FMC9 restriction endonucleases (15). One of DNA, they do not distinguish between
MAb in conjunction with an ELISA the restriction enzymes of the iso- cytosines methylated at different posi-
Review
tions in the pyrimidinic ring (13). How- task when such small amounts of DNA fied by the amplification of the diges-
ever, there are several restriction en- are handled. There is another method tion products with ligation-mediated
zymes that recognize the localization of that consists of measuring the conver- PCR and radioactive labeling. The ratio
the methyl group (64). sion of a large amplified DNA fragment of the two amplified fragments corre-
Once DNA has been digested with to a shorter DNA product that correlates lates with the degree of methylation at
methylation-sensitive endonucleases, with the amount of demethylation (65). the restriction site.
identification of the methylation status The procedure uses pairs of non- PCR methods involving prior diges-
of a gene can be accomplished by isoschizomeric enzymes to cleave ge- tion with CpG methylation-sensitive re-
Southern blot hybridization or PCR nomic DNA at closely spaced sites. The striction endonucleases can be compli-
procedures (Figure 3). In the first case, extent of cleavage by the methylation- cated by the inability to achieve 100%
digestion products are separated by gel sensitive restriction enzyme is quanti- digestion of the sensitive sites, even
electrophoresis, transferred to a nitro-
cellulose filter, and hybridized with a
radiolabeled probe (63). When DNA
digestion is accomplished by methyla-
tion-sensitive nucleases, obtaining a
DNA fragment that is larger than ex-
pected indicates methylation at one or
both restriction targets that flank the
homologous region of the DNA (Figure
3). An methylcytosine-positive will be
detected when at least 10% of the DNA
present that modification because of
their hybridization sensitivity.
When the amount of tissue is limit-
ing, detection of cytosine methylation Figure 3. Anticipated results when studying site-specific DNA methylation by Southern blot analy-
sis and PCR (two of the most important non-bisulfite-related methods) in the cases of methylation
can be achieved by PCR, which re- and unmethylation for a single CpG target. Black circles indicate cytosine methylation. Arrows sym-
quires less than 10 ng DNA, whereas bolize PCR primers. The discontinuous lines illustrate probes for Southern blot analysis.
Southern blot hybridization generally
requires up to 10 µg DNA. Further-
more, a methylcytosine-positive can be
detected when only 0.1% of the DNA
molecules present such base modifica-
tion (60). PCR primers must corre-
spond to flanking sequences of the re-
striction targets so that the absence of
methylation is revealed in the presence
of a PCR band (Figure 3) (87).
For quantitative analyses, Heiskanen
et al. (50) described a solid-phase quan-
titative primer extension method that
can be adapted to a microtitration well
format, thereby allowing the analysis of
a large series of samples. In particular,
the method produces linearity in a range
from 2% to 100% of malignant blasts
diluted with normal leukocytes. Quanti-
tative PCR can also be performed by a
modification (88) of the method de-
scribed by Singer-Sam et al. (87). The Figure 4. Methylation-specific PCR and bisulfate sequencing. (A) Methylation status of the p16INK4a
improvement, termed competitive CpG island in human primary breast tumors determined by methylation-specific PCR. PBR322/Msp di-
primer binding sites, includes an inter- gest (New England Biolabs, Beverly, MA, USA) is shown at left as the molecular weight marker. The pres-
ence of a visible PCR product in the lanes marked U indicates the presence of unmethylated genes, and the
nal standard sharing primer-binding presence of PCR product in the lanes marked M indicates the presence of methylated genes. In vitro methy-
sites with the genomic template so that lated DNA was used as a positive control for methylation, normal lymphocytes were used as a negative
demethylation results in a decrease of control for methylation, and water was used as a negative PCR control. The breast tumor BR2 is hyperme-
the PCR products. Nevertheless, a com- thylated at P16, while the remaining tumors (BR1, BR3, BR4, BR5, BR6, and BR7) are unmethylated. (B)
Graphical representation of the sodium bisulfite genomic sequencing of the CpG island of the APC pro-
parison of the samples entails the accu- moter. Each row represents an individual cloned and sequenced allele after sodium bisulfite DNA modifi-
rate quantification of the amount of cation. CpG sites are shown as circles and drawn to accurately reflect CpG density of the region. White
DNA, which is commonly a difficult dots are unmethylated CpGs, and black dots are methylated CpGs. IVD, in vitro methylated DNA.

640 BioTechniques Vol. 33, No. 3 (2002)

with multiple rounds of digestion using olution or contributes to the develop- adequate alkalization of the solution so
high concentrations of enzyme. Al- ment of a novel perspective on the re- that, if the pH is lower than nine, the
though PCR can amplify even a single sults. All of them work with PCR, rate of pyrimidine desulfonation is
target DNA molecule when sufficient which, among other advantages, is suit- rather slow, thereby rendering certain
rounds of amplification are employed, able for analyzing paraffin-embedded DNA polymerases incapable of repli-
misleading results can be obtained with tissues and poorly purified DNA. cating the template. However, the prob-
such methods if adequate controls are Bisulfite modification of DNA re- lem may be avoided by measuring the
not used. This is why cleavage with a quires prior DNA denaturation because pH after the alkalization step.
restriction isoschizomer of the methy- only methylcytosines that are located in It was originally believed that all the
lation-sensitive endonuclease is a com- single strands are susceptible to attack cytosines in densely methylated regions
monly used control (95). Even though (84). The partial denaturation of the in the chromosome replication origins
partially cleavage-resistant sites have DNA is a common event that causes one were methylated (92), but methylation
been commonly interpreted as being of the potential artifacts of the method. at non-CpG residues was later discov-
partially methylated, the importance of Renaturation induced by high-salt con- ered to be an artifact of partial bisulfite
the intrinsic resistance to cleavage of centrations could be an additional criti- conversion. Harrison et al. (48) attempt-
certain DNA targets should be stressed cal problem, although bisulfite is gener- ed to explain this observation by sug-
(59). This resistance can be overcome ally able to react with the DNA before gesting that a methylated cytosine could
by the addition of site-containing renaturation occurs. Nevertheless, false protect the neighboring unmethylated
oligonucleotide duplexes that drive the positives caused by the lack of the sin- outer cytosine from complete bisulfite
cleavage of the resistant sites. gle-stranded conformation have been conversion and that this is a sequence-
Non-bisulfite methods for the quan- reported (81). Incomplete desulfonation specific event. They also suggested that
tification of DNA methylation patterns after bisulfite treatment may also give the problem could be avoided by using
are simple, rapid, and can be used for rise to several problems (93). The pres- an internal standard that consists of
any known-sequence genomic DNA re- ence of residual bisulfite can avert the checking the complete bisulfite conver-
gion. These methods are extremely spe-
cific, but their limitation to specific re-
striction sites reduces their value.

