Bicyclic cyanothiazolidines as novel dipeptidyl peptidase 4 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.05.048

The synthesis and biochemical evaluation of novel cyanothiazolidine inhibitors of dipeptidyl peptidase 4 (DPP4) is described. Their main structural feature is a constrained bicyclic core that prevents the intramolecular formation of inactive cyclic species. The inhibitors show good to moderate biochemical potency against DPP4 and display distinct selectivity profiles towards DPP7, DPP8 and DPP9 depending on their substitution.

Full text of DOI:10.1016/j.bmcl.2009.05.048

Bioorganic & Medicinal Chemistry Letters 19 (2009) 4437–4440
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Bicyclic cyanothiazolidines as novel dipeptidyl peptidase 4 inhibitors
a
b
b
b
b
Juan M. Betancort a, , David T. Winn , Ruzhang Liu , Quansheng Xu , Junjuan Liu , Wensheng Liao ,
*
Shu-Hui Chen b, David Carney a, Denise Hanway a, James Schmeits a, Xinqiang Li a, Eric Gordon a,
David A. Campbell a,
*
a Phenomix Corporation, 5871 Oberlin Drive, Suite 200, San Diego, CA 92121, USA
b WuXi AppTec, Inc., No. 1 Building, 288 FuTe ZhongLu, Shanghai 200131, China
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and biochemical evaluation of novel cyanothiazolidine inhibitors of dipeptidyl peptidase 4
(DPP4) is described. Their main structural feature is a constrained bicyclic core that prevents the intra-
molecular formation of inactive cyclic species. The inhibitors show good to moderate biochemical
potency against DPP4 and display distinct selectivity profiles towards DPP7, DPP8 and DPP9 depending
on their substitution.
Received 30 April 2009
Revised 11 May 2009
Accepted 13 May 2009
Available online 18 May 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Diabetes
DPP4
Cyanopyrrolidine
Diabetes is a major health problem that is approaching pan-
demic proportions worldwide. The search for new therapies with
novel mechanistic approaches and improved safety profiles to con-
trol this chronic metabolic disease has come to the forefront of re-
search in medicinal chemistry.1 Type-2 diabetes mellitus accounts
for over 90% of the disease cases and is characterized by high levels
of glucose resulting from progressive insulin resistance. Glucagon-
like peptide-1 (GLP-1) is an incretin hormone secreted by intestinal
L-cells in response to food intake and is a key player in the modu-
lation of blood glucose levels.2 GLP-1 stimulates insulin secretion,
delays glucose absorption and inhibits glucagon secretion leading
to reduced hepatic glucose production. The primary physiological
route of GLP-1 degradation is cleavage by dipeptidyl peptidase 4
(DPP4), a ubiquitous and highly specific serine protease that pref-
erentially cleaves N-terminal dipeptides from polypeptides with
Most of the DPP4 inhibitors described to date have been dipep-
tidomimetics resembling the natural substrate preference of the
enzyme, in particular compounds containing L-proline mimetics
in the P1 position. A widely explored class of these inhibitors in-
cludes N-substituted cyanopyrrolidines such as NVP-728 (1),
LAF-237 (Vildagliptin, 2), BMS-477118 (Saxagliptin, 3) and ABT-
279 (4) (Fig. 1).6 The cyano group common to all of these inhibitors
acts as an electrophilic trap for the Ser 630 of the DPP4 catalytic
triad by formation of an imidate adduct.
It is well established that the P-2 site amine of these molecules
can intramolecularly attack the carbon of the nitrile group under
neutral and basic aqueous conditions to form an inactive cyclic
amidine (Fig. 2). In fact, this property has been closely monitored
OH
H
L-proline or
L-alanine at the penultimate position.
N
N
N
N
H
DPP4 inhibition, through the preservation of active GLP-1 levels,
N
CN
N
H
O
offers a number of potential advantages over existing diabetes
therapies including minimized risk for hypoglycemia and im-
proved b-cell survival.3 Clinical studies have shown that DPP4
inhibitors are well tolerated, lower blood glucose levels, improve
insulin response to oral glucose, and decrease HbA1c levels in pa-
tients with type 2 diabetes.4 The most advanced compound (Sitag-
NC
CN
O
1
2
N
NH2
HO
TM
HO2C
N
liptin, Januvia ) gained FDA approval in 2006 for the treatment of
N
type 2 diabetes.5
N
N
H
CN
O
CN
O
3
4
* Corresponding authors. Tel.: +1 858 731 5259; fax: +1 858 731 5225.
