Dynamic combinatorial mass spectrometry leads to inhibitors of a 2-oxoglutarate-dependent nucleic acid demethylase

Article abstract of DOI:10.1021/jm201417e

2-Oxoglutarate-dependent nucleic acid demethylases are of biological interest because of their roles in nucleic acid repair and modification. Although some of these enzymes are linked to physiology, their regulatory roles are unclear. Hence, there is a de

Full text of DOI:10.1021/jm201417e

Article
pubs.acs.org/jmc
Dynamic Combinatorial Mass Spectrometry Leads to Inhibitors of a
2-Oxoglutarate-Dependent Nucleic Acid Demethylase
Esther C. Y. Woon,†,∥ Marina Demetriades,†,⊥ Eleanor A. L. Bagg,†,⊥ WeiShen Aik,†,⊥
Svetlana M. Krylova,‡,⊥ Jerome H. Y. Ma,† MunChiang Chan,† Louise J. Walport,† David W. Wegman,‡
Kevin N. Dack,§ Michael A. McDonough,† Sergey N. Krylov,‡ and Christopher J. Schofield*,†
†Chemistry Research Laboratory, Department of Chemistry, University of Oxford, 12 Mansfield Road, Oxford, OX1 3TA, United
Kingdom
‡Department of Chemistry and Centre for Research on Biomolecular Interactions, York University, Toronto, Ontario, M3J 1P3,
Canada
§Department of World-Wide Medicinal Chemistry, Pfizer PharmaTherapeutics Division, Sandwich, CT 13 9NJ, United Kingdom
S
* Supporting Information
ABSTRACT: 2-Oxoglutarate-dependent nucleic acid demethy-
lases are of biological interest because of their roles in nucleic
acid repair and modification. Although some of these enzymes
are linked to physiology, their regulatory roles are unclear.
Hence, there is a desire to develop selective inhibitors for them;
we report studies on AlkB, which reveal it as being amenable to
selective inhibition by small molecules. Dynamic combinatorial
chemistry linked to mass spectrometric analyses (DCMS) led to
the identification of lead compounds, one of which was analyzed
by crystallography. Subsequent structure-guided studies led to
the identification of inhibitors of improved potency, some of which were shown to be selective over two other 2OG oxygenases.
The work further validates the use of the DCMS method and will help to enable the development of inhibitors of nucleic acid
modifying 2OG oxygenases both for use as functional probes and, in the longer term, for potential therapeutic use.
ethano adducts, such as 1,N6-ethenoadenine, 3,N4-ethenocyto-
INTRODUCTION
Alkylating agents, of both endogenous and exogenous origins,
are constantly modifying nucleic acids in cells, most commonly
■
sine, and 1,N2-ethenoguanine.
Eight human homologues (ALKBH1−8) of AlkB have been
identified,9 with ALKBH2 and ALKBH3 being shown to
by methylation. Some of these alkylating modifications are
catalyze nucleic acid N-demethylation;9 there is also evidence
damaging lesions and can lead to mutations, while others are
enzyme catalyzed, for example, the formation of 5-methyl-
that ALKBH2 functions as a repair enzyme in vivo.10 ALKBH8
catalyzes the modification of tRNA.11 The fat mass and obesity
gene (FTO) is linked to obesity by genome wide association
studies12 and has been shown to be a nucleic acid demethylase
acting on 3-methylthymine in single-stranded DNA13 and
RNA.14 Recently, the TET (Ten-Eleven-Translocation)
enzymes have been reported to catalyze hydroxylation of 5-
methylcytosine to give 5-hydroxymethylcytosine15 and 5-
carboxycytosine bases (Figure 1).16 In some CpG regions,
the level of TET-catalyzed modification is a substantial fraction
of the total 5-methylcytosine-modified DNA, leading to
speculation that TET-catalyzed modifications may be regu-
latory either by recruitment/loss of binding to regulatory
proteins and/or by a direct biophysical effect.17 Despite some
physiological connections (notably for FTO) with the
exception of AlkB itself and possibly ALKBH2, the exact
molecular roles of the nucleic acid-modifying 2OG-oxygenases
cytosine, and are regulatory.1,2 In recent years, it has become
apparent that a subfamily of Fe(II)- and 2-oxoglutarate (2OG)-
dependent oxygenases catalyzes the hydroxylation of the
methyl groups of methylated nucleic acids, which in the case
of N-methyl-modified DNA/RNA leads to demethylation and
thus repair. The first of these enzymes to be identified was
Escherichia coli AlkB,3 which is expressed in response to SN2
type alkylating agents.4 AlkB acts on a variety of lesions on
DNA and RNA but preferentially removes N-methyl groups
from 1-methyladenine or 3-methylcytosine in single-stranded
DNA.5 This is achieved by oxidizing the N-methyl group to a
hydroxymethyl group, which is coupled to the conversion of
2OG and O2 to succinate and CO2, respectively. The
hydroxymethyl intermediate then fragments to give form-
aldehyde and the unmodified base adenine and cytosine,
respectively (Figure 1).6 Structurally similar lesions are also
repaired by AlkB, albeit less efficiently,7,8 including 1-
methylguanine, 3-methylthymine, ethyl, propyl, hydroxyethyl,
and hydroxypropyl DNA adducts and exocyclic etheno and
Received: October 20, 2011
Published: January 23, 2012
© 2012 American Chemical Society
2173
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
Figure 1. Reactions catalyzed by 2OG-dependent oxygenases that modify nucleic acids. In the repair of 3-methylcytosine and 1-methyladenine
lesions by AlkB and ABH2/3, the methyl group is first oxidized to a hydroxymethyl group, which fragments to give formaldehyde and the
demethylated base. Methylated RNA as well as DNA may be a substrate in some cases.7,10 The insert shows a view from a crystal structure of
ALKBH2, showing the active site with Mn(II) substituting for Fe(II) (PDB ID: 3BUC).
are unclear. In many cases, they contain additional domains to
the catalytic domain (e.g., the C-terminal helical domain of
FTO and CXXC domain in the TET enzymes). Thus, the
dissection of the nucleic acid modulating roles by genetic
means might be problematic. We are therefore interested in
aiding the development of small molecule probes that target
their catalytic domains.
Here, we report the use of a combined approach employing
nondenaturing mass spectrometry binding assays, disulfide-
based dynamic combinatorial mass spectrometry (DCMS),
crystallographic analyses, fluorescence-based thermal shift
assays, and enzyme assays, which led to the identification of
N-oxalyl-L-cysteine derivatives 1 and 3-hydroxypyridine carbox-
amides 2 as potent inhibitors of AlkB. Some of these
compounds also demonstrated selectivity against other Fe(II)-
and 2OG-dependent oxygenases. The results should help to
enable the development of functional probes for nucleic acid-
modifying 2OG oxygenases.
mases;19,20 Poulsen has also applied DCMS to bovine carbonic
anhydrase II inhibition.21 In the DCMS technique, a “support-
ligand” that binds to the active site and contains a thiol side
chain is allowed to react reversibly (either at the active site or in
solution) with a set of thiols to form a mixture of disulfides.
Nondenaturing electrospray ionization mass spectrometry
(ESI-MS) of the protein−disulfide complexes is then used to
identify preferentially binding disulfides. For proteins that are
amenable to nondenaturing ESI-MS analysis, this method can
provide a rapid means of assessing the relative binding
strengths of ligands. In interpreting the nondenaturing ESI-
MS analyses, it is important to appreciate that different types of
noncovalent interaction survive the transition from solution to
gas phase differently.22 However, in the examples given above,
reasonable agreement between results from nondenaturing ESI-
MS binding data and those from solution-phase data was
observed.
N-Oxalylglycine (NOG) is a reasonably “generic” inhibitor of
2OG oxygenases, but there is evidence that N-oxalyl derivatives
of other amino acids can display selectivity for different
subfamilies.23,24 Hence, we began by using nondenaturing ESI-
MS to test for binding of NOG 3 and a set of L- (4a−d) and D-
(5a−d) enantiomers of N-oxalyl derivatives of amino acids
(Figures 2 and 3) to AlkB. Throughout this work, full-length E.
RESULTS AND DISCUSSION
■
To explore the tractability of the AlkB enzyme for small
molecule inhibition, we initially employed a disulfide-based
DCMS screen. We have employed this technique to 2OG-
dependent histone demethylases18 and metallo β-lacta-
2174
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
Figure 2. Structures of potential inhibitors investigated in this study.
