of 20
DNA Electrochemistry shows DNMT1 Methyltransferase
Hyperactivity in Colorectal Tumors
Ariel L. Furst
and
Jacqueline K. Barton
*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena,
CA 91125
Abstract
DNMT1, the most abundant human methyltransferase, is responsible for translating the correct
methylation pattern during DNA replication, and aberrant methylation by DNMT1 has been linked
to tumorigenesis. We have developed a sensitive signal-on electrochemical assay for the
measurement of DNMT1 activity in crude tissue lysates. We have further analyzed ten tumor sets
and found a direct correlation between DNMT1 hyperactivity and tumorous tissue. In the majority
of samples analyzed, the tumorous tissue has significantly higher DNMT1 activity than the
healthy adjacent tissue. No such correlation is observed in measurements of
DNMT1
expression
by quantitative PCR, DNMT1 protein abundance by Western blotting, or DNMT1 activity using a
radiometric DNA labeling assay. DNMT1 hyperactivity can result from both protein
overexpression and/or enzyme hyperactivity. DNMT1 activity measured electrochemically
provides a direct measure of activity in cell lysates and, as a result, provides a sensitive and early
indication of cancerous transformation.
Graphical Abstract
*
To whom correspondence should be addressed: jkbarton@caltech.edu.
Author Contributions.
A. L. F. conceived of experiments, acquired and analyzed data and wrote the manuscript. J. K. B. conceived
of experiments and wrote the manuscript.
Supporting Information Available. Additional experimental details as well as additional quantifications of the electrochemistry and
tritium data as well as images of Western blot bands used for quantification are available.
Publisher's Disclaimer:
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our
customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of
the resulting proof before it is published in its final citable form. Please note that during the production process errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
HHS Public Access
Author manuscript
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Published in final edited form as:
Chem Biol
. 2015 July 23; 22(7): 938–945. doi:10.1016/j.chembiol.2015.05.019.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
INTRODUCTION
Colorectal cancer is the third most prevalent cancer worldwide, causing approximately
700,000 mortalities annually (
Forman et al., 2013
). The study of the causes of this disease is
especially important, as its diagnosis is on the rise among people under the age of 50, and
one of the major causes of mortality from this disease is metastasis due to its late detection
(
Mandelblatt et al., 1996
). Many molecular factors have been found to contribute to the
onset of this disease, including a host of genetic mutations (
Fearon and Vogelstein, 1990
,
Lengauer et al., 1997
,
Liu et al., 1995
,
Vogelstein and Kinzler, 2004
) and epigenetic
modifications (
Jones and Laird, 1999
,
Feinberg and Tycko, 2004
,
Jones and Baylin, 2002
,
Frigola et al., 2006
), as well as the inactivation of DNA repair pathways (
Leach et al., 1993
,
Lynch and de la Chapelle, 2003
,
Jin and Robertson, 2012
).
While many factors likely contribute to the initiation and development of colorectal cancer,
epigenetic modifications are of special interest, as they are connected to the progression of a
variety of cancers (
Sharma et al., 2010
,
Esteller, 2008
). DNA methylation in particular has
garnered significant interest, as aberrant DNA methylation has been found to be a hallmark
of many cancers (
Baylin and Herman, 2000
,
Esteller, 2007
), including colorectal cancer
(
Toyota et al., 1999
). Genomic hypermethylation is often found in colorectal cancer and has
been linked to the methylation of tumor suppressor genes and genes for DNA repair
proteins, leading to their silencing and therefore tumorigenesis (
Esteller et al., 2001
,
Fearon
and Vogelstein, 1990
,
Esteller et al., 2000
).
In humans, there are two classes of methyltransferases:
de novo
methyltransferases
(DNMT3a, DNMT3b, and DNMT3L) and maintenance methyltransferases (DNMT1).
De
novo
methyltransferases are in relatively low copy number and are responsible for
establishing methylation patterns on the genome, meaning that they have a large preference
for unmodified DNA (
Okano et al., 1999
). In contrast, DNMT1, the most abundant
mammalian methyltransferase, is a maintenance methyltransferase responsible for
Furst and Barton
Page 2
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
transferring the genomic methylation pattern from the parent DNA strand to the daughter
strand during DNA replication (
Bestor, 2000
). Because of its vital role in maintaining
genomic methylation patterns during DNA replication, DNMT1 may be important in these
molecular transformations that lead to the development of colorectal cancer.
Despite the potential importance of DNMT1 activity in disease initiation and progression,
there is currently no clinical test for its activity. Generally, quantitative PCR (qPCR), which
can be used to quantify gene expression of this protein, is used as a correlative measurement
for the total amount of DNMT1 present (
El-Deiry et al., 1991
). Other methods, such as
bisulfite sequencing (
Toyota et al., 1999
,
Zou et al., 2002
), are used to detect specific,
disease-relevant methylation patterns for early clinical diagnosis. However, such techniques
are very costly and have limited efficacy (
Munteanu and Mastalier, 2014
). In order to obtain
a direct measure of methyltransferase activity, the current laboratory gold standard involves
radiolabeling DNA with a tritium-labeled methyl group (
Fraga and Esteller, 2002
). This
assay not only produces relatively high variability but also requires the use of radioactivity
and specialized instrumentation for measurement, making it impractical for clinical use.
