Small molecule inhibitors that discriminate between protein arginine N-methyltransferases PRMT1 and CARM1

Article abstract of DOI:10.1039/c1ob06100c

Protein arginine N-methyltransferases (PRMTs) selectively replace N-H for N-CH₃ at substrate protein guanidines, a post-translational modification important for a range of biological processes, such as epigenetic regulation, signal transduction and cancer progression. Selective chemical probes are required to establish the dynamic function of individual PRMTs. Herein, model inhibitors designed to occupy PRMT binding sites for an arginine substrate and S-adenosylmethionine (AdoMet) co-factor are described. Expedient access to such compounds by modular synthesis is detailed. Remarkably, biological evaluation revealed some compounds to be potent inhibitors of PRMT1, but inactive against CARM1. Docking studies show how prototype compounds may occupy the binding sites for a co-factor and arginine substrate. Overlay of PRMT1 and CARM1 binding sites suggest a difference in a single amino acid that may be responsible for the observed selectivity.

Full text of DOI:10.1039/c1ob06100c

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 7814
www.rsc.org/obc
PAPER
Small molecule inhibitors that discriminate between protein arginine
N-methyltransferases PRMT1 and CARM1†
James Dowden,*a Richard A. Pike,a Richard V. Parry,b Wei Hong,a Usama A. Muhsena and Stephen G. Wardb
Received 6th July 2011, Accepted 23rd August 2011
DOI: 10.1039/c1ob06100c
Protein arginine N-methyltransferases (PRMTs) selectively replace N–H for N–CH3 at substrate
protein guanidines, a post-translational modification important for a range of biological processes, such
as epigenetic regulation, signal transduction and cancer progression. Selective chemical probes are
required to establish the dynamic function of individual PRMTs. Herein, model inhibitors designed to
occupy PRMT binding sites for an arginine substrate and S-adenosylmethionine (AdoMet) co-factor
are described. Expedient access to such compounds by modular synthesis is detailed. Remarkably,
biological evaluation revealed some compounds to be potent inhibitors of PRMT1, but inactive against
CARM1. Docking studies show how prototype compounds may occupy the binding sites for a
co-factor and arginine substrate. Overlay of PRMT1 and CARM1 binding sites suggest a difference in
a single amino acid that may be responsible for the observed selectivity.
Introduction
Chemical modulators that address individual protein arginine
methyltransferases (PRMTs, EC 2.1.1.125) have potential appli-
cations for elucidating their dynamic contribution to cell biology,
or informing the development of novel therapeutics.1,2
PRMTs co-localise the versatile cellular co-factor S-adenosyl
methionine (AdoMet, 1, Scheme 1) and guanidine functionality of
the arginine-bearing target substrate to achieve site-specific guani-
dine N-methylation,3 leading to monomethylarginine (mMA).
Depending on the PRMT type, further methylation occurs at
either the same (N2), or adjacent (N3) nitrogen, to furnish
Scheme 1 Regiochemical classification of PRMTs and the structure of
co-substrate AdoMet 1, co-product AdoHcy 2 and inhibitors 3 and 4.
asymmetric or symmetrical dimethylarginine substitution (aDMA
or sDMA) respectively (Scheme 1).4
Irregular PRMT activity and arginine methylation patterns have
mixture.14 Recently, peptides featuring a reactive chloroacetami-
emerged as biomarkers for cardiovascular conditions,5,6 viruses,7
dine were demonstrated to give selective covalent inhibition of
immune disease8 and cancer.9 For example, aggressive prostate
PRMT1, but not PRMT3 or CARM1.15,16 Peptides incorporating
and breast cancers cells both show over-expression of PRMT 4
Nh-alkylated arginines have been shown to inhibit PRMTs 1 and
(also known as co-factor associated arginine methyltransferase,
6 about 10-fold more strongly than CARM1.17 The same group
CARM1).10–12
recently described further peptide inhibitors featuring Nh-arginine
Chemical interference has great potential for learning about
substituents resembling the homocysteine part of 1.18
PRMT functions and correcting aberrant processes mediated
In principle, small molecules may be readily adapted for
by them.1,2,13 PRMT-mediated peptide alkylation by an AdoMet
in vivo studies. The first reported PRMT inhibitor ‘AMI-1¢,19
derived nitrogen mustard gave potent inhibitors within the adduct
inspired reports of several related suramin-like analogues;20,21
most recently, the sulfonate functionality of AMI-1 has been
aSchool of Chemistry, University of Nottingham, University Park, Notting-
replaced by carboxylates while maintaining PRMT inhibition.22
ham, UK, NG7 2RD. E-mail: james.dowden@nottingham.ac.uk; Fax: +44
Virtual screening23 and hit-to-lead studies have revealed a number
115 9513565; Tel: +44 115 9513566
of potential inhibitor scaffolds from which pyrazole24–27 and
benzo[d]imidazole28 derivatives represent the most potent and
selective inhibitors so far. Interestingly, AdoHcy induced organ-
isation of the binding site was recently confirmed to be essential
for pyrazole compounds to achieve potency.29
bDepartment of Pharmacy and Pharmacology, University of Bath, Claverton
Down, Bath, UK, BA2 7AY
† Electronic supplementary information (ESI) available: characterisation
data for novel precursors, copies of NMR spectra and LC-MS data for
16–24; table of measured distances for docked structures and images for
docked structures of 19 and 20. See DOI: 10.1039/c1ob06100c
7814 | Org. Biomol. Chem., 2011, 9, 7814–7821
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Small molecules that target both AdoMet and arginine binding
sites offer an excellent platform to design selective PRMT in-
hibitors. We recently reported protoype PRMT1 inhibitors based
on the latter principle.30 Herein, we report comparison of a wider
panel of inhibitiors against PRMT1 and CARM1. Remarkably,
some of these compounds discriminate between PRMT1 and
CARM1. Docking studies and an overlay of crystal structures
that form a model that may explain the observed selectivity are
also reported.
Results and discussion
Chemistry
The inhibitor design extends non-reactive AdoMet analogues,
such as AzoAdoMet 3, to display functionality that can occupy
the arginine binding site. In principle inhibitor specificity may be
introduced by accessing non-covalent interactions with residues
unique to the individual PRMTs.29,31 A modular, three-step
synthetic strategy was developed to give expedient access to
AzoAdoMet analogues from readily available aldehydes.
