Synthesis and SAR of 1-hydroxy-1 H -benzo[ d ]imidazol-2(3 H)-ones as inhibitors of d -amino acid oxidase

Article abstract of DOI:10.1021/ml300212a

A series of 1-hydroxy-1H-benzo[d]imidazol-2(3H)-ones were synthesized and evaluated for their ability to inhibit human and porcine forms of d-amino acid oxidase (DAAO). The inhibitory potency is largely dependent on the size and position of substituents o

Full text of DOI:10.1021/ml300212a

Letter
pubs.acs.org/acsmedchemlett
Synthesis and SAR of 1‑Hydroxy‑1H‑benzo[d]imidazol-2(3H)‑ones as
Inhibitors of D‑Amino Acid Oxidase
James F. Berry,† Dana V. Ferraris,‡ Bridget Duvall,‡ Niyada Hin,‡ Rana Rais,† Jesse Alt,‡ Ajit G. Thomas,‡
Camilo Rojas,‡ Kenji Hashimoto,§ Barbara S. Slusher,†,‡ and Takashi Tsukamoto*,†,‡
‡
†Department of Neurology and Brain Science Institute, Johns Hopkins University, Baltimore, Maryland 21205, United States
§Division of Clinical Neuroscience, Chiba University Center for Forensic Mental Health, 1-8-1 Inohana, Chiba 260-8670, Japan
S
* Supporting Information
ABSTRACT: A series of 1-hydroxy-1H-benzo[d]imidazol-2(3H)-ones
were synthesized and evaluated for their ability to inhibit human and
porcine forms of D-amino acid oxidase (DAAO). The inhibitory
potency is largely dependent on the size and position of substituents on
the benzene ring with IC50 values of the compounds ranging from 70
nM to greater than 100 μM. Structure−activity relationships of this new
class of DAAO inhibitors will be presented in detail along with comparisons to previously published SAR data from other classes
of DAAO inhibitors. Two of these compounds were given to mice orally together with D-serine to assess their effects on plasma
D-serine pharmacokinetics.
KEYWORDS: D-amino acid oxidase (DAAO), D-serine, NMDA receptors, glucuronidation
umulative evidence suggests that allosteric activation of
the N-methyl-D-aspartate (NMDA) receptor provides a
C
new therapeutic avenue for the treatment of schizophrenia.1
NMDA receptor glycine modulatory site agonists (glycine, D-
serine, and D-alanine) were reported to improve not only
positive symptoms but also negative symptoms and cognitive
impairments where existing antipsychotics have failed to show
significant efficacy.2 D-Serine is particularly attractive since it is
more permeable than glycine to the blood−brain barrier.3
Furthermore, because there is no known signal transduction
site modulated by the low concentrations of D-serine other than
the glycine modulatory site, D-serine can facilitate NMDA
receptor function without affecting other central nervous
system receptors. Clinical development of D-serine, however,
is hampered by the high doses of D-serine (60−120 mg/kg)
required for optimal efficacy,4 which may not be tolerated due
to potential nephrotoxic effects.
Figure 1. Representative inhibitors of DAAO.
was dosed in conjunction with 1. The potency of D-serine was
significantly improved in a preclinical model of schizophrenia
when it was coadministered with 1.9 Thus, coadministration of
D-serine and a DAAO inhibitor may offer a new type of
treatment for schizophrenia complementary to existing
antipsychotics, which primarily target dopamine receptors.
In addition to 1, several structurally distinct DAAO inhibitors
have been indentified to date.10 These include 5-methylpyr-
azole-3-carboxylic acid 2,11 4H-furo[3,2-b]pyrrole-5-carboxylic
acid 3,12 4H-thieno[3,2-b]pyrrole-5-carboxylic acid 4,13 and 3-
hydroxyquinolin-2(1H)-one 5 (Figure 1).14 The discovery of 5
is of particular interest from the viewpoint of exploring new
pharmacophores for DAAO inhibitors as it presents possibilities
for other chemical moieties to serve as effective carboxylic acid
surrogates. The crystal structure of 5 bound to human DAAO
Two independent studies suggest that the underlying cause
of D-serine-induced nephrotoxicity can be attributed to
hydrogen peroxide generated by D-amino acid oxidase
(DAAO)-mediated metabolism of D-serine in the kidneys.5,6
DAAO is a flavoenzyme that catalyzes the oxidation of D-amino
acids to the corresponding imino acids and hydrogen peroxide
and is predominantly responsible for clearance of orally
administered D-serine in mice.7 Coadministration of D-serine
and a small molecule DAAO inhibitor may mitigate the
problem related to D-serine clinical use by reducing DAAO-
mediated metabolism of D-serine, thus enabling a dose
reduction as well as prevention of nephrotoxicity.