We have previously developed an electrochemical method for the assessment of DNMT1
activity from crude cultured cell and tissue lysate (
Furst et al., 2014
). This assay is
conducted on a multiplexed, two working electrode platform (Figure 1) that enables
electrochemical readout from disperse DNA monolayers with signal amplification and no
necessary background correction. With this platform, low-density DNA monolayers are
formed through electrochemical activation of an inert copper precatalyst into an active
catalyst (
Furst et al., 2013
). Electrochemical readout is accomplished through the
measurement of current generated from a catalytic cycle. In this cycle, DNA-mediated
charge transport reduces an intercalative redox probe, methylene blue, to leucomethylene
blue. This form of the probe has a reduced affinity for DNA, destacking from the helix and
entering solution. In solution, leucomethylene blue reduces the ferricyanide electron sink to
ferrocyanide, in turn becoming reoxidized to methylene blue. Amplified DNA-mediated
electrochemical signals are generated without necessitating background correction through
the detection of the current generated from the reoxidation of ferrocyanide at the secondary
electrode array, which is proportional to the amount of ferrocyanide present (Figure 2). This
cycling enables significant signal amplification, leading to very sensitive detection. Using
this sensitive detection platform, the methylation state of DNA can be assayed easily. Bound
to the electrode is a hemi-methylated oligonucleotide duplex. Methylation of this DNA
substrate is converted into an electrochemical signal through treatment of the hemi-
methylated DNA bound to the electrode with a methylation-specific restriction enzyme; if
the DNA oligonucleotide on the electrode has been fully methylated by DNMT1, restriction
is inhibited and the electrochemical signal remains large. On the other hand, if the DNA is
not further methylated by DNMT1, the restriction enzyme cuts the DNA, and the signal,
reflecting the amount of DNA, is significantly diminished (Figure 2) (
Muren and Barton,
2013
). Thus the assay is rapid, requires no background correction or complex
instrumentation, and requires only 500 μg of tissue per electrode, which is significantly less
sample than is currently required with a biopsy (
Furst et al., 2014
).
Furst and Barton
Page 3
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Using this platform, we evaluate DNMT1 activity in ten sets of tumor tissue and healthy
adjacent tissue, as well as in cultured colorectal carcinoma and normal colon cells.
Significantly more DNMT1 activity is observed electrochemically in the majority of the
tumor samples as compared to their healthy tissue counterparts, making this assay promising
as an early clinical diagnostic for cancerous transformations.
RESULTS
DNMT1 Activity Measured Electrochemically
The electrochemical assay for DNMT1 activity involves the signal-on detection of
methylation events on synthetic substrate DNA tethered to an electrode surface (
Muren and
Barton, 2013
). The process involves two steps: first, the pure methyltransferase or crude
lysate is added to the surface. Subsequently, a methylation-specific restriction enzyme is
used to cut any DNA that was not methylated during the previous treatment step. Successful
methylation, resulting in protection from a methylation-specific restriction enzyme,
maintains an electrochemical signal (Figure 2), while inactivity of the methyltransferase on
the DNA-modified surface results in a significantly diminished electrochemical signal
following restriction enzyme treatment.
When pure DNMT1 is titrated onto a DNA-modified electrode modified with hemi-
methylated DNA, a binding curve is generated, and a binding constant can be extracted
based on the percent of the electrochemical signal that remains following restriction of the
substrate DNA. A similar binding curve is generated from the addition of pure DNMT1 to
the lysate of 4000 HCT116
DNMT1
−/−
cultured cells (the previously determined optimal
lysate concentration for DNMT1 activity detection) (
Furst et al., 2014
). The titration data,
along with the curve fit to the Hill binding model, are shown in Figure 3. Based on the
inflection point of this curve, a K
D
of pure DNMT1 on this DNA-modified electrode is
found to be 31±1.3 nM, and 32±1.8 nM for pure DNMT1 in HCT116
DNMT1
−/−
lysate.
These values are in good agreement with previously determined values for the K
D
of
DNMT1 in solution (26 nM) (
Lee et al., 2005
). A surface K
D
on the same order as solution
values indicates that not only is our platform especially sensitive for the detection of this
protein but also that the morphology of the DNA on our surface mimics the native substrate
for the DNMT1 protein, allowing full access of the protein to the DNA bound to the surface.
Similarly, the differential signal protection from tumor and adjacent healthy tissue sets can
be measured electrochemically. Ten tumor sets (A–J), which were received snap frozen with
no further treatment, were evaluated, and an example (tissue set A) is shown in Figure 3.
Electrochemical data for this set show significant DNMT1 hyperactivity in the tumor sample
as compared to the normal adjacent tissue over several replicates. With the data shown, an
electrochemical assay using pure, isolated DNMT1 (green bar in Figure 3) is always
included as a positive control. Also it should be noted that the tumor lysate as well as the
adjacent healthy tissue lysate are tested for methyltransferase activity using both a hemi-
methylated and an unmethylated DNA substrate. It is clear, even without normalization to
the healthy adjacent tissue, that tumor A has significantly more DNMT1 activity than the
healthy tissue. Additionally, the incorporation of a comparison between the hemi-methylated
and unmethylated substrates ensures that the activity we monitor is specific to DNMT1, and
Furst and Barton
Page 4
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
not to any
de novo
methyltransferases, given the noted preference of DNMT1 for a hemi-
methylated substrate (
Hermann et al., 2004
).
The data for each tissue set tested (A–J) have been aggregated into a single graph indicating
the fold excess of DNMT1 activity in the tumor tissue as compared to the normal adjacent
tissue (Figure 4). To calculate fold excess, each tumor and adjacent normal tissue have been
normalized for the signal protection for the unmethylated DNA substrate to account for
methylation that is not DNMT1-specific. Subsequently, the tumor tissue value is normalized
to the adjacent healthy tissue, producing a ratio. If the fold excess is >1, the DNMT1 activity
in the tumor is higher than in the adjacent tissue. A fold excess of <1 indicates lower activity
in the tumor relative to the healthy tissue, while a value of 1 indicates equivalent expression.
As can be seen in Figure 4, none of the tumors exhibits lower DNMT1 activity than their
healthy tissue counterparts. In fact, the vast majority of the samples show significantly
higher activity of DNMT1 as compared to the healthy tissue.