Reductive alkylation of acetonide-protected 5¢-amino-5¢-
deoxyadenosine 832 with aldehyde 5, derived from aspartic acid,33
gave the AzoAdoHcy 9 in good yield (Scheme 2).30 Compounds
bearing tert-butyl butyrate 10, or N-boc propylamine 11, side
chains were made from aldehydes 634 and 735 respectively.
Scheme 3 Reagents and conditions: b, NaBH(OAc)3, ClCH2CH2Cl, rt; c,
TFA/H2O, rt, Amberlyst IRA-400(Cl-) (16–18),30 71% (19), 71% (20).
by further incorporation of amine 11 with carbamate 7; the related
symmetrical guanidine analogue 24 was similarly prepared using
aldehyde 12.36
In all cases, the global protecting group ablation proceeded
smoothly in aqueous trifluoroacetic acid, and the concentrated
products were subjected to ion exchange using Amberlyst IRA-
400(Cl-) and freeze drying to deliver the inhibitors 16–25 as
hygroscopic hydrochloride salts.
Biochemical evaluation
Previously, the natural product sinefungin 4 was confirmed to
be a potent but promiscuous inhibitor that inhibits PRMT1,
CARM1 and SET7 with similar potency (IC50 ~ 1–3 mM),30 thus
providing a positive control for inhibition. We also previously
reported a propargyl analogue of AzoAdoMet (not shown,
X = N, R = CH2C CH, Fig. 1 for general structure) as a
negative control that does not inhibit either methyltransferase,
indicating that additional functionality is required for inhibition.30
PRMT1 was inhibited by analogues 16–18 bearing guanidine
functionality regardless of the length of the alkyl linker, but the
lysine methyltransferase SET7 was not inhibited, with at least 50%
of enzyme activity remaining at 100 mM inhibitor concentrations.
Indeed, none of the compounds reported herein inhibit the lysine
methyltransferase.37 Evaluation of the same inhibitors against
CARM1 provided interesting new results. The analogue bearing
the guanidine via a three carbon linker, 16, inhibited CARM1
to approximately the same degree as PRMT1 (IC50 ~ 13.3
8.7 mM). Remarkably, the compounds that display guanidine via
a four or five carbon linker, 17 and 18 respectively, were very poor
Scheme 2 Reagents and conditions: a, NaBH(OAc)3, ClCH2CH2Cl, rt,
73% (9),30 23% (10), 45% (11).
Elaboration of intermediate secondary amine 9 by further
reductive alkylation with aldehydes 12,36 13,† 14† and global
deprotection gave guanidines 16–18 as previously described
(Scheme 3).30 Equivalent reactions with carbamate 7 gave analogue
19, featuring a primary amine rather than a guanidine group after
deprotection. Aldehyde 15† was subject to the same sequence
to give aryl guanidine 20. Amines 10 and 11 were similarly
reacted with guanidine 12 to give carboxylic acid 21 or amine 22,
respectively (Scheme 4). The symmetrical diamine 23 was obtained
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7814–7821 | 7815
©

View Article Online
Table 1 Inhibition constants for PRMT1, CARM1, and SET7
Compound Number
PRMT1a,b
CARM1a,c
SET7 a,c
Sinefungin 4
0.7 0.3
6.2 3.830
2.9 0.830
5.6 4.530
3.9 1.8
14.5 4.1
>100d
2.2 1.8
1.5 1.5
2.1 2.3
2.4 3.6
13.3 8.7
>100d
[2.5]30
>100d
>100d
16
17
18
19
20
21
22
23
24
e
e
—
—
—
—
—
—
—
—
—
—
e
e
e
e
e
e
e
3.3 3.0
1.1 1.3
4.6 6.3
e
e
—
a Mean IC50 standard deviation (mM) based on >triplicate experiments.
b Substrate = P3, the arginine glycine rich region of Src-associated protein
of 68 kDa (Sam68). c Substrate = Histone H3. d >50% enzyme activity at
100 mM. e No activity.
term goal of achieving greater cell permeability. Analogue 21,
that retained the carboxyl group but not the primary amine,
did not inhibit any methyltransferases. On the other hand, the
complementary analogue 22, that retained the primary amine
but not the carboxylic acid, inhibited PRMT1 and CARM1
with similar IC50 values (PRMT1 2.2 1.8 mM, CARM1 3.3
3.0 mM) to the prototype inhibitors. The latter analogue presents
amine and guanidine functional groups via propyl linkers, both
of which will be positively charged under the assay conditions. In
principle, symmetrical bis-amine 23 and bis-guanidine 24 derived
compounds might be expected to behave in a similar manner and
indeed both compounds inhibit PRMT1 and CARM1 with similar
IC50 values (1.1 1.3 and 2.1 2.3 mM for PRMT1 and 1.5 1.5
and 4.6 6.3 mM for CARM1, respectively, see Table 1).
Scheme 4 Reagents and conditions: b, NaBH(OAc)3, ClCH2CH2Cl, rt; c,
TFA/H2O, rt, then Amberlyst IRA-400(Cl-) 79% (21), 35% (22), 75% (23),
54% (24).
At present, the prototype compounds are somewhat flexible,
which allows them to adopt a range of conformations that might
fit a range of target enzymes in principle, thus, increasing rigidity
might deliver greater specificity in future. As a tentative step
toward exploring more rigid inhibitors, the AzoAdoMet analogue
featuring a benzylic linker 20 was synthesised; subsequent eval-
uation revealed that this compound appeared to inhibit PRMT1
(IC50 5.6 4.7 mM) and not CARM1 in a similar way to the related
analogues featuring four and five carbon flexible linkers.
Docking studies
Docking studies were performed to examine how the inhibitors
(i.e. 16) might be accommodated within a protein arginine methyl-
transferase active site. Unfortunately, PRMT1 crystal structures
(e.g. 1or8.pdb)38 are inappropriate for docking studies,20,23 since
they were obtained at a pH at which the enzyme is not active and
are also disordered at the N-terminus. A recent crystal structure
of CARM1 featuring AdoHcy 2 and a potent indole inhibitor
simultaneously bound at the active site has been described
(2Y1X.pdb).29 Unlike previously reported CARM1 structures, an
additional helix (aW) that interacts with the adenine group of
AdoHcy 2 was observed.
Fig. 1 Result of docking inhibitor 16 (white backbone) with CARM1
binding site (2Y1X.pdb, selected residues green backbone), also AdoHcy
2 (yellow), substrate arginine channel (green mesh).
inhibitors of CARM1, thus displaying remarkable discrimination
between these enzymes. Interestingly, analogue 19, that replaces
the guanidine with an amine, expected to protonate under the
assay conditions, was also a potent inhibitor of PRMT1, but not
CARM1.
