Development of rationally designed DNA N6 adenine methyltransferase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2012.03.072

A series of bisubstrate inhibitors for DNA N6 adenine methyltransferase (Dam) have been synthesized by linking an amine analogue of S-adenosylmethionine to an aryl moiety designed to probe the binding pocket of the DNA adenine base. An initial structure-activity relationship study has identified substituents that increase inhibitor potency to the ～10 μM range and improve selectivity against the human cytosine methyltransferase Dnmt1.

Full text of DOI:10.1016/j.bmcl.2012.03.072

Bioorganic & Medicinal Chemistry Letters 22 (2012) 3079–3082
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Development of rationally designed DNA N6 adenine
methyltransferase inhibitors
Gerard Hobley a, Jennifer C. McKelvie a, Jenny E. Harmer a, Jason Howe a, Petra C.F. Oyston b,
Peter L. Roach a,
⇑
a University of Southampton, School of Chemistry, University Road, Highfield, Southampton SO17 1BJ, UK
b Defense Science and Technology Laboratory (DSTL), Porton Down, Salisbury, Wiltshire SP4 0JQ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of bisubstrate inhibitors for DNA N6 adenine methyltransferase (Dam) have been synthesized by
linking an amine analogue of S-adenosylmethionine to an aryl moiety designed to probe the binding
pocket of the DNA adenine base. An initial structure–activity relationship study has identified substitu-
Received 7 October 2011
Revised 16 March 2012
Accepted 19 March 2012
Available online 27 March 2012
ents that increase inhibitor potency to the ꢀ10
l
M range and improve selectivity against the human
cytosine methyltransferase Dnmt1.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
DNA methyltransferases
Inhibitors
S-Adenosylmethionine
Methylation
Epigenetics
Antibacterial
Bacterial DNA adenine N6 methyltransferase (Dam) is an en-
zyme that catalyzes the transfer of a methyl group from the methyl
donor S-adenosylmethionine (AdoMet) to the N6 position of ade-
nine within the recognition sequence GATC. Dam has been shown
to play an important role in the virulence of pathogenic bacteria,1,2
with Dam mutant strains of Salmonella enterica serovar Typhimuri-
um and Yersinia pestis showing attenuation,3,4 and Dam overpro-
development of inhibitors of several human methyltransferases,
including catechol O-methyltransferase,14–16 spermine synthase,17
histone methyltransferases SET7/918 and DOT1L19,20 protein argi-
nine methyltransferases,21–24 DNA (cytosine-5)-methyltransferases
(Dnmt1–3)25 and indole N-methlytransferase.26 Enzymes targeted
from other organisms have included the plant sterol C-24 methyl-
transferase,27,28 Vaccinia virus RNA guanine 7-methyltransferase29
,
ducing strains of Yersinia pseudotuberculosis eliciting
a
fully
bacterial mycolic acid methyltransferase30 and the bacterial DNA
adenine methyltransferases CcrM31 and Dam.6 We have used this
rationale to investigate the selectivity of Dam inhibitors by incorpo-
rating a mimic of the nucleophilic adenine base into the design of
bisubstrate inhibitors based on the structure of AdoMet (Fig. 1).
Herein we report the synthesis and in vitro evaluation of a series
of these inhibitors.