Indeed, one of the early DAAO inhibitors, 6-chlorobenzo-
[d]isoxazol-3-ol (CBIO) 1 (Figure 1), was found to improve
the oral bioavailability of D-serine in rodents.8 In addition,
enhanced brain levels of D-serine were achieved when D-serine
Received: July 24, 2012
Accepted: September 16, 2012
Published: September 16, 2012
© 2012 American Chemical Society
839
dx.doi.org/10.1021/ml300212a | ACS Med. Chem. Lett. 2012, 3, 839−843

ACS Medicinal Chemistry Letters
Letter
highlights some key hydrogen-bonding interactions (Figure
2A).14 The 3-hydroxyl group is involved in two hydrogen
contribute to the high DAAO inhibitory potency of 6a, we
systematically modified particular functional groups of this
molecule and synthesized compounds 7−15 (Figure 3).
Displacement of the 1-hydroxyl group with a hydrogen atom
(compound 7) or amino group (compound 8) led to a
complete loss of activity. Methylation of the 1-hydroxyl group
(compound 9) also resulted in a complete loss of activity. This
is most likely caused by loss of hydrogen-bonding interactions
with Arg283 and Tyr228. These residues represent a key
component of the DAAO active site by interacting with a
carboxylate group of D-amino acid substrates. All known
competitive DAAO inhibitors take advantage of hydrogen-
bonding interactions with these two residues.
Methylation of the 3-position nitrogen (compound 10) as
well as displacement with a methylene group (compound 11)
caused a complete loss of potency. The 3-position nitrogen
group of 6a presumably forms a hydrogen bond with the
backbone carbonyl of Gly313 in a similar manner as the amino
group of 5 (Figure 2B). This explains the lack of potency in 10
and 11, which do not have the ability to form hydrogen-
bonding interactions with Gly313.
Figure 2. (A) Compound 5 (green) bound to the active site of DAAO
(3G3E). (B) Proposed binding mode of 6a (white) to the active site of
DAAO. Hydrogen-bonding interactions are shown as yellow dashed
lines.
We also synthesized three pyridine derivatives (compounds
13−15) to assess the effect of a nitrogen atom in the benzene
ring on the binding affinity to DAAO. One of these molecules,
compound 13, was slightly less potent than 6a, while the other
two derivatives were much weaker DAAO inhibitors.
Preliminary SAR indicates that the cyclic N-hydroxyurea is an
essential element for potent inhibition of DAAO. Furthermore,
the pyridinyl nitrogen appears to play only a small role in
enhancing potency. Hence, subsequent SAR studies were
directed toward a wide variety of 1-hydroxy-1H-benzo[d]-
imidazol-2(3H)-ones in which various substituents were
introduced onto the benzene ring at the 4-, 5-, 6-, and/or 7-
positions.
bonds, one with Tyr228 and the other with Arg283. Additional
interactions with the Arg283 residue are made by the 2-
carbonyl group, while the lactam NH donates a hydrogen bond
to the backbone carbonyl of Gly313. We hypothesized that the
lactam component of 5 can be replaced by other moieties as
long as these key interactions are preserved. Along this line, we
explored a variety of functional groups as alternatives to this
component and found that a cyclic N-hydroxyurea was an
effective ring system to constitute potent DAAO inhibitory
activity. Here, we disclose the synthesis and structure−activity
relationship studies of 1-hydroxy-1H-benzo[d]imidazol-2(3H)-
ones 6 as novel small molecule DAAO inhibitors (Figure 1).
