Applying the pro-drug approach to afford highly bioavailable antagonists of P2Y14

Article abstract of DOI:10.1016/j.bmcl.2010.12.113

Our series of competitive antagonists against the G-protein coupled receptor P2Y₁₄ were found to be highly shifted in the presence of serum (>99% protein bound). A binding assay using 2% human serum albumin (HSA) was developed to guide further SAR studies and led to the identification of the zwitterion 2, which is substantially less shifted (18-fold) than our previous lead compound 1 (323-fold). However, as the bioavailability of 2 was low, a library of ester pro-drugs was prepared (7a-7j) and assessed in vitro. The most interesting candidates were then profiled in vivo and led to the identification of the pro-drug 7j, which possesses a substantially improved pharmacokinetic profile.

Full text of DOI:10.1016/j.bmcl.2010.12.113

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4366–4368
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Applying the pro-drug approach to afford highly bioavailable antagonists
of P2Y14
⇑
Joël Robichaud , Jean-François Fournier, Sébastien Gagné, Jacques Yves Gauthier, Martine Hamel,
Yongxin Han, Martin Hénault, Stacia Kargman, Jean-François Levesque, Yaël Mamane,
Joseph Mancini, Nicolas Morin, Erin Mulrooney, Jin Wu, W. Cameron Black
Merck Frosst Centre for Therapeutic Research, 16711 TransCanada Hwy., Kirkland, Quebec, Canada H9H 3L1
a r t i c l e i n f o
a b s t r a c t
Article history:
Our series of competitive antagonists against the G-protein coupled receptor P2Y14 were found to be
highly shifted in the presence of serum (>99% protein bound). A binding assay using 2% human serum
albumin (HSA) was developed to guide further SAR studies and led to the identification of the zwitterion
2, which is substantially less shifted (18-fold) than our previous lead compound 1 (323-fold). However, as
the bioavailability of 2 was low, a library of ester pro-drugs was prepared (7a–7j) and assessed in vitro.
The most interesting candidates were then profiled in vivo and led to the identification of the pro-drug 7j,
which possesses a substantially improved pharmacokinetic profile.
Received 5 November 2010
Revised 17 December 2010
Accepted 21 December 2010
Available online 28 December 2010
Keywords:
P2Y14
GPR-105
Ó 2011 Published by Elsevier Ltd.
Antagonist
Pro-drug
Protein binding
The engineering of biologically active molecules is a complex,
multi-faceted process that encompasses multiple levels of fine-
tuning. Too often this endeavor is shunted in one of the last crucial
steps, which involves delivery of the active compound to the tar-
geted system via oral absorption.
While many of the strategies that have been devised to circum-
vent this issue revolve around the formulation of the active ingre-
dient, another approach that has gained considerable attention in
recent times involves the preparation of pro-drugs.1 These modi-
fied active compounds possess favorable physicochemical
properties that can significantly improve oral absorption and then
undergo efficient in vivo conversion to the active drug by various
metabolic processes.
During our investigations of the biological role of P2Y14,2a
widely expressed G-protein-coupled receptor that binds UTP-glu-
cose,3,4 we generated several series of potent non-nucleotide
antagonists.5,6 Many of these antagonists displayed both high
affinity for the P2Y14 receptor and good pharmacokinetic profiles
in mice. For example, compound 1 (see Fig. 1) has a Ki of 4.0 nM
in the chimpanzee P2Y14 binding assay,7 and is orally bioavailable
in mice (75 mg/kg, BID, 0.25% methocel, F = 67%, T1/2 = 3.0 h,
Cl = 1.6 mL/min/kg).
F3C
O
F3C
OH
O
OH
1
2
F
OH
F
Figure 1. Previously disclosed 4,7-disubsituted naphthoic acid antagonist of P2Y14
.
N
H
⇑
Corresponding author.
E-mail
addresses:
joel_robichaud@merck.com,
robichaud5@videotron.ca
Figure 2. Less shifted antagonist of P2Y14 identified using the binding assay
(J. Robichaud).
utilizing 2% HSA.
0960-894X/$ - see front matter Ó 2011 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2010.12.113