Flexible docking of 16 to a rigid CARM1 binding site from this
structure was carried out using AutoDock Vina39 within the PyRX
virtual screening tool.40
We next set out to investigate the effect of structural modifi-
cations to the prototype inhibitor design, specifically with a view
to reduce the overall polarity of the compound, with a longer
The most stable predicted binding mode revealed excellent
overlap between the docked position of prototype inhibitor 16
and that of AdoHcy 2 in the native crystal structure (Fig. 1).
7816 | Org. Biomol. Chem., 2011, 9, 7814–7821
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
˚
For example, interactions between the aminoacid component with
arginine R169 and aspartic acid D191 are maintained. These
models do not provide an obvious explanation for the difference
in inhibition observed for compounds featuring only carboxylate
21 or amine 22 functionality as yet. Recent studies of PRMT1
mutants suggest that the arginine (R54) equivalent to R169 is not
essential for binding of AdoMet, however.41
Most notable is that the remaining guanidine functionality of 16
extends into the channel believed to accommodate the substrate
arginine during catalysis. Furthermore, this guanidine is well
placed to gain optimal interactions with glutamic acid residues
E258 and E267 reported to be essential for binding and catalysis
respectively.38,41 In the docked structure, the distance between the
explained by this model since the glutamic acid (E258) is ~8.6 A
away in CARM1, but the glutamic acid of PRMT1 (Glu47) is
~4.5 A away in the overlay structure.† PRMT3 also features a
˚
carboxylate at this position (Glu229),42 thus it will be interesting
to learn whether these compounds inhibit this enzyme in future.
Analogue 19, with an amine in place of guanidine compared
to compound 16, docked with a best fit binding mode of similar
orientation to the latter compound in the CARM1 binding site.†
It is less obvious from this model of binding as to how selective
inhibition of PRMT1 but not CARM1 occurs for this compound.
As mentioned above, significant changes to the structure of
CARM1 occur upon binding AdoHcy 2.43 Indeed this reor-
ganisation is required for some inhibitors to achieve potency.29
Presumably, compounds based on an extended AdoHcy template,
especially 16 and 22–24, are able induce fit to the CARM1 binding
site. Molecular dynamics calculations that account for the motion
of individual residues and domains are desirable, but will also
require much more processor time. Future crystallographic and
NMR binding studies will be valuable for revealing information
about the structure and dynamics of non-covalent interactions
between protein and these inhibitors.
˚
guanidinium and these residues is ~3.2 A.† This analysis supports
the original design principle and will be a valuable tool for planned
virtual screening.
The most stable binding orientation predicted for the four and
five methylene linked compounds 17 (not shown) and 18 (Fig. 2),
are very similar to that for 16, but the longer linkers push the
guanidinium functional group beyond the glutamic acid residue
˚
˚
(E258), to ~5.9 A and ~6.9 A respectively,† which may explain the
reduced CARM1 inhibition observed for these compounds (Fig.
2). Overlay of the structure for PRMT1 with the docked structure
of 18 with CARM1 reveals different residues near the substrate
binding channel. While this does not represent a docked structure
with PRMT1, a hypothesis can be formed assuming that the
binding sites are similar. Specifically, PRMT1 features a glutamic
acid residue (Glu47) close to the putative substrate channel,
whereas CARM1 features an asparagine residue (Asn162). Thus,
electrostatic interactions between the guanidine of 17 and 18 and
Conclusions
These results demonstrate a common scaffold that can be readily
adapted to inhibit individual PRMTs. Small molecules designed
to occupy PRMT binding sites for both arginine substrate and
AdoMet co-factor have been demonstrated as inhibitors with
potency comparable to sinefungin. Remarkably, selective discrimi-
nation between individual PRMTs was observed, despite relatively
simple variations between the prototype inhibitor structures.
Confirmation of the bisubstrate mode of inhibition will require
more detailed binding and kinetic studies which are the subject
of ongoing studies. In the meantime, in silico docking studies
identify substantial overlap with the AdoMet for 16 with the
linker-supported guanidinium functional group extending into the
arginine substrate binding site as designed. The overlay of PRMT1
(1or8.pdb)38 and CARM1 (2Y1X.pdb)29 crystal structures high-
lighted differences in a key residue corresponding to Glu47 within
the PRMT1 binding site, Asn162 within CARM1, that may be
reached by the longer linkers 17 and 18, but not the three carbon
linkers 16 or 19. The modular synthesis demonstrated herein is well
suited to the efficient production of further analogues. Studies
to achieve selective inhibition by exploiting further sequence
variations between various PRMT binding sites are under way.
˚
˚
the carboxylate of Glu47, within approximately 4.4 A and 3.2 A
in the overlay structure,† may facilitate inhibition of PRMT1 by
17 and 18.
Experimental
Chemistry: general
All chemicals were purchased from commercial sources and used
as supplied unless stated. THF was freshly distilled over sodium
benzophenone ketyl, Et3N and CH2Cl2 were distilled from CaH2.
Thin layer chromatography was performed using silica gel 60 pre-
coated aluminium plates (0.20 mm thickness) from Macherey-
Nagel, with visualisation by UV light (254 nm) or exposure to
potassium permanganate solution. Column chromatography was
performed on silica gel (particle size 40–63 mm) from Fischer
Chemicals. Melting points were determined using a Stuart SMP3
Fig. 2 Result of docking inhibitor 18 (magenta backbone) with CARM1
binding site (2Y1X.pdb, selected residues green backbone), also AdoHcy
2 (yellow), substrate arginine channel (green mesh). PRMT1 residue
superimposed, Glu47 (orange, bold in parentheses). Note: different angle
with respect to Fig. 1 to emphasise proximity.