Figure 3. Nondenaturing ESI-MS-based binding assays of N-oxalylamino acids to AlkB at 50 V. AlkB·Fe(II) (labeled E) in the presence of (a) no
compound, and compounds (b) N-oxalyl-L-glycine (NOG) 3, (c) N-oxalyl-L-alanine 4a, (d) N-oxalyl-L-cysteine 4b, (e) N-oxalyl-L-leucine 4c, (f) N-
oxalyl-L-phenylalanine 4d, (g) N-oxalyl-D-alanine 5a, (h) N-oxalyl-D-cysteine 5b, (i) N-oxalyl-D-leucine 5c, and (j) N-oxalyl-D-phenylalanine 5d.
Molecular masses (Da) are given in parentheses. A 5-fold excess of Fe(II) was used to give a substantial peak for the AlkB·Fe(II) complex; m/z =
mass/charge (Da), z = 8.
coli AlkB was used for nondenaturing ESI-MS binding studies
while ΔN11 AlkB,25 a crystallizable and active form of the AlkB
lacking the first 11 N-terminal amino acids, was used for
enzyme assays, thermal shift assays, and crystallography. The
2175
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
ESI-MS results were in good agreement with differential
scanning fluorimetry (DSF or “thermal shift”) assays that
indirectly measure the effect of ligands on protein stability;26
that is, those compounds observed to bind strongly by ESI-MS
(e.g., 3, 4a,c,d, and 5a) gave higher thermal shifts than those
that were not observed to bind strongly or not at all (e.g., 4b,
5b−d) (Figure 3 and Table 1).
(Figure 4). A library of 37 thiols (each at 15 μM, Figure 4c)
was mixed with 4b (15 μM), AlkB (15 μM), and Fe(II) (75
μM) in ammonium acetate (15 mM, pH 7.5), under aerobic
conditions. The mixture was then analyzed for binding to AlkB
by nondenaturing ESI-MS over time. At t0, the major peak
apparent corresponds to the AlkB·Fe(II)·4b complex (Figure
5a). After a 1 h incubation, a group of peaks corresponding to
the binding of AlkB·Fe(II) to disulfides was apparent (Figure
5b). After 4 h, the AlkB·Fe(II)·4b complex almost disappeared,
and two main groups of peaks at 24443 and 24472 Da were
apparent, presumably reflecting a shift towards formation of the
more stable complexes (Figure 5e). These masses correspond
to the addition of thiols with molecular masses of ∼124 and
∼153 Da, respectively, and could represent 16 of the 37 thiols
in the library (Figure 4c, highlighted in red). Further ESI-MS
binding experiments with the 16 individual thiols determined
the identity of the bound disulfides to be 4b·3-nitrothiophenol
(12), 4b·2-hydroxythiophenol (13), and 4b·3-hydroxythiophe-
nol (14) (Figure 2 and Figure S2 in the Supporting
Information). The DCMS screen was then repeated using N-
oxalyl-D-cysteine 5b as the “support” ligand (Figure 5g). No
AlkB·disulfide complexes were observed, demonstrating stereo-
selectivity in disulfide recognition by AlkB, that is, a preference
for the N-oxalyl-L-amino acids. Moreover, AlkB has four
cysteine residues within or close to its active site,25 which can
potentially form disulfides with thiols. Therefore, an experiment
where the “support ligand” 4b was excluded from the reaction
mixture was performed. No binding of thiols was observed
(Figure 5h), demonstrating that the mass shifts observed are
likely due to binding of support ligand-based disulfides to AlkB,
rather than simultaneous separate binding of 4b and thiols on
the enzyme.
Table 1. Activity and Thermal Shifts of the Oxalyl Series
against Alkb
Stable carbon analogues, 15−17, of identified disulfide “hits”
12−14, respectively, wherein the disulfide bond is replaced
with a C−S bond, were then synthesized (Figure 2 and Scheme
1). ESI-MS binding experiments implied strong binding of all
three carbon analogues to AlkB, with 15 ranking first in binding
affinities among the three in pairwise competitive binding
experiments (Figure S3 in the Supporting Information). These
results are in agreement with those obtained from inhibition
and thermal shift assays, which give the following order of
inhibitions and melting temperature (Tm) shifts: 15 (IC50 = 5.2
μM, Tm shift = 4.0 °C) > 17 (IC50 = 48 μM, Tm shift = 1.8 °C)
> 16 (IC50 = 50.4 μM, Tm shift = 1.3 °C) (Table 1). Binding of
the carbon analogues to AlkB, as determined by nondenaturing
ESI-MS, was unaffected by an increase in the Fe(II)
concentration in the assays; hence, the observed inhibition is
unlikely to be a result of Fe(II) chelation in solution.
To investigate the mode of binding of 15 to AlkB, we then
carried out crystallographic analyses (PDB ID: 3T4H; Figures
6a and S4a in the Supporting Information). As in previous AlkB
structures, when complexed with 15, the active site iron is
coordinated by the side chains of His131 (2.2 Å), Asp133 (2.2
Å), His187 (2.2 Å), and a putative water molecule located trans
(2.1 Å) to His131. The iron is coordinated by the oxalyl group
of 15 in a bidentate manner, with the carboxamide oxygen of
15 (2.2 Å) trans to Asp133 and one of the carboxylate oxygens
of 15 (2.6 Å) trans to His187. The carboxylate oxygen, which is
not chelating the iron, is positioned to form hydrogen bonds
with the side chain nitrogen NH2 of Arg210 (3.3 Å) and the
side chain oxygen of Asn120 (3.2 Å). The other carboxylic acid
of 15 is positioned to form a salt bridge to Arg204 (2.9 and 3.0
Å) and to form hydrogen bonds with the hydroxyl group of
We then used a capillary electrophoresis-based assay27,28 to
measure IC50 values for the N-oxalyl amino acid derivatives
(Table 1). Again, a good correlation of the rankings of AlkB
inhibition potencies with the rankings for ESI-MS binding
strength [Kendall's tb = 0.81 (p < 0.0001), Spearman's ρ = 0.89
(p < 0.0001)] and the rankings for thermal shift [Kendall's tb =
0.64 (p < 0.0001), Spearman's ρ = 0.67 (p < 0.0001)] were
observed. Indeed, throughout the study, with a few exceptions,
a good correlation between the results of the three assay
methods (ESI-MS, thermal shift, and IC50 values) was observed
as shown by statistical analyses (Figure S1 in the Supporting
Information). These initial results were interesting because they
imply that when the Cα substituent is sufficiently large, only the
L-enantiomers bind efficiently to AlkB. The apparently
preferential binding of the L-enantiomer of N-oxalyl amino
acids to AlkB is notable because the 2OG-dependent
asparaginyl-hydroxylase FIH (factor inhibiting hypoxia indu-
cible factor)24 and, at least some of the JmjC Nε-methyl lysine
histone demethylases, selectively bind the D-enantiomers of N-
oxalyl amino acid derivatives.18,29 Accordingly, we concluded
that modifications at the Cα position of L-N-oxalyl amino acid
may enhance inhibition and enable selectivity.
To explore the extent of the subpocket that is accessible to
the N-oxalyl amino acid series, a DCMS screen was then
performed, using N-oxalyl-L-cysteine 4b as the “support ligand”
2176
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
Figure 4. DCMS approach applied to AlkB. (a) The N-oxalyl group of support ligand N-oxalyl-L-cysteine 4b anchors it into the active site of AlkB via
interaction with the Fe(II) ion (salmon), leaving the thiol side chain free for disulfide formation with thiols in solution (disulfides formed in solution
may also bind). (b) Selective formation of an AlkB−disulfide complex involving the thiol member (pink sphere) that fits best into the active site. (c)
Structures and molecular masses (in parentheses) of the thiols used. The thiols that have masses corresponding to the observed mass shifts relative
to the AlkB·4b complex, in the DC-MS analyses are in red.
Tyr122 (2.7 Å) and to the side chain oxygen of Asn206 (2.7 Å).
An overlay of this structure with that of AlkB in complex with
2OG (PDB ID: 3I3Q)30 reveals little difference in the overall
conformation of the active site residues [Figure S5 in the
Supporting Information, rmsd (all atoms) = 1.3 Å, rmsd (Cα) =
0.7 Å] except for the side chain position of Trp178, which is
observed to rotate slightly with respect to its position in the
2OG complex structure, presumably to enable an apparent π-
stacking interaction between the phenyl ring of 15 and the side
chains of His187 and Trp178, the former being one of the
residues involved in Fe(II) coordination (Figure S5 in the
Supporting Information). The navigation of the phenyl ring of
15 into a hydrophobic subpocket (defined by Ile143, Phe154,
Trp178, Ser182, and His187) is made possible through the
relatively flexible cysteinyl linker. The complex is further
stabilized by apparent hydrogen-bonding interactions between
the 3-nitro group of 15 and the backbone carbonyl oxygen of
Ser182 (3.3 Å). Modeling studies suggest that similar
interactions are not possible when the 3-nitro group is replaced
with a 2-hydroxyl group or a 3-hydroxyl group, possibly
explaining the lower inhibitory potencies of 16 (IC50 = 50.4
μM) and 17 (IC50 = 48 μM), respectively, as compared to 15
(IC50 = 5.2 μM).