E-mail addresses: juan.betancort@phenomixcorp.com (J.M. Betancort),
david.campbell@phenomixcorp.com (D.A. Campbell).
Figure 1. Cyanopyrrolidine inhibitors of DPP4.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.05.048

4438
J. M. Betancort et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4437–4440
R'
NH
R'
N
O
N
CN
O
CN
O
R
O
R
O
NH
H2N
H
N
N
H
CN
N
H
H2N
R'
N
S
N
N
S
R
5
IC50 > 1.6 uM
6
IC50 > 1.6 uM
trans-rotamer
cis-rotamer
O
Figure 2. Formation of cyclic amidine adducts.
CN
CN
O
H2N
N
N
H
H2N
by the medicinal chemistry groups working with this class of
inhibitors in order to minimize the formation of inactive species
and increase chemical stability.7
S
H
S
7
IC50 = 105 nM
8
IC50 > 1.6 uM
While increasing the size of the R or R0 substituent has been a
strategy pursued in some DPP4 programs, a straightforward alter-
native method of avoiding the formation of inactive cyclic species
Scheme 1. DPP4 activity of bicyclic cyanothiazolidine inhibitors.
could be the introduction of a short tether between the P-2 a-ami-
no acid and the five-membered P-1 heterocycle (Fig. 3). This tether
renders the amine unavailable for the intramolecular nucleophilic
attack on the nitrile group. Examination of the published X-ray
structures of DPP4 in complex with known inhibitors supported
this hypothesis due to the presence of the large S2 enzyme pocket
which would permit binding of the tethered molecule.8 The result-
ing rigid scaffold would in turn provide an entropically favored
platform to study the effect of the introduction of additional sub-
stituents on inhibitory potency and selectivity.9
O
CN
O
CN
N
H
N
H
HN
HN
S
S
9
IC50 > 16 uM
10 IC50 > 16 uM
Scheme 2. N-alkylated cyanothiazolidine inhibitors.
For the sake of synthetic ease, the thiazolidine ring (X = S) was
chosen as the P-1 proline mimetic. Initial efforts focused on deter-
mining the appropriate ring size required for activity (Scheme 1).
Since it was previously reported that the DPP4 enzyme could toler-
ate both natural and unnatural amino acids in its S2 pocket,10 com-
pounds 5 through 8 were synthesized following the methodology
developed by Baldwin et al. for related systems.11,12 The hexahy-
dropyrrolo[2,1-b]thiazole 7 possessing the unnatural stereochem-
istry in the P-2 amino acid exhibits the best potency against
DPP4.13
These preliminary results prompted us to explore lead optimi-
zation for the series. As exemplified by compounds 1, 2 and 4,
one of the most common strategies within the cyanopyrrolidine
series has been the alkylation of the amino group. Inhibitors 9
and 10 were prepared by reductive alkylation of amine 7 with
the corresponding carbonyl compounds.14 However, when evalu-
ated against DPP4, they showed a complete lack of activity
(Scheme 2). As a result, N-alkylated analogs were not further pur-
sued in our series.
pling conditions was carried out. The most efficient method
involved refluxing the mixture of thiazolidines in toluene in the
presence of pTSA.16 Aminolysis of methyl ester 16 followed by
dehydration of the resulting amide with TFAA yielded nitrile 17. Fi-
nal deprotection of the Boc group provided the free amine that was
isolated as its hydrochloride salt.
c-e
a, b
NH2
COOH
BocHN
Ph
Ph
N
COOMe
COOMe
Ph
Ph
11
12
13
OH
O
COOMe
f, g
h
HOOC
HN
BocHN
Ph
BocHN
Ph
S
O
14
15
In order to expand our SAR, we prepared a series of compounds
bearing a wide range of substituents in the
a-position of the P-2
amino acid. The desired compounds were synthesized according
to the procedure exemplified in Scheme 3 for the phenyl analog.
