Structure-activity relationships and hepatic safety risks of thiazole agonists of the thrombopoietin receptor

Article abstract of DOI:10.1016/j.bmcl.2010.05.087

5-F substitution of an aminothiazole moiety within a series of thrombopoietin receptor agonists leads to potent agents with an improved hepatic safety profile in rodent toxicology studies.

Full text of DOI:10.1016/j.bmcl.2010.05.087

Bioorganic & Medicinal Chemistry Letters 20 (2010) 4069–4072
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure–activity relationships and hepatic safety risks of thiazole agonists
of the thrombopoietin receptor
Amy S. Antipas, Laura C. Blumberg , William H. Brissette, Matthew F. Brown, Jeffrey M. Casavant,
Jonathan L. Doty, James Driscoll à, Thomas M. Harris, Christopher S. Jones, Sandra P. McCurdy,
§
*
Eric McElroy, Mark Mitton-Fry , Michael J. Munchhof, David A. Reim , Lawrence A. Reiter,
Sharon L. Ripp, Andrei Shavnya, Marc I. Smeets, Kristen A. Trevena
Pfizer Worldwide Research and Development, Groton, CT 06340, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
5-F substitution of an aminothiazole moiety within a series of thrombopoietin receptor agonists leads to
potent agents with an improved hepatic safety profile in rodent toxicology studies.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 12 March 2010
Revised 19 May 2010
Accepted 20 May 2010
Available online 26 May 2010
Keywords:
TPO
Thrombopoietin
Agonist
The term thrombocytopenia denotes a condition in which
platelet levels drop to below 50,000/ L. It occurs in immune
We now report that toxicological studies of 1 yielded additional
potential evidence of hepatic damage. When rats were subjected to
single oral doses ranging from 50 to 500 mg/kg, dramatic increases
in both AST and ALT were observed at all dose levels (Table 1).
Although histopathology revealed no correlated liver changes, the
brief exposure produced by a single dose may not have been suffi-
cient to induce any such changes. Additionally, when 1 was dosed
orally at 100 mg/kg in transgenic mice expressing the human TPO
receptor,7 mild elevations in AST were observed, although ALT lev-
els were unchanged. These and other studies revealing acute toxic-
ity at 250 mg/kg in rats indicated that 1 did not have a safety
profile that was commensurate with further development.
While these studies did not establish a causal link between the
5-unsubstituted-2-aminothiazole moiety and the observed hepato-
toxicity, we nevertheless sought to mitigate this possibility in fur-
ther analogs by eliminating this structural feature, either by
adding a substituent at C-5 of the thiazole or by altering the hetero-
cycle itself. In one example from the realm of non-steroidal anti-
inflammatory agents, incorporation of a substituent at C-5 of a
5-unsubstituted-2-aminothiazole converted the hepatotoxic
sudoxicam to the successful and non-hepatotoxic drug meloxicam.8
Accordingly, we introduced a 5-F group onto the thiazole ring
yielding 3. However, this analog was significantly less potent than
1 in the BaF3 reporter assay, which measures thrombopoietin
receptor agonist activity (Table 2).9 Previous related medicinal
chemistry efforts had revealed that co-planarity of the pendant
phenyl ring with the aminothiazole was desirable for maximal
l
thrombocytopenia purpura (ITP), in association with cancer
chemotherapy, and in instances of impaired liver function.1 The
potential for spontaneous bleeding can arise in all of these condi-
tions. Until recently the only available treatment was platelet infu-
sion, but the recent approvals of the oral drug eltrombopag2 and
the injectable romiplostim2,3 have offered promising new options.
The isonipecotic acid derivative 1 (Fig. 1) was recently reported
as an orally bioavailable agonist of the human thrombopoietin
(TPO) receptor.4 Although this compound had promising potency
and pharmacokinetic properties, the presence of a 5-unsubstitut-
ed-2-aminothiazole was of significant concern, owing to the pro-
pensity for such thiazoles to undergo metabolic activation to
generate potentially hepatotoxic species.5 As a result, the biotrans-
formation of 1 and its hydrolysis product 2 were examined.