Radiometric Assay for DNMT1 Activity
In addition to measuring DNMT1 activity with our electrochemical assay, the current
generally used radiochemical assay for DNMT1 activity (
Fraga and Esteller, 2002
) was used
to assess the ten tumor samples. This assay involves the addition of substrate DNA and
tritium-labeled S-Adenosyl methionine (SAM) cofactor, the source of methyl groups, to
crude lysate or pure DNMT1. Protein activity is extrapolated from the amount of radioactive
methylation added to the DNA following incubation, determined through scintillation
counting of the samples in triplicate. Just as with the electrochemical assay, both hemi-
methylated and unmethylated substrate DNA were tested with the crude lysate for
methylation. Analogously to the electrochemical assay, the resulting fold excess
determination is calculated based on the ratio of hemi-methylated substrate to unmethylated
substrate counts for the tumor sample normalized to the hemi-methylated divided by the
unmethylated counts for the normal tissue. As can be seen in Figure 4, although the trend in
fold excess activity is similar for tritium labeling as for electrochemistry, the extent of
hyperactivity is diminished across all samples. Additionally, the majority of the samples
have essentially statistically equivalent activity measured in this manner. This is, in part, due
to the correction for the activity on unmethylated DNA. In this assay, genomic DNA from
the lysate remains in the reaction mixture and is therefore capable of being methylated and
contributing to the overall activity, even though it is not the target substrate. This seems to
have led to higher activity in the unmethylated substrate samples for the tumors in many
cases, which, when corrected, may lead to overall lower fold excess activity. The values that
simply compare the hemi-methylated tumor counts to their normal tissue counterparts are
shown in Figures S1 and S2. Because of this large difference, tritium labeling of DNA
appears to be a less precise measurement of specific DNMT1 activity from lysate than
electrochemical measurements.
DNMT1
Expression Measured by RT-qPCR
Currently, the most common method to analyze methyltransferases from lysate is through
evaluation of genomic expression of the methyltransferase of interest. Thus, as a baseline
with which to compare our electrochemical data, the expression of the
DNMT1
gene in each
Furst and Barton
Page 5
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
tumor and adjacent healthy tissue sample was evaluated by reverse transcription-quantitative
PCR (RT-qPCR). A relative quantification method was used to determine the fold excess of
DNMT1
gene expression in the tumor sample as compared to the adjacent normal tissue
sample. For each set of samples, the relative abundance of
DNMT1
is first normalized to the
expression of
β
-Actin
in each sample, followed by comparison between the normal tissue
and tumor tissue. A fold excess of
DNMT1
from each tumor and normal tissue pair is
calculated, and a value of >1 indicates higher expression in the tumor as compared to the
normal tissue, <1 indicates lower expression in the tumor as compared to the healthy tissue,
and a value of 1 indicates equivalent expression.
Multiple primer sets were evaluated for consistency over multiple trials, and the most
consistent results were obtained with RealTime Primers, which were used for both
DNMT1
and
β
-Actin
quantification by RT-qPCR. As can be seen in Figure 5, there is a large degree
of variability among the ten tumor sets regarding the overexpression of
DNMT1
in the
tumors as compared to the normal adjacent tissues. Sample B has the most significant
upregulation, with a 22±5 fold excess of
DNMT1
expression, while sample G has significant
downregulation of
DNMT1
in the tumor sample, with a fold excess of 0.3±0.1. Overall, no
trends were observed with regards to the expression of
DNMT1
among these samples;
tumors A, B, E, F, as well as the HCT116 cells compared to the CCD-18Co cells all have
significant upregulation of
DNMT1
in the tumor tissue. Samples C, D, and I have equivalent
DNMT1
expression compared to adjacent tissue, while samples G, H, and J have significant
downregulation of
DNMT1
in the tumors.
Protein Content Measured by Western Blot
In addition to
DNMT1
expression evaluated by RT-qPCR, the total DNMT1 protein content
of each tissue sample was evaluated by Western blotting. Total nuclear protein (60 μg) was
added to each lane, and the amount of DNMT1 protein in the tumor tissue as compared to
the healthy adjacent tissue was determined through quantification of the bands resulting
from the Western blot (Figure 5). Again, a large amount of variability is observed in the
total amount of DNMT1 within each lysate sample. The bands for DNMT1 protein as well
as the nuclear protein Lamin A, used as a loading control, are given in Figure S3.
Importantly, the ratio of DNMT1 in the tumor as compared to the normal adjacent tissue
correlates directly to the fold excess of the
DNMT1
gene expression quantified by RT-
qPCR.
Figure 6 then compares our electrochemical measurement of activity for the various tumor
samples, normalized to the adjacent healthy tissue, with
DNMT1
expression, also
normalized. Remarkably, as is evident in the Figure, there is no correlation between
expression levels and resultant activity. While, for example, sample D has comparable
expression levels in the tumor and adjacent tissue, the protein is found to be hyperactive
electrochemically. In the case of sample B, the high activity seen electrochemically appears
instead to be a function of the very high expression levels. Thus the electrochemical
measurements allow a clear determination of methyltransferase activity associated with a
given sample, and comparisons with expression and/or Western blotting permit the
evaluation of whether the high activity in a given tumor sample results from high protein
Furst and Barton
Page 6
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
content or protein hyperactivity. The electrochemical assay is seen to provide the most direct
measure of DNMT1 activity, not simply its cellular abundance.
DISCUSSION
The electrochemical platform developed for the measurement of DNMT1 activity allows the
detection of this protein from crude cultured cell and colorectal tissue lysates. However, the
utility of this platform for potential clinical diagnostic applications rests in the analysis of
many samples. Here, ten tumor samples and adjacent healthy tissue have been evaluated.
Our electrochemical platform facilitates extremely sensitive protein detection, with
femtomole detection limits for the DNMT1 protein (
Furst et al., 2014
). Furthermore, the
electrochemical assay is a signal-on method of detection, ensuring specific DNMT1
discrimination from a host of proteins, DNA and RNA fragments found in crude lysates.
