Discovery of novel hydroxy-thiazoles as HIF-α prolyl hydroxylase inhibitors: SAR, synthesis, and modeling evaluation

Article abstract of DOI:10.1016/j.bmcl.2008.06.031

Inhibition of the PHD2 enzyme has been associated with increased red blood cell levels. From a screening hit, a series of novel hydroxyl-thiazoles were developed as potent PHD2 inhibitors.

Full text of DOI:10.1016/j.bmcl.2008.06.031

Bioorganic & Medicinal Chemistry Letters 18 (2008) 3925–3928
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel hydroxy-thiazoles as HIF-a prolyl hydroxylase inhibitors:
SAR, synthesis, and modeling evaluation
a
a
a
Christopher M. Tegley a, , Vellarkad N. Viswanadhan , Kaustav Biswas , Michael J. Frohn ,
Tanya A. N. Peterkin a, Catherine Chang a, Roland W. Bürli a, Jennifer H. Dao a, Henrike Veith b,
Norma Rogers b, Sean C. Yoder b, Gloria Biddlecome b, Philip Tagari a, Jennifer R. Allen a, Randall W. Hungate a
*
a Chemistry Research and Discovery, Amgen, Inc., One Amgen Center Drive, Thousand Oaks, CA 91320, USA
b 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:
Inhibition of the PHD2 enzyme has been associated with increased red blood cell levels. From a screening
hit, a series of novel hydroxyl-thiazoles were developed as potent PHD2 inhibitors.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 23 May 2008
Revised 10 June 2008
Accepted 10 June 2008
Available online 13 June 2008
Keywords:
Proyl hydroxylase inhibitors
HIF-1
a
Hypoxia
Ischemia
Anemia
Peripheral arterial disease
Myocardial infarction
Stroke
Diabetes
Hydroxy-thiazole
Hypoxia-inducible factor (HIF) is a transcription factor and key
regulator of cellular responses to hypoxia in a wide variety of
organisms.1 Among the genes regulated by HIF transcriptional
activity, many play crucial roles in angiogenesis, erythropoiesis,
energy utilization, vascular tone, apoptosis and cellular prolifera-
tion. HIF can also be involved in pathophysiological conditions like
cancers, where it is commonly elevated due to genetic alterations
or tumor hypoxia.2
is catalyzed by HIF prolyl hydroxylase domain enzymes (PHD1, 2,
3) in a reaction that requires Fe(II), 2-oxoglutarate (2-OG), and
O2.5 As pO2 falls, HIF-
a
is poorly hydroxylated, protein degradation
is blunted, and HIF-mediated transcription occurs.
It is thought that stabilization of HIF- by inhibition of PHD en-
a
zymes could be a viable therapy for anemia, for ischemic diseases
including myocardial infarction, stroke, peripheral arterial disease,
or for complications of diabetes.6 PHD2 is considered the key en-
The HIF transcription factor consists of an
a/b heterodimer that
zyme for HIF-a
regulation in a variety of cell lines7 and for stimu-
binds to hypoxia-response elements within HIF target genes to
lating erythropoiesis.8 For these reasons, we initiated a medicinal
chemistry program aimed at the identification and optimization
of potent PHD2 inhibitors (see Fig. 1).
control transcription.3 HIF-b is a constitutive nuclear protein that
dimerizes with oxygen-regulated HIF-
tutively produced protein whose stability is primarily regulated by
cellular pO2. At physiological pO2, HIF- is hydroxylated on con-
served proline residues (Pro402 and Pro564 on human HIF-1 ).
This proline hydroxylation creates recognition sites for the von
a subunits. HIF-a is a consti-
a
a
HO
N
OH
Hippel-Lindau protein which targets HIF-
a
for poly-ubiquitination
N
S
O
and subsequent proteasomal degradation.4 Proline hydroxylation
S
O
O
O
1
PHD2 IC50 = 0.072 μM
* Corresponding author. Tel.: +1 805 447 0622; fax +1 805 480 3016.
E-mail address: ctegley@amgen.com (C.M. Tegley).