BISULFITE METHODS FOR

MAPPING METHYLCYTOSINES
IN SPECIFIC DNA SEQUENCES

Although isoschizomer-based meth-

ods for the study of methycytosine oc-
currence at specific DNA regions dis-
play high sensitivity, they cannot
provide the critical information re-
quired for a complete understanding of
the role of methylcytosine in cell and
molecular biology. These aspects can
be addressed with the bisulfite modifi-
cation of the DNA (17).
There is an extensive range of meth-
ods for the quantification of the methy-
lation status of cytosines located in spe-
cific DNA regions based on the sodium
bisulfite treatment (Figure 1). Bisulfite
converts unmethylated cytosine to
uracil, while methylated cytosine does
not react (36). This reaction constitutes
the basis for discriminating between
methylated and unmethylated DNA.
Bisulfite transformation of DNA can be
followed by several methods, including
sequencing, methylation-specific PCR,
combined bisulfite restriction assays,
and others. In general, each modifica-
tion leads to greater sensitivity and res-

Vol. 33, No. 3 (2002) BioTechniques 641

Review
sion of a non-CpG site from an in vitro Ltd., Bundoora, Australia) that provides were set up to display methylcytosine
synthetic methylated oligonucleotide. a measure of the degree of methylation occurrence by the sequencing method,
As previously stated, the total con- at a particular target by comparing cyto- but only one termination reaction is
version of cytosines to uracils is critical sine and thymine signals that have been needed when the sequence of the target
to the success of the analyses. The labeled with the same fluorescent dyes. region is known (103). The choice of
maximum conversion rates of cytosine This method has been used to examine termination reaction depends on the type
occur at 55°C (4–18 h) and 95°C (1 h). the detailed methylation state of many of base conversion for which the se-
Special attention must be paid to the in- CpG island-containing genes in cancer quencing primers are designed, and the
fluence of the temperature and incuba- (66). Moreover, Radlinska et al. (79) results are obtained by comparing prod-
tion time on DNA integrity during the described an alternative method that al- ucts of the single termination reaction to
bisulfite treatment because, under these lows the direct localization of methyl- those of normal sequencing terminations
conditions, 84%–96% of the DNA is cytosines in the primary product of the from the control unmodified DNA.
degraded (47). Raizis et al. (80) also re- bisulfite treatment instead of the PCR
ported the occurrence of partial DNA amplification product. To achieve this, Methylation-Specific PCR
degradation during bisulfite treatment, they used a primer extension mixture
but in their case, it was due to low pH- containing only three deoxynucleotides Methylation-specific PCR is the
dependent depurination. A simple solu- (dATP, dCTP, and dTTP) and lacking most widely used technique for study-
tion to limit the degradation of the dGTP (IPEM) that produced an elonga- ing the methylation of CpG islands,
DNA template is provided by the au- tion stop at methylcytosine points. The which are common in the promoter re-
thors, who suggest reducing the time positions of the methylcytosine are rec- gions of many genes. Cytosines in CpG
required to complete the bisulfite reac- ognized by comparing the localization islands are usually unmethylated in
tion by means of increasing the bisul- of runoff products to products of se- normal tissues, whereas they become
fite concentration to 5 M. Under these quencing reactions performed on bisul- methylated in the promoter sequences
conditions, optimal PCR production fite-untreated template DNA. The of genes associated with certain abnor-
occurs after only 4 h, thereby minimiz- method, which gave no false-positive mal cellular processes such as cancer
ing DNA fragmentation. results at all, has a sensitivity of 4–7 ng (31). Methylation-specific PCR (51) is
(50 fmol of the methylated sites). one of the most effective choices for in-
Sequencing Normally, four termination reactions vestigating the methylation profile of

Sequencing bisulfite-altered DNA is

the most straightforward means of de-
tecting cytosine methylation. In gener-
al, after the denaturation and bisulfite
modification, dsDNA is obtained by
primer extension, and the fragment of
interest is amplified by PCR (17).
Methylcytosine can then be detected by
standard DNA sequencing of the PCR
products (Figure 4B). Cloning PCR
products into plasmid vectors followed
by the sequencing of individual clones
is an alternative method that, although
slower, could provide methylation
maps of single DNA molecules, instead
of the average values of the methylation
status in the population of molecules
provided by direct sequencing of the
PCR products. This approach has been
helpful in the study of the DNA methy-
lation of genes associated with cancer
such as APC (30) and Rb (89).
Consistent quantification of cytosine
methylation by direct sequencing has
been widely reported to entail the
means of fluorescence-based detection Figure 5. Methylation results for methylated, unmethylated, and mixtures of methylated/unmethy-
(78). Paul and Clark (75) described a lated alleles obtained by several bisulfite-related methods. MSP, methylation-specific PCR; Ms-
method of automated genomic sequenc- SnuPE, methylation-sensitive single nucleotide primer extension; Ms-SSCP, methylation-sensitive sin-
gle-stranded conformational polymorphism; MS-REs, methylation-sensitive restriction endonucleases;
ing with fluorescence detection U, unmethylated reaction; M, methylated reaction; C, cytosine band; and T, thymine band. Black circles
(GENESCAN; GeneScan Australia Pty. indicate cytosine methylation.

642 BioTechniques Vol. 33, No. 3 (2002)

these regions (28,31,32). specific primers and the primers for the
The differences between methylated unmethylated sequences has been com-
and unmethylated alleles that arise from pared. Non-quantitative, methylation-
sodium bisulfite treatment are the basis specific PCR results can be validated
of methylation-specific PCR (Figure 5) with a quantitative method on some of
and are especially valuable in CpG is- the samples (e.g., Southern blot analy-
lands because of the abundance of CpG sis with a CpG methylation-sensitive
sites. Primer design is a critical and restriction endonuclease). With previ-
complex component of the procedure ously unused primers, unmethylated
(Figure 5). Bisulfite-converted DNA and in vitro methylated control se-
strands are no longer complementary, quences should be tested.
so primer design must be customized To avoid such problems, Nuovo et al.
for each DNA chain. Methylation pat- (72) made a remarkable advance in the
terns of all sequences must be deter- method by combining methylation-spe-
mined in separate reactions (Figure cific PCR with in situ hybridization.
4A). To optimize the PCR amplification The modification allows for the methy-
step, the following critical requirements lation status of specific DNA sequences
must be considered when designing the to be visualized in individual cells. This
primers: (i) the annealing temperature powerful approach is mainly used to
of both primers must be similar and al- monitor methylation patterns in sample
ways between 55°C and 65°C; (ii) the tissues with complex mixtures of tu-
PCR product should be between 80 and mor-like and normal cells. Another
175 bp; (iii) each primer should contain method that was recently described
at least two CpG pairs; (iv) the sense combines methylation-specific PCR
primer should contain a CpG pair at its and denaturing HPLC (DHPLC), with
3′-end; and (v), to avoid false positives which even small cell mosaicisms of
(amplification of unmethylated, un- structurally normal and/or abnormal
modified DNA), primers should contain chromosomes can be detected (5). Fol-
non-CpG cytosines. lowing PCR amplification, the alleles
The great sensitivity of the method can be resolved from the two popula-
allows the methylation status of small tions of PCR products by DHPLC be-
samples of DNA, including those from cause they differ at several positions
paraffin-embedded or microdissected within the amplified sequence. Distinct
tissues (51). However, resolution at the clonotypic epigenotypes can also be
nucleotide level requires the sequenc- isolated and characterized by denatur-
ing of the PCR products, and quantita- ing gradient gel electrophoresis
tive data and identification of cellular (DGGE) (1), which allows the detection
heterogeneity in populations are still of almost any sequence change in every
not possible with this technique. If the DNA region, including differentially
PCR involves too many cycles of am- methylated sequences.
plification without ensuring that the re- Another quantitative approach is
action is in the linear response range called MethyLight (24), which uses flu-
with respect to the template concentra- orescence-based, real-time PCR tech-
tion, then large overestimates of the ex- nology and does not involve additional
tent of methylation can be obtained if PCR steps. The DNA is modified by
the sequence is amplifiable with both the bisulfite treatment and amplified by
the methylation-specific primers and fluorescence-based, real-time quantita-
the primers for the unmethylated se- tive PCR using locus-specific PCR
quence. In addition, the appropriate primers that flank an oligonucleotide
control DNA sequences should have probe with a 5′ fluorescence reporter
been shown first to give no PCR prod- dye and a 3′ quencher dye. The reporter
uct with the methylation-specific pri- is enzymatically released during the re-
mers. The conclusion about methyla- action, and fluorescence, which is pro-
tion could refer to only a small portional to the amount of PCR product
percentage of the examined cells in the and thus to the degree of DNA methy-
population unless the PCR is done lation, can be sequentially detected in
semi-quantitatively, varying the tem- an automated nucleotide sequencer de-
plate concentration or cycle number, vice. Fluorescence detection greatly in-
and the efficiency of the methylation- creases the sensitivity of the method,