protective immune response in vaccinated hosts.5 The response
of pathogenic bacteria to alterations in Dam concentrations has
led to an increased interest in the identification of Dam inhibitors
which, due to the lack of a functionally equivalent enzyme in hu-
mans, have potential as novel antibiotics.6,7
Analogues of AdoMet have long been targeted as a source of
methyltransferase inhibitors,8–12 although the wide range of meth-
yltransferases present in living organisms that share AdoMet as a
substrate has redefined the challenge to be the identification of
selective inhibitors. The concept of ‘bisubstrate inhibitors’, in which
the ligand combines features that mimic both the AdoMet and the
methyl acceptor substrate is an attractive route to achieving selec-
tivity.13 The ‘bisubstrate inhibitor’ approach has been applied to the
The bisubstrate AdoMet analogues were based on a modified
AdoMet scaffold in which the sulfonium ion was replaced with an
amine (AzaAdoMet, 7). Analogues 6–9 were prepared as described
by Wahhab et al.25 The analogues containing hydrogen 6 or methyl
7 moieties provided controls for evaluating the relative changes in
potency resulting from the incorporation of the adenine base ana-
logue substituents. Bisubstrate analogues 10–14 were synthesized
from a common amine scaffold which was prepared using a modifi-
cation of the method described by Wahhab et al.25 (Scheme 1) in
which the Cbz protecting group was substituted for a more labile
t-Boc group, facilitating deprotection under milder conditions.21
Abbreviations: TFA, trifluoroacetic acid; Dam, DNA adenine N6 methyltransfer-
ase; AdoMet, S-adenosylmethionine; AdoHcy, S-adenosylhomocysteine.
Briefly, the commercially available N-a-t-Boc-L-aspartic acid a-t-
⇑
Corresponding author. Tel.: +44 2380 59 5919; fax: +44 2380 59 3781.
E-mail address: plr2@soton.ac.uk (P.L. Roach).
butyl ester 4a was converted to the b-thioester 4b before reduction
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.03.072

3080
G. Hobley et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3079–3082
DNA
N
Tyr174
N
N
Pro173
Phe35
N
O
S
NH2
N
Ar
(CH2)n
NH2
NH2
N
N
O
H
N
N
n=0-2
N
N
Asp181
H
O
O
N
O
H
N
N
Ser40
Lys14
O
NH2
HO
Ile55
H
O
HO OH
O
H
O
H
N H
O
O
H
N
2
N H
Ala38
Trp10
O
O
Asp54
1
Figure 1. Schematic drawing of the substrate: active site interactions for E. coli Dam 1 and the family of compounds designed as bisubstrate inhibitors 2.
O
O
4a, X = OH
4b, X = SEt
a
b
Ac10-14, X = OH
Te10-14, X = SEt
a
b
R'HN
NH2
OtBu
X
4c
, X = H
Al10-14,
X = H
R
X
O
O
NHR'
H2N
NH2
O
N
N
O
c,
O
d, e
O
O
O
N
H
N
R
N
N
O
N
OtBu
N
N
OH
N
NH2
NH2
O
HO
N
OH
N
N
N
3
5
7-14
N
N
8
10
, R =
12
13
, R =
, R=
14, R=
6, R = H
O
S
N
N
H
H
7, R = CH3
9, R =
11, R =
, R=
N
H
Scheme 1. Synthesis of inhibitors. Reagents and conditions: (a) DCC, EtSH, DMAP (cat.), 2 h, 85–99%; (b) Et3SiH, Pd on C, 2 h, 27–94%; (c) NaBH(OAc)3, 210 min, 71%; (d) Al10–
14, NaBH(OAc)3, 12 h, 49-89%; (e) for compounds 8 and 9, R0 = Cbz, BF3ꢁEt2O, DCM, 0 °C, 8 h, 9–40%; for compounds 6, 10–14, R0 = Boc, TFA/H2O/p-anisole, 9:1:1, 3 h, 35–65%.
Inhibitor 6 was prepared by deprotection of compound 5 (R0 = Boc). All reaction were carried out at 18 °C.
to the aldehyde 4c.32 The 50-deoxy-50-amino-20,30-isopropylidene-
adenosine 3 was prepared as described by Reeve et al.33 and reacted
with 4c by a reductive amination to afford 5 in 71% yield. A series of
aryl substituted aldehydes Al10–14 were prepared32 from the cor-
responding acids Ac10–14 by reduction of a thioester intermediate
Te10–14 (Scheme 1). The reductive amination of 5 with aldehydes
Al10–14 in the presence of sodium triacetoxyborohydride yielded
protected tertiary amine adducts Ad10–14. Global deprotection of
Ad8–14 under acidic conditions yielded bisubstrate analogues
8–14. The switch of primary amine protecting group from Cbz25
(used for 8 and 9, 9–41%) to the more acid labile Boc21 allowed
faster, higher yielding deprotection (in the range 35–65% after
reverse phase chromatography for 10–14).