Importantly, measurements of methyltransferase activity using the hemi-methylated versus
unmethylated substrate distinguishes the methyltransferase activity as associated with
DNMT1 versus other methyltransferases, since only DNMT1 is preferentially active on the
hemi-methylated substrate (
Hermann et al., 2004
).
From the electrochemical analysis of the ten samples tested, in addition to lysate from
cultured colorectal carcinoma cells, a pattern was found to emerge. Hyperactivity in the
tumor samples as compared to the normal tissue is clearly found in the majority of tissue
specimens analyzed. Furthermore, the few samples that did not show hyperactivity in the
tumor tissue have equal activity between the tumor and healthy tissue. In fact, a one-way
ANOVA analysis of variance between the tumor and normal adjacent tissue DNMT1
activity was performed. Despite our small sample size (10), we found a confidence level of p
< 0.05, indicating a 95% confidence that the DNMT1 activity in the tumor tissues is
different from the healthy adjacent tissues.
Currently, the most prevalent method of methyltransferase activity measurement is a
radioactivity assay involving the application of tritium-labeled SAM cofactor to the
methyltransferase of interest in pure form or in a mixture from crude lysate (
Furst et al.,
2014
). Substrate DNA is added to the reaction, and activity is measured based on
scintillation counts resulting from tritium-labeled methyl groups added to DNA. This assay
is also used here as a comparison to our electrochemical assay. The results from this activity
assay, however, show less hyperactivity in the tumor samples as compared to the
electrochemical assay. Indeed, the radioactivity assay does not show for most samples a
statistically clear hyperactivity of DNMT1 in tumorous tissue. This lack of a clean
correlation is likely due to methylation of genomic DNA present in the lysate, which is
inseparable from measurements of the methylation of the synthetic, target DNA added to the
reaction. Tritium labeling, therefore, does not provide a pure measurement of DNMT1
activity on a target substrate when measured from lysate, as methylation of native DNA can
affect the levels of methylation measured and therefore, the ratio of methylation on hemi-
methylated versus unmethylated substrates. It is especially important to note that this is not a
problem with our electrochemical assay, as the only DNA methylation events that contribute
to the electrochemical signal are those that occur on the target DNA in the electrochemical
device. Noteworthy also is that the tritium-based assay is not suitable for clinical
Furst and Barton
Page 7
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
applications due to the large amounts of reagents required and the necessity of radioactivity
for detection.
Perhaps more important in the context of clinical relevance is the comparison of DNMT1
activity measurements to the primary method currently applied in analysis of clinical
samples: gene expression of the
DNMT1
gene by RT-qPCR. Many studies have focused on
correlations between the expression of this methyltransferase gene and tumorigenesis (
El-
Deiry et al., 1991
,
De Marzo et al., 1999
,
Peng et al., 2005
,
Girault et al., 2003
). However,
upon analysis of the ten sample sets involved in this study, no such correlation is observed.
There are just as many samples that have upregulation of
DNMT1
in tumors as equivalent
expression and underexpression. We also observe no correlation between the activity of
DNMT1 in a sample and its gene expression level as measured by RT-qPCR. Thus while
RT-qPCR has potential applications for general research regarding methyltransferases, in
and of itself, it is insufficient as a diagnostic to identify potentially tumorous
transformations.
Western blotting for total DNMT1 protein content in the lysate samples yields similar results
as RT-qPCR. The DNMT1 protein levels are found to correlate to the expression of
DNMT1
, but not to the activity of the protein measured electrochemically. This finding
indicates that it is not simply the amount of protein present that is responsible for the
hyperactivity in the tumor samples observed electrochemically. Electrochemical
measurement of DNMT1 activity is not a measure of the total amount of DNMT1, but
instead of the methylase activity of the DNMT1 protein on the hemi-methylated substrate
present in the sample. Based on the finding that the amount of DNMT1 does not directly
correlate to the activity of this protein present in the sample, it is likely that other post-
translational factors must influence the activity of DNMT1 (
Roundtree et al., 2000
,
Smallwood et al., 2007
).
If one considers the various sources of changes in protein activity, it is understandable that
the electrochemical assay, which directly measures the methyltransferase activity on a hemi-
methylated target, would yield the clearest diagnostic for DNMT1 as a source of epigenetic
change within a given tissue. Measurement of protein activity is closest in time and space to
the epigenetic changes responsible for cancerous transformation, certainly as compared to
the measurement of mRNA for the methyltransferase, which reflects effects on transcription,
and even measurements of the abundance of protein, reflecting the status of DNMT1
following translation. What is key for cancerous transformation is the activity itself, turning
on and off given genes as a result of methylation. The fact that only some of the tumorigenic
changes in a sample depend on the amount of DNMT1, measured by Western blot or
expression levels, underscores that point; tumorigenic changes are also seen to depend upon
enzymatic hyperactivity of a given DNMT1. In developing assays for epigenetic changes as
a source of tumorigenic change, this point needs to be kept in mind. In sum, then, using our
electrochemical data, it is clear that tumorigenesis does indeed correlate with DNMT1
hyperactivity, and hence the electrochemistry provides a useful early and sensitive
diagnostic for cancerous transformation.
Furst and Barton
Page 8
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
SIGNIFICANCE
We have developed an electrochemical platform based upon DNA charge transport for the
measurement of DNMT1 activity from crude lysate samples. Using this platform to analyze
ten colorectal carcinoma samples, as well as cultured colorectal carcinoma cells, we find a
direct correlation between hyperactivity of DNMT1 and tumorous tissue. Significant
hyperactivity of the protein is found in the majority of samples. This hyperactivity does not
correlate with either overexpression or total amount of DNMT1 within the sample. Instead,
significant enzyme hyperactivity is frequently evident. Furthermore, DNMT1 hyperactivity
as an indicator of cancerous transformation, measured electrochemically, is not cleanly
observed with the current standard techniques of DNMT1 analysis, including tritium
labeling, RT-qPCR, and Western blotting. Thus our electrochemical platform has the
potential to provide a sensitive method of detecting DNMT1-related cancerous
transformations and with greater reliability than current DNMT1 analysis techniques.