Vol. 33, No. 3 (2002) BioTechniques 643

Review
making it possible to detect a single agents. This involves tumor regression Methylation-Sensitive Single
methylated allele in 105 unmethylated because of the increase of the respon- Nucleotide Primer Extension
alleles. However, the major advantage siveness to alkylating agents caused by
is that sequence discrimination can be the lack of MGMT. Methylation-specif- Methylation-sensitive single nu-
achieved at any level of probe hy- ic PCR has also been useful in the de- cleotide primer extension employs
bridization or PCR amplification as a tection of the presence of tumoral DNA bisulfite-PCR combined with single
result of the particular primer design. in the serum of cancer patients (41). nucleotide primer extension to analyze
The quantitative nature of the assay is DNA methylation status quantitatively
based on real-time PCR and the inclu- COBRA at a particular DNA region without us-
sion of a methylated reference DNA. ing restriction enzymes (Figure 5). This
This assay allows the comparison of COBRA (102) constitutes a highly allows the analysis of almost all DNA
normal tissue samples with low levels specific approach releasing on the cre- regions (42). In this approach, the in-
of methylation at the test sequences and ation or detection of a target for restric- corporation of C instead of T results in
tumors with significantly more methy- tion endonuclease after bisulfite treat- methylation in the original DNA se-
lation. It also allows the quantitative as- ment. The major advantage of this quence (Figure 5).
sessment of DNA methylation at spe- method is that it provides semi-quanti- Following the bisulfite modification
cific sequences with high throughput, tative data regarding the methylation of the whole genomic DNA, the region
permitting the rapid analysis of many status at specific regions in any DNA of interest is amplified by PCR tech-
DNA samples at many different sites. sample (Figure 5). niques, using primers specifically de-
Because of the versatility of the DNA amplification products of spe- signed for the new bisulfite-modified
method, methylation-specific PCR has cific loci, previously modified by bisul- sequence of DNA. PCR-generated
been widely proposed as a rapid and fite, are digested with restriction en- products are also amplified using inter-
cost-effective clinical tool to use in the zymes that distinguish methylated from nal primers that terminate immediately
initial evaluation of a wide range of pa- unmethylated sequences so that the de- 5′ of the single nucleotide to be assayed.
tients with specific allele-dependent gree of DNA methylation is linearly cor- The relative average amount of allele
diseases. For instance, methylation- related with the relative amounts of di- methylation is quantified by monitoring
specific PCR is currently applied to gested and undigested products. The the incorporation of both [32P]dCTPs
evaluate the methylation status of the BstUI case may illustrate this point; its and [32P]dTTPs using a phosphorimag-
CpG island of the SNRPN gene and, cleavage site (CGCG) is resistant to ing device. Under these conditions, the
hence, for the rapid diagnosis of the bisulfite modification when it is methy- intensities of bands in the C tracks are
Prader-Willi and Angelman syndromes lated but is transformed to TGTG when proportional to the average amount of
(57). This approach also provides an it is unmethylated. Thus, DNA cleavage methylation at each tested CpG site.
accurate assay for the determination of after bisulfite treatment only occurs if Meanwhile, the intensity of the T tracks
X inactivation and can be carried out on the restriction target is methylated and is linearly correlated with the percent-
DNA samples that are unsuitable for re- the cleavage products are proportional to age of unmethylated cytosines.
striction digestion. Incidentally, these the degree of methylation of the ana- Similar to other quantitative bisul-
are the most widely used means to eval- lyzed sequence. The average relative fite-PCR methods, methylation-sensi-
uate the methylation status in active proportions of digestion products can be tive single nucleotide extension has a
and inactive X chromosomes. Uchida quantified by hybridization with 5′-end- broad range of clinical applications be-
et al. (96) described a slight modifica- labeled oligonucleotides and phospho- cause of its high sensitivity, specificity,
tion of this method, which they called rimager detection. In contrast to methy- and requirement for only small
human androgen-receptor gene-methy- lation-specific PCR, the PCR primers amounts of sample. Despite all of these
lation-specific PCR (HUMARA-MSP). should not contain CpG pairs to avoid advantages, PCR bias and analyses in
This involves the analysis of the X discrimination between different meth- CpG-rich regions can be a problem be-
chromosome inactivation pattern by an ylated templates. In general, this ap- cause of the difficulty of successfully
optimized methylation-specific PCR proach can be used when absolute designing primers lacking CpG dinu-
method that can identify the average percentages of methylated and unmeth- clotides. This is a key point if one con-
proportions of methylated and un- ylated alleles are required to guarantee siders the prior lack of knowledge of
methylated alleles through the exami- the final diagnosis. The major drawback their methylation status (82).
nation of a polymorphic short tandem is that BstUI will also cut if unconverted;
repeat near the 5′ promoter region. therefore, the use of this enzyme can Methylation-Sensitive Single-Strand
Methylation-specific PCR has also lead to the overestimation of methyla- Conformation Analysis
been successfully used to evaluate the tion. Monitoring conversion state with
responsiveness of human cancer pa- enzymes such as HpaII is needed (54). Single-stranded conformational poly-
tients to alkylating agents. By using Although it is specific, quantitative, morphism (SSCP) analyses were report-
methylation-specific PCR, Esteller et al. and sensitive, this approach entails com- ed several years ago in the broad context
(29) found hypermethylation of the plete bisulfite modification and cannot of detecting single-base polymorphisms
DNA-repair enzyme MGMT to be a key be used for all DNA sequences because as a means of resolving PCR products
factor in the resistance to alkylating it is confined to restriction targets. that differed by no more than a few