The benzaldimine of L-phenyl glycine methyl ester 12 was treated
with tBuOK and alkylated with allyl bromide. After hydrolysis of
the imine and protection of the resulting amine by a Boc group,
ozonolysis of the olefin followed by reductive work-up with PPh3
provided the corresponding hemiacetal.15 Condensation of 14 with
O
CO2Me
O
CN
N
i
j, k
N
Ph
BocHN
Ph
BocHN
S
S
H
H
16
17
O
CN
N
L-cysteine methyl ester hydrochloride gave a circa 1.6:1 mixture of
l-m
two pairs of the diastereomeric thiazolidines 15. In order to in-
crease the yield of isomer 16, a systematic investigation of cou-
Ph
HCl.H2N
S
H
18
O
CN
R
Scheme 3. Reagents and conditions: (a) SOCl2, MeOH, 0 °C, 2 h, 98%; (b) PhCHO,
TEA, MgSO4, CH2Cl2, rt, 18 h, 97%; (c) tBuOK, CH2CHCH2Br, THF, rt, 18 h; (d) 6 N HCl,
EtOAc, 1 h, 64%; (e) Boc2O, THF, reflux, 18 h, 87%; (f) O3, PPh3, CH2Cl2, ꢀ78 °C, 36%;
H
N
N
R'
X
(g) LiOH, THF, 1 h, 95%; (h)
L
-Cys-OMeꢁHCl, NaHCO3, EtOH, 16 h, 51%; (i) pTsOH,
( )n
Toluene, reflux, 1 h, 23%; (j) NH3, MeOH, rt, 2 h, 97%; (k) TFAA, TEA, CH2Cl2, rt, 1 h,
84%; (l) TFA, CH2Cl2, rt, 1 h, 14% after preparative HPLC; (m) HCl (g), CH2Cl2/Et2O,
2 h, 53%.
Figure 3. Constrained cyano derivatives.

J. M. Betancort et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4437–4440
4439
The substituted cyanothiazolidines 18–28 were evaluated for
Acknowledgements
DPP4 inhibitory activity by a fluorescence assay using Gly-Pro-
AMC as substrate.17 Inhibitors were also assayed for their selec-
tivity profiles against a variety of DPP4 homologues including
quiescent prolyl peptidase (QPP/DPPII/DPP7), DPP8 and DPP9
(Table 1).18
The authors would like to thank Drs. Michael Hepperle and
Marie O’Farrell for reviewing the manuscript.
References and notes
The substituted cyanothiazolidines exhibited potencies against
DPP4 ranging from moderate to good in the double digit nanomo-
lar range, with the phenyl substituted compound 18 displaying the
best potency at 26 nM. Selectivity profiles against the homologous
proteases were clearly modulated by the substitution present on
the lactam ring, with small alkyl groups simultaneously displaying
improved selectivities against DPP7 and DPP8. In general, the
cyanothiazolidines show the least selectivity against DPP9. This
is also the case for other cyanopyrrolidines described in the litera-
ture such as LAF-237 (2) and BMS-477118 (3). Larger substituents
on the lactam ring show decreased selectivity against DPP9. In fact,
compounds 25–27 exhibit greater potency against DPP9 than
DPP4. Though the X-ray structure of DPP9 is not available, detailed
homology models have recently been established to rationalize the
selectivity of known inhibitors against DPP4, DPP8 and DPP9.19
These studies show that the P2 pocket of DPP9 is larger than that
on DPP4. The SAR presented here for the cyanothiazolidines rein-
forces that observation. The phenyl substituted analog 18 shows
the best balance of potency and selectivity within this series. But
even small changes such as the introduction of a halogen substitu-
tion (compound 28) on the aromatic ring have a negative impact
on the overall profile of the compound.
1. (a) Ross, S. A.; Gulve, E. A.; Wang, M. Chem. Rev. 2004, 104, 1255; (b) Patel, M.;
Rybczynski, P. Exp. Opin. Invest. Drugs 2003, 12, 623.
2. (a) Arulmozhi, D. K.; Portha, B. Eur. J. Pharm. Sci. 2006, 28, 96; (b) Knudsen, L. B.
J. Med. Chem. 2004, 47, 4128; (c) Orsakov, C. Diabetologia 1992, 35, 701.
3. (a) Augustyns, K.; Van der Veken, P.; Senten, K.; Haemers, A. Curr. Med. Chem.
2005, 12, 971; (b) Weber, A. E. J. Med. Chem. 2004, 47, 4135; (c) Drucker, D. J.
Exp. Opin. Invest. Drugs 2003, 12, 87; (d) Pospisilik, A.; Martin, J.; Doty, T.; Ehses,
J. A.; Pamir, N.; Lynn, F. C.; Piteau, S.; Demuth, H.-U.; McIntosh, C. H. S.;
Pederson, R. A. Diabetes 2003, 52, 741.
4. (a) Herman, G. A.; Bergman, A.; Stevens, C.; Kotey, P.; Yi, B. M.; Zhao, P.;
Dietrich, B.; Golor, G.; Schroedter, A.; Keymeulen, B.; Lasseter, K. C.; Kipnes, M.