EXPERIMENTAL PROCEDURES
RT-qPCR
Total RNA was extracted from 30 mg of tissue and 3,000,000 cultured cells of each cell
type. Tissue samples were received snap frozen with no further treatment. Tissue samples
were homogenized in a total RNA extraction lysis buffer from a RNEasy kit (Qiagen). The
total RNA extracted using this kit was eluted into a 300 microliters of RNAse-free water.
Following RNA isolation, RT-PCR was performed to make a complete cDNA library using
a Transcription First Strand cDNA Synthesis kit (Roche) and standard RT-PCR
thermocycler conditions (C1000 Touch
Thermal Cycler, Bio-Rad). Total cDNA
concentrations were determined by nanodrop. For these samples, 250 ng/μL of total DNA
was used.
Three
DNMT1
primer sets were evaluated for consistency and reproducibility.
β
-Actin
expression was evaluated to normalize the
DNMT1
results with primers from RealTime
Primers;
GAPDH
was also evaluated as a housekeeping gene but was observed to be
extremely variable.
DNMT1
primers were obtained from RealTime Primers, qSTAR and
Genocopeiea.
DNMT1
primers from RealTime primers were found to have the most
consistent results over multiple trials, and were therefore used for all tissue sample
evaluation. Each sample was run in quadruplicate for RT-qPCR measurements. RT-qPCR
was performed on a CFX 96 Real-Time PCR Detection system (Bio-Rad). RT-qPCR results
were calculated using a relative quantification method, with the C
t
to determine fold excess
of the mRNA between the adjacent normal and tumor tissue. Error was propagated through
the experiment for four samples of each primer and sample set.
Cell and Tumor Preparation for
3
H and Electrochemistry
100 mg of each tissue sample was prepared as a crude cell lysate. Tissue samples were
homogenized before nuclear isolation using a 3 mL homogenizer with 100 strokes of the
pestle. Cultured cells were harvested upon confluence (approximately 7,000,000 cells) and
taken directly to rinsing. Both tissue and cultured cells were rinsed with 1 mL phosphate
Furst and Barton
Page 9
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
buffer (5 mM phosphate, 50 mM NaCl, pH 7.0) and centrifuged. The tissue and cells were
then prepared with a commercial NE-PER
TM
Nuclear and Cytoplasmic Extraction kit
(Thermo Scientific). Following nuclear lysis, the lysate was buffer exchanged with a 10 kDa
spin column into DNMT1 reaction buffer (50 mM Tris HCl, 1 mM EDTA, 5% glycerol, pH
7.8), flash frozen and stored at −80° C.
Once the nuclear lysate was aliquotted and frozen, total protein concentrations were
determined using a bicinchoninic acid (BCA) assay kit (Pierce) for protein concentration.
Concentrations of lysate were found to range from 1000–6000 μg/mL of protein. For all
assays, tissue samples were normalized to total protein concentration.
3
H Assay
Tritium assays were performed using the protocol previously established in our lab (
Muren
and Barton, 2013
). The DNA used as a substrate for electrochemical measurements, with
sequences provided in the Supplemental Information, including the hexynyl terminal
modification, was used in this radioactive methyltransferase activity assay. 0.5 μCi
3
H-SAM
with 20 μM DNA, 100 μg/mL of BSA, and lysate (final concentration of 500 μg/mL total
protein) were combined to a total volume of 20 μL in DNMT1 reaction buffer (50 mM Tris
HCl, 1 mM EDTA, 5% glycerol, pH 7.8). Samples with purified DNMT1 (BPS Biosciences)
were used as a positive standard, along with a negative standard with no methyltransferase.
Reactions were incubated for 2 h at 37 °C, followed by quenching with 30 μL of 10% TCA
in water. The solutions were then spotted onto DE81 filter paper (Whatman) and air-dried
for 15 minutes. Each filter paper was then washed by individually soaking it in 10 mL of 50
mM Na
2
HPO
4
for 15 minutes, followed by rinsing with 50 mM Na
2
HPO
4
and 95% ethanol.
Filter papers were heated to 37 °C to dry for 15 minutes before liquid scintillation counting.
Fold excess for radioactive measurements of a given tumor set was determined by taking the
ratio of the counts for the tumor lysate on the hemi-methylated substrate to the tumor lysate
on the unmethylated substrate divided by that same ratio for the normal tissue lysate.
Electrochemistry
Electrochemistry was performed on a bipotentiostat (CHInstruments 760E) with two
working electrodes, a platinum counter electrode and an AgCl/Ag reference electrode.
Constant potential amperometry electrochemical measurements were recorded for 90
seconds with an applied potential of 320 mV to the secondary electrode and −400 mV to the
primary electrode. Constant potential amperometry measurements were performed in Tris
buffer (10 mM Tris, 100 mM KCl, 2.5 mM MgCl
2
, 1 mM CaCl
2
, pH 7.6) with 4 μM
methylene blue and 300 μM potassium ferricyanide. Scans were taken individually at each
of the 15 secondary pin electrode sets. The percent signal remaining data are reported as
compared to pure DNMT1, with variation in the data representing the standard error across
three measurements of three electrodes, all at a given condition.