644 BioTechniques Vol. 33, No. 3 (2002)

nucleotides (77). Recently, Bianco et al. appropriate computer program (e.g., 32
(8) adapted this technique to the study of Karat Software; Beckman Coulter
methylation changes in specific DNA re- Spain S.A., Madrid, Spain). Under
gions. They proposed combining bisul- these conditions, the correlation coeffi-
fite modification of the DNA with SSCP cient in recovery experiments is
in a new method that they called methy- 0.9994. Although this is one of the
lation-sensitive single-strand conforma- most powerful approaches to the quan-
tion analysis (MS-SCCA). However, the titative analysis of the DNA methyla-
same method has also been called bisul- tion status of a defined DNA region,
fite-PCR-SSCP (62). unfortunately, there are currently no re-
Bisulfite modification of DNA gen- liable preparative facilities available.
erates accurate sequence disparities be-
tween methylated and unmethylated al-
leles, which can be resolved by SSCP OTHER METHYLATION-
(Figure 5). Methylcytosine detection DEPENDENT STRATEGIES
using conventional SSCP requires the NOT BASED ON BISULFITE
PCR amplification of the DNA frag-
ments of interest, denaturation of the Methylation-specific modification
double-stranded product, followed by of DNA can be achieved by the use of
non-denaturing slab gel electrophoresis hydrazine and permanganate. Hy-
and detection via silver staining by drazine can react with C and T but not
standard methods. with methylcytosine (16), whereas per-
MS-SSCA also yields semi-quantita- manganate reacts with methylcytosine
tive data. In particular, when one ana- and T but not with C (35). Both can re-
lyzes mixtures of methylated and un- act with ssDNA, but hydrazine can also
methylated DNA of a known ratio, the react with dsDNA.
relative intensities of the bands correlate Hydrazine and permanganate-modi-
with the relative degree of methylation fied nucleotides can be removed with
of the original sequence. However, de- piperidine and detected by a number of
spite the great sensitivity [e.g., success- sequencing methods such as ligation-
ful methylation analysis on microdis- mediated PCR (93). The major draw-
sected paraffin-embedded tissues (7)] back of both approaches is their poor
and rapidity of the method, the resolu- sensitivity because, even under optimal
tion for detecting single-base changes in conditions, usually at least 2 µg DNA
a DNA fragment usually requires at are required (65). Furthermore, at least
least 10% of the population to show the 10%–20% of the population should dis-
alteration. Nevertheless, this is no long- play a certain alteration to be detected
er a widespread problem because the by these methods (81).
modification of DNA normally involves Low sensitivity and resolution in
several base changes in a fragment, both methods, as well as the fact that
which implies that the efficiency of de- sequencing methods are extremely la-
tection may, in fact, reach almost 100%. bor-intensive, explain why hydrazine
In addition to classical non-denatur- and permanganate are only occasional-
ing slab gel electrophoresis, HPCE ly used to check data obtained by the
techniques can be considered an alter- bisulfite-modification technique.
native approach to resolving PCR poly-
morphisms produced after bisulfite
modification of the DNA (Figure 1) HOW TO FIND NEW
(90). This option has several advan- METHYLATION HOT SPOTS
tages over the classical method, such as
the simultaneous separation of alleles, Classical DNA methylation research
use of non-denaturing gel, and auto- helps to investigate the methylation sta-
mated means that allow sequential tus of cytosines that occur in known (or
analyses of a large number of samples. partially known) DNA sequences.
Moreover, the relative percentage of However, alternative ways of investi-
methylation can be accurately quanti- gating genome-wide methylation by
fied by calculating the area under searching for as yet unidentified spots
methylated and unmethylated peaks di- have been developed. All of these
rectly on the electropherogram with an methods rely on the distinctive proper-