The selective inhibition of PRMT1, but not CARM1 by
the analogue displaying the benzylic guanidine 20 may also be
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7814–7821 | 7817
©

View Article Online
melting point apparatus and are uncorrected. Optical rotations
were obtained at ambient temperature using a Jasco DIP-370
digital polarimeter. IR spectra were recorded in the solid phase
using a Thermo AVATAR 320 FT-IR or from CHCl3 using a
Bruker Tensor 27 spectrometer. 1H and 13C NMR spectra (d
[ppm], J [Hz]) were recorded on Bruker AV400, DPX400 and
AV500 spectrometers. Broadening of NMR signals was observed,
presumably due to slow conformational exchange. This effect
was most obvious in 13C spectra and could be recovered by
recording spectra at 70 ◦C on a JEOL EX270 spectrometer.†
Mass spectra were recorded using a service LC-TOF (Bruker
micrOTOF), running in an open-access mode. LC-MS data† were
obtained using a Varian “Prostar” System comprising: Varian
410 autosampler, 2 ¥ Varian 210 pump (binary system), Varian
Polaris 5 mm C18-A 250 mm ¥ 4.6 mm reverse phase column,
Varian 325 UV detector monitoring at 254 nm and a Varian 310-
MS TQ mass spectrometer operating in ESI mode. All runs were
performed with a flow rate of 1 ml min-1 and eluted with the
following linear gradient solvent system: t = 0 mins, 10% CH3CN:
90% H2O; t = 2 mins, 10% CH3CN: 90% H2O; t = 20 min, 100%
CH3CN: 0% H2O; t = 28 min, 100% CH3CN: 0% H2O; t = 30 min,
10% CH3CN: 90% H2O.
5¢-deoxy-5¢-[[(4-(1,1-dimethylethoxy)-4-oxobutyl]amino]-2¢,3¢-O-
(1-methylethylidene)-adenosine (10)
Method 1. White solid (0.59 g, 23%), mp 49–51 ◦C (CH2Cl2); [a]D19
-32.5 (c 1.05 in CHCl3); nmax/cm-1 3011 (m), 2984 (m), 2938 (m),
1720 (s), 1633 (s), 1589 (m), 1475 (m), 1457 (m), 1424 (m), 1370
(s), 1330 (m); 1H NMR dH (400 MHz; CDCl3) 8.35 (1H, s, ArH),
7.93 (1H, s, ArH), 6.00 (1H, d, J = 2.8, 1¢-H), 5.95 (2H, br s,
adenosine-NH2), 5.49-5.44 (1H, m, 2¢-H), 5.08-5.03 (1H, m, 3¢-
H), 4.42–4.36 (1H, m, 4¢-H), 3.00 (1H, br s, NH), 2.87–3.00 (2H,
m, 5¢-CH2), 2.72–2.56 (2H, m, CH2NH), 2.27 (2H, t, J = 7.3,
COCH2), 1.82–1.72 (2H, m, CH2CH2NH), 1.63 (3H, s, CH3),
1.44 (9H, s, C(CH3)3), 1.40 (3H, s, CH3); 13C NMR dC (100 MHz;
CDCl3) 172.8 (C), 155.6 (C), 153.0 (CH), 149.4 (C), 139.9 (CH),
120.4 (C), 114.6 (C), 91.0 (CH), 85.3 (CH), 83.4 (CH), 82.2 (CH),
80.2 (C), 51.1 (CH2), 49.0 (CH2), 33.2 (CH2), 28.1 (CH3), 27.3
(CH3), 25.4 (CH3), 25.1 (CH2); m/z (ESI) 449.3 (M + H+, 100%)
Found 449.2512 C21H33N6O5 requires 449.2507.
5¢-deoxy-5¢[(3-tert-butoxycarbonylaminopropyl)amino]-2¢,3¢-O-(1-
methylethylidene)-adenosine (11)
Method 1. White solid (0.40 g, 45%), mp 55–57 ◦C (CH2Cl2); [a]D19
-10.5 (c 0.35 in CHCl3); nmax/cm-1 3414 (w), 3305 (w), 2982 (m),
2931 (m), 2855 (w), 1707 (s), 1632 (s), 1588 (w), 1500 (m), 1473
(m), 1423 (w), 1369 (s); 1H NMR dH (400 MHz; CDCl3) 8.30 (1H,
s, ArH), 7.95 (1H, s, ArH), 6.29 (2H, br s, adenosine-NH2), 6.00
(1H, d, J = 3.1, 1¢-H), 5.46 (1H, dd, J = 6.3, 3.1 2¢-H), 5.33 (1H, br
s, NH), 5.03 (1H, dd, J = 6.3, 3.3, 3¢-H), 4.40-4.35 (1H, m, 4¢-H),
3.25–3.08 (1H, br m, CONHCH2), 2.96–2.84 (2H, m, 5¢-CH2),
2.78 (1H, s, NH), 2.74–2.66 (1H, m, CH2CHaHbNH), 2.66–2.57
(1H, m, CH2CHaHbNH), 1.68–1.61 (2H, m, CH2CH2NH), 1.62
(3H, s, CH3), 1.40 (9H, s, C(CH3)3), 1.39 (3H, s, CH3); 13C NMR
dC (100 MHz; CDCl3) 156.2 (C), 155.8 (C), 153.1 (CH), 149.3
(C), 139.9 (CH), 120.1 (C), 114.7 (C), 90.9 (CH), 85.1 (CH), 83.4
(CH), 82.2 (CH), 79.1 (C), 51.2 (CH2), 47.6 (CH2), 39.0 (CH2),
29.5 (CH2), 28.4 (CH3), 27.3 (CH3), 25.4 (CH3); m/z (ESI) 464.3
(M + H+, 100%) Found 464.2611 C21H34N7O5 requires 464.2616;
Calc. for C21H33N7O5.
General procedure for first reductive amination: method 1
5¢-Amino-5¢-deoxy-2¢,3¢-O,O-(1-methylethylidene)adenosine 832
(1.2 mol eq.) and aldehyde 5,33 634 or 735 (1.0 mol eq.) were
suspended in ClCH2CH2Cl (5 mL) at room temperature under Ar
and the mixture gently heated with vigorous stirring to obtain a
solution. NaBH(OAc)3 (1.4 mol eq.) was slowly added portionwise
and the reaction left for 2–4 h at room temperature under Ar.
The reaction was then quenched by the addition of saturated
aqueous Na2CO3 solution (5 mL). The mixture was extracted
with CH2Cl2 (3 ¥ 50 mL) and the combined organic layers were
dried over MgSO4 and concentrated in vacuo to yield the crude
product. The product was purified by column chromatography
on silica gel eluting with MeOH : CH2Cl2, 5 : 95 unless specified
otherwise.