(NOG, IC50 = 32 μM) (Table 1). A crystal structure of AlkB in
complex with the methylated trinucleotide substrate T-meA-T
(PDB ID: 2FD8)25 reveals close proximity of the nucleotide-
binding site (a deep, predominantly hydrophobic pocket) to
the 2OG-binding site, wherein the 1-methyladenine base is
sandwiched between Trp69 and His131 (Figure S6 in the
Supporting Information). Superimposition of the AlkB:T-meA-
T complex with that of AlkB:15 indicates that further
elaboration of the oxalyl group of 15, through replacement
with an appropriately substituted pyridyl or quinoline ring, to
partially occupy the nucleotide-binding site, would be likely to
interfere with the binding of both the nucleotide and the 2OG
simultaneously, thereby potentially improving inhibitory
potency. Hence, two further series of compounds, the pyridyl
(20−28) and quinoline (29−31) series, were investigated
(Figure 2, Table 2, and Schemes 2 and 3).
The stabilizing effect of the 2-naphthalene side chain is
apparent in the pyridyl and quinoline series. With the exception
of 23 (IC50 = 7.9 μM) and 22 (IC50 = 3.4 μM), compounds
with the 2-naphthalene side chain were found to be more
potent than those without; hence, 21 (IC50 = 17.8 μM), 25
(IC50 = 16.7 μM), 27 (IC50 = 14.7 μM), and 30 (IC50 = 22.3
μM) were more potent than 20 (IC50 >1 mM), 24 (IC50 = 51.3
μM), 26 (IC50 = 54.1 μM), and 29 (IC50 = 165 μM),
respectively (Table 2). The ranking of AlkB inhibition by
turnover assays correlates well with the rankings both for
thermal shift [Kendall's tb = 0.67 (p < 0.0001), Spearman's ρ =
0.84 (p < 0.0001)] and for ESI-MS binding strength [Kendall's
tb = 0.80 (p < 0.0001), Spearman's ρ = 0.90 (p < 0.0001),
Figure S1 in the Supporting Information], hence validating the
use of these techniques as efficient and rapid methods in
screening for 2OG oxygenase inhibitors.
To take advantage of the apparent hydrophobic subpocket
identified in the crystallographic analyses, analogues with
naphthalene side chains, 18 and 19, were proposed as improved
inhibitors (Table 1). These were synthesized according to
Scheme 1 (for synthetic details, see the Experimental Section
and Supporting Information). Consistent with this prediction,
the presence of a 2-naphthalene and 1-naphthalene side chain
in 18 (IC50 = 0.5 μM) and 19 (IC50 = 5.4 μM), respectively,
increases potency as compared to their parent compound 3
2177
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
linker of 18, however, adopts a different conformation from
that of 15, likely to align the bulkier 2-naphthalene side chain of
18 with the side chains of His187 and Trp178 for efficient π-
stacking interactions. Analysis of the AlkB:18 complex structure
suggests that the 1-naphthalene side chain of 19 is less able to
fit into the hydrophobic subpocket due to steric constraints,
accounting for its lower inhibitory activity (IC50 = 5.4 μM) as
compared to 18.
A crystal structure of AlkB in complex with 22 (PDB ID:
3T3Y; Figures 6c and S4c in the Supporting Information)
reveals a different mode of binding from those of 15 and 18,
wherein the pyridine ring, instead of occupying the same metal
coordinating position as the oxalyl groups in 15 and 18, “flips”
over with the pyridyl nitrogen (2.2 Å) trans to His131 to
partially occupy the hydrophobic subpocket. The side chain of
Trp178 is rotated (approximately 90° about the Cα−Cβ bond
and approximately 180° about the Cβ−Cγ bond) to form π-
stacking interactions with the pyridine ring. The 3-hydroxyl
group of 22 is positioned to participate in a hydrogen bond to
the side chain hydroxyl group of Ser145 (2.1 Å), possibly
accounting for the greater potencies of 22 (IC50 = 3.4 μM), 23
(IC50 = 7.9 μM), and 31 (IC50 = 42.5 μM) as compared to
analogues without the 3-hydroxyl group 20 (IC50 > 1 mM), 21
(IC50 = 17.8 μM), and 29 (IC50 = 165 μM), respectively.
To investigate the selectivity of the most potent AlkB
inhibitors against other Fe(II)- and 2OG-dependent oxy-
genases, three of the more potent inhibitors (15, 18, and 19)
were tested for inhibition against two physiologically important
human 2OG oxygenases PHD2 (hypoxia-inducible factor prolyl
hydroxylase 2, which is central to the hypoxic response) and
PHF8 (PHD finger protein 8, the mutations to which are
linked to midline defects).31 All three compounds showed IC50
> 1 mM for both PHD232 and PHF8,33 which represent
significant selectivity (200−2000-fold) towards AlkB.
Figure 5. Dynamic combinatorial mass spectrometric (DCMS)
analyses on AlkB. Nondenaturing ESI-MS spectra at 30 V showing
AlkB·Fe(II) (labeled E) in the presence of (a) N-oxalyl-L-cysteine 4b
and the thiol library at 0, (b) 1, (c) 2, (d) 3, (e) 4, and (f) 24 h; (g)
AlkB·Fe(II) in the presence of N-oxalyl-D-cysteine 5b and the thiol
library at 24 h; and (h) AlkB·Fe(II) in the presence of the thiol library
at 24 h without N-oxalyl-L-cysteine 4b. m/z = mass/charge (Da), z =
8.
CONCLUSIONS
■
Overall, the results demonstrate that AlkB, and by implication
other 2OG-dependent nucleic acid-modifying oxygenases, are
amenable to potent inhibition by small molecules. The
combined results, including crystallographic analyses, should
provide a basis for the development of potent and selective
AlkB/ALKBH inhibitors, suitable for use as functional probes
and, in the longer term, possibly for clinical use. Importantly,
the results demonstrate that a high degree of selectivity for
AlkB over other subfamilies of 2OG oxygenases should be
possible, as shown by a lack of inhibition of our most potent
AlkB inhibitors against two other 2OG oxygenases, PHD2 at 1
mM. Notably, in the N-oxalyl amino acid compounds, the L-
Interestingly, a crystal structure of the most potent
compound 18 (IC50 = 0.5 μM, Tm shift = 7.8 °C) in complex
with AlkB was then obtained (PDB ID: 3T4V; Figures 6b and
S4b in the Supporting Information). The oxalyl group of 18 is
observed to coordinate with iron in a manner analogous to that
of 2OG and 15 (Figures 6a and S3a in the Supporting
Information) with most apparent salt bridges and hydrogen
bonds in the AlkB:15 complex being conserved. The cysteinyl
a
Scheme 1. Synthesis of the N-Oxalyl Inhibitors
a
Conditions: (a) Monomethyl oxalyl chloride, Et3N, CH2Cl2. (b) RCH2Br, Et3N, CH2Cl2. (c) NaOH (1 N), MeOH.
2178
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
Figure 6. Active site views from structures of AlkB (green sticks) bound to (a) 15 (purple sticks; PDB ID: 3T4H), (b) 18 (yellow sticks; PDB ID:
3T4V), and (c) 22 (magenta sticks; PDB ID: 3T3Y). The experimental 2Fo − Fc electron density maps (contoured to 1.5σ), displayed as gray mesh,
are shown to the right.
stereochemistry at the Cα center gave preferential inhibition.
This contrasts with FIH and the JMJD2 family of histone
demethylases, which are selectively inhibited by the D-
stereochemistry of related N-oxalyl amino acid derivatives.18,24
Thus, consideration of the stereochemistry of the Cα center
(and equivalent in other series) is an important factor in
obtaining subfamily selectivity. It should be noted that there are
likely considerable variations in the active sites of the
ALKBH1−8 and related human enzymes (FTO, TET1−3);
thus, further discrimination within the group of nucleic acid-
modifying 2OG oxygenases is likely possible. This will be
important in developing 2OG oxygenase inhibitors for use as
functional probes.
binding results, IC50 values, and thermal shift assays. We
appreciate that there are limitations to the use of ESI-MS
binding assay results, both because not all proteins are suitable
for this type of assay and because the strength of binding as
observed by the ESI-MS appears to be dependent on the type
of interactions involved in binding,22 likely reflecting the
relative contribution of entropic and enthalpic components.
The latter consideration means that it can be difficult to
compare results between series of compounds that bind
differently. Nevertheless, in our work, we saw good correlation
between the binding strengths and the assay results and
anticipate that, as the empirical data sets accumulate,
confidence in the ESI-MS binding assays results will increase.