S.; Snyder, K.; Hilliard, D.; Tanen, M.; Cilissen, C.; De Smet, M.; Lepeleire, I.; Van
Dyck, K.; Wang, A. Q.; Zeng, W.; Davies, M. J.; Tanaka, W.; Holst, J. J.; Deacon, C.
F.; Gottesdiener, K. M.; Wagner, J. A. J. Clin. Endocrinol. Metab. 2006, 91, 4612;
(b) Ahren, B.; Pacini, G.; Foley, J. E.; Schweizer, A. Diabetes Care 2005, 28, 1936.
5. Thornberry, N. A.; Weber, A. E. Curr. Top. Med. Chem. 2007, 7, 557.
6. (a) Peters, J.-U. Curr. Top. Med. Chem. 2007, 7, 579; (b) Madar, D. J.; Kopecka, H.;
Pireh, D.; Yong, H.; Pei, Z.; Li, X.; Wiedeman, P. E.; Djuric, S. W.; Von Geldern, T.
W.; Fickes, M. G.; Bhagavatula, L.; McDermott, T.; Wittenberger, S.; Richards, S.
J.; Longenecker, K. L.; Stewart, K. D.; Lubben, T. H.; Ballaron, S. J.; Stashko, M. A.;
Long, M. A.; Wells, H.; Zinker, B. A.; Mika, Am. K.; Beno, D. W. A.; Kempf-Grote,
A. J.; Polakowski, J.; Segreti, J.; Reinhart, G. A.; Fryer, R. M.; Sham, H. L.;
Trevillyan, J. M. J. Med. Chem. 2006, 49, 6416; (c) Augeri, D. J.; Robl, J. A.;
Betebenner, D. A.; Magnin, D. R.; Khanna, A.; Robertson, J. G.; Wang, A.;
Simpkins, L. M.; Taunk, P.; Huang, Q.; Han, S.-P.; Abboa-Offei, B.; Cap, M.; Xin,
L.; Tao, L.; Tozzo, E.; Welzel, G. E.; Egan, D. M.; Marcinkeviciene, J.; Chang, S. Y.;
Biller, S. A.; Kirby, M. S.; Parker, R. A.; Hamann, L. G. J. Med. Chem. 2005, 48,
5025; (d) Villhauer, E. B.; Brinkman, J. A.; Naderi, G. B.; Burkey, B. F.; Dunning,
B. E.; Prasad, K.; Mangold, B. L.; Russell, M. E.; Hughes, T. E. J. Med. Chem. 2003,
46, 2774.
Although compounds 18–28 display different degrees of activ-
ity against the other proteases in the DPP family, they did not
exhibit significant inhibition against fibroblast activation protein
7. See for example Ref. 6d and Sakashita, H.; Akahoshi, F.; Kitajima, H.;
Tsutsumiuchi, R.; Hayashi, Y. Bioorg. Med. Chem. 2006, 14, 3662.
(FAP) (IC50 >10 lM), another proline-specific enzyme.
8. (a) Oefner, C.; D’Arcy, A.; Sweeney, A. M.; Pierau, S.; Gardiner, R.; Dale, G. E. Acta
Crystallogr., Sect. D 2003, 59, 1206; (b) Rasmussen, H. B.; Branner, S.; Wiberg, F.
C.; Wagtmann, N. Nat. Struct. Biol. 2003, 10, 19.
9. Campbell, D. A.; Betancort, J. M.; Winn, D. T. U.S. Patent Application 06/
009518A1.
10. Heins, J.; Welker, P.; Schonlein, C.; Born, I.; Hartrodt, B.; Neubert, K.; Tsuru, D.;
Barth, A. Biochim. Biophys. Acta 1988, 954, 161.
11. Baldwin, J. E.; Freeman, R. T.; Lowe, C.; Schofield, C. J.; Lee, E. Tetrahedron 1989,45, 4537.
12. Compound 6 was prepared from Fmoc-BTD purchased from Neosystems (Cat.
No. FB02601).
In summary, a novel series of bicyclic cyanothiazolidine inhibi-
tors of dipeptidyl peptidase 4 has been developed.20 The constrained
inhibitors were designed to avoid the formation of inactive cyclic
amidine species. They show good to moderate biochemical potency
and depending on their substitution exhibitdifferent selectivity pro-
files against closely related serine proteases. The incorporation of a
phenyl group in the P-2 amino acid provides the greatest potency
13. The amine and cyano groups adopt a relative spatial arrangement similar to the
one observed in the X-ray co-crystal structure of BMS-477118 (Saxagliptin, 3)
complexed with DPP4 Metzler, W. J.; Yanchunas, J.; Weigelt, C.; Kish, K.; Klei, H.