The inhibitory effect of compounds 6–14 and AdoHcy against Y.
pestis Dam was assessed using a real time fluorescence-based Dam
activity assay.34 Y. pestis Dam activity was measured in triplicate in
the presence 0–2 mM 6–14 or AdoHcy in activity assays containing
3.5 nM hemimethylated break-light oligonucleotide labelled with
fluorescein and dabcyl on adjacent ends, 16 lM AdoMet and
2 nM of the coupled restriction enzyme DpnI. The results of these
assays (Table 1) show compounds 6–14 to have IC50 values ranging
from around 10 lM to greater than 200 lM.
The results of the real time fluorescence-based activity assay
were validated using a gel-based assay in which AdoHcy and
compounds 6, 7 and 10–14 were incubated with Escherichia coli
Dam and a 10,664 base-pair unmethylated plasmid. The plasmid,

G. Hobley et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3079–3082
3081
Table 1
Inhibition of DNA methyltransferases by AdoMet analogues
NH2
O
O
N
R
N
N
IC50 (lM)
OH
Compound
NH2
HO
OH
N
N
R
Y. pestis Dama
Human Dnmt1c
0.941 0.076
Selectivity ratiod
AdoHcy
6
7, aza-AdoMet
—
H
CH3
11.9 4.8
84.0 20.9
237 68.0
0.1
1.3
302 29.4
N
8
>250b
>250b
N
9
10
11
12.3 0.8
13.1 2.7
N
H
N
H
12
13
14
39.2 2.6
28.9 8.9
31.3 13.0
O
S
132 9.36
4.6
HN
a
IC50 values are the concentration of compound which inhibits Y. pestis Dam activity by half and are expressed as mean standard deviation from two independent assays
of triplicate data points.
b
IC50 values were greater than the detection limit for the experiment.
IC50 values are the concentration of compound which inhibits Dnmt1 activity by half and are expressed as mean standard error from a single assay of triplicate data
c
points.
d
Selectivity ratio is defined as IC50(Dam)/IC50(Dnmt1).
pRL821 (Fig. S135), was used as it more closely resembles Dam’s
natural chromosomal DNA substrate and contains the Dam recog-
nition sequence (GATC) at 91 sites. For analysis, the plasmid was
linearized by the methylation insensitive restriction enzyme XhoI
and then digested with BclI, which only cleaves unmethylated
DNA. The restricted plasmid DNA was then visualized by agarose
gel electrophoresis, with a single DNA band indicating a fully
methylated plasmid and no inhibition of Dam; three bands indicat-
ing a mixture of fully and unmethylated DNA and partial inhibition
of Dam; and two bands indicating fully unmethylated DNA and
complete inhibition of Dam. The gel (Fig. 2) shows almost complete
methylation of the plasmid by E. coli Dam under assay conditions
(lane 3) and no methylation when AdoMet is excluded from the as-
say (lane 1), or by inclusion of the reaction product AdoHcy (2 mM,
lane 2) which inhibits Dam. The compounds identified as relatively
potent inhibitors (IC50 <250 lM) in the real time assay were ana-
lyzed by this method (compounds 6, 7, 10–14) and all were con-
firmed as Dam inhibitors using a plasmid DNA substrate.