5¢-[(3(S)-3-amino-3-carboxypropyl)][(3-aminopropyl)amino]-5¢-
deoxy-adenosine (19)
General procedure for second reductive amination: method 2
Method 2, th◦en 3. Pale yellow solid (0.19 g, 71% two steps),
Secondary amines 9,30 10 or 11 (1.0 mol eq.) and aldehydes 12,36
13–15† (1.1 mol eq.) with MgSO4 (10 mol eq.) were used, then as
method 1 above.
mp 137–138 C (dec.) (H2O); [a]2D6 +24.2 (c 1.00 in H2O);
n
max/cm-1 2970 (m), 2920 (m), 1686 (s), 1611 (w), 1505 (m),
1412 (m), 1322 (w), 1229 (m), 1128 (m), 1049 (s); 1H NMR
dH (400 MHz; D2O) 8.40 (2H, s, ArH), 6.10 (1H, d, J =
3.7, 1¢-H), 4.57–4.88 (1H, m, 2¢-H), 4.41–4.49 (2H, m, 3¢-
H and 4¢-H), 3.88–3.96 (1H, m, CHCH2), 3.63–3.80 (2H, m,
5¢-CH2), 3.41–3.58 (2H, m, CHCH2CH2), 3.29–3.39 (2H, m,
H2NCH2CH2CH2N), 2.95–3.05 (2H, m, H2NCH2CH2CH2N),
2.26–2.39 (1H, m, CHCHaHb), 2.12–2.23 (1H, m, CHCHaHb),
2.01–2.12 (1H, m, H2NCH2CH2CH2N); 13C NMR dC (100 MHz;
D2O) 171.6 (C), 149.9 (C), 148.1 (C), 144.4 (CH), 143.5 (CH), 119.3
(C), 90.0 (CH), 78.0 (CH), 73.0 (CH), 71.5 (CH), 55.3 (CH2), 51.5
(CH and CH2), 50.4 (CH2), 36.4 (CH2), 24.5 (CH2), 21.5 (CH2); 13C
NMR dC (68 MHz; D2O, 70 ◦C) 172.4, 150.9, 148.9, 145.4, 144.1,
120.0, 90.8, 78.9, 73.8, 72.3, 56.3, 52.9, 51.6, 37.3, 25.4, 22.4; m/z
General procedure for deprotection: method 3
Intermediates† from method 2 above were dissolved in TFA (4 mL)
and water (0.10 mL) and stirred overnight. The mixture was
concentrated in vacuo, dissolved in water (5 mL) and washed with
EtOAc (2 ¥ 5 mL). The aqueous layer was concentrated in vacuo
to ~1 mL then applied to a column of Amberlite IRA-400(Cl-)
ion exchange resin, eluted with water, and fractions containing
product were freeze dried to give the hydrochloride salts 16–18,30
19–24.
7818 | Org. Biomol. Chem., 2011, 9, 7814–7821
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
(ESI) 425.2 (M + H+, 100%), Found 425.2256 C17H29N8O5 requires
425.2255.
36.4 (CH2), 33.3 (CH2), 22.6 (CH2), 21.7 (CH2); m/z (ESI) 423.3
(M + H+, 100%) Found 423.2576 C17H31N10O3 requires 423.2575.
5¢-[di(3-aminopropyl)amino]-5¢-deoxy-adenosine (23)
5¢-[[(3(S)-3-amino-3-carboxypropyl][(4-guanidinophenyl)-
methyl]amino]-5¢-deoxy-adenosine (20)
Method 2, then 3. Pale yellow solid (0.28 g, 75% two steps), mp
◦
110–111 C (dec.) (H2O); [a]2D3 +16.5 (c 1.00 in H2O); nmax/cm-1
Method ◦2, then 3. Pale yellow solid (0.24 g, 71% two steps), mp
138-139 C (dec.) (H2O); [a]2D8 +34.9 (c 1.00 in H2O); nmax/cm-1
2973 (m), 2902 (m), 1682 (s), 1605 (w), 1577 (w), 1507 (w), 1408
2987 (m), 2971 (m), 2901 (m), 1684 (w), 1406 (m), 1394 (m),
1231 (w), 1077 (s), 1066 (s), 1047 (s), 949 (w); 1H NMR dH
(400 MHz; D2O) 8.42 (1H, s, ArH), 8.41 (1H, s, ArH), 6.14
(1H, d, J = 3.5, 1¢-H), 4.71–4.90 (1H, m, 2¢-H), 4.50 (1H, dd,
J = 6.8, 5.4, 3¢-H), 4.45 (1H, ddd, J = 9.2, 6.8, 2.4, 4¢-H),
3.76 (1H, dd, J = 14.2, 9.2, 5¢-CHaHb), 3.70 (1H, dd, J = 14.2,
2.4, 5¢-CHaHb), 3.37 (4H, t, J = 8.4, (H2NCH2CH2CH2)2N),
3.02 (4H, t, J = 7.7, (H2NCH2CH2CH2)2N), 2.00–2.18 (4H, m,
(H2NCH2CH2CH2)2N); 13C NMR dC (100 MHz; D2O) 150.1 (C),
148.1 (C), 144.6 (CH), 143.2 (CH), 119.3 (C), 90.1 (CH), 77.5
(CH), 73.1 (CH), 71.4 (CH), 55.1 (CH2), 50.7 (CH2), 36.4 (CH2),
21.5 (CH2); m/z (ESI) 381.2 (M + H+, 100%) Found 381.2362
C16H29N8O3 requires 381.2357.
1
(m), 1223 (m), 1055 (s), 925 (w), 853 (w), 828 (m), 744 (s); H
NMR dH (400 MHz; D2O) 8.35 (1H, s, ArH), 8.32 (1H, s, ArH),
7.47 (2H, d, J = 8.4, ArH), 7.24 (2H, d, J = 8.4, ArH), 6.10 (1H,
d, J = 3.2, 1¢-H), 4.57–4.82 (1H, m, 2¢-H), 4.37–4.52 (4H, m, 3¢-H,
4¢-H and ArCH2N), 3.90 (1H, dd, J = 8.8, 4.4, CHCH2), 3.61–
3.71 (2H, m, 5¢-CH2), 3.45–3.61 (2H, m, CHCH2CH2), 2.31–2.44
(1H, m, CHCHaHb), 2.16–2.30 (1H, m, CHCHaHb); 13C NMR dC
(100 MHz; D2O) 171.5 (C), 155.9 (C), 149.9 (C), 147.9 (C), 144.3
(CH), 143.4 (CH), 136.2 (C), 132.7 (CH), 127.6 (C), 125.4 (CH),
119.1 (C), 90.2 (CH), 78.0 (CH), 73.2 (CH), 71.6 (CH), 57.6 (CH2),
54.7 (CH2), 51.4 (CH), 51.3 (CH2), 24.6 (CH2); m/z (ESI) 559.2
(100%, [M + HCO2]+), 537.2 (38%, M + Na+), 515.2 (31%, M +
H+) Found 515.2462 C22H31N10O5 requires 515.2473.