Vol. 33, No. 3 (2002) BioTechniques 645

Review
ties of the CpG islands to find new primed PCR, is a simple DNA finger- though methylated CpG islands are se-
methylated sequences in the genome. printing technique that relies on arbitrar- lectively amplified, the cloning of truly
The restriction landmark genomic ily primed PCR amplification, followed CpG-rich DNA regions is frequently a
scanning (RLGS) (49) technique is one by digestion with restriction isoschizo- laborious task.
of the earliest ways reported for mers. Strain-specific arrays of DNA Another original approach to isolate
genome-wide, methylation-specific fragments are generated by PCR ampli- methylated CpG-rich regions has re-
searching (56). DNA is radioactively la- fication using arbitrary oligonucleotides cently been described (86). This method
beled at methylation-specific cleavage to prime DNA synthesis from genomic employs the affinity chromatography
sites and then size-fractionated in 1-D. sites that accidentally or roughly match. (20) of a fragment of the methyl-CpG
The products are then digested with any Usually, two cycles of PCR are per- binding domain of MeCP2 to purify
restriction endonuclease that is specific formed under low-stringency condi- methylated CpG-rich fragments from
for high-frequency targets. Fragments tions, followed by PCR at high strin- mixtures obtained by digestion with
are then separated in 2-D, which yields gency with specific primers. DNA methylation-specific restriction endonu-
several scattered methylation-related hot amplified in this manner is digested with cleases. Fragments of interest are then
spots. The location and strength of a spot a couple of methylation-sensitive iso- cloned into a λ Zap II vector (Strata-
reveal its locus and the copy number of schizomers, and fragments displaying gene, La Jolla, CA, USA), and the frag-
the corresponding restriction site. This differential methylation patterns are ments that are mostly rich in CpG dinu-
approach led to the discovery of several cloned and used as probes for Southern cleotides are isolated by the segregation
CpG islands of which there was no pre- blot analysis to corroborate the differen- of partially melted molecules in poly-
vious sequence knowledge (52,56). tial methylation of such DNA regions. acrylamide gels that contain a linear
Gonzalgo et al. (43) described anoth- Another approach is called CpG is- gradient of chemical denaturant. De-
er suitable tool for screening the genome land amplification (94). DNA is digest- spite the advantages of this approach,
for regions that display altered patterns ed with restriction isoschizomers, and the specificity of the methylated DNA
of DNA methylation. The method, the restriction products are PCR-ampli- binding column needs to be improved
termed methylation-sensitive arbitrarily fied after end-adaptor ligation. Al- for it to be a first-class method.
 Undoubtedly, one of the most effec- involving a multidisciplinary perspec- Hum. Mutat. 17:423-430.
tive means of genome-wide searching tive on the DNA methylation status. One 6.Bensaada, M., H. Kiefer, G. Tachdjian, J.M.
Lapierre, V. Cacheux, A. Niveleau, and P.
for CpG islands is the use of the novel approach is the quantification of the Metezeau. 1998. Altered patterns of DNA
CpG island arrays technology. Huang et overall degree of DNA methylation. methylation on chromosomes from leukemia
al. (53) proposed an array-based meth- This can be accomplished by high-per- cell lines: identification of 5-methylcytosines
od, termed differential methylation hy- formance separation techniques, by en- by indirect immunodetection. Cancer Genet.
bridization, which allows the simultane- zymatic/chemical means, and even by in Cytogenet. 103:101-109.
7.Bian, Y.S., P. Yan, M.C. Osterheld, C. Fon-
ous determination of the methylation situ hybridization using antibodies anti- tolliet, and J. Benhattar. 2001. Promoter
rate of greater than 276 CpG island loci. methylcytosine. To analyze the DNA methylation analysis on microdissected paraf-
CpG island library is obtained as previ- methylation status of a particular DNA fin-embedded tissues using bisulfite treatment
ously described by Cross et al. (20), and sequence in depth, methylation-sensitive and PCR-SSCP. BioTechniques 30:66-72.
8.Bianco, T., D. Hussey, and A. Dobrovic. 1999.
the DNA fragments are gridded on high- restriction endonucleases are commonly Methylation-sensitive, single-strand conforma-
density arrays. Genomic DNA from the used. In addition, once bisulfite modifi- tion analysis (MS-SSCA): a rapid method to
tissues of interest is digested with MseI, cation of the DNA has been accom- screen for and analyze methylation. Hum. Mu-
which yields a large number of frag- plished, it is possible to obtain quantita- tat. 14:289-293.
ments that contain intact CpG islands. tive or semi-quantitative data regarding 9.Bigger, C.H., K. Murray, and N.E. Murray.
1973. Recognition sequence of a restriction en-
Half of the subtracted DNA is then di- allele-specific methylation. Finally, sev- zyme. Nat. New Biol. 244:7-10.
gested with methylation-sensitive en- eral innovative techniques have been de- 10.Bird, A.P. 1986. CpG-rich islands and the
donuclease BstUI, whose sequence tar- veloped to investigate new, previously function of DNA methylation. Nature 321:209-
get occurs frequently within CpG unknown methylation hot spots within 213.
11.Bourc’his, D., D. Le Bourhis, D. Patin, A. Ni-
islands. The digestion products are used the whole genome. Future steps toward veleau, P. Comizzoli, J. Renard, and E. Vie-
as templates for linker PCR. Unmethy- automation and multi-assay arrays will gas-Pequignot. 2001. Delayed and incomplete
lated targets are differentiated from allow large numbers of samples to be reprogramming of chromosome methylation
methylated targets since the former are checked simultaneously. patterns in bovine cloned embryos. Curr. Biol.
cut and no PCR products are obtained, 11:1542-1546.
12.Busslinger, M., E. deBoer, S. Wright, F.G.
while the latter can be amplified by link- Grosveld, and R.A. Flavell. 1983. The se-
er PCR. Resulting oligonucleotides, ACKNOWLEDGMENTS quence GGCmCGG is resistant to MspI cleav-
termed pre-treated products, are used as age. Nucleic Acids Res. 11:3559-3569.
probes for screening hypermethylated We thank Dr. Esteban Ballestar for 13.Butkus, V., L. Petrauskiene, Z. Maneliene, S.
helpful discussion. This work was sup- Klimasauskas, V. Laucys, and A. Janulaitis.
sequences within the CpG island library. 1987. Cleavage of methylated CCCGGG se-
Differential methylation hybridization ported by I+D+I Project SAF grant no. quences containing either N4-methylcytosine
has been applied to the screening of 2001-0059 and The International Rett or 5-methylcytosine with MspI, HpaII, SmaI,
CpG methylation in cancer using CpG Syndrome Association. XmaI and Cfr9I restriction endonucleases. Nu-
arrays of 300 and 1104 targets, respec- cleic Acids Res. 15:7091-7102.
14.Catania, J., B.C. Keenan, G.P. Margison,
tively. Therefore, differential meth- and D.S. Fairweather. 1987. Determination of
REFERENCES
ylation hybridization can be used as a 5-methylcytosine by acid hydrolysis of DNA
sophisticated screening tool for select- 1.Abrams, E.S., S.E. Murdaugh, and L.S. Ler- with hydrofluoric acid. Anal. Biochem. 167:
ing putative DNA fragments susceptible man. 1990. Comprehensive detection of single 347-351.
base changes in human genomic DNA using 15.Cedar, H., A. Solage, G. Glaser, and A.
to analysis in greater depth by other Razin. 1979. Direct detection of methylated
more specific methods. A modification denaturing gradient gel electrophoresis and a
GC clamp. Genomics 7:463-475. cytosine in DNA by use of the restriction en-
of this method for the study of DNA 2.Adouard, V., R. Dante, A. Niveleau, E. De- zyme MspI. Nucleic Acids Res. 6:2125-2132.
methylation in cancer is the methyla- lain, B. Revet, and M. Ehrlich. 1985. The 16.Church, G.M. and W. Gilbert. 1985. The ge-
tion-specific oligonucleotide microarray accessibility of 5-methylcytosine to specific an- nomic sequencing technique. Prog. Clin. Biol.
tibodies in double-stranded DNA of Xantho- Res. 177:17-21.
(39). After bisulfite treatment and PCR 17.Clark, S.J., J. Harrison, C.L. Paul, and M.
monas phage XP12. Eur. J. Biochem. 152:115-
amplification, products are array-hy- 121. Frommer. 1994. High sensitivity mapping of
bridized. Methylation-specific oligo- 3.Annan, R.S., G.M. Kresbach, R.W. Giese, methylated cytosines. Nucleic Acids Res.
nucleotide microarrays are designed to and P. Vouros. 1989. Trace detection of modi- 22:2990-2997.
fied DNA bases via moving-belt liquid chro- 18.Costello, J.F. and C. Plass. 2001. Methylation
detect methylation at specific nucleotide matters. J. Med. Genet. 38:285-303.
matography-mass spectrometry using elec-
positions. Quantitative differences can trophoric derivatization and negative chemical 19.Cravo, M., R. Pinto, P. Fidalgo, P. Chaves, L.
be obtained by fluorescence detection. ionization. J. Chromatogr. 465:285-296. Gloria, C. Nobre-Leitao, and M.F. Costa.
4.Barbin, A., C. Montpellier, N. Kokalj-Vokac, 1996. Global DNA hypomethylation occurs in
A. Gibaud, A. Niveleau, B. Malfoy, B. Dutril- the early stages of intestinal type gastric carci-
laux, and C.A. Bourgeois. 1994. New sites of noma. Gut 39:434-438.
SUMMARY 20.Cross, S.H., J.A. Charlton, X. Nan, and A.P.
methylcytosine-rich DNA detected on meta-
phase chromosomes. Hum. Genet. 94:684-692. Bird. 1994. Purification of CpG islands using a
In contrast to genetic mutations, 5.Baumer, A., U. Wiedemann, M. Hergers- methylated DNA binding column. Nat. Genet.
DNA methylation alters gene expression berg, and A. Schinzel. 2001. A novel MSP/ 6:236-244.
without DNA base changes. Currently, DHPLC method for the investigation of the 21.de Capoa, A., F. Menendez, I. Poggesi, P. Gi-
methylation status of imprinted genes enables ancotti, C. Grappelli, M.R. Marotta, M. Di
DNA methylation can be studied using a Leandro, C. Reynaud, et al. 1996. Cytologi-
the molecular detection of low cell mosaicisms.
great variety of experimental techniques, cal evidence for 5-azacytidine-induced demeth-

