Discovery of thiadiazole amides as potent, S1P3-sparing agonists of sphingosine-1-phosphate 1 (S1P1) receptor

Article abstract of DOI:10.1016/j.bmcl.2012.02.016

High-throughput screening of GSK compound collection led to the discovery of a novel series of thiadiazole amides as potent and S1P₃-sparing sphingosine-1-phosphate 1 (S1P₁) receptor agonists. Synthesis, structure and activity relationship, selectivity, and some developability properties are described.

Full text of DOI:10.1016/j.bmcl.2012.02.016

Bioorganic & Medicinal Chemistry Letters 22 (2012) 2456–2459
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of thiadiazole amides as potent, S1P3-sparing agonists
of sphingosine-1-phosphate 1 (S1P1) receptor
Heng Xu, Haibo Zhang, Linbo Luan, Yan Xu, Chengyong Li, Yonghui Wang, Fangbin Han, Ting Yang,
Feng Ren, Jia-Ning Xiang, John D. Elliott, Yonggang Zhao, Taylor B. Guo, Hongtao Lu, Wei Zhang,
⇑
David Hirst, Matthew Lindon, Xichen Lin
GlaxoSmithKline Pharmaceuticals, R&D China, 898 Halei Road, Zhangjiang Hi-Tech Park, Pudong, Shanghai 201023, China
a r t i c l e i n f o
a b s t r a c t
Article history:
High-throughput screening of GSK compound collection led to the discovery of a novel series of thiadi-
azole amides as potent and S1P3-sparing sphingosine-1-phosphate 1 (S1P1) receptor agonists. Synthesis,
structure and activity relationship, selectivity, and some developability properties are described.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 21 November 2011
Revised 21 January 2012
Accepted 6 February 2012
Available online 15 February 2012
Keywords:
S1P1 receptor agonists
Thiadiazole amides
SAR
Multiple sclerosis (MS)
The sphingosine-1-phosphate (S1P) receptors have been receiv-
ing greater attention as a therapeutic target with the recent FDA
approval of FTY720 for oral treatment of remitting–relapsing mul-
tiple sclerosis (RRMS).1 S1P is a bioactive lysolipid with pleiotropic
functions that mediates a large variety of biological effects via
extracellular signaling through five distinct G-protein coupled
receptors, numbered S1P1 through S1P5.2 Agonism of S1P receptors
prevents the migration of autoimmune lymphocytes from lymph
nodes and other secondary lymphoid organs into the periphery
including the central nervous system.1 This interruption of lym-
phocyte migration, ostensibly without affecting cell types of the in-
nate immune system (e.g., neutrophils or macrophages) as well as
affecting cellular reactivity of lymphocytes to antigen challenge,
promises a new immunomodulatory therapeutic principle for a
variety of autoimmune diseases.
FTY720 is a prodrug and its phosphorylated form FTY720-P ex-
erts in vivo pharmacological effects.3 FTY720-P is poorly selective
and proves to be a potent agonist of four of the five S1P subtypes
(S1P1, S1P3–5).4 S1P1 receptor agonism has been shown to correlate
with lymphocyte recirculation, while S1P3 receptor agonism has
been linked to a number of side effects in preclinical development
and clinical trials.5–7 To attenuate the potential risk of adverse ef-
fects associated with activation of S1P3 receptor and therefore to
improve the safety profile, more selective compounds for S1P1
are needed. Herein, we describe the discovery of a novel series of
thiadiazole amides as potent and selective S1P1 receptor agonists.
A high throughput screening (HTS) using a GTPc
S assay8 was
employed to search the GSK compound collection for S1P1 receptor
agonists. The thiadiazole amides such as 3 and 4 were identified as
one of the most interesting hit series with EC50s of ꢀ1
lM (Fig. 1).
The hit structures feature a tertiary amide directly linked to a thia-
diazole core and represent a novel chemistry starting point. To fol-
low up the hit series with aim of improving the potency as well as
other drug-like properties, the structure–activity relationship
(SAR) was explored based on the core template of thiadiazole
amides.
The general synthetic methods employed to prepare thiadiazole
amides are summarized in Scheme 1. Thiadiazole 6 was prepared
by reaction of acid chlorides with amino thiourea, followed by
cyclization under acid conditions. Thiadiazole 6 was then alkylated
with different alkyl aldehydes using standard reductive animation
procedures to give amines 7, which was coupled with various acid
chlorides to afford the targeted thiadiazole amides 8. One interest-
ing finding during the course of the synthesis is that if amides were
formed first, alkylation reaction took place at the nitrogen on the
thiadiazole ring rather than amide nitrogen as illustrated in
Scheme 2. Therefore, alkylated thiadiazole 10 was formed as the
major product with a trace amount of 8. To the best of our knowl-
edge, very few examples were reported in the literature demon-
strating alkylation reactions for thiadiazole rings.9–12
With the robust synthesis route developed, we first explored
the SAR of the right-hand side (RHS) moiety by preparing analogs
⇑
Corresponding author. Tel.: + 86 21 61590718; fax: +86 21 61590730.
E-mail address: Xichen.2.Lin@gsk.com (X. Lin).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.02.016

H. Xu et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2456–2459
2457
O
O
N
N
N
N
F
N
N
R
S
S
O
Cl
NH2
N
OH
1
2
3
4
R = H, R = PO3H2
Figure 1. Structures of FTY720 (1), its metabolite FTY720-phosphate (2) and two HTS hits (3 and 4).
O
N
N
O
H
N
S
NH2
F
S
F
F
N
H
b
S
Cl
a
H2N
NHNH2
N
NH2
5
6
O
R2
N
N
N
NH
c
d
N
F
F
R1
S
R1
S
7
8
Scheme 1. Reagents and conditions: (a) pyridine, RT, overnight; (b) H2SO4, 40 °C, 30 min, 30% from two steps; (c) NaBH(OAc)3, aldehydes, THF, 50 °C, overnight, 35–45%; (d)
acid chlorides, Et3N, microwave, 120 °C, 2 h, ꢀ50%.
O
N
R1
O
N
R2
R2
N
N
H2N
N
N
N
F
a
b
S
NH
F
F
S
S
9
10
Scheme 2. Reagents and conditions: (a) acid chlorides, pyridine, RT, 4 h, 35–45%; (b) alkyl bromides, K2CO3, DMF, ꢀ30%.
Table 2
Table 1
S1P1 and S1P3 pEC50 values for compounds 3 and 12a–c
S1P1 and S1P3 pEC50 values for compounds 3 and 11a–l
a
a
a
a
Compd
R
S1P1 pEC50
S1P3 pEC50
Compd
R
S1P1 pEC50
S1P3 pEC50
O
O
11a
11b
11c
11d
11e
11f
11g
11h
11i
3
H
2-Cl
2-F
2-OMe
2-Me
3-Cl
6.0
6.3
6.5
6.2
6.2
6.1
5.9
5.9
6.2
6.6
5.9
5.9
6.7
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
3
6.6
<5.0
<5.0
12a
<5.0
N
O
3-F
O
3-OMe
3-Me
4-Cl
4-F
4-OMe
4-Me
12b
12c
S
<5.0
5.2
<5.0
<5.0
11j
11k
11l
a
S1P1 and S1P3 assays were performed as described in Ref. 13.
a
S1P1 and S1P3 assays were performed as described in Ref. 13.
fonamide, and amine moieties yielded much less active or
completely inactive compounds. This amide moiety seems to be
part of the pharmacophore and remains in future compound
design.
Subsequently, the efforts were focused on the alkyl moiety on
the amide nitrogen. As shown in Table 3, the size of the alkyl moi-
ety is important for S1P1 potency. S1P1 potency (pEC50) were all
less than 5.5 with H, Me and Et and started to increase with the
length of the alkyl moieties from n-Pr to n-Bu, while potency
containing different small substituents on the phenyl ring. As
shown in Table 1, various groups were tolerated in the RHS (S1P1
pEC50s: 5.5–6.7). All compounds listed in Table 1 showed good
selectivity for S1P1 over S1P3 (10-fold at least) and 100-fold selec-
tivity could be achieved with some compounds.
In order to evaluate the importance of the amide function, var-
ious replacements were then synthesized. As shown in Table 2, sul-

2458
H. Xu et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2456–2459
Table 3
S1P1 receptors since other secondary forces could not be reason-
ably projected for these pure alkyl groups.