5¢-deoxy-5¢-[di(3-guanidinopropyl)amino]adenosine (24)
Method 1, 2, then 3. Pale yellow solid (0.11 g, 54%), mp 80–82 ◦C
(H2O); [a]2D3 +11.9 (c 1.00 in H2O); nmax/cm-1 3671 (w), 2987 (s),
2970 (s), 2901 (s), 1665 (m), 1406 (m), 1394 (m), 1381 (m), 1249
5¢-[(3-carboxypropyl)][(3-guanidinopropyl)amino]-5¢-deoxy-adeno-
sine (21)
1
(m), 1230 (m), 1074 (s), 1066 (s), 1052 (s), 1028 (s), 898 (w); H
Method 2, then 3. Pale yellow solid (0.24 g, 79% two steps), mp
115–6 ◦C (dec.) (H2O); [a]D28 +13.7 (c 1.00 in H2O); nmax/cm-1 2986
(m), 2971 (m), 2935 (w), 1685 (s), 1626 (m), 1509 (w), 1409 (m),
1396 (m), 1321 (w), 1224 (m), 1126 (m), 1048 (s); 1H NMR dH (400
MHz, D2O) 8.41 (1H, s, ArH), 8.40 (1H, s, ArH), 6.11 (1H, d, J =
4.1, 1¢-H), 4.56–4.87 (1H, m, 2¢-H), 4.43–4.50 (2H, m, 3¢-H and 4¢-
H), 3.71–3.81 (1H, m, 5¢-CHaHb), 3.59–3.68 (1H, m, 5¢-CHaHb),
3.13–3.33 (6H, m, COCH2CH2CH2 and NHCH2CH2CH2), 2.31–
2.46 (2H, m, COCH2), 1.81–2.02 (4H, m, COCH2CH2CH2 and
NHCH2CH2CH2); 13C NMR dC (100 MHz, D2O) 176.5 (C), 156.7
(C), 150.1 (C), 148.0 (C), 144.7 (CH), 143.5 (CH), 119.3 (C),
90.1 (CH), 78.2 (CH), 72.9 (CH), 71.7 (CH), 54.7 (CH2), 50.8
(CH2 ¥ 2), 38.0 (CH2), 30.2 (CH2), 22.6 (CH2), 18.4 (CH2); m/z
(ESI) 452.2 (M + H+, 100%) Found 452.2341 C18H30N9O5 requires
452.2364.
NMR dH (400 MHz; D2O) 8.44 (1H, s, ArH), 8.43 (1H, s, ArH),
6.14 (1H, d, J = 3.7, 1¢-H), 4.59–4.75 (1H, m, 2¢-H), 4.41–4.55 (1H,
m, 3¢-H and 4¢-H), 3.81 (1H, dd, J = 14.2, 10.0, 5¢-CHaHb), 3.61–
3.74 (1H, m, 5¢-CHaHb), 3.12–3.37 (8H, m, NH2CH2CH2CH2N ¥
2), 1.84–2.05 (4H, m, NH2CH2CH2CH2N ¥ 2); 13C NMR dC (100
MHz; D2O) 156.7 (C), 150.0 (C), 148.0 (C), 144.5 (CH), 143.5
(CH), 119.3 (C), 90.3 (CH), 78.2 (CH), 73.1 (CH), 71.6 (CH), 54.7
(CH2), 51.2 (CH2), 38.1 (CH2), 22.9 (CH2); m/z (ESI) 465.3 (M +
H+, 100%) Found 465.2793 C18H33N12O3 requires 465.2793.
Enzymatic studies
Preparation of GST-linked fusion proteins. The coding se-
quences of required proteins were cloned into a pGEX-5X-1
expression vector (Amersham Pharmacia Biotech). GST fused
protein was expressed in DH5a cells grown in LB-medium. 200 mL
cultures with an A600 of 0.6 were induced with 0.5 mM IPTG and
grown for 3 h at 30 ◦C. After centrifugation of cells at 4000g, the
pellet was resuspended in 5 mL of GST-binding buffer (PBS, pH
7.3) containing one protease inhibitor tablet (Complete, Roche,
Mannheim, Germany) for 50 mL of buffer. Cells were lysed using
a high-pressure cell disruptor (Constant Cell Disruption Systems).
The resulting lysate was centrifuged at 20000g for 20 min at 4 ◦C.
GST fusion protein was purified from soluble extracts by binding
to glutathione agarose beads for 2 h at 4 ◦C.
5¢-deoxy-5¢-[(3-guanidinopropyl)][(3-aminopropyl)amino]adenosine
(22)
Method 2, then 3. Pale yellow solid (0.15 g, 35% two steps), mp
◦
115–116 C (dec.) (H2O); [a]2D6 +12.2 (c 1.10 in H2O); nmax/cm-1
(neat) 3149 (w), 3054 (w), 2989 (w), 1680 (s), 1624 (m), 1507 (w),
1419 (w), 1322 (w), 1226 (w), 1127 (m), 1053 (m); 1H NMR dH (400
MHz; D2O) 8.40 (1H, s, ArH), 8.39 (1H, s, ArH), 6.11 (1H, d, J =
3.5, 1¢-H), 4.75–4.84 (1H, m, 2¢-H), 4.39–4.51 (2H, m, 3¢-H and 4¢-
H), 3.73–3.82 (1H, m, 5¢-CHaHb), 3.62–3.70 (1H, m, 5¢-CHaHb),
3.26–3.37 (4H, m, H2NCH2CH2CH2N and H2NCH2CH2CH2N),
Preparation of enzyme substrates. P3 is a protein sequence
encoding the arginine glycine rich region of Src-associated protein
of 68 kDa (Sam68). For PRMT-1 methylation assays, GST-P3
substrate protein was released from the glutathione agarose beads
by three incubations with 200 mL of 20 mM glutathione in PBS
(pH7.4). After each hour rotating at 4 ◦C beads were sedimented
by centrifugation in a microfuge at 1500 rpm for 5 min and
3.08–3.21
(2H,
m,
C(N)NHCH2CH2CH2N),
2.96–
3.04 (2H, m, H2NCH2CH2CH2N), 1.98–2.14 (2H, m,
C(N)NHCH2CH2CH2N), 1.81–1.98 (2H, m, C(N)NHCH2-
CH2CH2N); 13C NMR dC (100 MHz; D2O) 156.7 (C), 150.1 (C),
148.0 (C), 144.7 (CH), 143.3 (CH), 119.3 (C), 90.2 (CH), 79.4
(CH), 73.1 (CH), 71.5 (CH), 54.8 (CH2), 51.1 (CH2), 38.0 (CH2),
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7814–7821 | 7819
©

View Article Online
supernatant removed and aliquots retained. CARM1 and SET-
7 enzyme assays used 1 mg mL-1 Histone H3 (Sigma Aldrich).