Figure 1. Screening hit hydroxy-thiazole 1.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.06.031

3926
C. M. Tegley et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3925–3928
Table 2
We undertook a high throughput screen (HTS) of our internal
Structure–activity relationship of amide side-chaina
compound collection utilizing a homogeneous time-resolved fluo-
rescence resonance energy transfer (TR-FRET) assay to test for
PHD2 inhibition. This assay detects PHD2-dependent hydroxyl-
OH
N
Ph
N
ation of Pro564 in a peptide derived from human HIF-1a 9
. We
X
CO2H
S
O
identified compound 1 as a possible starting point for SAR studies
to optimize the potency of this novel class of PHD2 inhibitor.
Table 1 highlights modifications to the carboxylate side-chain
and thiazole ring. The importance of both the hydroxy-thiazole
moiety and the carboxylate for PHD2 inhibition are evident in
compounds 2, 3, and 4 which lack activity. Mono-methylation of
the carboxylate side-chain resulted in a >18-fold decrease in activ-
ity relative to compound 1 as shown by compound 11, whereas the
gem-dimethyl analog 12 resulted in a complete loss of inhibitory
activity. In addition, methylation of the thiazole ring (compound
13) and homolog 14 resulted in a >5-fold decrease in potency for
both compounds when compared to the lead compound 1.
With preliminary SAR established, our strategy was to maintain
the hydroxy-thiazole carboxylate as our core scaffold and examine
substitution effects on the amide side-chain and sulfonyl group
substitution.
b
Compound
X
PHD2 IC50 (lM)
1
15
16
17
18
SO2
S
O
NH
CH2
0.072 0.026
2.3 1.6
4.4 0.68c
1.7 0.19c
>40
a
IC50 values represent an average of at least three titrations SD.
Assay details to be published elsewhere (Ref. 9).
Mean value from two experiments.
b
c
Table 3
Structure–activity relationship of sulfone substitutiona
OH
N
R
N
S
CO2H
We attempted to replace the sulfonyl group with an alternate
linker to the distal phenyl group. As shown in Table 2, replacement
of the sulfone with a sulfide 15 resulted in a greater than 30-fold
decrease in potency. Substitution of the sulfone with an ether 16
or amine 17 retained weak activity whereas the phenethyl analog
18 resulted in a complete loss of inhibitory activity.
O
S
O
O
b
Compound
R
PHD2 IC50 (lM)
1
19
20
21
22
23
24
25
26
27
28
Ph
Me
0.072 0.026
0.600 0.089c
0.050 0.023
0.021 0.009
0.072 0.010
0.098 0.013
0.033 0.001
0.027 0.012
0.021 0.008
0.011 0.009
0.003 0.002
4-Me-Ph
4-tBu-Ph
2-Cl-Ph
3-Cl-Ph
4-Cl-Ph
2-Naphthyl
4-Ph-Ph
Bn
The structure–activity relationships of phenyl group substitu-
tion are highlighted in Table 3. Replacement of the phenyl group
Table 1
Structure–activity relationship of hydroxy-thiazole carboxylatea
4-tBu-Bn
b
Compound
Structure
PHD2 IC50
(l
M)
a
IC50 values represent an average of at least three titrations SD.
Assay details to be published elsewhere (Ref. 9).
Mean value from two experiments.
OH
N
b
c
Ph
S
N
1
0.072 0.026
CO2H
CO2Et
O
O
S
S
O
O
H
N
N
Ph
S
with a methyl group shown by compound 19 resulted in an 8-fold
decrease in potency relative to compound 1. Substitution around
the distal phenyl group was varied to understand the role of elec-
tronic and hydrophobic substitution effects on inhibitory activity.
Electron-donating groups such as the para-methyl analog 20 had
a modest effect on activity whereas the para-chloro analog 24
was slightly more potent than compound 1 and may be slightly
more potent than the ortho-chloro 22, and meta-chloro 23 analogs.
The potency of the 4-tert-butyl-phenyl analog 21 (IC50 = 21 nM) led
us to further extend the hydrophobic substituents. The 2-naphthyl
analog 25 and 4-biphenyl analogs 2 showed at least a 2-fold
improvement in potency relative to compound 1. A significant in-
crease in potency was seen with the benzyl sulfone analog 27
which showed a 7-fold improvement in potency compared to com-
pound 1. Further optimization of compound 27 was achieved as
shown by compound 28 (IC50 = 3 nM) which demonstrated a great-
er than 20-fold increase in potency compared to compound 1.
Altogether the SAR studies showed that changes to the hydro-
xy-thiazole and to the side chain resulted in a decrease in potency
whereas extension of the distal hydrophobic substituents resulted
in a significant improvement in potency.
2
3
>40
O
O
OH
N
Ph
S
N
>40
CO2Et
CO2H
O
O
S
O
O
H
N
N
Ph
S
4
>40
O
O
S
OH
N
Me
CO2H
Ph
S
N
N
11
12
1.3 0.58
>40
O
O
S
S
O
O
OH
N
Me
Me
Ph
S
CO2H
O
O
OH
N
Ph
S
N
N
13
0.40 0.23
0.41 0.22
CO2H
O
O
S
O
Me
To enable SAR exploration of this novel class of PHD2 inhibitors,
three synthetic routes (Schemes 1–3) were developed. Analogs
were prepared according to Scheme 1 when the sulfonyl acetic acid
intermediates were commercially available. Ethyl 2-aminothia-
zole-4-acetate 5 was coupled to phenyl sulfonyl acetic acid to af-
ford the thiazole ethyl ester 2 which was subsequently oxidized
OH
N
Ph
S
14
CO2H
O
O
S
O
a
IC50 values represent an average of at least three titrations SD.
Assay details to be published elsewhere (Ref. 9).
b

C. M. Tegley et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3925–3928
3927
with OxoneÒ to give the hydroxy-thiazole ester 3 and finally
hydrolyzed to give the hydroxy-thiazole carboxylate 1. Lastly,
compound 2 was hydrolyzed to afford the thiazole carboxylate 4.
When the sulfonyl acetic acid intermediates were not available,
the synthetic route depicted in Scheme 2 was utilized to take
advantage of commercially available substituted thio-phenols.
MCPBA oxidation of ethyl 2-(2-chloroacetamido)-4-thiazoleacetate
6 afforded the intermediate hydroxy-thiazole 7.10 Hydroxy-thia-
zole 7 was then reacted with substituted aryl thiols to give analogs
10 after hydrolysis. The sulfide-analogs could also be oxidized with
OxoneÒ to give the sulfone-analogs 8 after hydrolysis.
N
S
CO2Et
CO2Et
H2N
5
a
OH
N
N
O
N
CO2Et
O
O
b
Ph
S
S
S
Ph
N
H
S
O
O O
3
2
Scheme 3 shows the synthetic schemes used to prepare each of
the amide side-chain analogs 16, 17, and 18. Ethyl 2-aminothia-
zole-4-acetate 5 was coupled with either phenoxyacetic acid, N-
Boc-N-phenyl glycine or 3-phenyl-n-propionic acid then oxidized
with OxoneÒ and hydrolyzed. The Boc-protected precursor to ana-
log 17 was de-protected with HCl/dioxane in the final step.
To leverage our medicinal chemistry effort we decided to use
our recently published X-ray crystal structure of the catalytic do-
main of human PHD2 co-crystallized with a 2-OG-competitive iso-
quinoline inhibitor.11 The active site is lined by a number of
hydrophobic residues. In addition, several polar residues and water
molecules line the pocket and interact with the ligand and Fe(II) in
the active site. Ample evidence,12 suggests coordination of Fe(II)
and a salt bridge involving Arg383 as key elements for potent
PHD2 inhibition.
As a first step to assess the binding modes of hydroxy-thiazoles
1, a conformational analysis was performed and showed two min-
ima equal in energy with the barrier to rotation (in the absence of
Fe) about the N-carbonyl bond of 18 kcal/mol. Potential binding
modes corresponding to these minima were considered, based on
overlays with the X-ray structure of the isoquinoline compound.
Although, there is no difference in energy between the two con-
formers (Fig. 2), the geometry of the hydroxyl thiazolylidene acet-
amide represented in binding mode 1, corresponded to the single
crystal X-ray structure of compound 1 (Fig. 3).13
c
c
OH
N
CO2H
O
N
O
N
O
CO2H
S
Ph
O O
S
S
Ph
N
S
O
H
1
4
Scheme 1. Reagents and conditions: (a) PhSO2CH2CO2H, EDC, Et3N, DCM, rt, 80%;
(b) OxoneÒ, MeOH, H2O, rt, 40%; (c) aq NaOH, THF, rt, 100%.
O
CO2Et
Cl
Cl
R
N
N
H
S
6
a
OH
N
OH
N
N
N
b, c, d
R
CO2H
CO2Et
S
S
S
O
O O
O
8
7
b
Figures 4 and 5 show the putative binding modes corresponding
to the two different conformations of the core group, following
optimization of the two complexes.14 Both binding modes con-
firmed the salt bridge between the carboxylate group and
Arg383, penetrating a tight pocket lined by several hydrophobic
residues. The binding mode 1 (Fig. 4) showed the aryl group at-
tached to the sulfone making hydrophobic contacts with Ile256,
Trp258, Met299 and Tyr310. The binding mode 2 (Fig. 5) also re-
vealed contacts with Tyr310 and other hydrophobic residues. An
important distinction between the two putative binding modes,
OH
N
OH
d
N
N
N
R
CO2H
S
CO2Et
S
S
S
O
O
9
10
Scheme 2. Reagents and conditions: (a) MCPBA, CHCl3, rt, 90%; (b) R-SH, Et3N, THF,
40–99%; (c) OxoneÒ, EtOH, H2O, rt; (d) aq NaOH, THF, rt, 100%.
OH
N
Ph
Ph
N
O
CO2H
CO2H
FeII
S
O
a, b, c
FeII
O
16
CO2H
CO2H
O
N
O
O
O
N
O
O
S
N
OH
N
S
S
N
S
N
N
S
CO2Et
e, b, c, d
O
N
H
H2N
Binding mode 1
Binding mode 2
S
O
Figure 2. Two possible binding modes of compound 1.
17
OH
5
N
Ph
N
f, b, c
CO2H
S
O
18
Scheme 3. Reagents and conditions: (a) PhOCH2CO2H, HATU, DIPEA, DMF, rt, 94%;
(b) OxoneÒ, MeOH, H2O, rt, 20–36%; (c) aq NaOH, THF, rt, 38–88%; (d) HCl, dioxane,
rt, 88%; (e) PhN(Boc)CH2CO2H, HATU, DIPEA, DMF, rt, 58%; (f) Ph(CH2)2CO2H, TBTU,
DIPEA, DMF, rt, 100%.
Figure 3. Single crystal X-ray structure of 1.

3928
C. M. Tegley et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3925–3928
hydrophobic substituents, can be attributed to favorable interac-
tions with lipophilic residues that line the active site.
In summary, we discovered novel PHD2 inhibitor
a
1
(IC50 = 72 nM) via high throughput screening. Through SAR studies,
we successfully optimized the potency of the lead compound 1 by
over 20-fold as shown by compound 28 (IC50 = 3 nM). In addition,
modeling studies provided insight into the possible binding mode
of this novel class of PHD2 inhibitors. Description of in vivo prop-
erties of analogs in this series will be the subject of future
publications.
Acknowledgement
We thank Dr. Richard J. Staples for the single crystal X-ray of
compound 1.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.06.031.
Figure 4. Binding mode 1 of the lead compound 1, shown in pink. The inner pocket
containing the acid group of the ligand, making a salt bridge with Arg383. The inner
pocket is shown as a Connoly semi-transparent surface for the residues lining the
pocket, color coded in atom colors of the closest atom type. Fe(II) ion is identified as
an orange sphere and crystallographic waters within the active site are identified as
red spheres. Some important protein residues are identified. The protein main chain
is shown, rendered by its secondary structure.
References and notes
1. (a) Bruick, R. K. Genes Dev. 2003, 17, 2614; (b) Safran, M.; Kaelin, W. G. J. Clin.
Invest. 2003, 111, 779; (c) Semenza, G. L. Physiology 2004, 19, 176; (d) Schofield,
C. J.; Ratcliffe, P. J. Nat. Rev. Mol. Cell Biol. 2004, 5, 343.
2. (a) Pugh, C. W.; Ratcliffe, P. J. Nat. Med. 2003, 9, 677; (b) Semenza, G. L. Nat. Rev.
Cancer 2003, 3, 721.