Vol. 33, No. 3 (2002) BioTechniques 647

Review
ylation of the heterochromatic regions of Pinus radiata trees, p. 495-506. In E. Ritter and using restriction sites as landmarks. Proc. Natl.
human chromosomes. Chromosome Res. 4: S. Espinel (Eds.), Applications of Biotechnolo- Acad. Sci. USA 88:9523-9527.
271-276. gy to Forest Genetics. A.F. Albundia, Vitoria. 50.Heiskanen, M., A.C. Syvanen, H. Siitari, S.
22.Diaz-Sala, C., M. Rey, A. Boronat, R. Bes- 35.Fritzsche, E., H. Hayatsu, G.L. Igloi, S. Iida, Laine, and A. Palotie. 1994. A novel method
ford, and R. Rodriguez. 1995. Variations in and H. Kossel. 1987. The use of permanganate to quantitate methylation of specific genomic
the DNA methylation and polypeptide patterns as a sequencing reagent for identification of 5- regions. PCR Methods Appl. 4:26-30.
of adult hazel (corylus avellana L.) associated methylcytosine residues in DNA. Nucleic 51.Herman, J.G., J.R. Graff, S. Myohanen, B.D.
with sequential in vitro subcultures. Plant Cell Acids Res. 15:5517-5528. Nelkin, and S.B. Baylin. 1996. Methylation-
Rep. 15:218-221. 36.Furuichi, Y., Y. Wataya, H. Hayatsu, and T. specific PCR: a novel PCR assay for methyla-
23.Duthie, S.J., S. Narayanan, S. Blum, L. Pirie, Ukita. 1970. Chemical modification of tRNA- tion status of CpG islands. Proc. Natl. Acad.
and G.M. Brand. 2000. Folate deficiency in Tyr-yeast with bisulfite. A new method to mod- Sci. USA 93:9821-9826.
vitro induces uracil misincorporation and DNA ify isopentenyladenosine residue. Biochem. 52.Hirotsune, S., K. Hirose, H. Kataoka, J.
hypomethylation and inhibits DNA excision re- Biophys. Res. Commun. 41:1185-1191. Kuromitsu, Y. Furuichi, M. Muramatsu, Y.
pair in immortalized normal human colon ep- 37.Galm, O., M. Rountree, E. Bachman, K.-W. Matsuda, and Y. Hayashizaki. 1994. Spot
ithelial cells. Nutr. Cancer 37:245-251. Jair, S.B. Baylin, and J. Herman. 2001. En- mapping on the standard profile of restriction
24.Eads, C.A., K.D. Danenberg, K. Kawakami, zymatic regional methylation assay: a novel landmark genomic scanning (RLGS) of sorted
L.B. Saltz, C. Blake, D. Shibata, P.V. Danen- method to quantify regional CpG methylation chromosome 20 using methylation-insensitive
berg, and P.W. Laird. 2000. MethyLight: a density. Genome Res. 12:153-157. enzyme. Genomics 24:593-596.
high-throughput assay to measure DNA methy- 38.Gama-Sosa, M.A., R.M. Midgett, V.A. 53.Huang, T.H., M.R. Perry, and D.E. Laux.
lation. Nucleic Acids Res. 28:E32. Slagel, S. Githens, K.C. Kuo, C.W. Gehrke, 1999. Methylation profiling of CpG islands in
25.Ehrlich, M., M.A. Gama-Sosa, L.H. Huang, and M. Ehrlich. 1983. Tissue-specific differ- human breast cancer cells. Hum. Mol. Genet.
R.M. Midgett, K.C. Kuo, R.A. McCune, and ences in DNA methylation in various mam- 8:459-470.
C. Gehrke. 1982. Amount and distribution of mals. Biochim. Biophys. Acta 740:212-219. 54.Jackson, P., D. Millar, E. Kingsley, G. Yard-
5-methylcytosine in human DNA from differ- 39.Gitan, R.S., H. Shi, C.M. Chen, P.S. Yan, and ley, K. Ow, S. Clark, and P.J. Russell. 2000.
ent types of tissues of cells. Nucleic Acids Res. T.H. Huang. 2002. Methylation-specific oligo- Methylation of a CpG island within the pro-
10:2709-2721. nucleotide microarray: a new potential for moter region of the KAI1 metastasis suppressor
26.Ehrlich, M. and K.C. Ehrlich. 1993. Effect of high-throughput methylation analysis. Genome gene is not responsible for down-regulation of
DNA methylation on the binding of vertebrate Res. 12:158-164. KAI1 expression in invasive cancers or cancer
and plant proteins to DNA. EXS 64:145-168. 40.Gloria, L., M. Cravo, A. Pinto, L.S. de cell lines. Cancer Lett. 157:169-176.
27.Eick, D., H.J. Fritz, and W. Doerfler. 1983. Sousa, P. Chaves, C.N. Leitao, M. Quina, 55.Jones, P.L. and A.P. Wolffe. 1999. Relation-
Quantitative determination of 5-methylcytosine F.C. Mira, et al. 1996. DNA hypomethylation ships between chromatin organization and
in DNA by reverse-phase high-performance and proliferative activity are increased in the DNA methylation in determining gene expres-
liquid chromatography. Anal. Biochem. 135: rectal mucosa of patients with long-standing ul- sion. Semin. Cancer Biol. 9:339-347.
165-171. cerative colitis. Cancer 78:2300-2306. 56.Kawai, J., S. Hirotsune, K. Hirose, S. Fushi-
28.Esteller, M., S.R. Hamilton, P.C. Burger, S.B. 41.Goessl, C., M. Muller, R. Heicappell, H. ki, S. Watanabe, and Y. Hayashizaki. 1993.
Baylin, and J.G. Herman. 1999. Inactivation Krause, and K. Miller. 2001. DNA-based de- Methylation profiles of genomic DNA of
