A new 1′-methylenedisulfide deoxyribose that forms an efficient cross-link to DNA cytosine-5 methyltransferase (DNMT)

Full text of DOI:10.1021/ja8064304

Published on Web 12/08/2008
A New 1′-Methylenedisulfide Deoxyribose that Forms an Efficient Cross-Link
to DNA Cytosine-5 Methyltransferase (DNMT)
Uddhav Kumar Shigdel and Chuan He*
Department of Chemistry, The UniVersity of Chicago, 929 East 57th Street, Chicago, Illinois 60637
Received August 13, 2008; E-mail: chuanhe@uchicago.edu
Although only four bases, adenine, guanine, cytosine, and thymine,
encode all genetic information in DNA, there is a heritable “fifth” base,
5-methylcytosine, which can induce epigenetic changes that alter
chromatin structures. Methylation at the 5-position of cytosine in a
CpG dinucleotide is catalyzed by a conserved group of proteins called
DNA cytosine-5 methyltransferases (DNMTs) by using S-adenosyl-
methionine (SAM) as the cofactor (Figure 1A). This modification has
a profound effect on gene expression. Many cancer cells are character-
ized by abnormal DNA methylation. Repetitive DNA sequences and
some genes are hypomethylated and transcriptionally active, whereas
many genes are hypermethylated and transcriptionally inactive.1
Inhibition of human DNMTs has been shown to be an effective strategy
to treat various cancers.1e,f
Despite our understanding of the DNA methylation status of
many genes in cancer, we know little about which DNMTs mediate
these effects at particular promoters, and even less about the protein
complexes in which these enzymes function. In order to identify
Figure 1. Probes used in trapping cytosine-5 methyltransferases (DNMTs).
(A) Methylation of cytosine C5 by DNMTs using S-adenosyl-methionine
particular DNMTs, the design of a new probe that can efficiently
trap these enzymes on specific DNA sequence is necessary.
Although probes such as 5-azacytidine (5-Aza-dC) and 5-fluoro-
cytidine (5-FdC), F, have been extensively used to trap DNMTs
onto DNA (Figure 1, parts B and C),2 both probes bear some
shortcomings. The activation of the C5 carbon of the target cytosine
in 5-FdC is an inefficient, slow process, and 5-Aza-dC can be
hydrolyzed. Taking advantage of DNMTs’ invariant active-site Cys
residue and their base-flipping property,3 we report here a new
disulfide-based DNA probe, B, (Figure 1D and Figure 2) which
provides a much faster trapping of DNMTs to DNA.
(SAM). (B) 5-Azacytidine, a widely used DNMTs trapping probe. (C)
Trapping of DNMTs using 5-fluorocytidine (F). (D) The 1′-methylenedis-
ulfide deoxyribose (B) probe that can trap DNMTs through the active-site
Cys residue.
with bases G and A opposite to F in the complementary strand,
respectively (Figure 2A).
To test the cross-linking efficacy of the new base probe B (in dsDNA)
to DNMTs we used a bacterial methyltransferase M. HhaI from Hemo-
philus hemolyticus. M. HhaI methylates the internal C in the target
sequence 5′-GCGC-3′. This sequence was incorporated into our synthetic
DNA with the target cytosine replaced with either B or F.
Wild-type M. HhaI was overexpressed and purified,6a and its
cross-linking to the modified DNA was tested. We incubated 3 equiv
of DNA (3-8) with 1 equiv of M. HhaI at 16 °C for 0.25-16 h.
The reaction was quenched with methyl methanethiolsulfonate, and
the cross-linked products were examined using nonreducing SDS-
polyacrylamide gel electrophoresis. To our delight, all the probes
containing B cross-linked efficiently with M. HhaI as shown by
the appearance of a new band with retarded mobility on the SDS-
PAGE gel (Figure 2B).
The 1′-methylenedisulfide deoxyribose probe was designed to
trap DNMTs on DNA through formation of a disulfide bond with
the protein’s active-site Cys residue. Even though disulfide-based
probes have been used to trap other proteins to DNA,4 this probe
is the first one designed to trap DNMTs. The synthetic scheme of
the corresponding phosphoramidite is shown in Scheme 1. Briefly,
Hoffer’s chlorosugar (3,5-di-O-toluoyl-R,ꢀ-1-chloro-2-deoxy-D-
ribofuranose)5a and ꢀ-cyanosugar (3,5-di-O-toluoyl-ꢀ-1-cyano-2-
deoxy-D-ribofuranose)5b (1) were prepared as described previously.
This compound was then converted to the 3′- phosphoramidite
derivative 7 in 8 steps.