As shown in Figure 3, 1-hydroxy-1H-benzo[d]imidazol-
2(3H)-one 6a inhibited a human form of DAAO with an
IC50 value of 0.6 μM. To identify structural features that
The general synthetic route to 1-hydroxy-1H-benzo[d]-
imidazol-2(3H)-ones is illustrated in Scheme 1. A key
intermediate 19 was prepared either from phenylisocyanate
16 or aniline 17.15 Oxidative ring closure of 19 was carried out
with lead tetraacetate,16 providing the bicyclic core structure 20.
Scheme 1. Synthesis of 1-Hydroxy-1H-benzo[d]imidazol-
a
2(3H)-ones
a
Reagents and conditions: (a) BnONH2, DCM, rt, 80−100%. (b)
Et3N, DCM, rt, 64−84%. (c) Pb(OAc)4, CHCl3, rt, 9−83%. (d) H2,
10% Pd/C, EtOH, 13−90% or HBr, AcOH, reflux, 28%.
Figure 3. Preliminary SAR studies of analogues of 6a.
840
dx.doi.org/10.1021/ml300212a | ACS Med. Chem. Lett. 2012, 3, 839−843

ACS Medicinal Chemistry Letters
Letter
Subsequent removal of the benzyl protecting group by catalytic
hydrogenation afforded the desired 1-hydroxy-1H-benzo[d]-
imidazol-2(3H)-ones 6.
Inhibitory potencies of the 1-hydroxy-1H-benzo[d]imidazol-
2(3H)-ones were measured in both human and porcine forms
of DAAO using previously reported assay methods.8 Using this
assay, compound 5 was found to inhibit human and porcine
DAAO with IC50 values of 0.02 and 0.08 μM, respectively.
Table 1 summarizes the in vitro DAAO inhibitory data of the 1-
hydroxy-1H-benzo[d]imidazol-2(3H)-ones 6.
superior to 6. In contrast, bulkier substitutions (6f−j) resulted
in significant loss of inhibitory potency. While some of the 5- or
6-substituted analogues were found to be potent DAAO
inhibitors, all of the 7-substituted compounds exhibited weaker
inhibitory potency. Disubstitution at the 4-, 5-, and/or 6-
positions was well tolerated as exemplified by compounds 6l
and 6o. Among those tested, four compounds (6b, 6e, 6l, and
6o) exhibited potent DAAO inhibitory activity with IC50 values
equal to or less than 100 nM.
The overall SAR trends are largely consistent with those of
analogues derived from 1 and 5. In general, bulky substituents
on the aryl ring of 6 are not tolerated at the active site of
DAAO. This is consistent with the findings from the
crystallographic studies indicating the limited space available
within the substrate binding site, which is parallel to the flavin
ring. A more thorough SAR analysis revealed that substitution
patterns preferred for 1-hydroxy-1H-benzo[d]imidazol-2(3H)-
ones14 better correlate with those of 3-hydroxyquinolin-2(1H)-
ones than those of benzo[d]isoxazol-3-ols.8 For example,
substitution at the corresponding 4-position of the benzo[d]-
isoxazol-3-ol series by a fluorine or methyl group resulted in a
substantial decline in the inhibitory potency as opposed to
compounds 6b and 6c, displaying improved potency over 6a.
Although speculative, it is conceivable that a similar binding
mode is in effect for 1-hydroxy-1H-benzo[d]imidazol-2(3H)-
ones and 3-hydroxyquinolin-2(1H)-ones, which share the
common α-hydroxycarbonyl moiety. However, none of the
hydroxy-1H-benzo[d]imidazol-2(3H)-ones tested was as po-
tent as compound 5. This could be due to the optimal
interaction of compound 5 with the carboxylate binding site of
DAAO, which cannot be reproduced by the N-hydroxylurea
moiety of 6.
It was previously reported that some DAAO inhibitors are
more potent against the human form of DAAO than the rat
form.14 In contrast, as shown in Table 1, there was little
difference in potency when 1-hydroxy-1H-benzo[d]imidazol-
2(3H)-ones were tested against human and porcine forms of
the enzyme. It should be noted that weaker binding to the rat
enzyme might be attributed to the conformational changes of
rat DAAO due to the deletion of Leu27.14 Interestingly, the
porcine enzyme possesses the Leu27 residue, and this may be
the reason for similar potency seen in both human and porcine
forms of DAAO.
To determine the effects of the N-hydroxyureas on D-serine
plasma levels, mice (n = 3 per time point) were dosed with
compounds 6e or 6o (30 mg/kg, po) along with D-serine (30
mg/kg, po) with 10% DMSO/90% saline as a vehicle. As
shown in Figure 4, compound 6e showed no effect on plasma
D-serine pharmacokinetics. Indeed, subsequent studies revealed
that compound 6e has negligible oral bioavailability in mice
(data not shown). When D-serine was coadministered with
compound 6o, plasma levels of D-serine were higher as
compared to those of treated with D-serine alone (*p < 0.05
at 30 min after administration). The pharmaco-enhancing
effects of 6o, however, only lasted for a short period of time (<2
h), presumably due to its poor oral bioavailability.
In an attempt to elucidate the mechanism behind the poor
oral bioavailability of 6e and 6o, their metabolic stability was
measured in mouse plasma and liver microsomes (Table 2).
Both compounds showed high stability in mouse plasma. In
liver microsomes, compound 6e was substantially metabolized
in the presence of NADPH. In contrast, compound 6o
appeared to be resistant to NADPH-dependent CYP450-
Table 1. Inhibition of DAAO by 1-Hydroxy-1H-
benzo[d]imidazol-2(3H)-ones
a
a
IC50 (μM)
cmpd
R4
R5
R6
R7
human
porcine
6a
6b
6c
6d
6e
6f
H
F
H
H
H
0.6
0.1
0.2
80
1
0.5
0.4
20
0.4
50
50
10
30
40
1
H
H
H
H
H
H
H
H
H
CH3
F
H
H
H
H
H
H
H
H
H
H
H
H
CH3
F
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
F
CH3
CH2CH3
OCH3
0.1
10
CH2Ph
6g
6h
6i
OCH2CH3
50
i-Pr
20
n-Pr
20
6j
sec-butyl
>100
20
6k
6l
CH3
CH3
0.07
0.8
20
0.7
0.2
4
6m
6n
6o
6p
6q
6r
OCH2−CH2
CH3
H
F
H
0.08
0.9
4
0.4
1
H
H
H
H
H
H
H
H
H
H
H
H
H
F
H
H
F
CH3
F
4
70
30
1
6s
CH3
F
2
6t
CH2−CH2
30
10
0.6
2
6u
6v
6w
6x
6y
6z
6aa
6ab
H
H
H
H
H
H
H
H
F
0.4
9
Cl
CH3
F
30
7
8
20
>100
>100
20
>100
CH2−CH2
F
>100
>100
3
CH3
H
H
F
CH3
40
a
Assay methods are described in the Supporting Information.
When each of the four aromatic hydrogen atoms was
substituted by a fluorine atom, the resulting compounds (6b,
6p, 6u, or 6aa) exhibited no significant loss of potency except
for compound 6aa. When each of the four aromatic hydrogen
atoms was substituted by a methyl group, it was only tolerated
at the 4-position (6c), and the rest of the compounds (6q, 6w,
and 6ab) exhibited significantly higher IC50 values. In general,
smaller substituents are more preferred at all four positions,
although the 4-position appears more tolerant of slightly larger
substituents. Indeed, fluorine-, methyl-, and methoxy-substi-
tuted analogues (6b, 6c, and 6e) exhibited inhibitory potency
841
dx.doi.org/10.1021/ml300212a | ACS Med. Chem. Lett. 2012, 3, 839−843

ACS Medicinal Chemistry Letters
Letter
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthesis, characterization, and methods of DAAO assay, D-
serine pharmacokinetics, and metabolic stability studies. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Tel: 410-614-0982. E-mail: ttsukamoto@jhmi.edu.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Figure 4. Effects of DAAO inhibitors on plasma D-serine levels
following oral coadministration.
We thank Yuko Fujita (Chiba University) for her technical
assistance in the measurement of plasma D-serine levels. This
work was in part supported by National Institutes of Health
(R01MH091387 to T.T.) and the Johns Hopkins Brain Science
Institute NeuroTranslational Drug Discovery program.
Table 2. Metabolic Stability of Compounds 6e and 6o
compd
matrix
plasma
cofactor
% remaining (after 1 h)
6e
none
90
ABBREVIATIONS
DAAO, D-amino acid oxidase; NMDA, N-methyl-D-aspartate;
CBIO, 6-chlorobenzo[d]isoxazol-3-ol
■
microsomes
microsomes
microsomes
plasma
none
∼100
60
NADPH
UDPGA
none
a
ND
6o
∼100
∼100
∼100
0.6
REFERENCES
microsomes
microsomes
microsomes
none
■
NADPH
UDPGA
(1) Moghaddam, B.; Javitt, D. From revolution to evolution: The
glutamate hypothesis of schizophrenia and its implication for
treatment. Neuropsychopharmacology 2012, 37, 4−15.
(2) Javitt, D. C. Glutamatergic theories of schizophrenia. Isr. J
Psychiatry Relat. Sci. 2010, 47, 4−16.
(3) Oldendorf, W. H. Brain uptake of radiolabeled amino acids,
amines, and hexoses after arterial injection. Am. J. Physiol. 1971, 221,
1629−1639.
a
Not detected.
mediated oxidation, presumably due to the electron-with-
drawing effect of the two fluorine atoms. This may explain the
ability of compound 6o to slightly enhance D-serine levels in
vivo. Interestingly, both compounds were completely con-
sumed in the presence of UDPGA. The results clearly show
that both compounds undergo phase II glucuronidation in the
liver. The most likely site for glucuronidation in these
compounds is the 1-hydroxyl group. Unfortunately, as evident
from the loss of potency in 7 and 9, this hydroxyl group is
essential for high affinity binding to the DAAO active site.
Further structural optimization of the 1-hydroxy-1H-benzo[d]-
imidazol-2(3H)-one scaffold will require fine-tuning of both the
steric and the electronic environment surrounding the 1-
hydroxyl group in a way that does not compromise the
inhibitory potency while minimizing the degree of glucur-
onidation.
In conclusion, we have identified a new series of DAAO
inhibitors based on 1-hydroxy-1H-benzo[d]imidazol-2(3H)-
one scaffold. Through systematic SAR studies, we identified
potent DAAO inhibitors including compounds 6e and 6o.
While 6o showed marginal effects on plasma D-serine
concentrations, no effects were observed with 6e. Subsequent
studies revealed that these compounds were subjected to phase
I and II metabolism. The crystal structure of DAAO in complex
with imino-DOPA revealed that its catechol moiety takes a
position nearly perpendicular to the flavin ring.17 This
hydrophobic cavity, distinct from the active site, has not been
utilized by most DAAO inhibitors and could serve as an
additional binding pocket to enhance the affinity to DAAO.
Further structural optimization, however, should be attentively
directed toward improving metabolic stability while enhancing
potency.
(4) Kantrowitz, J. T.; Malhotra, A. K.; Cornblatt, B.; Silipo, G.; Balla,
A.; Suckow, R. F.; D'Souza, C.; Saksa, J.; Woods, S. W.; Javitt, D. C.
High dose D-serine in the treatment of schizophrenia. Schizophr. Res.
2010, 121, 125−130.
(5) Williams, R. E.; Lock, E. A. Sodium benzoate attenuates D-serine
induced nephrotoxicity in the rat. Toxicology 2005, 207, 35−48.
(6) Maekawa, M.; Okamura, T.; Kasai, N.; Hori, Y.; Summer, K. H.;
Konno, R. D-Amino-acid oxidase is involved in D-serine-induced
nephrotoxicity. Chem. Res. Toxicol. 2005, 18, 1678−1682.
(7) Rais, R.; Thomas, A. G.; Wozniak, K.; Wu, Y.; Jaaro-Peled, H.;
Sawa, A.; Strick, C. A.; Engle, S. J.; Brandon, N. J.; Rojas, C.; Slusher,
B. S.; Tsukamoto, T. Pharmacokinetics of Oral D-Serine in D-Amino
Acid Oxidase Knockout Mice. Drug Metab. Dispos. Published online
July 26, 2012; DOI: 10.1124/dmd.112.046482.
(8) Ferraris, D.; Duvall, B.; Ko, Y. S.; Thomas, A. G.; Rojas, C.;
Majer, P.; Hashimoto, K.; Tsukamoto, T. Synthesis and biological
evaluation of D-amino acid oxidase inhibitors. J. Med. Chem. 2008, 51,
3357−3359.
(9) Hashimoto, K.; Fujita, Y.; Horio, M.; Kunitachi, S.; Iyo, M.;
Ferraris, D.; Tsukamoto, T. Co-administration of a D-amino acid
oxidase inhibitor potentiates the efficacy of D-serine in attenuating
prepulse inhibition deficits after administration of dizocilpine. Biol.
Psychiatry 2009, 65, 1103−1106.
(10) Ferraris, D. V.; Tsukamoto, T. Recent advances in the discovery
of D-amino acid oxidase inhibitors and their therapeutic utility in
schizophrenia. Curr. Pharm. Des. 2011, 17, 103−111.
(11) Adage, T.; Trillat, A. C.; Quattropani, A.; Perrin, D.; Cavarec, L.;
Shaw, J.; Guerassimenko, O.; Giachetti, C.; Greco, B.; Chumakov, I.;
Halazy, S.; Roach, A.; Zaratin, P. In vitro and in vivo pharmacological
profile of AS057278, a selective D-amino acid oxidase inhibitor with
potential anti-psychotic properties. Eur. Neuropsychopharmacol. 2008,
18, 200−214.
842
dx.doi.org/10.1021/ml300212a | ACS Med. Chem. Lett. 2012, 3, 839−843

ACS Medicinal Chemistry Letters
Letter
(12) Sparey, T.; Abeywickrema, P.; Almond, S.; Brandon, N.; Byrne,
N.; Campbell, A.; Hutson, P. H.; Jacobson, M.; Jones, B.; Munshi, S.;
Pascarella, D.; Pike, A.; Prasad, G. S.; Sachs, N.; Sakatis, M.; Sardana,
V.; Venkatraman, S.; Young, M. B. The discovery of fused pyrrole
carboxylic acids as novel, potent D-amino acid oxidase (DAO)
inhibitors. Bioorg. Med. Chem. Lett. 2008, 18, 3386−3391.
(13) Smith, S. M.; Uslaner, J. M.; Yao, L.; Mullins, C. M.; Surles, N.
O.; Huszar, S. L.; McNaughton, C. H.; Pascarella, D. M.; Kandebo, M.;
Hinchliffe, R. M.; Sparey, T.; Brandon, N. J.; Jones, B.; Venkatraman,
S.; Young, M. B.; Sachs, N.; Jacobson, M. A.; Hutson, P. H. The
behavioral and neurochemical effects of a novel D-amino acid oxidase
inhibitor compound 8 [4H-thieno [3,2-b]pyrrole-5-carboxylic acid]
and D-serine. J. Pharmacol. Exp. Ther. 2009, 328, 921−930.
(14) Duplantier, A. J.; Becker, S. L.; Bohanon, M. J.; Borzilleri, K. A.;
Chrunyk, B. A.; Downs, J. T.; Hu, L. Y.; El-Kattan, A.; James, L. C.;
Liu, S.; Lu, J.; Maklad, N.; Mansour, M. N.; Mente, S.; Piotrowski, M.
A.; Sakya, S. M.; Sheehan, S.; Steyn, S. J.; Strick, C. A.; Williams, V. A.;
Zhang, L. Discovery, SAR, and pharmacokinetics of a novel 3-
hydroxyquinolin-2(1H)-one series of potent D-amino acid oxidase
(DAAO) inhibitors. J. Med. Chem. 2009, 52, 3576−3585.
(15) Parrish, D. A.; Zou, Z.; Allen, C. L.; Day, C. S.; King, S. B. A
convenient method for the synthesis of N-hydroxyureas. Tetrahedron
Lett. 2005, 46, 8841−8843.
(16) Cooley, J. H.; Jacobs, P. T. Oxidative ring closure of 1-
benzyloxy-3-arylureas to 1-benzyloxybenzimidazolones. J. Org. Chem.
1975, 40, 552−557.
(17) Kawazoe, T.; Tsuge, H.; Imagawa, T.; Aki, K.; Kuramitsu, S.;
Fukui, K. Structural basis of D-DOPA oxidation by D-amino acid
oxidase: Alternative pathway for dopamine biosynthesis. Biochem.
Biophys. Res. Commun. 2007, 355, 385−391.
843
dx.doi.org/10.1021/ml300212a | ACS Med. Chem. Lett. 2012, 3, 839−843