5 A. O. Yildirim, P. Bulau, D. Zakrzewicz, K. E. Kitowska, N. Weiss-
mann, F. Grimminger, R. E. Morty and O. Eickelberg, Am. J. Respir.
Cell Mol. Biol., 2006, 35, 436–443.
6 X. Chen, F. Niroomand, Z. F Liu, A. Zankl, H. A. Katus, L. Jahn and
C. P. Tiefenbacher, Basic Res. Cardiol., 2006, 101, 346–353.
7 M. C. Boulanger, C. Liang, R. S. Russell, R. T. Lin, M. T. Bedford, M.
A. Wainberg and S. Richard, J. Virol., 2005, 79, 124–131.
8 F. Blanchet, A. Cardona, F. A. Letimier, M. S. Hershfield and O. Acuto,
J. Exp. Med., 2005, 202, 371–377.
9 M. Yoshimatsu, G. Toyokawa, S. Hayami, M. Unoki, T. Tsunoda, H. I.
Field, J. D. Kelly, D. E. Neal, Y. Maehara, B. A. J. Ponder, Y. Nakamura
and R. Hamamoto, Int. J. Cancer, 2011, 128, 562–573.
Inhibition assays. In vitro methylation reactions were per-
formed according to published procedures.19 Briefly, methylation
reactions were carried out in the presence of [3H]AdoMet (79 Ci
mmol-1 from a 12.6 mM stock solution in dilute HCl/ethanol
9 : 1, pH 2.0–2.5, Amersham Biosciences) and reaction buffer
(200 mM NaCl, 20 mM Tris pH 8.0). Various concentrations of
each compound were incubated with either GST-CARM1 or GST-
SET-7 with histone H3 substrate, or GST-PRMT1 with GST-
P3 substrate. Substrates (5 mg) were incubated with recombinant
enzymes (5 mg) in the presence of 0.5 mM [3H]AdoMet for 90 min
at 30 ◦C in a final volume of 50 mL. Reactions were terminated by
addition of 20 mL SDS-PAGE sample buffer and boiling for 5 min,
and reaction product was resolved by 10% SDS-PAGE. Proteins
were transferred to nitrocellulose membrane, and visualised with
Ponceau S stain. Substrate proteins were excised and incorporated
tritium quantified by scintillation counting (Perkin Elmer Tri-carb
2800TR). Results were plotted as % enzyme activity of control
samples, and IC50 values were determined from at least three
separate assays and are presented as mean values S.E.M.
10 D. B. Seligson, S. Horvath, T. Shi, H. Yu, S. Tze, M. Grunstein and S.
K. Kurdistani, Nature, 2005, 435, 1262–1266.
11 (a) S. Majumder, Y. Liu, O. H. Ford, J. L. Mohler and Y. E. Whang,
Prostate, 2006, 66, 1292–1301; (b) H. Hong, C. H. Kao, M. H. Jeng, J.
N. Eble, M. O. Koch, T. A. Gardner, S. B. Zhang, L. Li, C. X. Pan, Z.
Q. Hu, G. T. MacLennan and L. Cheng, Cancer, 2004, 101, 83–89.
12 S. El Messaoudi, E. Fabbrizio, C. Rodriguez, P. Chuchana, L. Fauquier,
D. H. Cheng, C. Theillet, L. Vandel, M. T. Bedford and C. Sardet, Proc.
Natl. Acad. Sci. U. S. A., 2006, 103, 13351–56.
13 E. M. Bissinger, R. Heinke, W. Sippl and M. Jung, Med. Chem.
Commun., 2010, 1, 114–124.
14 T. C. Osborne, R. L. Roska, S. R. Rajski and P. R. Thompson, J. Am.
Chem. Soc., 2008, 130, 4574–4575.
15 K. L. Bicker, O. Obianyo, H. L. Rust and P. R. Thompson, Mol.
BioSyst., 2011, 7, 48–51.
16 O. Obianyo, C. P. Causey, T. C. Osborne, J. E. Jones, Y. H. Lee, M. R.
Stallcup and P. R. Thompson, ChemBioChem, 2010, 11, 1219–1223.
17 T. M. Lakowski, P. ’t Hart, C. A. Ahern, N. I. Martin and A. Frankel,
ACS Chem. Biol., 2010, 5, 1053–1063.
Docking studies
18 P. ’t Hart, T. M. Lakowski, D. Thomas, A. Frankel and N. I. Martin,
ChemBioChem, 2011, 12, 1427–1432.
19 D. Cheng, N. Yadav, R. W. King, M. S. Swanson, E. J. Weinstein and
M. T. Bedford, J. Biol. Chem., 2004, 279, 23892–23899.
20 Y. Feng, M. Y. Li, B. H. Wang and Y. G. Zheng, J. Med. Chem., 2010,
53, 6028–6039.
21 K. Bonham, S. Hemmers, Y. H. Lim, D. M. Hill, M. G. Finn and K.
A. Mowen, FEBS J., 2010, 277, 2096–2108.
22 S. Castellano, C. Milite, R. Ragno, S. Simeoni, A. Mai, V. Limongelli, E.
Novellino, I. Bauer, G. Brosch, A. Spannhoff, D. Cheng, M. T. Bedford
and G. Sbardella, ChemMedChem, 2010, 5, 398–414.
23 R. Ragno, S. Simeoni, S. Castellano, C. Vicidomini, A. Mai, A. Caroli,
A. Tramontano, C. Bonaccini, P. Trojer, I. Bauer, G. Brosch and G.
Sbardella, J. Med. Chem., 2007, 50, 1241–1253.
24 A. V. Purandare, Z. Chen, T. Huynh, S. Pang, J. Geng, W. Vaccaro,
M. A. Poss, J. O’Connell, K. Nowak and L. Jayaraman, Bioorg. Med.
Chem. Lett., 2008, 18, 4438–4441.