To incubate electrodes with desired proteins, a 1.25 mm deep Teflon spacer was attached to
the primary electrode surface with clips to form an isolated 4 μL volume well around each
electrode. Generally, three electrodes on the device were incubated with 65 nM DNMT1 and
160 μM SAM and 100 μg/mL BSA as a positive control. For electrodes incubated with
Furst and Barton
Page 10
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
lysate, the lysate was diluted to a final concentration of ~200 μg/mL and directly combined
with SAM to a final SAM concentration of 160 μM with 50 μg/mL BSA. Three electrodes
modified with hemi-methylated DNA were treated with tumor lysate and three treated with
adjacent normal tissue lysate. Similarly, three electrodes modified with unmethylated DNA
were incubated with tumor lysate and three with normal tissue lysate. Each electrode had 4
μL of the desired solution added to the well. The platform was incubated at 37 °C for 1.5 h
in a humidified container. The DNA monolayers were then treated with 1 μM protease
solution in phosphate buffer (5 mM phosphate, 50 mM NaCl, pH 7.0) for 1 h, followed by
thorough rinsing with phosphate buffer (5 mM phosphate, 50 mM NaCl, pH 7.0) and
scanning by constant potential amperometry. The restriction enzyme
BssH
II was then added
at a concentration of 1500 units/mL for 1.5 h at 37 °C.
BssH
II was prepared by buffer
exchange into DNMT1 reaction buffer (50 mM Tris HCl, 1 mM EDTA, 5% glycerol, pH
7.8) using a size exclusion column (10 kDa, Amicon). The electrodes were again rinsed with
phosphate buffer and scanned. Results from three trials per tumor sample were aggregated
and averaged. Fold excess for electrochemical measurements of a given tumor set was
determined by taking the ratio of the percent signal remaining for the tumor lysate on the
hemi-methylated substrate to the tumor lysate on the unmethylated substrate divided by that
same ratio for the normal tissue lysate.
Western Blot Analysis of Lysate for DNMT1
The relative amount of DNMT1 protein in each tumor set was established by Western blot.
Samples were diluted to a final loading of 60 μg of protein per lane with DNMT1 reaction
buffer (50 mM Tris HCl, 1 mM EDTA, 5% glycerol, pH 7.8) and Lammli reagent with
betamercaptoethanol. Samples were probe sonicated for 10 seconds at 20% power, followed
by boiling at 90 °C for 5 minutes. Samples were loaded onto 4–12% polyacrylamide gels in
MOPS SDS buffer (50 mM MOPS, 50 mM Tris base, 0.1% SDS, 1 mM EDTA, pH 7.7) and
run at 175 mV for 2.5 hours at 4 °C. Gels were subsequently transferred to membranes with
a dry transfer procedure for 1.5 h. Membranes were blocked with 5% milk in TBST (20 mM
Tris base, 150 mM NaCl, 0.1% Tween-20, pH 7.6) at room temperature for 1 hour, followed
by overnight incubation with a 1° antibody in 5% milk in TBST (1:1000 for DNMT1 (R &
D) and 1:1000 for Lamin A (Santa Cruz)). The membranes were then rinsed with TBST
buffer. Lamin A membranes were incubated with goat anti-rabbit 2° antibody (Abcam)
(1:5000 in 5% milk with 0.02% SDS in TBST) or Donkey Anti Sheep for DNMT1 (Santa
Cruz) (1:5000) for 1 hour and then rinsed with TBST. Membranes were scanned on a Li-Cor
Odyssey CLx infrared gel scanner, and bands were quantified using Image Studio software.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
We are grateful to the NIH (GM61077) for their financial support of this research.
Furst and Barton
Page 11
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
References
Baylin SB, Herman JG. DNA hypermethylation in tumorigenesis: epigenetics joins genetics. Trends in
Genetics. 2000; 16:168–174. [PubMed: 10729832]
Bestor TH. The DNA methyltransferases of mammals. Human Mol Genet. 2000; 9:2395–2402.
[PubMed: 11005794]
De Marzo AM, Marchi VL, Yang ES, Veeraswamy R, Lin X, Nelson WG. Abnormal Regulation of
DNA Methyltransferase Expression during Colorectal Carcinogenesis. Cancer Res. 1999; 59:3855–
3860. [PubMed: 10463569]
El-Deiry WS, Nelkin BD, Celano P, Yen RWC, Falco JP, Hamilton SR, Baylin SB. High expression
of the DNA methyltransferase gene characterizes human neoplastic cells and progression stages of
colon cancer. Proc Natl Acad Sci USA. 1991; 88:3470–3474. [PubMed: 2014266]
Esteller M, Toyota M, Sanchez-Cespedes M, Capella G, Peinado MA, Watkins DN, Issa JPJ,
Sidransky D, Baylin SB, Herman JG. Inactivation of the DNA Repair Gene
O
6
-
Methylguanine-
DNA Methyltransferase
by Promoter Hypermethylation Is Associated with G to A Mutations in
K-
ras
in Colorectal Tumorigenesis. Cancer Res. 2000; 60:2368–2371. [PubMed: 10811111]
Esteller M, Corn PG, Baylin SB, Herman JG. A Gene Hypermethylation Profile of Human Cancer.
Cancer Res. 2001; 61:3225–3229. [PubMed: 11309270]
Esteller M. Cancer epigenomics: DNA methylomes and histone-modification maps. Nat Rev Genet.
2007; 8:286–298. [PubMed: 17339880]
Esteller M. Epigenetics in Cancer. N Engl J Med. 2008; 358:1148–1159. [PubMed: 18337604]
Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell. 1990; 61:759–767.
[PubMed: 2188735]
Feinberg AP, Tycko B. The history of cancer epigenetics. Nat Rev Cancer. 2004; 4:143–153.
[PubMed: 14732866]
Forman, D.; Bray, F.; Brewster, DH.; Gombe, MC.; Kohler, B.; Piñeros, M.; Steliarova-Foucher, E.;
Swaminathan, R.; Ferlay, J., editors. Cancer Incidence in Five Continents, Vol. X electronic
version. Lyon: IARC; 2013.