J. Robichaud et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4366–4368
4367
However, further studies revealed that compound 1 had a very
high affinity for serum proteins and was >99% bound in the
presence of 4.5% human serum albumin (HSA). Such high levels
of protein binding often negatively impact in vivo activity due to
the low level of free drug in circulation.
A binding assay using 2% HSA was thus developed to rapidly
assess the extent of protein binding for various antagonists. Under
these newly developed assay conditions, compound 1 displayed a
Ki of 1292 nM, which represented a 323-fold shift from the binding
assay in the absence of HSA. Armed with this new tool, an SAR
campaign was initiated to identify less shifted compounds. These
efforts ultimately led to the identification of the zwitterion 2
(see Fig. 2) which showed good affinity for the P2Y14 receptor
(Ki = 1.9 nM in the chimpanzee P2Y14 binding assay), and was only
18-fold shifted in the presence of 2% HSA (Ki = 35 nM).
elements. In this case, the oral bioavailability of compound 2 (mice,
50 mg/kg, 0.25% methocel) suffered considerably (F = 5%, C24h
0.68 M). Aware that the polar zwitterionic properties of com-
pound 2 might be responsible for the low bioavailability, we rea-
soned that using an ester pro-drug to cap the acid moiety might
improve oral absorption.
=
l
This strategy was thus applied to antagonist 2 with the aim of
delivering a highly bioavailable pro-drug that would be efficiently
converted to high circulating levels of the active compound. We
decided that the most efficient manner to identify the best pro-
drug would be to prepared a library of esters whose potential
would first be assessed in vitro, followed by in vivo profiling of
the most promising candidates.
The synthetic route devised (see Scheme 1) began from the pre-
viously described boronic ester 3,4 which was cross-coupled with
the aryl bromide 4 to afford the desired biaryl which was con-
verted to compound 5 by hydrogenation of the double bond. The
zwitterionic compound 2 could be prepared by hydrolysis of the
ethyl ester using lithium hydroxide (LiOH) in a mixture of tetrahy-
drofuran and methanol (THF/MeOH) followed by benzyloxycar-
bonyl (BOC) deprotection using trifluoroacetic acid (TFA) in
dichloromethane (DCM). The ethyl ester pro-drug 7b was prepared
by simple BOC deprotection using similar conditions.
However, as is often the case, the improvement of one mole-
cular property inadvertently leads to the demise of other crucial
F3C
F3C
O
O
a
b
OEt
OEt
3
While the preparation of the BOC-protected piperidine 6a
required a simple hydrolysis of the ethyl ester using LiOH in
THF/H2O at 45 °C, the synthesis of the CBZ-protected piperidine
6b required a three-step sequence. First, the removal of the BOC
protecting group was again accomplished using TFA in DCM, the
piperidine was then treated with benzyl chloroformate (CBZ-Cl)
and sodium carbonate (Na2CO3) in THF/H2O, and the ethyl ester
was finally hydrolyzed as described above to afford compound 6b.
While the BOC-protected piperidine 6a was utilized in the
alkylation of the free carboxylic acid using benzyl chloride, the
CBZ-protected piperidine 6b was utilized in the alkylation reaction
with all other electrophiles. The BOC and CBZ protecting groups of
this library of compounds were then removed respectively using
TFA in DCM and using hydrogen with palladium on charcoal, to
afford the desired pro-drug compounds 7a–7j.7
B
5
O
+
O
Br
4
N
O
for 6a: c
for 6b: d,e,c
for 2: c,d
N
for 7b:
d
O
O
O
F3C
O
F3C
2: R = H
O
7a: R = Me
OH
7b: R = Et
OR
7c: R = i-Pr
7d: R = Bn
7e: R = c-Hex
7f: R = CH2CH2OH
7g: R = CH2OCOMe
f
b or d
6a: R = tert-Bu
6b: R = Bn
This library of pro-drugs was first evaluated for metabolic
stability in vitro by parallel incubation of compounds 7a–7j in
whole blood or liver microsomes + NADPH (see Fig. 3). These
experiments demonstrated that pro-drugs 7g, 7h, 7i and 7j were
among the most labile and gave high conversion to the parent drug
2 in the presence of whole blood.8
7h: R = CH(Me)OCH2COi-Pr
7i: R = CH2CH2NMe2
N
7j: R = CH2CONMe2
N
H
O
OR
Scheme 1. Reagents and conditions: (a) PdCl2(dppf)ÁCH2Cl2, aq Na2CO3, DMF, 85 °C;
(b) H2, Pd/C*, EtOAc; (c) LiOH, THF/MeOH, 45 °C; (d) TFA, DCM; (e) ClCO2Bn, Na2CO3,
THF/H2O; (f), R–Cl, K2CO3, DMF, rt ? 45 °C.
In vivo profiling of pro-drug 7h revealed inefficient hydrolysis
of the ester functionality and afforded low levels of the parent drug
100
O
Whole blood
Liver Microsomes
N
O
O
O
O
O
N
90
80
70
60
50
40
30
20
10
0
O
F
O
O
F
F
O
O
O
O
7
O
N
H
O
O
O
O
O
O
O
OH
O
O
O
O
c
b
e
i
g
f
a
d
h
j
Figure 3. In vitro screening of the pro-drug library in whole blood and in liver microsomes (1 h incubation, % conversion to zwitterion 2).