Although 1 did not produce any detectable glutathione adducts,
the treatment of human liver microsomes with 2 led to the forma-
tion of such adducts, indicating that oxidative bioactivation had
occurred.6
* Corresponding author. Tel.: +1 860 686 2123.
E-mail address: mark.mitton-fry@pfizer.com (M. Mitton-Fry).
Present address: Alkermes, Inc., 88 Sidney Street, Cambridge, MA 02139, USA.
Current address: Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080,
à
USA.
§
Current address: SNBL USA, Ltd, 6605 Merrill Creek Parkway, Everett, WA 98203,
USA.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.05.087

4070
A. S. Antipas et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4069–4072
Cl
HO
O
O
Cl
N
S
N
S
N
F
F
HO
O
O
HN
H2N
N
N
CF3
+
CF3
OH
N
2
1
Figure 1.
Table 1
Pharmacokinetic and serum chemistry analysis from single oral doses of compounds 1 and 5a
b
b
Compound
Dose (mg/kg)
Mean AST (IU/L)
Mean ALT (IU/L)
Mean Cmax
(l
g/mL)
Mean AUC0–24 (lg h/mL)
1
1
1
1
5
5
5
5
0
50
250
500
0
50
250
500
96
1937
2952
3495
102
87
897
1899
55
442
450
579
56
58
202
337
0
20
48
84
0
50
120
280
0
84
832
1486
0
548
1818
4736
a
These were conducted as separate studies, each with their own control group (entries 1 and 5)
All concentrations refer to total concentrations.
b
Table 2
Cl
Cl
N
HO
O
O
HO
O
O
F
R
N
N
N
S
N
S
N
X
F
HN
HN
N
CF3
CF3
F
3
4-11
Cl
HO
O
O
F
HN
N
N
N
CF3
12
Compound
R
X
BaF3ÁEC50 (lM)
1
3
4
5
6
7
8
9
0.033
0.702
0.1
0.089
0.066
0.086
0.16
H
F
CH3
Cl
CH2CH3
C
C
C
C
C
C
C
N
(CH2)2CH3
CN
0.13
10
11
12
>0.898
0.574
>0.811
potency.10 Thus, adding a substituent at C-5 of the thiazole, as in 3,
likely leads to an unfavorable biaryl ring orientation owing to a
steric and/or electronic clash between the C-5 thiazole and the
C-2 phenyl fluoro substituents. In order to relieve this undesirable
conformation effect, the C-2 phenyl substituent would have to be
removed. Previous efforts in a pyrimidine benzamide series of
TPO receptor agonists had revealed that the 4-F-3-CF3-phenyl
substitution pattern was only about 2-fold less active than the
2-F-3-CF3 substitution pattern.11 Indeed, incorporation of this sub-
stitution pattern into the isonipecotic acid series led to a com-
pound (4) only about 3-fold less potent than 1. Doing so in
conjunction with a thiazole C-5 fluoro group likewise yielded a po-
tent agonist, 5. A number of other nonpolar thiazole C-5 substitu-
ents led to similarly potent compounds, 6–9. Of the substituents
examined only the polar cyano group led to a compound with sig-
nificantly reduced potency (10). Altering the heterocycle to either a
thiadiazole (11) or a pyrazole (12) also led to significantly lower
Table 3
Cl
HO
O
O
F
N
S
N
HN
5, 13-20
N
R1
R2
Compound
R1
R2
BaF3ÁEC50 (lM)
5
13
14
15
16
17
18
19
20
CF3
Cl
Br
F
F
F
Cl
F
F
F
F
F
0.089
0.48
0.67
>1.0
0.62
0.95
>0.97
0.22
0.058
Cl
OCF3
OCH2CF3
CH2OEt
Butyl
Isobutyl

A. S. Antipas et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4069–4072
4071
Approach 1: direct fluorination
F
S
N
S
N
H2N
H2N
Selectfluor®
CF3
CF3
CH3CN
0oC to rt
F
21
F
22, 42%
Approach 2: de novo ring synthesis
H2N
N
Cl
S
Br
F
S
i.