Potency data (Table 1) indicates that replacing the sulfur atom
in AdoHcy with an unsubstituted nitrogen atom (compound 6) re-
sults in an eightfold decrease in potency. As observed for AdoMet
analogues with a wide range of other methyltransferases,18,21,25
Figure 2. 0.8% Agarose gel electrophoresis of restriction digest products. Lanes: L—
ladder; 1—cleavage control (no AdoMet); 2—inhibition control (containing 2 mM
AdoHcy); 3—methylation control; 4—compound 6; 5—compound 7; 6—compound
12; 7—compound 13; 8—compound 14; 9—compound 10; 10—compound 11.
Compounds were included at a concentration of 2 mM.
the presence of the nitrogen atom does not lead to a complete loss
of enzyme inhibition and a nitrogen centered scaffold provides a

3082
G. Hobley et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3079–3082
starting point for the development of Dam inhibitors. However,
comparison of the IC50 for AdoHcy and 6 suggests there is still a
need to develop improved replacements for the sulfonium group
which provide metabolic stability but also maintain or enhance po-
tency. Replacement of the hydrogen (6) with a methyl group (7)
decreases the inhibitory effect, with a threefold decrease in po-
tency compared to 6 and a 20-fold decrease compared to AdoHcy.
Modification of the nitrogen with groups containing a single het-
eroaromatic ring, such as the pyridin-2-ylmethyl 8 and 3-(pyri-
din-2-yl)propyl 9 analogues resulted in very weak inhibitors (IC50
Acknowledgements
This work was supported by the Transformational Medical
Technologies program contract W911NF-08-C-0004 from the
Department of Defense Chemical and Biological Defense program
through the Defense Threat Reduction Agency (DTRA). We thank
the Engineering and Physical Sciences Research Council for financial
support (to J.E.H.). We thank Dr. F. Turlais and colleagues at Cancer
Research Technology for the gift of Dnmt1 and helpful discussions.
>250 lM). In contrast, analogues containing bicyclic heteroaro-
Supplementary data
matic substituents showed a two to eightfold increase in potency
relative to 6, with indole analogues 10 and 11 showing similar lev-
els of potency to AdoHcy.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.03. 072.
Increasing the length of the linker between the aryl substituent
and the nitrogen of the AdoMet mimic from two to three carbons
had little effect on potency, suggesting a degree of flexibility within
the binding site accommodating the adenine mimic. This result is
surprising as it contrasts with the effects seen for compounds 6–
9, where increases in the steric bulk of the substituent correlated
with a decrease in potency. The increase in potency observed for
analogues containing even bulkier aromatic heterocyclic groups
(10–14) suggests that these compounds may employ an alternative
binding conformation to AdoMet and its closest analogues, or may
exploit a previously unexplored binding pocket.
Attempts at probing this binding interaction by varying the het-
eroatom have shown a slight decrease in potency when a nitrogen
is substituted for an oxygen or sulfur. In addition changing the
point of attachment in the case of the indole analogue from the
C3 (8) to the C4 (12) position leads to only a minor decrease in po-
tency. For these bicyclic heteroaromatic groups, a hydrophobic
binding pocket might be anticipated, although hydrogen bonding
of the indole N–H may also contribute to binding.
The potential to identify selective inhibitors was the principle
motivation for preparing bisubstrate inhibitors of Dam. The selec-
tivity of the compounds was investigated by comparing the IC50 for
Dam with the IC50 for the human cytosine methyltransferase
Dnmt1. The inhibitory effect of AdoHcy, 7 and 13 against human
Dnmt1 was measured using a real time fluorescence-based Dnmt1
activity assay.36 There were differences in the conditions used to
assay the activity of Dam and Dnmt1 (including pH, buffer, salt
and substrate concentrations) and as such, comparisons between
the IC50 values for different enzymes must be interpreted carefully.
References and notes
1. Marinus, M. G.; Casadesus, J. FEMS Microbiol. Rev. 2009, 33, 488.
2. Heusipp, G.; Falker, S.; Schmidt, M. A. Int. J. Med. Microbiol. 2007, 297, 1.
3. Heithoff, D. M.; Sinsheimer, R. L.; Low, D. A.; Mahan, M. J. Science 1999, 284,
967.