Importantly, time-course experiments demonstrated that our
probe B cross-linked to M. HhaI much faster than the commercially
available F (compare gel analyses B and C to gel analysis D in
Figure 2). Even after lowering the protein and DNA concentrations
to 2.5 µM, respectively, and performing reaction on ice, the cross-
linking of M. HhaI with the B-containing probe 3 completes within
one minute (Figure 2C), while it takes 16 h to obtain comparable
cross-linking with the F-modified probe. Cross-linking of M. HhaI
to the F-modified DNA worked better when the complementary
strand contains G opposite to F (Figure 2D, lanes 5 and 6), whereas
for DNA containing B there seems to be less preference for the
base opposite to B (Figure 2B, lanes 5-8). Since B cannot form
canonical DNA base pairs, it may be preferentially recognized by
DNMTs in an unstable base pair.
Using solid-phase DNA synthesis phosphoramidite 7 was
incorporated into an oligonucleotide to give ssDNA-1 (Figure 2A).
MALDI-TOF MS of ssDNA-1 before and after tris(2-carboxyeth-
yl)phosphine (TCEP) treatment validated the incorporation of 7 into
DNA as well as the presence of the disulfide group in the
oligonucleotide (Figure S13, Supporting Information). Annealing
ssDNA-1 with bases G, A, T, and C opposite to the modified B
gave double-stranded dsDNA-3, dsDNA-4, dsDNA-5, and dsDNA-
6, respectively (Figure 2A). For the purpose of comparison, we
also synthesized ssDNA-2 which contained commercially available
5-FdC (F). Double-stranded dsDNA-7 and dsDNA-8 were prepared
9
17634 J. AM. CHEM. SOC. 2008, 130, 17634–17635
10.1021/ja8064304 CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
To confirm that the cross-linking is occurring to our probe through
the active-site Cys81, we mutated this residue to Ser, and then
overexpressed and purified the mutant enzyme.6b Both dsDNA-3 and
dsDNA-7 failed to show any cross-linking with the mutant C81S M.
HhaI (Figure 2F). This result suggests that the cross-linking between
the wild-type M. HhaI and the B-containing DNA is occurring
primarily through the active-site Cys81. The binding affinity of the
mutant C81S M. HhaI to the B-containing dsDNA-3 in the absence
of reducing agents was measured. Kd was determined to be ∼20 µM
(Figure S15, Supporting Information), which is higher than that
reported between a 37 mer regular DNA and M. HhaI.6c To further
validate that our probe works for cytosine-5 methyltransferases in
general, we tested cross-linking between M. SssI and our probes. A
very good cross-linking yield was obtained between M. SssI and
dsDNA-3 (Figure S16, Supporting Information), further demonstrating
that B can be a general and efficient cross-linker to trap cytosine-5
methyltransferases onto DNA.
In conclusion, we demonstrate here a newly designed 1′-methyl-
enedisulfide deoxyribose that can efficiently trap DNMTs onto DNA.
Considering the importance of DNMTs that control various biological
processes, the development of this new probe will greatly aid the
preparation of DNMT-DNA complexes for various studies. Although
the current 1′-methylenedisulfide deoxyribose lacks base recognition
by DNMTs and cannot be incorporated into DNA by polymerase, the
fast and highly efficient cross-linking reaction will expedite the pre-
paration of DNMT-DNA complexes for various applications. The
probe may be incorporated into a specific sequence of promoter DNA
for trapping and identifying DNMT that works on that sequence and
the complex can be used as the bait to fish out partner proteins that
help to direct DNA methylation of a specific promoter sequence. Beside
proteomic studies, the high yields of DNMT-DNA complexes can be
prepared for structural characterization as well. All these studies are
currently undergoing in our laboratory.
Figure 2. Cross-linking of a methyltransferase (M.) HhaI with various
oligonucleotides. All the experiments contain 300 µM of SAM. Experiment
C was carried out on ice, whereas all other experiments were carried out at
16 °C. All the reactions were quenched with 20 mM methyl methanethi-
olsulfonate, and SDS gel was run without DTT. (A) Oligonucleotides used
in M. HhaI cross-linking; B indicates 1′-methylenedisulfide deoxyribose,
F indicates 5-fluorocytidine. (B) Coomassie Blue-stained nonreducing SDS
gel analysis of cross-linking between WT M. HhaI (10 µM) and DNA 3-6
(30 µM). The slow mobility band is the cross-linked complex band. Lanes
2-5 show the time-course experiments of the reaction between M. HhaI
and DNA-3 at 16 °C. (C) Silver stained nonreducing SDS gel analysis of
cross-linking between M. HhaI (2.5 µM) and dsDNA-3 (2.5 µM) after 1,
3 and 5 min on ice. (D) SDS gel analysis of the cross-linking between M.