of the DNA repair gene O6-methylguanine- tection of prostate cancer in blood, urine, and mouse developmental brain detected by restric-
DNA methyltransferase by promoter hyperme- ejaculates. Ann. NY Acad. Sci. 945:51-58. tion landmark genomic scanning (RLGS)
thylation is a common event in primary human 42.Gonzalgo, M.L. and P.A. Jones. 1997. Rapid method. Nucleic Acids Res. 21:5604-5608.
neoplasia. Cancer Res. 59:793-797. quantitation of methylation differences at spe- 57.Kosaki, K., M.J. McGinniss, A.N. Veraksa,
29.Esteller, M., J. Garcia-Foncillas, E. Andion, cific sites using methylation-sensitive single W.J. McGinnis, and K.L. Jones. 1997. Prad-
S.N. Goodman, O.F. Hidalgo, V. Vanaclocha, nucleotide primer extension (MS-SNuPE). Nu- er-Willi and Angelman syndromes: diagnosis
S.B. Baylin, and J.G. Herman. 2000. Inacti- cleic Acids Res. 25:2529-2531. with a bisulfite-treated methylation-specific
vation of the DNA-repair gene MGMT and the 43.Gonzalgo, M.L., G. Liang, C.H. Spruck, III, PCR method. Am. J. Med. Genet. 73:308-313.
clinical response of gliomas to alkylating J.M. Zingg, W.M. Rideout, III, and P.A. 58.Kuo, K.C., R.A. McCune, C.W. Gehrke, R.
agents. N. Engl. J. Med. 343:1350-1354. Jones. 1997. Identification and characterization Midgett, and M. Ehrlich. 1980. Quantitative
30.Esteller, M., A. Sparks, M. Toyota, M. of differentially methylated regions of genomic reversed-phase high performance liquid chro-
Sanchez-Cespedes, G. Capella, M.A. Peina- DNA by methylation-sensitive arbitrarily matographic determination of major and modi-
do, S. Gonzalez, G. Tarafa, et al. 2000. primed PCR. Cancer Res. 57:594-599. fied deoxyribonucleosides in DNA. Nucleic
Analysis of adenomatous polyposis coli pro- 44.Gowher, H., O. Leismann, and A. Jeltsch. Acids Res. 8:4763-4776.
moter hypermethylation in human cancer. Can- 2000. DNA of Drosophila melanogaster con- 59.Kupper, D., M. Reuter, A. Meisel, and D.H.
cer Res. 60:4366-4371. tains 5-methylcytosine. EMBO J. 19:6918- Kruger. 1997. Reliable detection of DNA CpG
31.Esteller, M., P.G. Corn, S.B. Baylin, and J.G. 6923. methylation profiles by the isoschizomers
Herman. 2001. A gene hypermethylation pro- 45.Grigg, G. and S. Clark. 1994. Sequencing 5- MspI/HpaII using oligonucleotide stimulators.
file of human cancer. Cancer Res. 61:3225- methylcytosine residues in genomic DNA. BioTechniques 23:843-847.
3229. Bioessays 16:431-436. 60.Kutueva, L.I., V.V. Ashapkin, and B.F. Van-
32.Esteller, M., M.F. Fraga, M. Zhou-Guo, J. 46.Gruenbaum, Y., T. Naveh-Many, H. Cedar, yushin. 1996. The methylation pattern of a cy-
Garcia-Foncillas, A.K. Godwin, J. Trojan, I. and A. Razin. 1981. Sequence specificity of tosine DNA-methyltransferase gene in Arabi-
Hedenfalk, C. Vaurs-Barrihre, et al. 2001. methylation in higher plant DNA. Nature dopsis thaliana plants. Biochem. Mol. Biol. Int.
Aberrant CpG-island methylation profile and 292:860-862. 40:347-353.
5-methylcytosine DNA content in hereditary 47.Grunau, C., S.J. Clark, and A. Rosenthal. 61.Leonard, S.A., S.C. Wong, and J.W. Nyce.
human cancer mimics sporadic tumorigenesis. 2001. Bisulfite genomic sequencing: systemat- 1993. Quantitation of 5-methylcytosine by one-
Hum. Mol. Genet. 10:3001-3007. ic investigation of critical experimental para- dimensional high-performance thin-layer chro-
33.Fraga, M.F., R. Rodriguez, and M.J. Canal. meters. Nucleic Acids Res. 29:E65. matography. J. Chromatogr. 645:189-192.
2000. Rapid quantification of DNA methyla- 48.Harrison, J., C. Stirzaker, and S.J. Clark. 62.Maekawa, M., K. Sugano, H. Kashiwabara,
tion by high performance capillary elec- 1998. Cytosines adjacent to methylated CpG M. Ushiama, S. Fujita, M. Yoshimori, and T.
trophoresis. Electrophoresis 21:2990-2994. sites can be partially resistant to conversion in Kakizoe. 1999. DNA methylation analysis us-
34.Fraga, M.F., M.L. Centeno, A.E. Valdés, P. genomic bisulfite sequencing leading to methy- ing bisulfite treatment and PCR-single-strand
Moncaleán, M.J. Canal, and R. Rodriguez. lation artifacts. Anal. Biochem. 264:129-132. conformation polymorphism in colorectal can-
2000. Genomic DNA methylation and poly- 49.Hatada, I., Y. Hayashizaki, S. Hirotsune, H. cer showing microsatellite instability. Biochem.
amines levels as key processes in plant aging: Komatsubara, and T. Mukai. 1991. A ge- Biophys. Res. Commun. 262:671-676.
applications for clonal multiplication of mature nomic scanning method for higher organisms 63.Maniatis, T., E.F. Fritsch, and J. Sambrook.