4. Robinson, V. L.; Oyston, P. C.; Titball, R. W. FEMS Microbiol. Lett. 2005, 252, 251.
5. Julio, S. M.; Heithoff, D. M.; Provenzano, D.; Klose, K. E.; Sinsheimer, R. L.; Low,
D. A.; Mahan, M. J. Infect. Immun. 2001, 69, 7610.
6. Mashhoon, N.; Carroll, M.; Pruss, C.; Eberhard, J.; Ishikawa, S.; Estabrook, R. A.;
Reich, N. J. Biol. Chem. 2004, 279, 52075.
7. Naumann, T. A.; Tavassoli, A.; Benkovic, S. J. Chembiochem 2008, 9, 194.
8. Borchardt, R. T.; Huber, J. A.; Wu, Y. S. J. Med. Chem. 1976, 19, 1094.
9. Borchardt, R. T. J. Med. Chem. 1980, 23, 347.
10. Coward, J. K.; Bussolotti, D. L.; Chang, C. D. J. Med. Chem. 1974, 17, 1286.
11. Kumar, R.; Srivastava, R.; Singh, R. K.; Surolia, A.; Rao, D. N. Bioorg. Med. Chem.
2008, 16, 2276.
12. Blanchard, P.; Dodic, N.; Fourrey, J. L.; Lawrence, F.; Mouna, A. M.; Robert-Gero,
M. J. Med. Chem. 1991, 34, 2798.
13. Broom, A. D. J. Med. Chem. 1989, 32, 2.
14. Lerner, C.; Masjost, B.; Ruf, A.; Gramlich, V.; Jakob-Roetne, R.; Zurcher, G.;
Borroni, E.; Diederich, F. Org. Biomol. Chem. 2003, 1, 42.
15. Matuszewska, B.; Borchardt, R. T. J. Neurochem. 1983, 41, 113.
16. Ellermann, M.; Paulini, R.; Jakob-Roetne, R.; Lerner, C.; Borroni, E.; Roth, D.;
Ehler, A.; Schweizer, W. B.; Schlatter, D.; Rudolph, M. G.; Diederich, F. Chemistry
2011, 17, 6369.
17. Woster, P. M.; Black, A. Y.; Duff, K. J.; Coward, J. K.; Pegg, A. E. J. Med. Chem.
1989, 32, 1300.
18. Mori, S.; Iwase, K.; Iwanami, N.; Tanaka, Y.; Kagechika, H.; Hirano, T. Bioorg.
Med. Chem. 2010, 18, 8158.
19. Yao, Y.; Chen, P.; Diao, J.; Cheng, G.; Deng, L.; Anglin, J. L.; Prasad, B. V.; Song, Y.
J. Am. Chem. Soc. 2011, 133, 16746.
20. Daigle, S. R.; Olhava, E. J.; Therkelsen, C. A.; Majer, C. R.; Sneeringer, C. J.; Song,
J.; Johnston, L. D.; Scott, M. P.; Smith, J. J.; Xiao, Y.; Jin, L.; Kuntz, K. W.;
Chesworth, R.; Moyer, M. P.; Bernt, K. M.; Tseng, J. C.; Kung, A. L.; Armstrong, S.
A.; Copeland, R. A.; Richon, V. M.; Pollock, R. M. Cancer Cell 2011, 20, 53.
21. Dowden, J.; Hong, W.; Parry, R. V.; Pike, R. A.; Ward, S. G. Bioorg. Med. Chem.
Lett. 2010, 20, 2103.
22. Obianyo, O.; Causey, C. P.; Osborne, T. C.; Jones, J. E.; Lee, Y. H.; Stallcup, M. R.;
Thompson, P. R. Chembiochem 2010, 11, 1219.