HhaI and DNA 7-8 (lanes 5-6); lanes 2-4 show the time course
experiments. Even after 1 h incubation, we could barely observe cross-
linking between DNA-7 and M. HhaI. (E) Silver-stained SDS gel analysis
of the cross-linking reaction between M. HhaI (2.5 µM) and DNA-3 (2.5
µM) competed with reducing agent (DTT). (F) The C81S mutant M. HhaI
did not show any cross-linking with DNA-3 and DNA-7.
Acknowledgment. We thank Dr. L. H. Lu, Mr. Z. Li, Ms. S.
Wegner, Mr. P. Brugarolas for helpful discussion and New England
Biolabs, in particular Dr. Sanjay Kumar, for generously providing
the construct of M. HhaI. This research was supported by the W. M.
Keck Foundation (C.H.), a Concept Award from the U.S. Army
Medical Research Office, a Cottrell Scholar Award (C.H.) from
the Research Corporation and a Camille & Dreyfus Foundation
Teaching Scholar Award (C.H.).
Scheme 1. Synthesis of 1′-Methylenedisulfide Deoxyribose
Phosphoramidite
Supporting Information Available: Experimental details, Figures
S13-S16. This material is available free of charge via the Internet at
http://pubs.acs.org.
References
(1) (a) Jones, P. A.; Baylin, S. B. Nat. ReV. Genet. 2002, 3, 415–428. (b) Herman,
J. G.; Baylin, S. B. N. Engl. J. Med. 2003, 349, 2042–2054. (c) Robertson,
K. D. Nat. ReV. Genet. 2005, 6, 597–610. (d) Goll, M. G.; Bestor, T. H.
Annu. ReV. Biochem. 2005, 74, 481–514. (e) Jones, P. A.; Baylin, S. B.
Cell 2007, 128, 683–692. (f) Gilbert, J.; Gore, S. D.; Herman, J. G.; Carducci,
M. A. Clin. Cancer Res. 2004, 10, 4589–4596.
(2) (a) Yoder, J. A.; Soman, N. S.; Verdine, G. L.; Bestor, T. H. J. Mol. Biol.
1997, 270, 385–395. (b) Reinisch, K. M.; Chen, L.; Verdine, G. L.;
Lipscomb, W. N. Cell 1995, 82, 143–153. (c) Klimasauskas, S.; Kumar, S.;
Roberts, R. J.; Cheng, X. Cell 1994, 76, 357–369. (d) Osterman, D. G.;
DePillis, G. D.; Wu, J. C.; Matsuda, A.; Santi, D. V. Biochemistry 1998,
27, 5204–5210. (e) Christman, J. K. Oncogene 2002, 21, 5483–5495.
(3) (a) Roberts, R. J.; Cheng, X. Annu. ReV. Biochem. 1998, 67, 181–198. (b)
Verdine, G. L. Cell 1994, 76, 197–200.
(4) (a) Verdine, G. L.; Norman, D. P. G. Annu. ReV. Biochem. 2003, 72, 337–
366. (b) Banerjee, A.; Santos, W. L.; Verdine, G. L. Science. 2006, 311,
1153–1157. (c) Yang, C.-G.; Yi, C.; Duguid, E. M.; Sullivan, C. T.; Jian,
X.; Rice, P. A.; He, C. Nature 2008, 452, 961–965.
(5) (a) Rolland, V.; Kotera, M.; Lhomme, J. Synth. Commun. 1997, 27, 3505–
3511. (b) Jazouli, M.; Guianvarc’h, D.; Soufiaoui, M.; Bougrin, K.; Vierling,
P.; Benhida, R. Tetrahedron Lett. 2003, 44, 5807–5810.
(6) (a) Kumar, S.; Cheng, X.; Pflugrath, J. W.; Roberts, R. J. Biochemistry 1992,
31, 8648–8653. (b) Wu, J. C.; Santi, D. V. Nucleic Acids Res. 1988, 16,
703–717. (c) Mi, S.; Roberts, R. J. Nucleic Acids Res. 1993, 21, 2459–2464.
We tested the cross-linking reaction between M. HhaI and DNA
in the presence of varying amounts of DTT. Even with up to 7
mM of DTT, the cross-linking yield was not affected, highlighting
the extremely high stability of the disulfide bond between the protein
and the DNA probe (Figure 2E). The cross-linking is also selective
to the B-containing probe as a different disulfide-modified DNA
gave almost no cross-linking under the same reaction conditions
(Figure S14, Supporting Information).
JA8064304
9
J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008 17635