25 M. Allan, S. Manku, E. Therrien, N. Nguyen, S. Styhler, M. F. Robert,
A. C. Goulet, A. J. Petschner, G. Rahil, A. Robert Macleod, R. De´ziel,
J. M. Besterman, H. Nguyen and A. Wahhab, Bioorg. Med. Chem.
Lett., 2009, 19, 1218–1223.
AutoDock Vina (version 1.1.2)39 was used within the PyRX virtual
screening tool (version 0.8)40 using default settings. Docking was
performed at the binding site of CARM1 chain B (2Y1X.pdb)29
with ligands removed. The receptor was treated as rigid within a
3
˚
25 A box centred upon the location of the sulfur atom of SAH
in the original structure. Default settings were used, which treated
all bonds as rotatable except for the furan conformation which
was fixed. Ligand files (*.pdbqt) were prepared for docking within
PyRX from minimised 3D ligand co-ordinates (*.mol2) generated
in OpenBabel (version 2.3.0)44,45 from 2D ligands co-ordinates
(*.mol) with chirality determined from parity flags. Graphical
outputs, including overlay, were generated using PyMol (version
0.99);46 pathways from the binding site were calculated using the
CAVER plugin (version 2.1)47 starting from the same co-ordinates
as the sulfur atom of AdoHcy 2.
26 T. Huynh, Z. Chen, S. Pang, J. Geng, T. Bandiera, S. Bindi, P. Vianello,
F. Roletto, S. Thieffine, A. Galvani, W. Vaccaro, M. A. Poss, G.
L. Trainor, M. V. Lorenzi, M. Gottardis, L. Jayaraman and A. V.
Purandare, Bioorg. Med. Chem. Lett., 2009, 19, 2924–2927.
27 H. Wan, T. Huynh, S. Pang, J. Geng, W. Vaccaro, M. A. Poss, G.
L. Trainor, M. V. Lorenzi, M. Gottardis, L. Jayaraman and A. V.
Purandare, Bioorg. Med. Chem. Lett., 2009, 19, 5063–5066.
28 E. Therrien, G. Larouche, S. Manku, M. Allan, N. Nguyen, S. Styhler,
M. F. Robert, A. C. Goulet, J. M. Besterman, H. Nguyen and A.
Wahhab, Bioorg. Med. Chem. Lett., 2009, 19, 6725–6732.
29 J. S. Sack, S. Thieffine, T. Bandiera, M. Fasolini, G. J. Duke, L.
Jayaraman, K. F. Kish, H. E. Klei, A. V. Purandare, P. Rosettani,
S. Troiani, D. Xie and J. A. Bertrand, Biochem. J., 2011, 436, 331–339.
30 J. Dowden, W. Hong, R. V. Parry, R. A. Pike and S. G. Ward, Bioorg.
Med. Chem. Lett., 2010, 20, 2103–2105.
Acknowledgements
Financial support from the Medical Research Council Grant
ref: G0700840, the University of Nottingham (Studentship to
W.H.) and The Higher Committee for Education Development
Iraq (Studentship to UAM) is gratefully acknowledged. We thank
Oreste Acuto (University of Oxford, UK) for the gift of GST-P3
plasmid DNA and helpful discussions. We thank Mark Bedford
(University of Texas, MD Anderson Cancer Center, USA) for
the gift of GST-PRMT1, GST-CARM1 and GST-SET-7 plasmid
DNA.
31 V. Campagna-Slater, M. W. Mok, K. T. Nguyen, M. Feher, R.
Najmanovich and M. Schapira, J. Chem. Inf. Model., 2011, 51, 612–
623.
Notes and references
32 A. M. Reeve and C. A. Townsend, Tetrahedron, 1998, 54, 15959–15974.
33 R. M. Werner, O. Shokek and J. T. Davis, J. Org. Chem., 1997, 62,
8243–8246.
1 R. A. Copeland, M. E. Solomon and V. M. Richon, Nat. Rev. Drug
Discovery, 2009, 8, 724–732.
2 P. A. Cole, Nat. Chem. Biol., 2008, 4, 590–597.
34 M. Kolb, J. Barth, J.-G. Heydt and M. J. Jung, J. Med. Chem., 1987,
3 B. C. Smith and J. M. Denu, Biochim. Biophys. Acta., 2009, 1789, 45–57.
4 M. T. Bedford, J. Cell Sci., 2007, 120, 4243–4246.
30, 267–272.
7820 | Org. Biomol. Chem., 2011, 9, 7814–7821
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
35 S. Hourcade, A. Ferdenzi, P. Retailleau, S. Mons and C. Marazano,
Eur. J. Org. Chem., 2005, 1302–1310.
36 N. S. Freeman,Y. Tal-Gan, S. Klein, A. Levitzki and C. Gilon, J. Org.
Chem., 2011, 76, 3078–3085.
37 S. Mori, K. Iwase, N. Iwanami, Y. Tanaka, H. Kagechika and T. Hirano,
Bioorg. Med. Chem., 2010, 18, 8158–8166.
38 X. Zhang and X. D. Cheng, Structure, 2003, 11, 509–520.
39 O. Trott and A. J. Olson, J. Comput. Chem., 2010, 31, 455–
461.
42 X. Zhang, L. Zhou and X. D. Cheng, EMBO J., 2000, 19, 3509–3519.
43 N. Troffer-Charlier, V. Cura, P Hassenboehler, D. Moras and J.
Cavarelli, EMBO J., 2007, 26, 4391–4401.
44 R. Guha, M. T. Howard, G. R. Hutchison, P. Murray-Rust, H. Rzepa,
C. Steinbeck, J. K. Wegner and E. Willighagen, J. Chem. Inf. Model.,
2006, 46, 991–998.
45 The
Open
Babel
Package,
version
2.3.0
http://
openbabel.sourceforge.net/ (accessed Jan 2011).
46 W. L. DeLano The PyMOL Molecular Graphics System. (2008)
DeLano Scientific LLC, Palo Alto, CA, USA.
40 L. K. Wolf, Chem. Eng. News, 2009, 87, 31.
41 H. L. Rust, C. I. Zurita-Lopez, S. Clarke and P. R. Thompson,
Biochemistry, 2011, 50, 3332–3345.
47 M. Petrek, M. Otyepka, P. Bana´s, P. Kosinova´, J. Koca and J.
Damborsky´, BMC Bioinformatics, 2006, 7, 316.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7814–7821 | 7821
©