The work also further validates the use of the DCMS method
for identifying lead compounds useful for inhibitor develop-
ment. An advantage of the DCMS binding assay method over
most other techniques for the analysis of compound binding is
that it directly analyzes the mass of protein−ligand complexes.
We observed good correlation between the ESI-MS-based
EXPERIMENTAL SECTION
■
Chemical Synthesis. Reagents and solvents were from Aldrich or
Alfa Aesar. Reactions were monitored by TLC, which was performed
on precoated aluminum-backed plates (Merck, silica 60 F254).
Melting points were determined using a Leica Galen III hot-stage
2179
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
N-Oxalyl-L-cysteine (40). A solution of aqueous NaOH (1 N, 10.00
mL, 10.0 mmol) was added dropwise to a mixture of 34 (0.57 g, 1.7
mmol) in MeOH (20 mL) at 0 °C. The mixture was stirred for 1 h and
then allowed to warm to room temperature overnight, after which it
was evaporated in vacuo. The resulting residue was resuspended in
H2O, acidified with aqueous HCl (1 N) to pH 2, and extracted with
EtOAc. The organic extracts were washed (saturated NaHCO3, H2O,
and brine), dried (Na2SO4), and evaporated in vacuo. This gave 40 as
a white solid (0.38 g, 73%); mp 159−160 °C. Rf = 0.2 (MeOH/
CH2Cl2 3:7). IR (neat) νmax/cm−1 3010−3364 (NH and OH), 1645
(amide CO), 1524 (carboxylate CO). 1H NMR (500 MHz, MeOD): δ
3.01 (1H, dd, J = 7.0, 14.0 Hz, CH2SH), 3.09 (1H, dd, J = 5.0, 14.0 Hz,
CH2SH), 4.67 (1H, dd, J = 4.5, 6.5 Hz, CH). 13C NMR (500 MHz,
MeOD): δ 35.4 (CH2SH), 55.9 (CH), 165.5 (COCO2H), 166.6
(COCO2H), 177.4 (CO2H). HRMS (ESI, negative ion) C5H6NO5S
[M − H]− requires 191.9972; found, 191.9968.
Table 2. Activity and Thermal Shifts of the Pyridyl Series
(20−28, 32, and 33) and Quinoline Series (29−31) against
AlkB
Synthesis of N-Oxalyl Inhibitor 15 and 18. N-Methyloxalyl-
S-(3-nitrobenzyl)-L-cysteine Methyl Ester (35). Triethylamine (2.50
mL, 17.9 mmol) was added dropwise to a solution of 34 (3.04
g, 13.8 mmol) and 3-nitrobenzyl brominde (3.3 g, 15.1 mmol)
in anhydrous CH2Cl2 (50 mL) at 0 °C. The mixture was stirred
for 2 h and then allowed to warm to room temperature
overnight, after which it was evaporated in vacuo. The resulting
residue was dissolved in EtOAc, washed (saturated NaHCO3,
H2O, and brine), dried (Na2SO4), and evaporated in vacuo.
Chromatography (EtOAc/hexane 2:3) gave 35 as a colorless oil
(2.45 g, 50%). Rf = 0.3. IR (neat) υ/cm−1: 3340 (NH), 1735
1
(CO ester), 1684 (CO amide) 1525, 1352 (NO2). H NMR
(500 MHz, CDCl3): δ 2.89 (1H, dd, J = 6.0, 14.0 Hz, cysteinyl
CH2), 2.97 (1H, dd, J = 5.0, 14.5 Hz, cysteinyl CH2), 3.78 (3H,
s, CO2CH3), 3.82 (2H, s, benzyl CH2), 3.94 (3H, s,
COCO2CH3), 4.76−4.84 (1H, m, CH), 7.51 (1H, t, J = 8.0
Hz, H5′), 7.66 (1H, d, J = 8.0 Hz, H4′), 7.73 (1H, d, J = 8.0 Hz,
NH), 8.13 (1H, d, J = 8.0 Hz, H6′), 8.19 (1H, s, H2′). 13C
NMR (500 MHz, CDCl3): δ 33.2 (cysteinyl CH2), 35.8 (benzyl
CH2), 52.0 (CH), 53.1 (CO2CH3), 53.9 (COCO2CH3), 122.4
(C6′), 123.8 (C2′), 129.6 (C5′), 135.0 (C4′), 139.6 (C1′), 148.4
(C3′), 155.9 (COCO2CH3), 160.2 (COCO2CH3), 169.9
(CO2CH3). HRMS (ESI, positive ion) C14H16N2NaO7S [M +
Na]+ requires 379.0570; found, 379.0574.
melting point apparatus and microscope. Infrared spectra were
recorded from Nujol mulls between sodium chloride discs, on a
Bruker Tensor 27 FT-IR spectrometer. NMR spectra were acquired
using a Bruker DPX500 NMR spectrometer. Chemical shifts (δ) are
given in ppm, and the multiplicities are given as singlet (s), doublet
(d), triplet (t), quartet (q), multiplet (m), and broad (br). Coupling
constants J are given in Hz ( 0.5 Hz). High resolution mass spectra
(HRMS) were recorded using Bruker MicroTOF. The purity of all
compounds synthesized was ≥95% as determined by analytical
reverse-phase HPLC (Ultimate 3000). The synthesis and character-
ization of compound 31 and 29 were as reported in refs 34 and 35,
respectively. The syntheses of compounds 16−17, 19−21, 23−28, 30,
32, and 33 are given in the Supporting Information.
N-Oxalyl-S-(3-nitrobenzyl)-L-cysteine (15). A solution of aqueous
NaOH (1N, 13.00 mL, 13.0 mmol) was added dropwise to a solution
of 35 (0.77 g, 2.2 mmol) in THF (20 mL) at 0 °C. The mixture was
stirred for 1 h and then allowed to warm to room temperature
overnight, after which it was evaporated in vacuo. The resulting residue
was resuspended in H2O, acidified with aqueous HCl (1 N) to pH 2,
and extracted with EtOAc. The organic layer was washed (saturated
NaHCO3, H2O, and brine), dried (Na2SO4), and evaporated in vacuo.
Chromatography (MeOH/CH2Cl2 3:7) gave 15 as an off-white solid
(0.43 g, 61%); mp 114−115 °C. Rf = 0.4. IR (neat) νmax/cm−1 3016−
3353 (NH and OH), 1738 (amide CO), 1654 (carboxylate CO), 1525,
1
1352 (NO2). H NMR (500 MHz, MeOD): δ 2.93 (1H, dd, J = 7.0,
14.2 Hz, cysteinyl CH2), 3.05 (1H, dd, J = 5.0, 14.0 Hz, cysteinyl
CH2), 3.93 (1H, s, benzyl CH2), 4.62−4.69 (1H, m, CH), 7.58 (1H, t,
J = 8.0 Hz, H5′), 7.78 (1H, d, J = 8.0 Hz, H4′), 8.14 (1H, d, J = 8.0 Hz,
H6′), 8.26 (1H, s, H2′). 13C NMR (500 MHz, MeOD): δ 33.6
(cysteinyl CH2), 36.3 (benzyl CH2), 53.8 (CH), 123.2 (C6′), 124.8
(C2′), 130.9 (C5′), 136.5 (C4′), 142.2 (C1′), 149.8 (C3′), 165.3
(COCO2H), 166.2 (COCO2H), 177.5 (CO2H). HRMS (ESI, positive
ion) C12H12N2NaO7S [M + Na]+ requires 351.0257; found, 351.0264.
N-Methyloxalyl-S-(2-naphthalenemethyl)-L-cysteine Methyl Ester
(38). Triethylamine (0.4 mL, 2.94 mmol) was added dropwise to a
solution of 34 (0.50 g, 2.26 mmol) and 2-bromomethylnaphthalene
(0.55 g, 2.49 mmol) in anhydrous CH2Cl2 (40 mL) at room
temperature. The mixture was stirred at room temperature for 18 h,
washed (H2O), dried (MgSO4), and evaporated in vacuo.
Chromatography (hexane/EtOAc 6:4) gave 38 as a white solid
(0.36 g, 43%); mp 87−90 °C. Rf = 0.35. [α]D = −67.1 (c 2.56 × 10−3).
IR (neat) ν/cm−1: 3303 (NH), 1735 (CO ester), 1687 (CO amide).
1H NMR (500 MHz, CDCl3): δ 2.91 (1H, qd, J = 5.0, 16.0 Hz,
N-Methyloxalyl-L-cysteine Methyl Ester (34). Monomethyl oxalyl
chloride (2.94 mL, 32.0 mmol) was added dropwise to a mixture of L-
cysteine methyl ester hydrochloride (5.00 g, 29.1 mmol) and Et3N
(8.9 mL, 64.0 mmol) in anhydrous CH2Cl2 (100 mL) at 0 °C. The
mixture was then stirred at room temperature for overnight, after
which it was evaporated in vacuo. The resulting residue was dissolved
in EtOAc, washed (saturated NaHCO3, H2O, and brine), dried
(Na2SO4), and evaporated in vacuo. Chromatography (EtOAc/hexane
1:1) gave 34 as a pale yellow oil (6.43 g, quantative yield). Rf = 0.4. IR
1
(neat) υ/cm−1: 3346 (NH), 1743 (CO ester), 1615 (CO amide). H
NMR (500 MHz, CDCl3): δ 2.90 (1H, dd, J = 6.5, 14.0 Hz, CH2SH),
2.95 (1H, dd, J = 5.0, 14.0 Hz, CH2SH), 3.82 (3H, s, CO2CH3), 3.93
(3H, s, COCO2CH3), 4.70−4.75 (1H, m, CH), 7.88 (1H, br s, NH).
13C NMR (500 MHz, CDCl3): δ 26.3 (CH2SH), 53.1 (CH), 53.8
(CO2CH3), 54.1 (COCO2CH3), 155.9 (COCO2CH3), 160.2
(COCO2CH3), 169.3 (CO2CH3). HRMS (ESI, positive ion)
C7H11NNaO5S [M + Na]+ requires 244.0250; found, 244.0253.
2180
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Pyridyl Inhibitors
a
Conditions: (a) Compound 55, diisopropylethylamine, hydroxybenzotriazole, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide, CH2Cl2. (b) LiOH
(1 N), 1,4-dioaxane. (c) Glycine methyl ester hydrochloride, diisopropylethylamine, hydroxybenzotriazole, 1-ethyl-3-(3-dimethylaminopropyl)-
carbodiimide, CH2Cl2. (d) L-Alanine methyl ester hydrochloride or L-tryptophan methyl ester hydrochloride for compounds 50 and 51, respectively,
diisopropylethylamine, hydroxybenzotriazole, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide, CH2Cl2..
a
Scheme 3. Synthesis of Quinoline Inhibitors
a
Conditions: (a) Glycine (40 for compound 53), diisopropylethylamine, hydroxybenzotriazole, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide,
CH2Cl2. (b) LiOH (1 N), 1,4-dioaxane.
cysteinyl CH2), 3.75 (3H, s, CO2CH3), 3.88−3.91 (2H, m, cysteinyl
CH2), 3.90 (3H, s, COCO2CH3), 4.81−4.85 (1H, m, CH), 7.45−7.51
(3H, m, ArCH), 7.71 (1H, d, J = 6.5 Hz, ArCH), 7.81−7.83 (3H, m,
ArCH). 13C NMR (500 MHz, CDCl3): δ 32.85 (CH2), 36.8 (cysteinyl
CH2), 52.0 (CH), 52.9 (CO2CH3), 53.8 (COCO2CH3), 126.0 (Ar
CH), 126.3 (Ar CH), 126.8 (Ar CH), 127.6 (Ar CH), 127.7 (Ar CH),
127.7 (Ar CH), 128.6 (Ar CH), 132.6 (Ar C), 133.2 (Ar C), 134.6 (Ar
C), 155.9 (COCO2CH3), 160.2 (COCO2CH3), 170.0 (CO2CH3).
HRMS (ESI+) C18H19NNaO5S [M + Na]+ requires 384.0876; found,
384.0877.
N-Oxalyl-S-(2-naphthalenemethyl)-L-cysteine (18). A solution of
aqueous NaOH (1 N, 0.9 mL, 0.9 mmol) was added dropwise to a
solution of 38 (0.17 g, 0.47 mmol) in MeOH (10 mL) and was stirred
at room temperature overnight, after which it was evaporated in vacuo.
The resulting residue was resuspended in H2O, acidified with aqueous
HCl (1 N) to pH 2, and extracted with EtOAc. The organic layer was
washed (H2O, brine), dried (MgSO4), and evaporated in vacuo.
Washing with Et2O (3 mL) gave 18 as a white solid (0.07 g, 50%); mp
159−160 °C. [α]D = −43.6 (c 1.01 × 10−3). IR (neat) νmax/cm−1 3309
(N−H), 3235 (COO-H), 1748 (HOCO), 1726 (CO), 1639
1
(NCO). H NMR (500 MHz; MeOD-d4): δ 2.89−3.05 (2H, m,
SCH2), 3.95 (2H, s, CH2), 4.68−4.71 (1H, m, CH), 7.46−7.51 (3H,
m, ArCH), 7.76 (1H, s, Ar CH), 7.81−7.84 (3H, m, Ar CH). 13C
NMR (125 MHz, MeOD-d4): δ 33.1 (CH2), 37.2 (CH2), 53.9 (CH),
126.9 (Ar CH), 127.1 (Ar CH), 127.3 (Ar CH), 128.2 (Ar CH), 128.7
(Ar CH), 129.3 (Ar CH), 129.5 (Ar CH), 134.1 (Ar C), 134.7 (Ar C),
136.6 (Ar C), 172.6 (CO2H). HRMS (ESI−) C16H14NO5S [M − H]−
requires 332.0598; found, 332.0595.
Synthesis of Pyridyl Inhibitor 22. Methyl 2-(3-
Hydroxypicolinamido)acetate (46). To a stirred solution of
picolinic acid (0.50 g, 3.59 mmol) in dry CH2Cl2 (20 mL)
under N2, hydroxybenzotriazole (0.58 g, 4.31 mmol), 1-ethyl-3-
2181
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
(3-dimethylaminopropyl) carbodiimide (0.76 mL, 4.31 mmol),
and diisopropylethylamine (1.25 mL, 7.18 mmol) were added.
The mixture was stirred for 10 min, and glycine methyl ester
hydrochloride (0.45 g, 3.59 mmol) was added. The reaction
mixture was stirred for 36 h until consumption of starting
material. The mixture was washed with water, and the organic
layer was dried with MgSO4 and concentrated under reduced
pressure. The yellow oil was chromatographed (hexane/EtOAc
7:3 to 5:5). Chromatography gave compound 46 as white
AlkB Expression and Purification. AlkB and ΔN11 AlkB
proteins were expressed and purified as described.23,25 BL21 (DE3)
E. coli transformed with pET24a AlkB or ΔN11 AlkB were grown at
37 °C and 220 rpm to an OD600 of 0.6. Protein expression was induced
by the addition of 0.2 mM IPTG (Melford Laboratories Ltd.). Growth
was continued at 28 °C for 4 h (AlkB) or 15 °C for 16 h (ΔN11
AlkB), and then, cells were harvested by centrifugation. The resulting
cell pellet was stored at −80 °C. Cell pellets were resuspended to
homogeneity in 0.1 M MES, pH 5.8, 1 mM MgCl2, and 1× Roche
complete EDTA-free protease inhibitor cocktail. Cells were lysed on
ice by sonication, and the lysate was cleared by centrifugation and
filtration. AlkB was purified from the crude cell lysate by cation
exchange chromatography using a 50 mL S sepharose column, with
elution achieved by application of a gradient to 1 M NaCl. Further
purification was achieved by gel filtration using a 300 mL Superdex 75
column (Pharmacia) in a buffer of 50 mM HEPES, pH 7.5.
Nondenaturing ESI-MS. All 37 thiols used for DCL generation
were from Sigma-Aldrich or Alfa Aesar. AlkB was desalted using a Bio-
Spin 6 Column (Bio-Rad, Hemel Hempstead, United Kingdom) in 15
mM ammonium acetate (pH 7.5). The stock solution was diluted with
the same buffer to a final concentration of 100 μM. FeSO4·7H2O was
dissolved in 20 mM HCl at a concentration of 100 mM. This was then
diluted with Milli-Q water to give final working concentrations of 500
μM. The protein was mixed with Fe(II) and an inhibitor to give final
concentrations of 15 μM AlkB, 75 μM Fe(II), and 15 μM inhibitor.
The solution was then incubated for 30 min at room temperature prior
to ESI-MS analysis.
Mass spectrometric data were acquired using a Q-TOF mass
spectrometer (Q-TOF micro, Micromass, Altrincham, United King-
dom) interfaced with a NanoMate (Advion Biosciences, Ithaca, NY)
with a chip voltage of 1.70 kV and a delivery pressure 0.25 psi. The
sample cone voltage was typically 80 V with a source temperature of
40 °C and with an acquisition/scan time of 10 s/1 s. Calibration and
sample acquisition were performed in the positive ion mode in the
range of 500−5000 m/z. The pressure at the interface between the
atmospheric source and the high vacuum region was fixed at 6.60
mbar. External instrument calibration was achieved using sodium
iodide. Data were processed with the MassLynx 4.0 (Waters).
DSF. DSF was performed using a MiniOpticon Real-Time PCR
Detection System (Bio-Rad), monitoring protein unfolding using
SYPRO orange (Invitrogen) according to reported method.26 FAM
(492 nm) and ROX (610 nm) filters were used for excitation and
emission, respectively. Reaction mixes contained 2 μM protein, 50 μM
Mn(II), 200 μM compounds, and 1× SYPRO orange in a final volume
of 50 μL. Reagents were prepared in HEPES buffer except metals,
which were dissolved as 100 mM stocks in 20 mM HCl, and then
further diluted in Milli-Q water. Compounds tested were prepared in
100% DMSO and added such that the final concentration of DMSO
was 5% (v/v).
Fluorescence readings were taken every 1 °C in the range 25−95
°C, with the temperature increased linearly by 1 °C min−1. The
software provided was used to perform global minimum subtraction.
The inflection point, representing Tm, was calculated by fitting the
Boltzmann equation to the sigmoidal curves obtained; data were
processed using GraphPad Prism 5.0. The Tm shift caused by the
addition of small molecules/fragments was determined by subtraction
of the “reference” Tm (protein incubated with metal and 5% DMSO)
from the Tm obtained in the presence of the compound.35 Conditions
were tested in triplicate, with standard deviations typically <1 °C.
Inhibition Assays for AlkB.27,28 Synthetic fluorescently labeled
DNA substrate (5′-TTCmTTTTTTTTTTTT-3′-fluorescein) and
product (5′-TTCTTTTTTTTTTTT-3′-fluorescein) were produced
by ATDBio (University of Southampton, United Kigndom). All other
chemicals were purchased from Sigma-Aldrich (Toronto, ON). An
uncoated fused-silica capillary was purchased from Polymicro
Technologies (Phoenix, AZ). All solutions were made using deionized
water filtered through a 0.22 μm filter (Millipore, Nepean, ON).
All experiments were conducted using an uncoated fused silica
capillary with a total length of 50 cm (40 cm to the detection window),
inner diameter of 75 μm, and outer diameter of 365 μm. The capillary
needles (0.35 g, 47%); mp 54−55 °C. Rf = 0.36. IR (neat) νmax
/
cm−1 3355 (O−H, N−H), 1749(OCO), 1651 (NCO). 1H
NMR (500 MHz; CDCl3): δ 3.79 (3H, s, CH3), 4.23 (2H, d, J
= 6.0, CH2), 7.28−7.35 (2H, m, Ar CH), 8.07 (1H, dd, J = 4.0
1.5, Ar CH), 8.46 (1H, s, NH), 11.76 (1H, s, OH). 13C NMR
(125 MHz, CDCl3): δ 40.7 (CH2), 52.5 (CH3), 126.05 (Ar
CH), 128.9 (Ar CH), 131.0 (Ar C), 139.8 (Ar CH), 157.7 (Ar
C), 169.0 (CONH), 169.6 (CO2CH3). HRMS (ESI+)
C9H10N2NaO4 [M + Na]+ requires 233.0533; found, 233.0525.
2-(3-Hydroxypicolinamido)acetic Acid (22). To a solution of 46
(0.02 g, 0.095 mmol) in 1,4-dioxane (5 mL), 1 N lithium hydroxide
(0.2 mL, 0.2 mmol) was added. The mixture was stirred at room
temperature for 24 h until the consumption of starting material and
acidified with acetic acid to pH 3 and diluted with CH2Cl2. The
solution was washed with water, dried (MgSO4), and concentrated in
vacuo. Concentration gave compound 22 as a white solid (2.0 mg,
20%); mp 169−170 °C. IR (neat) νmax/cm−1 3354 (O−H, N−H),
1
2950 (COO-H), 1745 (OCO), 1678 (NCO). H NMR (500
MHz; DMSO-d6): δ 4.00 (2H, d, J = 6.0, CH2), 7.45 (1H, dd, J = 8.5
1.5, Ar CH), 7.56 (1H, q, J = 4.0, Ar CH), 8.20 (1H, dd, J = 4.5 1.5, Ar
CH), 9.33 (1H, t, J = 6.0, NH), 12.30 (1H, s, OH). 13C NMR (125
MHz, DMSO-d6): δ 40.6 (CH2), 125.95 (Ar CH), 129.35 (Ar CH),
130.9 (Ar C), 140.0 (Ar CH), 157.2 (Ar C), 169.0 (CONH), 170.5
(CO2H). HRMS (ESI−) C8H7N2O4 [M − H]− requires 195.0411;
found, 195.0417.
Synthesis of Quinoline Inhibitor 29. Methyl 2-(Quinoline-2-
carboxamido)acetate (52). Triethylamine (3.31 mL, 23.75
mmol) was added dropwise to a solution of glycine methyl
ester hydrochloride (0.13 g, 1.04 mmol) and quinaldoyl
chloride (0.20 g, 1.04 mmol) in anhydrous CH2Cl2 (20 mL)
at room temperature. The mixture was stirred at room
temperature for 2 days, washed (H2O), dried (MgSO4), and
evaporated in vacuo. Chromatography (hexane/EtOAc 7:3 to
6:4) gave 52 as a white solid (0.12 g, 50%); mp 98−99 °C. Rf =
0.35. IR (neat) νmax/cm−1 3328 (N−H), 1746 (CH3OCO),
1
1653 (NCO). H NMR (500 MHz; CDCl3): δ 3.83 (3H, s,
CH3), 4.35 (2H, d, J = 6.0, CH2), 7.62−7.65 (1H, m, Ar CH),
7.76−7.80 (1H, m, Ar CH), 7.88 (1H, d, J = 4.0, ArCH), 8.15
(1H, dd, J = 8.8, 1.0, Ar CH), 8.31 (2H, dd, J = 15.0, 8.5, Ar
CH), 8.71 (1H, t, J = 5.0, NH). 13C NMR (125 MHz, CDCl3):
δ 41.3 (CH2), 52.4 (CH3), 118.8 (Ar CH), 127.7 (Ar CH),
128.0 (Ar CH), 129.4 (Ar CH), 129.8 (Ar CH), 130.1 (Ar CH),
137.5 (Ar C), 146.5 (Ar C), 149.0 (Ar C), 164.8 (CONH),
170.2 (CO2CH3). HRMS (ESI−) C13H12N2O3 [M + Na]−
requires 267.0740; found, 267.0731.
2-(Quinoline-2-carboxamido)acetic Acid (29).35 To a solution of
52 (25 mg, 0.10 mmol) in 1,4-dioxane (3 mL), 1 N lithium hydroxide
(0.20 mL, 0.20 mmol) was added. The mixture was stirred at room
temperature for 18 h until the consumption of starting material, and
then, it was acidified with acetic acid to pH 3 and diluted with CH2Cl2.
The solution was washed with water, dried (MgSO4), and
concentrated in vacuo. Concentration gave compound 29 as a white
solid (15 mg, 66%); mp 183−184 °C. IR (neat) νmax/cm−1 3357 (N−
1
H), 2915 (COO−H), 1749 (HOCO), 1624 (NCO). H NMR
(500 MHz; MeOD-d4): δ 4.25 (2H, s, CH2), 7.68−7.71 (1H, m, Ar
CH), 7.82−7.85 (1H, m, Ar CH), 8.00 (1H, d, J = 8.0, Ar CH), 8.19
(2H, t, J = 8.5, Ar CH), 8.47 (1H, d, J = 8.5, Ar CH). 13C NMR (125
MHz, DMSO-d6): δ 42.1 (CH2), 119.5 (Ar CH), 129.0 (Ar CH),
129.4 (Ar CH), 130.8 (Ar CH), 130.85, 131.5 (Ar CH), 138.9 (Ar
CH), 148.1, 150.5, 167.2 (CONH), 172.9 (CO2H). HRMS (ESI−)
C12H10N2O3 [M − H]− requires 229.0619; found, 229.0611.
2182
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
was mounted on a P/ACE MDQ capillary electrophoresis (CE)
instrument (Beckman Coulter, Fullerton, CA) with temperature
control to keep the capillary at 15 °C. A sample was introduced into
the capillary by a pressure pulse of 0.5 psi for 5 s. The reaction product
(P) and substrate (S) were then separated by CE at 20 kV and
quantitated with laser-induced fluorescence (LIF) detection (fluo-
rescence excitation at 488 nm and fluorescence detection at 520 nm).
The CE run buffer for kinetic and inhibition studies was 20 mM Borax
and 60 mM SDS at pH 8.0.
ASSOCIATED CONTENT
* Supporting Information
Chemical synthesis and compound characterizations, statistical
analyses on correlations between ESI-MS, thermal shift and
inhibition data, nondenaturing ESI-MS binding experiments,
and crystallographic structure solution methods. This material
is available free of charge via the Internet at http://pubs.acs.org.
■
S
Accession Codes
The enzymatic reaction was initiated by the addition of AlkB
protein to a mixture containing the incubation buffer (50 mM Tris-
The coordinates for AlkB in complex with 15, 18, and 22 have
been deposited in the RCSB Protein Data Bank as PDB IDs
3T4H, 3T4V, and 3T3Y, respectively.
HCl, pH 7.5),
4 mM AA, 160 μM 2OG, 80 μM
(NH4)2SO4·FeSO4·6H2O, 100 nM 5′-TTCmTTTTTTTTTTTT-3′-
fluorescein, and 2154 units of catalase. The relative activity of AlkB
was measured using 5 nM AlkB and 100 nM oligonucleotide substrate,
while varying the inhibitor concentration in the range 1−1000 μM.
The enzymatic reactions were incubated for 4.5 min at room
temperature and stopped by adding a prechilled EDTA solution (5
mM final concentration). The demethylated DNA product formed was
separated from the unreacted substrate by CE and quantified with LIF.
The relative activity of AlkB at different inhibitor concentrations was
plotted against the inhibitor concentration, and the IC50 values were
calculated as the concentration of inhibitor reducing the enzymatic
activity to half its maximal level, using the OriginPro 8.0 software.
Inhibition Assays for PHF833 and PHD2. Reactions consisted of
PHF8 (2 μM), Fe(II) (10 μM), ascorbate (100 μM), 2OG (20 μM)
H3(1−14)K4me3K9me2 peptide (20 μM), and inhibitor (1 mM) in
1% DMSO, 500 mM NaCl, 100 mM HEPES (pH 7.5). PHF8, Fe(II),
ascorbate, and inhibitors were preincubated at 37 °C for 15 min before
the addition of 2OG and peptide. Reactions were incubated at 37 °C
for 20 min prior to 1:1 quenching with methanol. Product formation
was assessed by MALDI-TOF; 1 μL of quenched reaction was mixed
with 1 μL of α-cyano-4-hydroxycinnamic acid and spotted onto a
MALDI-TOF plate for analysis.33 Inhibition levels were measured
relative to an inhibitor free reaction. Inhibition assay methods for
PHD2 will be reported elsewhere. Details of the N terminally
truncated PHD2 was prepared as reported.36,37
Protein Crystallography. Crystals of AlkB in complex with 15,
18, and 22 were grown in sitting drops at 293 K using vapor diffusion
methods. The ratio of protein to reservoir solution for the AlkB:15 and
AlkB:18 cocrystallization drops was 2:1 (300 nL total drop volume)
and for AlkB:22 was 1:1 (200 nL total drop volume), and the reservoir
volume was 80 μL. The AlkB protein solution contained 10.1 mg/mL
protein, 50 mM HEPES, pH 7.5, 2.2 mM ammonium iron(II) sulfate,
and 5.7 mM 15 or 1 mM 18. The protein solution for the AlkB:22
complex contained the same except that the ammonium iron(II)
sulfate concentration was 0.44 mM and 22 was 1 mM. The reservoir
solution for AlkB:15 contained 0.2 M sodium chloride, 0.1 M HEPES,
pH 7.5, and 25% w/v polyethylene glycol (PEG) 3350; for AlkB:18,
0.2 M ammonium sulfate, 0.1 M tris-hydrochloride, pH 8.5, and 25%
w/v PEG 3350; for AlkB:22, 0.1 M bis-tris, pH 6.5, and 25% w/v PEG
3350. The crystals were cryocooled using well solution diluted to 25%
v/v glycerol and flash cooled in an Oxford Cryosystems nitrogen gas
stream. Data were collected from a single crystal at 100 K using a
Rigaku FR E+ Superbright diffractometer equipped with a copper
rotating anode, Osmic HF optics, and a Saturn 944+ CCD detector.
The data were indexed, integrated, and scaled using HKL2000,38 and
the structure was determined by molecular replacement using the
AutoMR (PHASER)39 subroutine in PHENIX40 with 2FDJ (PDB
ID)25 as a search model for AlkB:15 and 3T4H (PDB ID) as a search
model for AlkB:18 and AlkB:22. Iterative rounds of model building
and refinement using COOT41 and PHENIX41 were performed until
the decreasing R and Rfree no longer converged. The final R factors for
the models were as follows: AlkB:15, R = 15.9% and Rfree = 18.7%;
AlkB:18, R = 16.1% and Rfree = 20.2%; and AlkB:22, R = 18.2% and
Rfree = 22.1%.
AUTHOR INFORMATION
Corresponding Author
*Tel: +44(0)1865 275 625. Fax: +44(0)1865 285 022. E-mail:
Christopher.schofield@chem.ox.ac.uk.
■
Present Address
∥Department of Pharmacy, National University of Singapore,
18 Science Drive 4, Singapore 117543, Singapore.
Author Contributions
⊥These authors contribute equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Pfizer Limited, the Biotechnology and Biological
Sciences Research Council, and the Wellcome Trust for
funding our work.
ABBREVIATIONS USED
■
2OG, 2-oxoglutarate; FTO, fat mass and obesity protein; TET,
Ten-Eleven-Translocation protein; PHF8, PHD finger protein
8; PHD2, hypoxia-inducible factor prolyl hydroxylase 2;
DCMS, dynamic combinatorial mass spectrometry; ESI-MS,
electrospray ionization mass spectrometry; DSF, differential
scanning fluorimetry; Tm, melting temperature; rmsd, root-
mean-square deviation
REFERENCES
■
(1) Wu, S. C.; Zhang, Y. Active DNA demethylation: many roads
lead to Rome. Nat. Rev. Mol. Cell Biol. 2010, 11, 607−620.
(2) Sedgwick, B.; Bates, P. A.; Paik, J.; Jacobs, S. C.; Lindahl, T.
Repair of alkylated DNA: Recent advances. DNA Repair 2007, 6, 429−
442.
(3) Sancar, A. DNA repair in humans. Annu. Rev. Genet. 1995, 29,
69−105.
(4) Vaughan, P.; Sedgwick, B.; Hall, J.; Gannon, J; Lindahl, T.
Environmental mutagens that induce the adaptive response to
alkylating agents in Escherichia coli. Carcinogenesis 1991, 12, 263−268.
(5) Trewick, S. C.; Henshaw, T. F.; Hausinger, R. P.; Lindahl, T.;
Sedgwick, B. Oxidative demethylation by E. coli alkB directly reverts
DNA base damage. Nature 2002, 419, 174−178.
(6) Schofield, C. J.; Zhang, Z. Structural and mechanistic studies on
2-oxoglutarate-dependent oxygenases and related enzymes. Curr. Opin.
Struct. Biol. 1999, 9, 722−731.
(7) Delaney, J. C.; Essigmann., J. M. Mutagenesis, genotoxicity, and
repair of 1-methyladenine, 3-alkylcytosines, 1-methylguanine, and 3-
methylthymine in alkB Escherichia coli. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 14051−14056.
(8) Delaney, J. C.; Smeester, L.; Wong, C.; Frick, L. E.; Taghizadeh,
K.; Wishnok, J. S.; Drennan, C. L.; Samson, L. D.; Essigmann, J. M.
AlkB reverses etheno DNA lesions caused by lipid oxidation in vitro
and in vivo. Nat. Struct. Mol. Biol. 2005, 12, 855−860.
2183
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184

Journal of Medicinal Chemistry
Article
(9) Tsujikawa, K.; Koike, K.; Kitae, K.; Shinkawa, A.; Arima, H.;
Suzuki, T.; Tsuchiya, M.; Makino, M.; Furukawa, T.; Konishi, N.;
Yamamoto, H. Expression and sub-cellular localization of human ABH
family molecules. J. Cell. Mol. Med. 2007, 11, 1105−1116.
(10) Ringvoll, J.; Nordstrand, L. M.; Vagbo, C. B.; Talstad, V.; Reite,
K.; Aas, P. A.; Lauritzen, K. H.; Liabakk, N. B.; Bjork, A.; Doughty, R.
W.; et al. Repair deficient mice reveal mABH2 as the primary oxidative
demethylase for repairing 1meA and 3meC lesions in DNA. EMBO J.
2006, 25, 2189−2198.
(26) Niesen, F. H.; Berglund, H.; Vedadi, M. The use of differential
scanning fluorimetry to detect ligand interactions that promote protein
stability. Nat. Protoc. 2007, 2, 2212−2221.
(27) Karkhanina, A. A.; Mecinovic, J.; Musheev, M. U.; Krylova, S.
M.; Petrov, A. P.; Hewitson, K. S.; Flashman, E.; Schofield, C. J.;
Krylov, S. N. Direct Analysis of Enzyme-Catalyzed DNA Demethy-
lation. Anal. Chem. 2009, 81, 5871−5875.
(28) Krylova, S. M.; Karkhanina, A. A.; Musheev, M. U.; Bagg, E. A.;
Schofield, C. J.; Krylov, S. N. DNA aptamers for as analytical tools for
the quantitative analysis of DNA-dealkylating enzymes. Anal. Biochem.
2011, 414, 261−265.
(29) Rose, N. R; McDonough, M. A.; King, O. N. F.; Kawamura, A.;
Schofield, C. J. Inhibition of 2-oxoglutarate dependent oxygenases.
Chem. Soc. Rev. 2011, 40, 4364−4397.
(11) Born, E. V. D.; Vagbø, C. B.; Songe-Møller, L.; Leihne1, V.;
̊
Lien, G. F.; Leszczynska, G.; Malkiewicz, A.; Krokan, H. E.; Kirpekar,
F.; Klungland, A.; Falnes, P. Ø. ALKBH8-mediated formation of a
novel diastereomeric pair of wobble nucleosides in mammalian tRNA.
Nat. Commun. 2011, 2, article number 172.
(12) Frayling, T. M.; et al. A Common Variant in the FTO Gene Is
Associated with Body Mass Index and Predisposes to Childhood and
Adult Obesity. Science 2007, 316, 889−894.
(13) Gerken, T.; Girard, C. A.; Tung, Y. C.; Webby, C. J.; Saudek, V.;
Hewitson, K. S.; et al. The obesity-associated FTO gene encodes a 2-
oxoglutarate-dependent nucleic acid demethylase. Science 2007, 318,
1469−1472.
(14) Jia, G.; Yang, C.-G.; Yang, S.; Jian, X.; Yi, C.; Zhou, Z.; He, C.
Oxidative demethylation of 3-methylthymine and 3-methyluracil in
single-stranded DNA and RNA by mouse and human FTO. FEBS Lett.
2008, 582, 3313−3319.
(15) Tahiliani, M.; Koh, K. P.; Shen, Y.; Pastor, W. A.; Bandukwala,
H.; Brudno, Y.; Agarwal, S.; Iyer, L. M.; Liu, D. R.; Aravind, L.; Rao, A.
Conversion of 5-Methylcytosine to 5-Hydroxymethylcytosine in
Mammalian DNA by MLL Partner TET1. Science 2009, 324, 930−
935.
(16) Ito, S.; Shen, L.; Dai, Q.; Wu, S. C.; Collins, L. B.; Swenberg, J.
A.; He, C.; Zhang, Y. Tet Proteins Can Convert 5-Methylcytosine to
5-Formylcytosine and 5-Carboxylcytosine. Science DOI: 10.1126/
science.1210597.
(17) Thalhammer, A.; Hansen, A. S.; El-Sagheer, A. H.; Brown, T.;
Schofield, C. J. Hydroxylation of methylated CpG dinucleotides
reverses stabilisation of DNA duplexes by cytosine 5-methylation.
Chem. Commun. 2011, 47, 5325−5327.
(18) Rose, N. R.; Woon, E. C. Y.; Kingham, G. L.; King, O. N. F;
Mecinovic, J.; Clifton, I. J.; Ng, S. S.; Talib-Hardy, J.; Oppermann, U.;
́
McDonough, M. A.; Schofield, C. J. Selective Inhibitors of the JMJD2
Histone Demethylases: Combined Nondenaturing Mass Spectromet-
ric Screening and Crystallographic Approaches. J. Med. Chem. 2010,
53, 1810−1818.
(19) Lienard, B. M. R.; Huting, R.; Lassaux, P.; Galleni, M.; Frere, J.-
́ ̀
̈
M.; Schofield, C. J. Dynamic combinatorial mass spectrometry leads to
(30) Yu, B.; Hunt, J. F. Enzymological and structural studies of the
mechanism of promiscuous substrate recognition by the oxidative
DNA repair enzyme AlkB. Proc. Natl. Acad. Sci. U.S.A. 2009, 106,
14315−14320.
(31) Loenarz, C.; Schofield, C. J. Expanding chemical biology of 2-
oxoglutarate oxygenases. Nat. Chem. Biol. 2008, 4, 152−156.
(32) Schofield, C. J.; Ratcliffe, P. J. Oxygen sensing by HIF
hydroxylases. Nat. Rev. Mol. Cell Biol. 2004, 5, 343−354.
(33) Loenarz, C.; Ge, W.; Coleman, M. L.; Rose, N. R.; Cooper, C.
D. O.; Klose, R. J.; Ratcliffe, P. J.; Schofield, C. J. PHF8, a gene
associated with cleft lip/palate and mental retardation, encodes for an
Nε-dimethyl lysine demethylase. Hum. Mol. Genet. 2010, 19, 217−222.
(34) Weidmann, K.; Baringhaus, K.-H.; Tschank, G. Bickel, 3-
hydroxypyridine-2-carboxamidoesters, their preparation and their use
as pharmaceuticals. M. Eur. Pat. Appl. EP 765871 A1 19970402, 1997.
(35) Davis, J. W. Jr. Studies with quinolines. I. Synthesis of quinaldic
acid and some of its amide derivatives. J. Org. Chem. 1959, 24, 1691−
1694.
(36) McDonough, M. A.; Li, V.; Flashman, E.; Chowdhury, R.; Mohr,
C.; Lienard, B. M. R.; Zondlo, J.; Oldham, N. J.; Clifton, I. J.; Lewis, J.;
McNeill, L. A.; Kurzeja, R. J. M.; Hewitson, K. S.; Yang, E.; Jordan, S.;
Syed, R. S.; Schofield, C. J. Cellular oxygen sensing: Crystal structure
of hypoxia-inducible factor prolyl hydroxylase (PHD2). Proc. Natl.
Acad. Sci. U.S.A. 2006, 103, 9814−9819.
(37) McNeill, L. A.; Flashman, E.; Buck, M. R. G.; Hewitson, K. S.;
Clifton, I. J.; Jeschke, G.; Claridge, T. D. W.; Ehrismann, D.; Oldham,
N. J.; Schofield, C. J. Hypoxia-inducible factor prolyl hydroxylase 2 has
a high affinity for ferrous iron and 2-oxoglutarate. Mol BioSyst. 2005, 1,
321−324.
(38) Otwinowski, Z.; Minor, W. In Methods in Enzymology; Elsevier:
1997; Vol. 276, pp 307−326.
(39) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M.
D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software. J. Appl.
Crystallogr. 2007, 40, 658−674.
metallo-beta-lactamase inhibitors. J. Med. Chem. 2008, 51, 684−688.
(20) Lienard, B. M. R.; Selevsek, N.; Oldham, N. J.; Schofield, C. J.
́
́
(40) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis,
I. W.; Echols, N.; Headd, J. J.; Hung, L.-W.; Kapral, G. J.; Grosse-
Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.
J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H.
PHENIX: A comprehensive Python-based system for macromolecular
structure solution. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66,
213−221.
(41) Emsley, P.; Cowtan, K. Coot: Model-building tools for
molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126−2132.
Combined mass spectrometry and dynamic chemistry approach to
identify metalloenzyme inhibitors. ChemMedChem 2007, 2, 175−179.
(21) Poulsen, S.-A. Direct screening of a dynamic combinatorial
library using mass spectrometry. J. Am. Soc. Mass Spectrom. 2006, 17,
1074−1080.
(22) Daniel, J. M.; Friess, S. D; Rajagopalan, S.; Wendt, S.; Zenobi, R.
Quantitative determination of noncovalent binding interactions using
soft ionization mass spectrometry. Int. J. Mass Spectrom. 2002, 216, 1−
27.
(23) Welford, R. W. D.; Schlemminger, I.; McNeill, L. A.; Hewitson,
K. S.; Schofield, C. J. The Selectivity and Inhibition of AlkB. J. Biol.
Chem. 2003, 278, 10157−10161.
(24) McDonough, M. A.; McNeill, L. A.; Tilliet, M.; Papamicael, C.
A.; Chen, Q. Y.; Banerji, B.; Hewitson, K. S.; Schofield, C. J. Selective
inhibition of factor inhibiting hypoxia-inducible factor. J. Am. Chem.
Soc. 2005, 127, 7680−7681.
(25) Yu, B.; Edstrom, W. C.; Benach, J.; Hamuro, Y.; Weber, P. C.;
Gibney, B. R.; Hunt, J. F. Crystal structures of catalytic complexes of
the oxidative DNA/RNA repair enzyme AlkB. Nature 2006, 439, 879−
884.
2184
dx.doi.org/10.1021/jm201417e | J. Med. Chem. 2012, 55, 2173−2184