23. Bicker, K. L.; Obianyo, O.; Rust, H. L.; Thompson, P. R. Mol. Biosyst. 2011, 7, 48.
24. Osborne, T. C.; Obianyo, O.; Zhang, X.; Cheng, X.; Thompson, P. R. Biochemistry-
Us 2007, 46, 13370.
In the case of Dnmt1, the IC50 for AdoHcy of 0.9
l
M was close to
the reported literature values (0.8
l
M37 and 0.3
l
M38 respectively)
and for the analogue inhibitors, the introduction of the 2-
(benzo[b]thiophen-3-yl)ethyl group (13) gave an IC50 measure-
ment comparable (within error) to the methyl analogue 7.
However, comparison of IC50s for 7 and 13 with Dam shows a more
than fourfold preference for 13, with a 40-fold difference between
Dam and Dnmt1 in selectivity ratio between AdoHcy and 13 (Table
1). This modest but measurable preference of Dam for inhibitors
bearing the bicyclic heteroaromatic substituent provides a lead
for structure–activity relationship studies to develop more potent
and selective Dam inhibitors.
In summary, this study has confirmed the value of using the
more acid labile Boc protecting group for the efficient synthesis
of bisubstrate AdoMet analogues. The potential of rationally de-
signed bisubstrate AdoMet analogues as Dam inhibitors has also
been demonstrated with several compounds showing IC50 values
in the micromolar range and comparable to the product AdoHcy.
A study of a small library of tertiary amine AdoMet analogues iden-
tified bicyclic aromatic substituents joined by a two or three
carbon linker as combining improved potency with a degree of
selectivity.
25. Wahhab, A.; Besterman, J.; Isakovic, L.; Llewellyn, D.; Rahil, J.; Saavedra, O.;
Dezziel, R. Patent WO/2006/078,752; 2006.
26. Benghiat, E.; Crooks, P. A. J. Med. Chem. 1983, 26, 1470.
27. Janssen, G. G.; Nes, W. D. J. Biol. Chem. 1992, 267, 25856.
28. Nes, W. D.; Janssen, G. G.; Bergenstrahle, A. J. Biol. Chem. 1991, 266, 15202.
29. Benghiat, E.; Crooks, P. A.; Goodwin, R.; Rottman, F. J. Pharm. Sci. 1986, 75, 142.
30. Vaubourgeix, J.; Bardou, F.; Boissier, F.; Julien, S.; Constant, P.; Ploux, O.; Daffe,
M.; Quemard, A.; Mourey, L. J. Biol. Chem. 2009, 284, 19321.
31. Wahnon, D. C.; Shier, V. K.; Benkovic, S. J. J. Am. Chem. Soc. 2001, 123, 976.
32. Han, Y. L.; Chorev, M. J. Org. Chem. 1972, 1999, 64.
33. Reeve, A. M.; Breazeale, S. D.; Townsend, C. A. J. Biol. Chem. 1998, 273, 30695.
34. Wood, R. J.; Maynard-Smith, M. D.; Robinson, V. L.; Oyston, P. C.; Titball, R. W.;
Roach, P. L. PLoS ONE 2007, 2, e801.
35. Leonardi, R. Ph.D. Thesis, University of Southampton, 2004.
36. Wood, R. J.; McKelvie, J. C.; Maynard-Smith, M. D.; Roach, P. L. Nucleic Acids Res.
2010, 38, e107.
37. Isakovic, L.; Saavedra, O. M.; Llewellyn, D. B.; Claridge, S.; Zhan, L.; Bernstein,
N.; Vaisburg, A.; Elowe, N.; Petschner, A. J.; Rahil, J.; Beaulieu, N.; Gauthier, F.;
MacLeod, A. R.; Delorme, D.; Besterman, J. M.; Wahhab, A. Bioorg. Med. Chem.
Lett. 2009, 19, 2742.
38. Lee, W. J.; Zhu, B. T. Carcinogenesis 2006, 27, 269.