O
F
F
i. i-PrMgCl
H2N
NH2
THF
DMF, rt
CF3
ii.
O
CF3
CF3
ii. aqueous workup
iii. AcOH, 65oC
MeO
Cl
F
N
F
23
F
25, 73%
24
22, 90%
Scheme 1.
agonist activity, a finding consistent with work in the pyrimidine
benzamide series.11
knock-out/human knock-in of exon 8–10 of the TPO receptor,
spanning the transmembrane domain, was created to generate a
‘humanized’ TPOr transgenic mouse.7 The human TPO receptor
transgenic mice exhibited dose-dependent increases in platelets
measured on day 6 after two doses of compound on days 0 and
1, whereas wild-type mice did not. The EC50 for compound 1 was
Earlier work had established that the optimal agonist activity
was obtained with a lipophilic group at C-3 of the phenyl ring
and a 2,3-di-, 3,4-di-, or 2,3,4-tri-substitution pattern.11 We thus
examined a limited range of 3,4-disubstitution variations in order
to improve potency further. However, with the exception of the
3-isobutyl-4-F derivative 20, these analogs were significantly less
potent (Table 3). Since 20 introduced additional potential sites
for metabolism and was more lipophilic than 5 (C Log P = 5.61
and 4.54, respectively), we chose to focus our efforts on 5.
In order to synthesize 5, we developed two approaches for the
efficient preparation of key intermediate 22 (Scheme 1). In the first
of these, aminothiazole 21 could be directly fluorinated in modest
yield using SelectfluorÒ.6 In the second approach, 22 could be syn-
thesized using a de novo synthesis of the heterocyclic ring. In this
case the bromoarene 23 is first magnesiated and then reacted with
the Weinreb amide 24.12 The resulting ketone (25) was reacted
with thiourea13 and the crude product heated with glacial acetic
acid to afford the desired thiazole product in high yield.14
As has already been reported,6 neither 5 nor 22 undergo meta-
bolic activation in an in vitro system. This differentiation was fur-
ther manifested in lowered hepatic effects in in vivo studies with 5
(Table 1). In contrast to 1, no changes in either AST or ALT were
observed at an oral dose of 50 mg/kg, despite significantly higher
exposures of 5. At higher doses, the increases in AST and ALT were
of diminished magnitude relative to those seen with 1, again
despite higher total exposures of 5 versus 1 at each dose level. In
four-day studies, 10, 50 and 150 mg/kg daily oral dosing in rats
led to no clinical pathology, and only mild increases in ALT (one
animal of four on day 1) and bilirubin (one animal of four on day
5) were observed. No histopathological changes were found at
any dose.
27
lg h/mL, and for compound 5, it was 7
lg h/mL.
These positive results demonstrate that appropriate substitu-
tion at C-5 of the thiazole along with concomitant alteration of
the substitution on the phenyl ring can provide potent, orally ac-
tive TPO receptor agonists7 with reduced potential for hepatotox-
icity in rodent toxicology studies.
Acknowledgments
The authors would like to thank Drs. Anthony Marfat and Amit
Kalgutkar for helpful discussions, as well as Dr. Shane Eisenbeis for
his related work on the fluorinated aminothiazole synthesis. We
also thank the reviewers for their insightful comments on the
Letter.
References and notes
1. Ikeda, Y.; Miyakawa, J. Thrombosis Haemostasis 2009, 7, 239.
2. Nurden, A. T.; Viallard, J.-F.; Nurden, P. Lancet 2009, 373, 1562.
3. Patel, H.; Patel, N.; Vyas, A.; Patel, M.; Pandey, S. J. Clin. Diagn. Res. 2009, 3, 1690.
4. Munchhof, M. J.; Abramov, Y. A.; Antipas, A. S.; Blumberg, L. C.; Brissette, W. H.;
Brown, M. F.; Casavant, J. M.; Doty, J. L.; Driscoll, J.; Harris, T. M.; Wolf-Gouveia,
L. A.; Jones, C. S.; Li, Q.; Linde, R. G.; Lira, P. D.; Marfat, A.; McElroy, E.; Mitton-
Fry, M.; McCurdy, S. P.; Reiter, L. A.; Ripp, S. L.; Shavnya, A.; Shepard, R. M.;
Sperger, D.; Thomasco, L. M.; Trevena, K. A.; Zhang, F.; Zhang, L. Bioorg. Med.
Chem. Lett. 2009, 19, 1428.
5. Kalgutkar, A. S.; Gardner, I.; Obach, R. S.; Shaffer, C. L.; Callegari, E.; Henne, K. R.;
Mutlib, A. E.; Dalvie, D. K.; Lee, J. S.; Nakai, Y.; O’Donnell, J. P.; Boer, J.; Harriman,
S. P. Curr. Drug Metab. 2005, 6, 161.
6. Kalgutkar, A. S.; Driscoll, J.; Zhao, S. X.; Walker, G. S.; Shepard, R. M.; Soglia, J. R.;
Atherton, J.; Yu, L.; Mutlib, A. E.; Munchhof, M. J.; Reiter, L. A.; Jones, C. S.; Doty,
J. L.; Trevena, K. A.; Shaffer, C. L.; Ripp, S. L. Chem. Res. Toxicol. 2007, 20, 1954.
7. 7.Brissette, W. H.; Lira, P. D.; McCurdy, S. P.; Nelson, R. T.; Neote, K.; Stock, J. L.;
Driscoll, J. P.; Trevena, K. A.; Shepard, R. M.; Jones, C. S.; Munchhof, M. J.; Reiter,
L. A.; Ripp, S. L. Unpublished results.
In order to put these results into proper context, an evaluation
of in vivo efficacy was warranted. Our proprietary series of small
molecules, and those of others, are functional agonists of the
human TPO receptor. However, they have been found to be inactive
when tested against non-human animal species of bone marrow
cells, due to a unique histidine found only in the human receptor’s
transmembrane domain.15 These findings precluded the usual pre-
clinical pharmacological evaluation of these compounds in stan-
dard laboratory animal species. In order to develop an animal
model for in vivo pharmacologic testing of compounds, a murine
8. Obach, R. S.; Kalgutkar, A. S.; Ryder, T. F.; Walker, G. S. Chem. Res. Toxicol. 2008,
21, 1890.
9.
A murine hematopoietic BaF3 cell line with the transfected human TPO
receptor (hTPOr) and the STAT1/3 responsive b-lactamase reporter were
maintained in culture with rTPO and washed and placed in TPO free media 18
hours prior to assay. Compound and rTPO dilutions were prepared in triplicate
in assay media and delivered to a 96 well Costar black, clear bottom plate
(Corning Life Sciences, Acton, MA) using the BioMek 2000 (Beckman-Coulter,

4072
A. S. Antipas et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4069–4072
14. To water bath-cooled solution of isopropylmagnesium chloride in THF
Fullerton, CA). Twenty microliters of cells (10,000 cells per well) were added
a
using a Multi-drop (ThermoElectron Corp, Milford, MA) and the plates were
incubated for 5 h at 37 °C, 5% CO2. Controls included wells with cells plus
150 ng/mL hTPO in assay media, cells plus assay media and assay media only.
Ten microliters of LiveBLAzer™-FRET B/G (CCF4-AM) was added to each well
and the plates were incubated in the dark, at room temperature for 1 h. The
plates were read on a LJL Analyst (Molecular Devices Corporation, Sunnyvale,
CA) equipped with the 405–20 excitation filter, two emission filters (blue
(33.6 mL, 2.0 M, 67 mmol, 1.0 equiv) was added aryl bromide 23 (9.5 mL,
67 mmol, 1.0 equiv) under N2 at a rate such that the internal temperature of
the reaction did not exceed 35 °C. The reaction was allowed to stir 3 h at
ambient temperature, then cooled in a brine/ice bath. The amide 24 (10.17 g,
65.38 mmol, 1.0 equiv) was then added dropwise by cannula such that the
temperature did not exceed 10 °C. THF (5 mL) was used to wash the flask
previously containing 24, and this was also added by cannula. The reaction was
allowed to stir 13 h, warming slowly to ambient temperature. The reaction was
then quenched by addition of 100 mL satd aq NH4Cl and extracted 4 Â 50 mL
chloroform. The combined organic layers were dried over MgSO4, filtered, and
concentrated to give an orange oil. Purification by flash chromatography
(gradient elution from 0% to 30% ethyl acetate in hexanes) afforded the desired
product 25 as a yellow oil (12.42 g, 73%). This oil was dissolved in DMF (15 mL)
and thiourea (11.16 g, 146.6 mmol, 3.0 equiv) was added. The solids dissolved
within a few minutes, and the reaction vessel became warm to the touch. The
resulting solution was stirred overnight and then subjected to aqueous work-
up, drying over MgSO4, filtration, and concentration to afford a yellow oil. To
this oil was added glacial acetic acid, and the vessel was placed in a sand batch
(105 °C) for 20 min, achieving an internal temperature of 65 °C). The reaction
was then concentrated and purified by flash chromatography to afford the
desired product 22 as a grey solid (12.11 g, 90%).
channel 460–40 and green channel 530–10) and
a 425 dichroic filter. A
stimulation index (SI) was calculated for each compound using the formula
[(460/530 ratio drug samples divided by 460/530 no drug or TPO control
ratio)] À 1. An EC50 was calculated by plotting SI drug ratio drug to SI hTPO
control ratio. Full experimental details are described in WO2007036769.
10. Reiter, L. A.; Jones, C. S.; Abramov, Y. A.; Bordner, J.; Brissette, W. H.; DiCapua, F.
M.; Lira, P. D.; McCurdy, S. P.; Rescek, D.; Samardjiev, I.; Smeets, M. I.; Withka, J.
Bioorg. Med. Chem. Lett. 2008, 18, 3000.
11. Reiter, L. A.; Subramanyam, C.; Mangual, E. J.; Jones, C. S.; Smeets, M. I.;
Brissette, W. H.; McCurdy, S. P.; Lira, P. D.; Linde, R. G.; Li, Q.; Zhang, F.; Antipas,
A. S.; Blumberg, L. C.; Doty, J. L.; Driscoll, J. P.; Munchhof, M. J.; Ripp, S. L.;
Shavnya, A.; Shepard, R. M.; Sperger, D.; Thomasco, L. M.; Trevena, K. A.; Wolf-
Gouveia, L. A.; Zhang, L. Bioorg. Med. Chem. Lett. 2007, 17, 5447.
12. Kumar, V.; McCloskey, P.; Bell, M. R. Tetrahedron Lett. 1994, 35, 833.
13. Yamawaki, K.; Nomura, T.; Yasukata, T.; Uotani, K.; Miwa, H.; Takeda, K.;
Nishitani, Y. Bioorg. Med. Chem. Lett. 2007, 15, 6716.
15. Nakamura, T.; Miyakawa, Y.; Miyamura, A.; Yamane, A.; Suzuki, H.; Ito, M.;
Ohnishi, Y.; Ishiwata, N.; Ikeda, Y.; Tsuruzoe, N. Blood 2006, 107, 4300.