Fraga MF, Esteller M. DNA Methylation: A Profile of Methods and Applications. BioTechniques.
2002; 33:632–649. [PubMed: 12238773]
Frigola J, Song J, Stirzaker C, Hinshelwood RA, Peinado MA, Clark SJ. Epigenetic remodeling in
colorectal cancer results in coordinate gene suppression across an entire chromosome band. Nat
Genet. 2006; 38:540–549. [PubMed: 16642018]
Furst AL, Landefeld S, Hill MG, Barton JK. Electrochemical Patterning and Detection of DNA Arrays
on a Two-Electrode Platform. J Am Chem Soc. 2013; 135:19099–19102. [PubMed: 24328227]
Furst AL, Muren NB, Hill MG, Barton JK. Label-free electrochemical detection of human
methyltransferase from tumors. Proc Natl Acad Sci USA. 2014; 111:14985–14989. [PubMed:
25288757]
Girault I, Tozlu S, Lindereau R, Bieche I. Expression Analysis of DNA Methyltransferases 1, 3A, and
3B in Sporadic Breast Carcinomas. Clin Cancer Res. 2003; 9:4415–4422. [PubMed: 14555514]
Hermann A, Goyal R, Jeltsch A. The DNMT1 DNA-(cytosine-C5)-methyltransferase Methylates DNA
Processively with High Preference for Hemi-methylated Target Sites. J Biol Chem. 2004;
279:48350–48359. [PubMed: 15339928]
Jin, B.; Robertson, KD. DNA Methyltransferases, DNA Damage Repair, and Cancer. In: Karpf, AR.,
editor. Epigenetic Alterations in Oncogenesis. New York, United States: Springer New York;
2012. p. 3-29.
Jones PA, Laird PW. Cancer-epigenetics comes of age. Nat Genet. 1999; 21:163–167. [PubMed:
9988266]
Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet. 2002;
3:415–428. [PubMed: 12042769]
Leach FS, Nicolaides NC, Papadopoulos N, Liu B, Jen J, Parsons R, Peltomaki P, Sistonen P,
Aaltonen LA, Nystron-Lahti M, Guan XY, Zhang J, Meltzer PS, Yu JW, Kao FT, Chen DJ,
Cerosaletti KM, Fournier REK, Todd S, Lewis T, Leach RJ, Naylor SL, Weissenbach J, Mecklin
Furst and Barton
Page 12
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
JP, Jarvinen H, Petersen GM, Hamilton SR, Green J, Jass J, Watson P, Lynch HT, Trent JM, de la
Chapelle A, Kinzler KW, Vogelstein B. Mutations of a
mutS
homolog in hereditary nonpolyposis
colorectal cancer. Cell. 1993; 75:1215–1225. [PubMed: 8261515]
Lee BH, Yegnasubramanian S, Lin X, Nelson WG. Procainamide is a specific inhibitor of DNA
methyltransferase 1. J Biol Chem. 2005; 280:40749–40756. [PubMed: 16230360]
Lengauer C, Kinzler KW, Vogelstein B. Genetic instability in colorectal cancers. Nature. 1997;
386:623–627. [PubMed: 9121588]
Liu B, Nicolaides NC, Markowitz S, Willson JKV, Parsons RE, Jen J, Papadopolous N, Peltomäki P,
de la Chapelle A, Hamilton SR, Kinzler KW, Vogelstein B. Mismatch repair gene defects in
sporadic colorectal cancers with microsatellite instability. Nat Genet. 1995; 9:48–55. [PubMed:
7704024]
Lynch HT, de la Chapelle A. Hereditary Colorectal Cancer. N Engl J Med. 2003; 348:919–932.
[PubMed: 12621137]
Mandelblatt J, Andrews H, Kao R, Wallace R, Kerner J. The late-stage diagnosis of colorectal cancer:
demographic and socioeconomic factors. Am J Public Health. 1996; 86:1794–1797. [PubMed:
9003140]
Munteanu I, Mastalier B. Genetics of colorectal cancer. J Med Life. 2014; 7:507–511. [PubMed:
25713610]
Muren NB, Barton JK. Electrochemical Assay for the Signal-On Detection of Human DNA
Methyltransferase Activity. J Am Chem Soc. 2013; 135:16632–16640. [PubMed: 24164112]
Okano M, Bell DW, Haber DA, Li E. DNA Methyltransferases Dnmt3a and Dnmt3b Are Essential for
De Novo Methylation and Mammalian Development. Cell. 1999; 99:247–257. [PubMed:
10555141]
Peng DF, Kanai Y, Sawada M, Ushijima S, Hiraoka N, Kosuge T, Hirohashi S. Increased DNA
methyltransferase 1 (DNMT1) protein expression in precancerous conditions and ductal
carcinomas of the pancreas. Cancer Sci. 2005; 96:403–408. [PubMed: 16053511]
Roundtree MR, Bachman KE, Baylin SB. DNMT1 binds HDAC2 and a new co-repressor, DMAP1, to
form a complex at replication foci. Nat Genet. 2000; 25:269–277. [PubMed: 10888872]
Sharma S, Kelley TS, Jones PA. Epigenetics in cancer. Carcinogenesis. 2010; 31:27–36. [PubMed:
19752007]
Smallwood A, Esteve PO, Pradhan S, Carey M. Functional cooperation between HP1 and DNMT1
mediates gene silencing. Genes & Dev. 2007; 21:1169–1178. [PubMed: 17470536]
Toyota M, Ahuja N, Ohe-Toyota M, Herman JG, Baylin SB, Issa JPJ. CpG island methylator
phenotype in colorectal cancer. Proc Natl Acad Sci, USA. 1999; 96:8681–8686. [PubMed:
10411935]
Toyota M, Ho C, Ahuja N, Jair KW, Li Q, Ohe-Toyota M, Baylin SB, Issa JPJ. Identification of
Differentially Methylated Sequences in Colorectal Cancer by Methylated CpG Island
Amplification. Cancer Res. 1999; 59:2307–2312. [PubMed: 10344734]
Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med. 2004; 10:789–799.