4368
J. Robichaud et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4366–4368
3.0
2.5
2.0
1.5
1.0
0.5
0.0
parent compound 2 (AUC = 2.2 vs 54.4
els of the pro-drug 7j at 24 h of less than 0.1
l
M h), with circulating lev-
M (see Fig. 4).
No side effects observed
LOQ L713 : 1nM
l
LOQ L542 : 1nM
In summary, our newly developed binding assay for the
G-protein coupled receptor P2Y14 utilizing 2% HSA has enabled
us to identify zwitterion 2 (Ki = 35 nM, 18-fold shifted), which is
significantly less shifted in the presence of protein. This advance
was limited by the erosion of bioavailability when compared with
our initial lead 1. Solving this issue required the preparation of a
small library of ester pro-drugs (7a–7j) that were first evaluated
in vitro in the presence of whole blood and liver microsomes, fol-
lowed by in vivo profiling of the three most promising candidates.
This strategy led to the discovery of pro-drug 7j which possesses a
substantially enhanced pharmacokinetic prolife over its parent
compound 2, and which opens the path for further in vivo studies
aimed at the elucidation of the biological role of the P2Y14 receptor.
PErosdterrugprLo-7d1ru3g(7nj=2, AUC= 2.2)
Parent zwitterion 2
Parent L-542. (n=2, AUC= 54.4)
F3C
F3C
O
O
N
O
OH
O
in vivo
7j
2
N
H
N
H
References and notes
1. Ettmayer, P.; Amidon, G. L.; Clement, B.; Testa, B. J. Med. Chem. 2004, 47, 2393.
2. Jacobson, K. A.; Ivanov, A. A.; de Castro, S.; Harden, T. K.; Ko, H. Purinergic
Signalling 2009, 5, 75.
0
4
8
12
Time (h)
16
20
24
3. Chambers Chambers, J. K.; Macdonald, L. E.; Sarau, H. M.; Ames, R. S.; Freeman,
K.; Foley, J. J.; Zhu, Y.; McLaughlin, M. M.; Murdock, P.; McMillan, L.; Trill, J.;
Swift, A.; Aiyar, N.; Taylor, P.; Vawter, L.; Naheed, S.; Szekeres, P.; Hervieu, G.;
Scott, C.; Watson, J. M.; Murphy, A. J.; Duzic, E.; Klein, C.; Bergsma, D. J.; Wilson,
S.; Livi, G. P. J. Biol. Chem. 2000, 175, 10767.
Figure 4. Pharmacokinetic profile of ester pro-drug 7j (blue line) and its conversion
to zwitterion 2 (red line).
4. Gauthier, J. Y.; Bayly, C. I.; Belley, M.; Deschênes, D.; Fournier, J.-F.; Gagné, S.;
Gareau, Y.; Hamel, M.; Han, Y.; Hénault, M.; Hydjazie, H.; Kargman, S.; Lavallé,
G.; Lesvesque, J.-F.; Li, L.; Mamane, Y.; Mancini, J.; Morin, N.; Mulrooney, E.;
Plamondon, S.; Robichaud, J.; Thérien, M.; Tranmer, G.; Wang, Z.; Wu, J.; Black,
W. C. Bioorg. Med. Chem. Lett. 2011, 21, 2836.
5. Guay, D.; Beaulieu, C.; Belley, M.; Crane, S. N.; DeLuca, J. C.; Gareau, Y.; Hamel,
M.; Henault, M.; Hyjazie, H.; Kargman, S.; Chan, C. C.; Xu, L.; Gordon, R.; Li, L.;
Mamane, Y.; Morin, N.; Mancini, J.; Therien, M.; Truong, V. L.; Wang, Z.; Black, W.
C. Bioorg. Med. Chem. Lett. 2011, 21, 2832.
6. Hamel, M.; Henault, M.; Hyjazie, H.; Morin, M. N.; Bayly, C.; Skorey, K.; Therien,
A.; Mancini, J.; Brideau, C.; Kargman, S. J. Biomol. Screening., Accepted for
Publication.
7. The mesylate salts of all compounds (2 and 7a–7j) were prepared for the
purpose of in vivo evaluation.
2. On the other hand, pro-drugs 7i and 7j were efficiently con-
verted in vivo to the parent zwitterion 2, with pro-drug 7j display-
ing the highest levels of compound 2 and was identified as the
most promising candidate.
The pharmacokinetic profile of pro-drug 7j (mice, 50 mg/kg,
0.25% methocel) afforded plasma levels (Fvirtual = 67%, C24h
2.28 M, AUC = 54 M h, Cl = 1.6 mL/min/kg) of the parent drug 2
that were substantially higher than when compound 2 was admin-
istered directly as a zwitterion (F = 5%, C24h = 0.68 M, AUC =
23 M h). Furthermore, in vivo conversion of pro-drug 7j to parent
=
l
l
l
l
compound 2 was found to be very efficient as the measured plasma
levels of the pro-drug 7j represented less than 5% of the AUC of the
8. Pro-drug 7g was not further profiled in vivo as it was found to be chemically
unstable.