S1P1 and S1P3 pEC50 values for compounds 13a–g
Based on the structural features of FTY720 as well as other pub-
lished S1P receptor agonists,9,14–17 hydrophilic portions (e.g., acid,
amine, alcohol) are typically needed in the molecule to gain the po-
tency. With this in mind, we then focused our efforts in exploring
the left-hand side (LHS) moiety by introducing polar moieties. As
shown in Table 4, analogs with different hydrophilic groups in
the para position of the phenyl ring (see 14b, 14d and 14h) pro-
vided a significant increase in S1P1 potency to EC50 10–100 nM.
Particularly, EC50 values for analogs (14f and 14p–r) bearing a car-
boxylic acid group fell into nanomolar range. However, having po-
lar groups at the meta position of the phenyl ring did not enhance
potency. Besides substituted phenyls, indole as an example of bicy-
clic ring was found to be a good moiety for S1P1 potency enhance-
ment. Further derivatization of the indole nitrogen or introduction
of halogen substitutions also yielded potent compounds (see 14j–
l). Notably, when thiadiazole portion was shifted from 5 to 7-posi-
tion of the indole ring, potency was dramatically decreased by
a
a
Compd
R
S1P1 pEC50
S1P3 pEC50
13a
13b
13c
13d
13e
13f
H
Me
Et
n-Pr
<5.0
<5.0
<5.5
6.2
6.9
6.6
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.5
n-Bu
Cyclopropyl-CH2-
n-Pentyl
13g
6.2
a
S1P1 and S1P3 assays were performed as described in Ref. 13.
dropped after bigger/longer n-Pentyl group was introduced
(H ꢁ Me ꢁ Et <n-Pr <n-Bu >n-Pentyl, 13a–13g). We hypothesize
this might be due to the optimal hydrophobic interactions with
Table 4
S1P1 and S1P3 pEC50 values for compounds 14a–r
a
a
a
a
Compd
R
S1P1 pEC50
S1P3 pEC50
Compd
R
S1P1 pEC50
S1P3 pEC50
N
HO
14a
7.2
7.9
<5.0
<5.0
14j
8.1
7.5
<5.5
<5.0
Cl
HO
N
14b
14c
14k
14l
N
F
N
N
5.4
<5.0
8.2
<5.0
N
N
14d
14e
7.3
6.6
<5.0
5.6
14m
14n
5.6
7.1
<5.0
<5.5
N
O
O
O
N
HOOC
HOOC
S
S
N
O
O
N
14f
14g
14h
8.9
6.2
7.5
6.3
<5.0
<5.0
14o
14p
14q
6.6
10.4
9.4
<5.0
5.8
HOOC
HOOC
N
N
N
<5.0
N
Cl
HOOC
N
N
14i
8.9
<5.5
14r
8.8
<5.0
a
S1P1 and S1P3 assays were performed as described in Ref. 13.

H. Xu et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2456–2459
2459
leading to the identification of the thiadiazole amide series with
nanomolar potency. Although the series has high clearance result-
ing in poor pharmacokinetic profiles, the series appears to be an
acceptable starting point for further lead optimization. Ongoing
work is to design and synthesize cyclic amides as well as other het-
erocyclics as replacements or bioisosteres of the tertiary amide
moiety to increase their metabolic stability while retaining S1P1
potency.
Acknowledgments
We thank Drs. Maite De Los Frailes and Greg Osborne for their
initial work during the HTS, and Drs. Eric Yang, Hong Lu, Zongping
Zhang, Jiansong Yang for their helpful comments and discussions.
Figure 2. Reduction of blood lymphocyte counts by IP dose of selected compounds.
Compounds were evaluated for its ability to induce peripheral lymphocyte
reduction following a single administration. Naïve mice were given compounds
through ip (2 mg/kg) and blood samples were taken at 4 h after dosing. Lympho-
cytes in mice were counted by FACS within the lymphocyte gate. Percentages
indicate cell counts relative those of the control.
References and notes
1. Brinkmann, V.; Billich, A.; Baumruker, T.; Heining, P.; Schmouder, R.; Francis,
G.; Aradhye, S.; Burtin, P. F. Nat. Rev. 2010, 9, 883.
2. (a) Chun, J.; Brinkmann, V. Discovery Med. 2011, 12, 213; (b) Brinkmann, V.;
Cyster, J. G.; Hla, T. Am. J. Transplant. 2004, 4, 1019.
3. (a) Adachi, K. Yuki Gosei Kagaku Kyokaishi 2011, 69, 904; (b) Chiba, K.; Kataoka,
H.; Seki, N.; Maeda, Y.; Sugahara, K.; Fingolimod Inflamm. Regen. 2011, 31, 167.
4. Brinkmann, V.; Davis, M. D.; Heise, C. E.; Albert, R.; Cottens, S.; Hof, R.; Bruns,
C.; Prieschl, E.; Baumruker, T.; Hiestand, P.; Foster, C. A.; Zollinger, M.; Lynch, K.
R. J. Biol. Chem. 2002, 277, 21453.
Table 5
Intrinsic clearance with liver microsomes of selected compoundsa
5. Kovarik, J. M.; Lu, M.; Riviere, G.; Barbet, I.; Maton, S.; Goldwater, D. R.;
Schmouder, R. L. Eur. J. Clin. Pharmacol. 2008, 64, 457.
Compd Mouse (
l
L/
Rat (
l
L/
Dog (
l
L/
Human (lL/
min  mg
min  mg
min  mg
min  mg
6. Forrest, M.; Sun, S.-Y.; Hajdu, R.; Bergstrom, J.; Card, D.; Doherty, G.; Hale, J.;
Keohane, C.; Meyers, C.; Milligan, J.; Mills, S.; Nomura, N.; Rosen, H.;
Rosenbach, M.; Shei, G.-J.; Singer, I. I.; Tiam, M.; West, S.; White, V.; Xie, J.;
Proia, R. L.; Mandala, S. J. Pharmacol. Exp. Ther. 2004, 309, 758.
7. Sanna, M. G.; Liao, J.; Jo, E.; Alfonso, C.; Ahn, M.-Y.; Peterson, M. S.; Webb, B.;
Lefebvre, S.; Chun, J.; Gray, N.; Rosen, H. J. Biol. Chem. 2004, 279, 13839.
8. Hall, D. A.; Beresford, I. J. M.; Browning, C.; Giles, H. Br. J. Pharmacol. 1999, 126,
810.
9. Clerici, F.; Pocar, D.; Guido, M.; Loche, A.; Perlini, V.; Brufani, M. J. Med. Chem.
2001, 44, 931.
10. Nagao, Y.; Hirata, T.; Goto, S.; Sano, S.; Kakehi, A.; Iizuka, K.; Shiro, M. J. Am.
Chem. Soc. 1998, 120, 3104.
protein)
protein)
protein)
protein)
14b
14f
14h
14i
712.4
137.4
478.4
63.0
278.6
99.6
539.0
43.2
43.0
370.2
595.8
1151.0
823.4
464.2
625.2
61.2
13.8
109.4
77.2
48.6
118.8
14l
59.4
14p
91.2
80.6
aIntrinsic clearance assays were performed as described in Ref. 18.
11. Antonaroli, S.; Bianco, A.; Brufani, M.; Cellai, L.; Baido, G. L.; Potier, E.; Bonomi,
L.; Perfetti, S.; Fiaschi, A. L.; Segre, G. J. Med. Chem. 1992, 35, 2697.
12. Kubota, S.; Ueda, Y.; Fujikane, K.; Toyooka, K.; Shibuya, M. J. Org. Chem. 1980,
45, 1473.
13. The pEC50 of S1P1 receptor agonists for compounds were determined by
Invitrogen’s b-arrestin-mediated recombinant hS1P1 internalization assay
based on a b-lactamase reporter in the U2OS cell line (TangoÒ). The pEC50 of
S1P3 receptor agonists for compounds were determined by using Invitrogen’s
Ca2+-based GeneBLAzerÒ assay system in HEK293T cell line that stably
overexpress hS1P3.
14. (a) Cusack, K. P.; Stoffel, R. H. Curr. Opin. Drug Discovery Dev. 2010, 13, 481; (b)
Bolli, M. H.; Lescop, C.; Nayler, O. Curr. Top. Med. Chem. 2011, 11, 726.
15. Pennington, L. D.; Sham, K. K. C.; Pickrell, A. J.; Harrington, P. E.; Frohn, M. J.;
Lanman, B. A.; Reed, A. B.; Croghan, M. D.; Lee, M. R.; Xu, H.; McElvain, M.; Xu,
Y.; Zhang, X.; Fiorino, M.; Horner, M.; Morrison, H. G.; Arnett, H. A.; Fotsch, C.;
Wong, M.; Cee, V. J. ACS Med. Chem. Lett. 2011, 2, 752.
16. Lanman, B. A.; Cee, V. J.; Cheruku, S. R.; Frohn, M.; Golden, J.; Lin, J.; Lobera, M.;
Marantz, Y.; Muller, K. M.; Neira, S. C.; Pickrell, A. J.; Rivenzon-Segal, D.; Schutz,
N.; Sharadendu, A.; Yu, X.; Zhang, Z.; Buys, J.; Fiorino, M.; Gore, A.; Horner, M.;
Itano, A.; McElvain, M.; Middleton, S.; Schrag, M.; Vargas, H. M.; Xu, H.; Xu, Y.;
Zhang, X.; Siu, J.; Burli, R. W. ACS Med. Chem. Lett. 2011, 2, 102.
17. Li, Z.; Chen, W.; Hale, J. J.; Lynch, C. L.; Mills, S. G.; Hajdu, R.; Keohane, C. A.;
Rosenbach, M. J.; Milligan, J. A.; Shei, G. J.; Chrebet, G.; Parent, S. A.; Bergstrom,
J.; Card, D.; Forrest, M.; Quackenbush, E. J.; Wickham, L. A.; Vargas, H.; Evans, R.
M.; Rosen, H.; Mandala, S. J. Med. Chem. 2005, 48, 6169.
ꢀ1000-fold (comparing 14m with 14i). Electronic deficient ring
seems to provide less potent compound, which was exemplified
by 14l compared to 14i with a half log decrease of potency. Dihe-
dral angle between thiadiazole and indole moieties can further
negatively impact S1P1 activity. This was supported by the evi-
dence that analogs with Cl substituent on the 3-position of the in-
dole were found to be more than 10-fold less potent than those
without this substituent due to the combination of electronic and
steric effects (comparing 14k, 14r with 14i, 14p, respectively).
Compounds 14b, 14f, 14h, 14i, 14l, 14p were further evaluated
in in vivo lymphopenia model. Hence, the number of lymphocytes
circulating in blood was measured at 4 h after intraperitoneal (ip)
administration. Given the variability of the lymphocyte count, the
reduction was only considered to be very mild (14b, 14f and 14l) or
none (14h, 14i and 14p) (Fig. 2). The lack of pharmacodynamic ef-
fect (i.e., lymphopenia) may be attributed to their poor pharmaco-
kinetic profiles. A generally high in vitro intrinsic clearance in liver
microsomes was observed in these molecules (Table 5). Compound
14l, with the lowest in vitro intrinsic clearance in mouse, gave the
most significant lymphopenia. The screening pharmacokinetic (PK)
model (ip, 3 time points) also demonstrated low blood exposures
within time of period of 0–4 h for 14i and 14p (133 and 114 h
ng/mL for AUC0–4 of 14i and 14p, respectively). These results con-
firmed our hypothesis that the mild lymphopenia was due to poor
drug exposure. More efforts are being focused on improving the PK
profile of the thiadiazole amides, and progress will be reported in
due course.
18. In vitro metabolic incubations were carried out in 96-well polypropylene
plates on a waterbath shaker at 37 °C. Reaction mixtures consisted of 0.5 lM
compound, 0.5 mg/mL of microsomes and 50 mM potassium phosphate buffer
(pH 7.4). To initiate the reaction, beta-NADP (0.44 mM in incubation), G6P
(5.2 mM in incubation) and G6PDH (1.2 U/mL in incubation) were added to
give a final volume of 0.8 mL. An initial time-point (t0) was collected and
subsequent aliquots were collected at 10, 20, 30, 40, 50 and 60 min. Reactions
were terminated by transferring 100 lL of the incubation mixture into 200 lL
of ice-cold acetonitrile:methanol:acetic acid containing internal standard
(80:20:1, v/v/v). Precipitated protein was pelleted by centrifugation and the
resultant supernatant was transferred to a new 96-well polypropylene plate
for LC/MS/MS analysis. Compound remaining as a ratio to internal standard is
measured using specific HPLC-MS/MS methods. The first order elimination rate
constant (k) is determined from non-linear regression analysis of peak area
ratio versus time.
In summary, a series of thiadiazole amides as novel and selec-
tive S1P1 receptor agonists were identified from the high through-
put screening. Based on the template, an initial SAR was explored,