3. Wenger, R. H.; Stiehl, D. P.; Camenisch, G. Science’s STKE 2005, 306,
re12.
4. (a) Jaakkola, P.; Mole, D. R.; Tian, Y. M.; Wilson, M. I.; Gielbert, J.; Gaskell, S. J.;
von Kriegsheim, A.; Hebestreit, H. F.; Mukherji, M.; Schofield, C. J.; Maxwell, P.
H.; Pugh, C. W.; Ratcliffe, P. J. Science 2001, 292, 468; (b) Ivan, M.; Kondo, K.;
Yang, H. F.; Kim, W.; Valiando, J.; Ohh, M.; Salic, A.; Asara, J. M.; Lane, W. S.;
Kaelin, W. G. Science 2001, 292, 464; (c) Semenza, G. L. Nat. Rev. Cancer 2003, 3,
721.
5. Epstein, A. C. R.; Gleadle, J. M.; McNeill, L. A.; Hewitson, K. S.; O’Rourke, J.; Mole,
D. R.; Mukherji, M.; Metzen, E.; Wilson, M. I.; Dhanda, A.; Tian, Y.-M. ; Masson,
N.; Hamilton, D. L.; Jaakkola, P.; Bartstead, R.; Hodgkin, J.; Maxwell, P. H.; Pugh,
C. W.; Schofield, C. J.; Ratcliffe, P. J. Cell 2001, 107, 43.
6. Hewitson, K. S.; McNeill, L. A.; Schofield, C. J. Curr. Pharm. Des. 2004, 10, 821.
7. 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.
8. (a) 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, 654; (b) Percy, M.
J.; Furlow, P. W.; Beer, P. A.; Lappin, T. R. J.; McMullin, M. F.; Lee, F. S. Blood
2007, 110, 2193; (c) Minamishima, Y. A.; Moslehi, J.; Bardeesy, N.; Cullen, D.;
Bronson, R. T.; Kaelin, W. G., Jr. Blood 2008, 111, 3236.
9. Dao, J. H.; Kurzeja, R. J. M.; Morachis, J. M.; Veith, H.; Lewis, J.; Yu, V.; Tegley, C.;
Tagari, P. Anal. Biochem., submitted for publication.
10. Perrone, E.; Alpegiani, M.; Giudici, F.; Bedeschi, A.; Pellizzato, R.; Nannini, G.
J. Heterocycl. Chem. 1984, 21, 1097.
11. 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.; Jordon, S.; Syed, R. S.; Schofield, C. J. Proc. Natl. Acad.
Sci. U.S.A. 2006, 103, 9814.
12. (a) Mole, D. R.; Schlemminger, I.; McNeill, L. A.; Hewitson, K. S.; Pugh, C. W.;
Ratcliffe, P. J. Bioorg. Med. Chem. Lett. 2003, 13, 2677; (b) Warshakoon, N. C.;
Wu, S.; Boyer, A.; Kawamoto, R.; Sheville, J.; Renock, S.; Xu, K.; Pokross, M.;
Zhou, S.; Winter, C.; Walter, R.; Mekel, M.; Evdokimov, A. G. Bioorg. Med. Chem.
Lett. 2006, 16, 5517; (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; (d) 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; (e) 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.
13. Crystallographic data has been deposited in the Cambridge Crystallographic
Data Centre for small molecules and allocated the deposition number
CCDC688161. Copies of data can be obtained free of charge on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK.
14. Modeling notes and references are contained in the supplementary
material.
Figure 5. Binding mode 2 of the lead compound 1, shown in gray.
is a potential interaction of Arg389 with the sulfone in binding
mode 1 that is absent in binding mode 2. Due to these reasons,
we speculated that the binding mode with the exocyclic nitrogen
facing Fe (II) was more probable.
Our modeling and SAR studies are in agreement that the hydro-
xy-thiazole and carboxylate moieties are key binding elements
that mimic 2-OG in the binding site and are important elements
for PHD2 inhibition. Changes to the lead compound 1 that disrupt
either the coordination with Fe(II) or the salt bridge with Arg383
leads to a significant decrease in potency. The importance of the
sulfone moiety hypothesized in the binding model may be noted
and is in agreement with the observed SAR. The significant
improvement in potency achieved, via extension of the distal