648 BioTechniques Vol. 33, No. 3 (2002)

 1982. Molecular Cloning: A Laboratory Manu- bining PCR and single-strand conformation A549 cells. J. Chromatogr. A 924:377-386.
al. CSH Laboratory Press, Cold Spring Harbor, polymorphism (SSCP) analysis. Am. J. Hum. 92.Tasheva, E.S. and D.J. Roufa. 1994. Densely
NY. Genet. 49:106-111. methylated DNA islands in mammalian chro-
64.McClelland, M., M. Nelson, and E. Raschke. 78.Quivy, J.P. and P.B. Becker. 1994. Direct mosomal replication origins. Mol. Cell. Biol.
1994. Effect of site-specific modification on re- dideoxy sequencing of genomic DNA by liga- 14:5636-5644.
striction endonucleases and DNA modification tion-mediated PCR. BioTechniques 16:238- 93.Thomassin, H., E.J. Oakeley, and T. Grange.
methyltransferases. Nucleic Acids Res. 22: 241. 1999. Identification of 5-methylcytosine in
3640-3659. 79.Radlinska, M. and K. Skowronek. 1998. complex genomes. Methods 19:465-475.
65.McGrew, M.J. and N. Rosenthal. 1993. Novel procedure for the detection of 5-methyl- 94.Toyota, M., C. Ho, N. Ahuja, K.W. Jair, Q.
Quantitation of genomic methylation using lig- cytosine. Acta Microbiol. Pol. 47:327-334. Li, M. Ohe-Toyota, S.B. Baylin, and J.P. Issa.
ation-mediated PCR. BioTechniques 15:722- 80.Raizis, A.M., F. Schmitt, and J.P. Jost. 1995. 1999. Identification of differentially methylat-
729. A bisulfite method of 5-methylcytosine map- ed sequences in colorectal cancer by methylat-
66.Melki, J.R., P.C. Vincent, and S.J. Clark. ping that minimizes template degradation. ed CpG island amplification. Cancer Res.
1999. Concurrent DNA hypermethylation of Anal. Biochem. 226:161-166. 59:2307-2312.
multiple genes in acute myeloid leukemia. 81.Rein, T., D.A. Natale, U. Gartner, M. Nigge- 95.Turk, P.W. and S.A. Weitzman. 1995. Free
Cancer Res. 59:3730-3740. mann, M.L. DePamphilis, and H. Zorbas. radical DNA adduct 8-OH-deoxyguanosine af-
67.Miller, O.J., W. Schnedl, J. Allen, and B.F. 1997. Absence of an unusual “densely methy- fects activity of HpaII and MspI restriction en-
Erlanger. 1974. 5-methylcytosine localized in lated island” at the hamster dhfr ori-β. J. Biol. donucleases. Free Radic. Res. 23:255-258.
mammalian constitutive heterochromatin. Na- Chem. 272:10021-10029. 96.Uchida, T., H. Ohashi, E. Aoki, Y. Nakahara,
ture 251:636-637. 82.Rein, T., M.L. DePamphilis, and H. Zorbas. T. Hotta, T. Murate, H. Saito, and T. Ki-
68.Miniou, P., M. Jeanpierre, V. Blanquet, V. 1998. Identifying 5-methylcytosine and related noshita. 2000. Clonality analysis by methyla-
Sibella, D. Bonneau, C. Herbelin, A. Fischer, modifications in DNA genomes. Nucleic Acids tion-specific PCR for the human androgen-re-
A. Niveleau, et al. 1994. Abnormal methyla- Res. 26:2255-2264. ceptor gene (HUMARA-MSP). Leukemia
tion pattern in constitutive and facultative (X 83.Saluz, H. and J.P. Jost. 1993. Major tech- 14:207-212.
inactive chromosome) heterochromatin of ICF niques to study DNA methylation. EXS 64:11- 97.Veilleux, C., J. Bernardino, A. Gibaud, A.
patients. Hum. Mol. Genet. 3:2093-2102. 26. Niveleau, B. Malfoy, B. Dutrillaux, and C.A.
69.Mizugaki, M., K. Itoh, T. Yamaguchi, S. 84.Shapiro, R., V. DiFate, and M. Welcher. Bourgeois. 1995. [Changes in methylation of
Ishiwata, T. Hishinuma, S. Nozaki, and N. 1974. Deamination of cytosine derivatives by tumor cells: a new in situ quantitative approach
Ishida. 1996. Preparation of a monoclonal anti- bisulfite. Mechanism of the reaction. J. Am. on interphase nuclei and chromosomes]. Bull.
body specific for 5-methyl-2′- deoxycytidine Chem. Soc. 96:206-212. Cancer 82:939-945.
and its application for the detection of DNA 85.Sharma, M., R. Jain, E. Ionescu, and H.K. 98.Wagner, I. and I. Capesius. 1981. Determina-
methylation levels in human peripheral blood Slocum. 1995. Capillary electrophoretic sepa- tion of 5-methylcytosine from plant DNA by
cells. Biol. Pharm. Bull. 19:1537-1540. ration and laser-induced fluorescence detection high-performance liquid chromatography.
70.Montpellier, C., C.A. Burgeois, N. Kokalj- of the major DNA adducts of cisplatin and car- Biochim. Biophys. Acta 654:52-56.
Vokac, M. Muleris, A. Niveleau, C. Rey- boplatin. Anal. Biochem. 228:307-311. 99.Wolffe, A.P., P.L. Jones, and P.A. Wade.
naud, A. Gibaud, B. Malfoy, et al. 1994. 86.Shiraishi, M., Y.H. Chuu, and T. Sekiya. 1999. DNA demethylation. Proc. Natl. Acad.
Detection of methylcytosine-rich heterochro- 1999. Isolation of DNA fragments associated Sci. USA 96:5894-5896.
matin on banded chromosomes. Application to with methylated CpG islands in human adeno- 100.Worth, C.C., O.J. Schmitz, H.C. Kliem,
cells with various status of DNA methylation. carcinomas of the lung using a methylated and M. Wiessler. 2000. Synthesis of fluores-
Cancer Genet. Cytogenet. 78:87-93. DNA binding column and denaturing gradient cently labeled alkylated DNA adduct standards
71.Norwood, C.B., E. Jackim, and S. Cheer. gel electrophoresis. Proc. Natl. Acad. Sci. USA and separation by capillary electrophoresis.
1993. DNA adduct research with capillary elec- 96:2913-2918. Electrophoresis 21:2086-2091.
trophoresis. Anal. Biochem. 213:194-199. 87.Singer-Sam, J., M. Grant, J.M. LeBon, K. 101.Wu, J., J.P. Issa, J. Herman, D.E. Bassett,
72.Nuovo, G.J., T.W. Plaia, S.A. Belinsky, S.B. Okuyama, V. Chapman, M. Monk, and A.D. Jr., B.D. Nelkin, and S.B. Baylin. 1993. Ex-
Baylin, and J.G. Herman. 1999. In situ detec- Riggs. 1990. Use of a HpaII-polymerase chain pression of an exogenous eukaryotic DNA
tion of the hypermethylation-induced inactiva- reaction assay to study DNA methylation in the methyltransferase gene induces transformation
tion of the p16 gene as an early event in onco- Pgk-1 CpG island of mouse embryos at the of NIH 3T3 cells. Proc. Natl. Acad. Sci. USA
genesis. Proc. Natl. Acad. Sci. USA time of X-chromosome inactivation. Mol. Cell. 90:8891-8895.
96:12754-12759. Biol. 10:4987-4989. 102.Xiong, Z. and P.W. Laird. 1997. COBRA: a
73.Oakeley, E.J., A. Podesta, and J.P. Jost. 1997. 88.Singer-Sam, J., J.M. LeBon, R.L. Tanguay, sensitive and quantitative DNA methylation as-
Developmental changes in DNA methylation and A.D. Riggs. 1990. A quantitative HpaII- say. Nucleic Acids Res. 25:2532-2534.
of the two tobacco pollen nuclei during matura- PCR assay to measure methylation of DNA 103.Zhou, Y., J.M. Magill, R.J. Newton, and C.
tion. Proc. Natl. Acad. Sci. USA 94:11721- from a small number of cells. Nucleic Acids Magill. 1997. Use of a single sequencing ter-
11725. Res. 18:687. mination reaction to distinguish between cyto-
74.Oakeley, E.J., F. Schmitt, and J.P. Jost. 1999. 89.Stirzaker, C., D.S. Millar, C.L. Paul, P.M. sine and 5-methylcytosine in bisulfite-modified
Quantification of 5-methylcytosine in DNA by Warnecke, J. Harrison, P.C. Vincent, M. DNA. BioTechniques 22:850-854.
the chloroacetaldehyde reaction. BioTech- Frommer, and S.J. Clark. 1997. Extensive
niques 27:744-752. DNA methylation spanning the Rb promoter in Address correspondence to:
75.Paul, C.L. and S.J. Clark. 1996. Cytosine retinoblastoma tumors. Cancer Res. 57:2229-
methylation: quantitation by automated ge- 2237. Dr. Manel Esteller
nomic sequencing and GENESCAN analysis. 90.Suzuki, H., F. Itoh, M. Toyota, T. Kikuchi, H. Cancer Epigenetics Laboratory
BioTechniques 21:126-133. Kakiuchi, Y. Hinoda, and K. Imai. 2000. Program of Molecular Pathology
76.Piyathilake, C.J., A.R. Frost, W.C. Bell, D. Quantitative DNA methylation analysis by flu- Centro Nacional de Investigaciones
Oelschlager, H. Weiss, G.L. Johanning, A. orescent polymerase chain reaction single- Oncológicas (CNIO)
Niveleau, D.C. Heimburger, et al. 2001. Al- strand conformation polymorphism using an C/Melchor Fernández Almagro, no. 3. E-28029,
tered global methylation of DNA: an epigenetic automated DNA sequencer. Electrophoresis Madrid, Spain
difference in susceptibility for lung cancer is 21:904-908. e-mail: [email protected]
associated with its progression. Hum. Pathol. 91.Tan, W.G., T.J. Carnelley, P. Murphy, H.
32:856-862. Wang, J. Lee, S. Barker, M. Weinfeld, and
77.Poduslo, S.E., M. Dean, U. Kolch, and S.J. X.C. Le. 2001. Detection of DNA adducts of
O’Brien. 1991. Detecting high-resolution benzo[a]pyrene using immunoelectrophoresis
polymorphisms in human coding loci by com- with laser-induced fluorescence. Analysis of

Vol. 33, No. 3 (2002) BioTechniques 649