[PubMed: 15286780]
Zou HZ, Yu BM, Wang ZW, Sun JY, Cang H, Gao F, Li DH, Zhao R, Feng GG, Yi J. Detection of
Aberrant p16 Methylation in the Serum of Colorectal Cancer Patients. Clin Cancer Res. 2002;
8:188–191. [PubMed: 11801557]
Furst and Barton
Page 13
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
HIGHLIGHTS
DNMT1 activity is effectively measured electrochemically from tissue samples.
Greater DNMT1 activity is observed in colorectal tumors than in healthy tissue.
Higher DNMT1 activity in tumors does not correlate to the measured amount of
DNMT1.
Furst and Barton
Page 14
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 1.
Electrochemical array for DNMT1 activity detection. The array contains two sets of fifteen
gold electrodes, each embedded in a Teflon plate. Each electrode has a 1 mm diameter. The
two complementary Teflon arrays are assembled with a 150 μm spacer between them, which
was previously determined to be the optimal distance such that signals are not diffusion-
limited.
29
The electrodes of the primary (bottom) array are modified with DNA of the
desired sequences such that DNA-mediated charge transport is detectable. The electrodes of
the secondary (top) array are bare for electrochemical detection.
Furst and Barton
Page 15
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 2.
Signal-on electrochemical assay for DNMT1 detection. Left: The bottom electrode modified
with a dilute DNA monolayer is responsible for generating electrochemical signals through
DNA-mediated charge transport (CT) amplified by electrocatalysis. Methylene blue (MB+),
a DNA intercalating redox probe, is reduced by DNA CT to leucomethylene blue (LB),
where it can interact with an electron sink, ferricyanide. Upon interaction, ferricyanide is
reduced to ferrocyanide, reoxidizing leucomethylene blue to methylene blue in the process.
Current is generated and detected at the secondary electrode from the reoxidation of
ferrocyanide. The current generated is proportional to the amount of ferrocyanide oxidized,
which depends on the amount of methylene blue reduced by DNA CT and dissociated. To
detect DNMT1, crude lysate (multicolored shapes in background) is added to the electrode.
If DNMT1 (blue heart) is capable of methylating the hemi-methylated DNA substrate (green
arrows), the DNA on the electrode becomes fully methylated. If the protein is not active, the
DNA remains hemi-methylated (red arrows). The lysate is washed away prior to the addition
of the restriction enzyme. A methylation-specific restriction enzyme (
BssH
II, brown heart)
is then added that cuts the hemi-methylated DNA (red arrow), decreasing the amount of
bound methylene blue and significantly attenuating the electrochemical signal, while leaving
the fully methylated DNA untouched. Constant potential amperometry (right) is used to
measure the percent change before and after restriction enzyme treatment. If the restriction
enzyme did not affect the DNA (top), the signals overlay. If, however, the restriction
enzyme cuts the DNA, the signal is significantly attenuated (bottom). Constant potential
amperometry is run for 90 s with a 320 mV potential applied to the secondary electrode and
a −400 mV potential applied to the primary electrode relative to an AgCl/Ag reference. All
scans are in Tris buffer (10 mM Tris, 100 mM KCl, 2.5 mM MgCl
2
, 1 mM CaCl
2
, pH 7.6)
with 4 μM methylene blue and 300 μM potassium ferricyanide.
Furst and Barton
Page 16
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 3.
Detection of DNMT1 in pure form and from crude lysate. A titration of pure DNMT1
protein (
left
) demonstrates the sensitivity of this method of detection. In blue is shown the
titration of pure DNMT1 on our electrodes, while in red is shown pure DNMT1 added to
HCT116
DNMT1
−/−
cultured cell lysate. When the data are fit to a Hill binding model (fits
shown as solid traces in plot), a K
D
of 31±1.3 nM protein is extracted for pure DNMT1 and
32 nM±1.8 nM for DNMT1 added to lysate. The data from an array used to measure the
DNMT1 activity from tumor A (
right
) show the differential between active lysate on
electrodes and inactive lysate. The green bar shows electrodes treated with 65 nM pure
DNMT1 as a positive control. The blue bars show electrodes treated with tumor A lysate on
hemi-methylated substrate (solid) and unmethylated substrate (dashed). The red bars show
electrodes treated with adjacent normal tissue A lysate on hemi-methylated DNA (solid) and
unmethylated DNA (dashed). As can be seen, a significantly higher amount of signal
protection is observed for the tumor tissue on the hemi-methylated substrate than for the
adjacent normal tissue on that substrate. The error bars show standard error across three
electrodes.
Furst and Barton
Page 17
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 4.
DNMT1 activity measured electrochemically and radioactively. The fold excess activity
measured electrochemically (left) shows hyperactivity (fold excess > 1) in all but two of the
tissue samples. Those that do not show hyperactivity show equivalent DNMT1 activity
between tumor and normal tissue (fold excess ~1). When DNMT1 activity is measured with
radioactive labeling (right), the same hyperactivity is not observed likely because the
measurement is convoluted by genomic DNA in the lysate samples. In both cases, the data
for both the tumor and normal tissue on the hemi-methylated substrate are first normalized
to that of the unmethylated substrate, and the data for the tumor tissue are then normalized to
the normal adjacent tissue. Data for hemimethylated substrates without normalization to
unmethylated substrates are shown in Figures S1 and S2. Each of the letters represents one
of the tumor and healthy adjacent tissue sets, and the bar denoted ‘cells’ represents the result
from the comparison between HCT116 colorectal carcinoma and healthy CCD-18Co
cultured cells.
Furst and Barton
Page 18
Chem Biol
. Author manuscript; available in PMC 2016 July 23.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript