Structure-guided design of substituted aza-benzimidazoles as potent hypoxia inducible factor-1α prolyl hydroxylase-2 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.08.012

We report the structure-based design and synthesis of a novel series of aza-benzimidazoles as PHD2 inhibitors. These efforts resulted in compound 22, which displayed highly potent inhibition of PHD2 function in vitro.

Full text of DOI:10.1016/j.bmcl.2008.08.012

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5023–5026
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure-guided design of substituted aza-benzimidazoles as potent
hypoxia inducible factor-1a prolyl hydroxylase-2 inhibitors
a
a
a
a
Mike Frohn a, , Vellarkad Viswanadhan , Alexander J. Pickrell , Jennifer E. Golden , Kristine M. Muller ,
Roland W. Bürli a, Gloria Biddlecome b, Sean C. Yoder b, Norma Rogers b, Jennifer H. Dao a, Randall Hungate a,
Jennifer R. Allen a
*
a Department of Chemistry Research and Discovery, Amgen, Inc., One Amgen Center Drive, Thousand Oaks, CA 91320, USA
b Department of Hematology, Amgen, Inc., One Amgen Center Drive, Thousand Oaks, CA 91320, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We report the structure-based design and synthesis of a novel series of aza-benzimidazoles as PHD2
inhibitors. These efforts resulted in compound 22, which displayed highly potent inhibition of PHD2 func-
tion in vitro.
Received 12 June 2008
Revised 1 August 2008
Accepted 5 August 2008
Available online 9 August 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
PHD-2
HIF-1a
Prolyl hydroxylase
Hypoxia
Normoxia
Aza-benzimidazole
Higher eukaryotes have developed sophisticated mechanisms
to enhance uptake and distribution of oxygen in response to hy-
poxia (below normal cellular pO2). A central component of this
process is oxygen-regulated gene activation mediated by hypoxia
inducible transcription factor (HIF).1 HIF is a heterodimer com-
to respond to changing oxygen levels. HIF prolyl hydroxylases
(PHD-1, -2, and -3) are part of a superfamily of non-heme, Fe(II),
and 2-oxoglutarate dependent dioxygenases. The PHD enzymes
are exquisite oxygen sensors that are responsible for HIF-
a stabil-
ity via O2-dependent proline hydroxylation. HIF-1 , for instance,
a
posed of HIF-
for the HIF-
in mammalian cells but only HIF-1
binding domains essential for transcriptional activity. While HIF-
and HIF-b are constitutively expressed genes, only HIF- protein
levels are regulated by changes in cellular oxygen tension. Under
hypoxic conditions, HIF-1 protein is relatively stable, enabling
dimerization with HIF-b and subsequent transcription of genes
that contain the hypoxia-responsive element (HRE). These genes
encode proteins important for erythropoiesis and angiogenesis,
as well as various glycolytic enzymes. Together, the regulated
expression of these proteins represents a systemic response to re-
duced oxygen availability.
a
and HIF-b subunits. Three separate genes encoding
subunit (HIF-1 , -2 , and -3 ) have been discovered
and -2 contain the DNA
becomes hydroxylated on prolines 402 and 564 under normoxia,
creating recognition sites for the Von Hippel-Lindau protein com-
a
a
a
a
a
a
plex, which then targets HIF-a for poly-ubiquitination and subse-
quent proteosomal degradation. As hypoxia ensues, PHD
a
a
enzymes lack adequate O2 for catalysis, and HIF-a proline hydrox-
ylation is impaired.
a
Since HIF stabilization under hypoxia results in expression of
genes with possible therapeutic utility, HIF-1 stabilization during
a
normoxia offers an attractive strategy for the treatment of anemia
and ischemic diseases.2 We recently became interested in achiev-
ing this goal through the small-molecule inhibition of PHD2, the
most broadly expressed isoform and the one most clearly associ-
ated with a role in erythropoiesis.3 The present discussion de-
scribes one effort toward rational design of PHD2 inhibitors
using a structure-guided approach.
Interestingly, during normoxia HIF-a protein is continually syn-
thesized and then degraded — a seemingly wasteful process, but
one which allows this important regulator to be readily available
The X-ray crystal structure of isoquinoline 1 bound to PHD2 has
recently been reported, which provided the starting point for struc-
ture-based design efforts.4 A schematic representation of critical
polar interactions between inhibitor and PHD2 active site residues
is shown in Figure 1. The salt bridge between the carboxylate moi-
* Corresponding author. Tel.: +1 805 447 2761, +1 805 551 4632; fax: +1 805 480
1337, +1 805 498 0882.
E-mail address: mfrohn@amgen.com (M. Frohn).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.08.012

5024
M. Frohn et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5023–5026
A, Fig. 1). This structure retains the carboxyl group and Fe(II) bind-
O
Fe(II)
N
OH
ing element of 1, but replaces the hydroxyl with an imidazole
nitrogen. Alkyl substitution(s) at the indicated positions would
provide access to hydrophobic areas within the active site. To sup-
port the design computationally, a minimally substituted aza-
benzimidazole (R1 = R2 = R3 = H, 14) was optimized by molecular
mechanics. Overlay of the optimized ground state conformation
of this structure with the binding conformation of 1 (as observed
in the co-crystal structure), followed by docking of the structure
into the active site of PHD2 crystal structure and energetic evalua-
tion showed the potential of this new chemotype.
Importantly, a more detailed molecular modeling analysis of
this compound revealed the possibility of two distinct binding
modes, each possessing a different orientation of the bicyclic sys-
tem relative to the amide carbonyl (Fig. 2). The accessibility of
the two binding modes appears to be motivated by Fe(II) coordina-
tion, with each mode employing a different ring nitrogen for this
function. The structure in Figure 2(a) relies on an imidazole nitro-
gen, while that in Figure 2(b) coordinates through the pyridine
nitrogen similar to isoquinoline 1. Quantum mechanical calcula-
tions with the template N-methyl-1H-imidazo[4,5-c]pyridine-4-
carboxamide showed that the minima corresponding to the two
N
H
O
O
O
I
N
O
O
OH
N
H
I
O
H
NH
1
O
H
Arg 383
N
H
N
H
O
R3
N
OH
N
H
N
Tyr 303
N
R1
R2
A
Figure 1. Polar interactions between 1 and PHD2. General structure of designed
aza-benzimidazole scaffold.
modes are energetically very close, with
a mild preference
(ꢀ0.6 kcal) in the gas phase for the conformation shown in Figure
2(a). This energy difference is consistent with molecular mechanics
calculations. With this information in hand, a structure–activity
relationship (SAR) study was initiated.
Compounds for this study were prepared according to Schemes
1–5. The aza-benzimidazole core was typically constructed by a
three-step sequence beginning with 4-chloro-3-nitropyridine 2.
SNAr displacement of the chloride with the requisite amine,6
reduction of the nitro group, and cyclization by heating in the pres-
ence of a carboxylic acid resulted in the bicyclic core structure
(general structure I, Scheme 1). Synthesis of the C-4 side chain be-
gan with the regioselective installation of a nitrile group (general
structure III), which proceeded by formation of the pyridine
N-oxide followed by a modified Reissert–Henze reaction.7 None
of the regioisomeric nitrile was isolated from these reactions.8
Figure 2. Optimized binding modes of 14 in the active site of PHD2. Fe(II) is shown
in pink.
ety of 1 and the guanidino side chain of Arg383, as well as Fe(II)
coordination through the isoquinoline nitrogen and amide car-
bonyl functionalities, appears to be crucial for binding. In addition,
Tyr303 serves as an H-bond donor to the isoquinoline hydroxyl.5
The aza-benzimidazole scaffold was designed with the goal of
discovering structurally diverse PHD2 inhibitors (general structure
R2
N
R2
N
R2
R2
N
3
NO2
O
N
N
N
N
R1
R1
R1
R1
4 CN (f), (g), (h)
N
1
Cl
OH
(d)
(e)
(a), (b), (c)
N
H
N
O
N
N
N
N
O
2
I
III
14-20
II
Scheme 1. Synthesis of compounds 14–20. Reagents and conditions: (a) R1NH2, concd HCl, H2O, 2-methoxyethanol, reflux or R1NH2, EtOH, rt, 72–99%. (b) Pd-C, H2, MeOH,
95–99%. (c) R2CO2H, 80–120 °C, 78–96%. (d) m-CPBA, CH2Cl2, rt, 56–89%. (e) N,N-dimethylcarbamoyl chloride, TMSCN, CH2Cl2, rt, 63–91%. (f) concd HCl or 60% H2SO4:dioxane
(1:1, v/v), 80–95 °C, 23–95%. (g) H2NCH2CO2EtÁHCl, H2NCH2CO2BnÁHCl, or H2NCH2CO2tBuÁHCl, HBTU, NEt3 or (iPr)2NEt, DMF, rt, 47–89%. (h) NaOH (aq), THF or Pd-C, H2, MeOH
or TFA, 9–63%.
O
N
N
OH
OH
(e)
N
H
N
O
from 5a
Cl
21
O
O
N
N
N
N
N
N
N
N
CN
CN (c), (d)
OR
(a), (b)
(f), (g)
N
H
N
H
N
N
N
O
from 5b
N
O
Cl
Cl
Ph
5a: R = tert-butyl
5b: R = benzyl
22
3
4
Scheme 2. Synthesis of 21, 22. Reagents and conditions: (a) m-CPBA, CH2Cl2, rt, 50%. (b) POCl3, 100 °C, 95%. (c) 60% aq H2SO4: dioxane (1:1, v/v), 90 °C, 40%. (d)
H2NCH2CO2tBuÁHCl, 49% or H2NCH2CO2BnÁHCl, HBTU, (iPr)2NEt, DMF, rt, 74% (e) TFA, rt, 66%. (f) phenylboronic acid, Pd(tBu2PhP)2Cl2, KOAc, EtOH, 90 °C. (g) NaOH (aq), THF, rt,
6% for 2 steps.

M. Frohn et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5023–5026
5025
amino-3-nitrobenzoate (9) using standard chemistry. N-arylation
under the influence of Cu(OAc)2 with phenylboronic acid gave a
mixture of regioisomeric products, of which 11 was isolated as
the major component.10 Elongation of the C-4 side chain provided
24 (Scheme 4). Lastly, thieno-pyridine 25 was synthesized from
commercially available chloride 12 by palladium-catalyzed car-
bonylation and side chain construction (Scheme 5).
NO2
N
NO2
N
H
N
CO2Et
Cl
CO2Et
(a)
N
N
Cl
6
Cl
7
(b)
NH2
N
O
H
N
N
CO2Et
Analogs were evaluated for inhibitory potency in our PHD2 en-
OH
(c), (d), (e)
N
H
zyme assay, which detects proline hydroxylation of a HIF-1a pep-
N
N
N
N
O
tide (Table 1).11 Encouragingly, the initially modeled aza-
23
8
benzimidazole 14 (R1 = R2 = R3 = H) proved to have similar potency
to isoquinoline 1 (PHD2 IC50 of 1 = 1.40 lM; 14 = 2.82 lM). Since
Scheme 3. Synthesis of 23. Reagents and conditions: (a) aniline, CH2Cl2, rt, 48%. (b)
Pd (black), H2 (1 atm), EtOH, rt, 61%. (c) HC(OEt)3, p-TSA, 70 °C, 79%. (d)
H2NCH2CO2tBuÁHCl, (iPr)2NEt, dioxane, 90 °C, 82%. (e) TFA, rt, 51%.
the presumed critical polar elements were already present in this
molecule, subsequent analogs were designed to increase potency
via addition of lipophilic substituents at R1, R2, and R3.
SAR studies began with a brief scan of substituents at R1.
Replacing the N–H of 14 with N–CH3 in 15 immediately led to an
NH2
N
improvement in potency (IC50 = 0.57 lM). Unfortunately, signifi-
cantly increasing the size of this group further (16–18) did not lead
(a), (b)
HN
O2N
CO2Me
CO2Me
to additional improvements in activity. Ultimately, phenyl-substi-
9
10
tuted analog 16 (IC50 = 0.29
lM) was chosen as the optimal group
(c)
for subsequent investigation.
The effects of substitutions at R2 were investigated next.
Improvements in potency via modifications at this position proved
difficult. For example, the most active analog prepared in this ser-
ies, CH3-substituted compound 19, was roughly equipotent to 16
N
N
O
N
N
(d), (e), (f)
CO2Me
OH
N
H
O
24
11
Scheme 4. Synthesis of 24. Reagents and conditions: (a) Pd-C, H2, MeOH, 97%. (b)
HC(OEt)3, p-TSA, reflux, 89%. (c) Phenylboronic acid, Cu(OAc)2, NEt3, CH2Cl2, rt, 13%.
(d) KOH (aq), 60 °C, 64%. (e) H2NCH2CO2EtÁHCl, HBTU, (iPr)2NEt, DMF, rt, 44%. (f)
NaOH (aq), THF, rt, 32%.
Table 1
PHD2 inhibition by aza-benzimidazole analogs 14–22
R2
N
O
S
S
S
O
R1
N
(b), (c), (d)
OH
Cl
CO2Me
OH
(a)
N
H
N
H
N
O
N
N
N
O
R3
12
13
25
Compound
R1
R2
R3
PHD2a (IC50
, lM)
Scheme 5. Synthesis of 25. Reagents and conditions: (a) PdCl2dppp, CO (25 psi),
NEt3, DMF/MeOH (2:1, v/v), 100 °C, 31%. (b) NaOH (aq), dioxane, rt, 50%. (c)
H2NCH2CO2tBuÁHCl, HBTU, (iPr)2NEt, DMF, rt, 46%. (d) TFA, rt, 65%.
14
15
16
17
18
19
20
21
22
H
H
H
H
H
H
CH3
CF3
H
H
H
H
H
H
H
H
Cl
Ph
2.82 0.08
0.57 0.05
0.29 0.19
0.32 0.13
0.33 0.11
0.40 0.04
5.54 0.95
CH3
Ph
Bn
4-BrPh
Ph
Ph
Ph
Ph
Nitrile hydrolysis under acidic conditions, amide coupling with a
protected amino acid, and liberation of the carboxylic acid (reac-
tion conditions dependent upon identity of the ester) completed
the synthesis of compounds 14–20.9
0.066 0.004
0.003 0.001
H
a
IC50 values are the average of at least 4 separate experiments.
Analogs carrying substituents at R3 (21 and 22) were constructed
in a similar fashion (Scheme 2). Beginning with nitrile 3 (prepared
using Scheme 1), a chloride was introduced regioselectively by pyr-
idine N-oxidation and subsequent treatment with POCl3 to give 4.
Selective hydrolysis of the nitrile was accomplished by treatment
with H2SO4 (aq) in dioxane at 90 °C; the chloride proved stable to
these conditions. Amide coupling with a protected amino acid pro-
vided an intermediate ester (5a or 5b) that was either converted di-
rectly to the corresponding acid (R = tert-butyl) to give 21, or was
arylated via Suzuki cross coupling followed by ester saponification
(R = Bn) to give 22. Purine 23 was constructed from ethyl 2,6-di-
chloro-5-nitropyrimidine-4-carboxylate (6). Chemoselective dis-
placement of the C-6 chloride with aniline followed by tandem
nitro reduction/chloride removal using heterogeneous hydrogena-
tion conditions (palladium black) resulted in pyrimidine 8. Forma-
tion of the imidazole ring with subsequent side chain elaboration
as detailed above finished the synthesis (Scheme 3).
Benzimidazole 24 was prepared by a slightly different strategy.
First, the N–H benzimidazole 10 was assembled from methyl 2-
Figure 3. Optimized binding mode of 22 in the active site of PHD2. Fe(II) is shown
in pink.

5026
M. Frohn et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5023–5026
N
O
gained considerable insight into the structural requirements for
efficient PHD2 inhibition.
N
O
S
O
N
N
OH
OH
OH
N
H
N
H
N
H
N
N
O
O
N
O
23
24
25
References and Notes
IC50 > 40 μM
PHD2 (IC50) = 0.73 μM
PHD2 (IC50) >40 μM
1. For recent reviews, see: (a) Schofield, C. J.; Ratcliffe, P. J. Nat. Rev. Mol. Cell. Biol.
2004, 5, 343; (b) Semenza, G. L. Physiology (Bethesda) 2004, 19, 176; (c) Safran,
M., ; Kaelin, W. G., Jr J. Clin. Invest. 2003, 111, 779; (d) Bruick, R. K. Genes Dev.
2003, 17, 2614. and references therein.
2. (a) Warshakoon, N. C.; Wu, S.; Boyer, A.; Kawamoto, R.; Renock, S.; Xu, K.;
Pokross, M.; Evdokimov, A. G.; Zhou, S.; Winter, C.; Walter, R.; Mekel, M. Bioorg.
Med. Chem. Lett. 2006, 16, 5687; (b) Warshakoon, N. C.; Wu, S.; Boyer, A.;
Kawamoto, R.; Sheville, J.; Bhatt, R. T.; Renock, S.; Xu, K.; Pokross, M.; Zhou, S.;
Walter, R.; Mekel, M.; Evdokimov, A. G.; East, S. Bioorg. Med. Chem. Lett. 2006,
16, 5616; (c) Warshakoon, N. C.; Wu, S.; Boyer, A.; Kawamoto, R.; Sheville, J.;
Renock, S.; Xu, K.; Pokross, M.; Evdokimov, A. G.; Walter, R.; Mekel, M. Bioorg.
Med. Chem. Lett. 2006, 16, 5598. and references therein.
3. (a) Dann, C. E., III; Bruick, R. K. Biochem. Biophys. Res. Commun. 2005, 338, 639;
(b) Schofield, C. J.; Ratcliffe, P. J. Biochem. Biophys. Res. Commun. 2005, 338, 617;
(c) Berra, E.; Benizri, E.; Ginouves, A.; Volmat, V.; Roux, D.; Pouyssegur, J. EMBO
J. 2003, 22, 4082; (d) Appelhoff, R. J.; Tian, Y. M.; Raval, R. R.; Turley, H.; Harris,
A. L.; Pugh, C. W.; Ratcliffe, P. J.; Gleadle, J. M. J. Biol. Chem. 2004, 279, 38458; (e)
Minamishima, Y. A.; Moslehi, J.; Bardeesy, N.; Cullen, D.; Bronson, R. T.; Kaelin,
W. G., Jr Blood 2007, 10, 117812; (f) Percy, M. J.; Furlow, P. W.; Beer, P. A.;
Lappin, T. R. J.; McMullin, M. F.; Lee, F. S. Blood 2007, 110(6), 2193; (g) Percy, M.
J.; Zhao, Q.; Flores, A.; Harrison, C.; Lappin, T. R. J.; Maxwell, P. H.; McMullin, M.
F.; Lee, F. S. Proc. Natl. Acad. Sci. U.S.A. 2006, 103(3), 654.
4. McDonough, M. A.; Li, V.; Flashman, E.; Chowdhury, R.; Mohr, C.; Liénard, 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. Proc. Nat. Acad. Sci.
2006, 103, 9814.
5. The isoquinoline hydroxyl is desolvated and not satisfied by any proton
acceptors.
6. Palmer, B. D.; Smaill, J. B.; Boyd, M.; Boschelli, D. H.; Doherty, A. M.; Hamby, J.
M.; Khatana, S. S.; Kramer, J. B.; Kraker, A. J.; Panek, R. L.; Lu, G. H.; Dahring, T.
K.; Winters, R. T.; Showalter, H. D. H.; Denny, W. A. J. Med. Chem. 1998, 41, 5457.
7. Fife, W. K. J. Org. Chem. 1983, 48, 1375.
8. Penning, T. D.; Chandrakumar, N. S.; Desai, B. N.; Djuric, S. W.; Gasiecki, A. F.;
Malecha, J. W.; Miyashiro, J. M.; Russell, M. A.; Askonas, L. J.; Gierse, J. K.;
Harding, E. I.; Highkin, M. K.; Kachur, J. F.; Kim, S. H.; Villani-Price, D.; Pyla, E.
Y.; Ghoreishi-Haack, N. S.; Smith, W. G. Bioorg. Med. Chem. Lett. 2003, 13, 1137.
9. Final compounds were in excess of 95% purity as measured by 1H NMR and
Figure 4. PHD2 inhibition of core-modified analogs 23–25.
(IC50 = 0.40
activity (20, IC50 = 5.54
l
M). Incorporating a CF3 group led to significant loss in
M). This can be rationalized by either ste-
l
ric conflicts of this larger substituent with the protein or a weak-
ened ability to engage Fe(II) or Tyr303 due to electron
withdrawal from the imidazole nitrogen(s).
In contrast to R2, modification of R3 proved to be beneficial for
potency. Chloro-substituted analog 21 (R1 = Ph, R2 = H, R3 = Cl)
showed a fourfold increase in potency vs 16 (IC50 = 0.066 lM), pro-
viding motivation to continue the SAR study. Gratifyingly, it was
found that highly potent PHD2 inhibitors could be obtained with
aryl substitution. Phenyl-substituted analog 22 proved to be the
most active compound prepared in the series (IC50 = 0.003 lM).
Molecular modeling studies of diphenyl-substituted analog 22
suggest that in contrast to 14, only a single preferred binding con-
formation is accessible. Coordination of Fe(II) is achieved through
the imidazole nitrogen, which allows relatively strong van der
Waals and hydrophobic contacts with Ile256, Met299, Tyr303,
and Asp254 (Fig. 3). The alternative binding mode results in strong
steric repulsions between the R3 phenyl and His313 in this model.
Various alternative core analogs that have weak (or missing)
interactions with Fe(II) or Tyr303 were also prepared, and three
of these are shown in Figure 4. All analogs lose substantial activity
when compared with their aza-benzimidazole counterpart.
Molecular modeling studies. The co-crystal structure of HIF-PHD2
with an isoquinoline inhibitor 1 was used as the starting point for
molecular mechanics and dynamics calculations. For docking aza-
benzimidazole 14 and analogs, the X-ray structure was used as
the template. FLAME12 was used to generate alignment-based
models of 14 starting from the two different conformations corre-
sponding to the minima of the lead. These were obtained by ab ini-
tio quantum mechanical calculations, performed for each
conformer at B3LYP/6-31G* level, as implemented in Gaussian98
program. Thus, two sets of alignments were obtained, with up to
ten different conformations in each set. These were used as the
starting points for docking and binding energy assessment using
molecular mechanics and dynamics,13 in conjunction with the AM-
BER forcefield14 for the protein and GAFF forcefield15 for the li-
HPLC. HPLC: (Phenomenex, MAX RP, 4
l, 50 Â 2.0 mm, 1 mL/min, A: 0.1% TFA in
H2O, B: 0.1% TFA in MeCN, 10–100% B in 10 min), 1H NMR (Bruker, 400 MHz) in
DMSO-d6 or CDCl3.
10. Lam, P. Y. S.; Clark, C. G.; Saubern, S.; Adams, J.; Winters, M. P.; Chan, D. M. T.;
Combs, A. Tetrahedron Lett. 1998, 39, 2941. Regioisomers were separated and
identified by NOESY correlations.
11. HIF-PHD2 activity was measured utilizing homogenous time-resolved
fluorescence energy transfer technology by detecting the hydroxyl
modification of the proline residue of the P564-HIF-1
a
(Biotin-
DLEMLAPYIPMDDDFQL) peptide substrate resulting in recognition by the
Europium-tagged Von Hippel-Lindau, Elongin B and Elongin C heterotrimeric
complex (VCB-Eu complex). Compound inhibitor potency was determined
using 1 nM HIF-PHD2, 100 nM P564-HIF-1
a peptide, 0.25 lM 2-OG, 100 lM
FeCl2, and 2 mM ascorbic acid in Reaction Buffer (30 mM MES, pH 6, 10 mM
NaCl, 10 mM CaCl2, 0.25% Brij-35). The reaction was terminated after 1 h with
50 mM succinic acid in Detection Buffer (50 mM Tris–HCl, pH 8.0, 100 mM
NaCl, 0.05% Tween 20, 0.5% NaN3) containing a final concentration of 25 nM
streptavidin-APC and 2.5 nM VCB-Eu. The POC (percentage of control) was
determined by comparing the signal from hydroxylated peptide substrate in
the enzyme reaction containing inhibitor compound with that from PHD2
enzyme with DMSO vehicle alone, and no enzyme. The data were fit to the 4-
gands. Binding affinity calculations were performed using
a
previously described procedure that treats the protein as flexible.
All the protein residues within the first shell (66 Å of any atom)
of the docked ligand were allowed to be flexible. These include
Asp254, Lys255, Ile256, Trp258, Met299, Ala301, Tyr303, Tyr310,
Ile327, Tyr329, Leu343, Val376, Arg383, and Trp389 residues. Crys-
tallographic waters were included in the simulations. A mild re-
straint was placed on the ligand. Top ranking poses were
compared visually and energetically.
parameter model using
algorithm.
a Levenburg–Marquardt non-linear regression
12. Cho, S.-J.; Sun, Y. J. Inf. Model 2006, 46, 298. FLAME is implemented as a C++
program, using OpenEye toolkits, OEChem and CASE, available at http://
www.eyesopen.com.
13. Lee, M.; Sun, Y. J. Chem. Theor. Comput. 2007, 3, 1106.
14. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.;
Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995,
117, 5179.
In summary, the structure-based design and synthesis of a ser-
ies of aza-benzimidazoles as PHD2 inhibitors is presented. These
efforts resulted in compound 22, which displayed highly potent
15. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. J. Comput. Chem.
2004, 25, 1157.
inhibition of PHD2 function (IC50 = 0.003 lM). In addition, we have