E.; Xie, D.; Zhang, Y.; Corbett, M.; Tamura, J. K.; He, B.; Hamann, L. G.; Kirby, M.
S.; Marcinkeviciene, J. Protein Sci. 2008, 17, 412.
combined with increased selectivity. While substitutionin the a-po-
sition of the P-2 amino acid is well tolerated, N-alkylation had a
negative effect on potency.
Table 1
14. Szardenings, A. K.; Burkoth, T. S.; Look, G. C.; Campbell, D. A. J. Org. Chem. 1996,
61, 6720.
DPP4 inhibition and selectivity profile of substituted bicyclic cyanothiazolidines
15. Fuchi, N.; Doi, T.; Harada, T.; Urban, J.; Cao, B.; Kahn, M.; Takahashi, T.
Tetrahedron Lett. 2001, 42, 1305.
16. Extensive NOE studies were carried out in conjunction with X-ray structure
determination of a key intermediate in order to assign the stereochemistry.
See: Liu, R.; Wu, Y.; Li, Q.; Liao, W.; Chen, S.-H.; Li, G.; Betancort, J. M.; Winn, D.
T.; Campbell, D. A. Tetrahedron 2008, 64, 4363.
O
CN
N
H
R
HCl.H2N
S
17. The reported IC50 values are an average of at least two replicates. Human DPP4
was purchased from Research Diagnostics, Inc., Concord, MA. The full length
cDNAs for human DPP family members DPP7, DPP8, DPP9, and fibroblast
activation protein (FAP) were obtained from Open Biosystems, Huntsville, AL.
The cDNAs were cloned into the pFastBac vector with the addition of an N-
terminal FLAG tag on DPP7, C-terminal 6xHis tags on DPP8 and DPP9, and an
N-terminal 6xHis tag on FAP. Baculovirus was prepared using the Bac-to-Bac
Baculovirus Expression System (Invitrogen, Carlsbad, CA). The cDNAs in the
final baculovirus constructs were sequence verified. S9 insect cells (Invitrogen,
Carlsbad, CA) were grown to mid-log phase at 27 °C with shaking at 125 rpm
and then adjusted to 2 x 10E6/ml just prior to infection with DPP7, DPP8, DPP9,
and FAP baculoviral stocks at an MOI of 4. The infected cells were grown for
48 h and the cell pellets harvested and frozen until purification. DPP7 was
purified using anti-FLAG immunoaffinity gel (Sigma, St. Louis, MO) according
to the manufacturer’s instructions. DPP8, DPP9, and FAP were purified using a
B-PER 6xHis Fusion Protein Column Purification Kit (Pierce, Rockford, IL). Test
compounds were dissolved in DMSO or in 50 mM glycine buffer (pH 3.0).
Recombinant human DPPs were incubated with inhibitor at concentrations of
Compounds
R
DPP4 IC50 nM
DPP7 SIa
DPP8 SIa
DPP9 SIa
2b
3b
7
3
3
105
26
56
33
63
53
109
89
59
>16,000
>3333
>476
101
129
277
6
835
114
40
100
30
31
23
15
41
7
13
22
3
10
3
4
2
3
36
1
H
18
19
20
21
22
23
24
25
26
27
28
Ph
Et
iPr
iBu
sBu(S)
Bn
59
23
32
35
3
1
17
Cp
Cy
BnCH2
CyCH2
4-FPh
4
5
6
44
0.5
0.2
0.4
4
205
118
152
0.5 nM to 50
lM in reaction buffer at 27 °C for 10 min before addition of
a
substrates Gly-Pro AMC (7-amido-4-methylcoumarin) for DPP4, Lys-Pro AMC
for DPP7, and Ala-Pro AMC for DPP8, DPP9 and FAP. Cleavage of AMC was
monitored using an excitation wavelength of 360 nm and an emission
Selectivity Index (SI) determined by dividing enzyme IC50 by DPP4 IC50
.
b
Data generated in our laboratories. Though absolute numbers differ from
published data, the relative selectivity and ranking order is the same.

4440
J. M. Betancort et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4437–4440
wavelength of 460 nm. Reaction buffers were 25 mM (2-(4-Morpholino)-
19. Rummey, C.; Metz, G. Proteins 2007, 66, 160.
Ethane Sulfonic Acid), pH 5.5 for DPP7; 25 mM Tris, pH 8, 1% Triton X-100,
100 mM NaCl for DPP8; and 25 mM Tris pH 8 for DPP9 and FAP. IC50 and IC90
calculations were performed by nonlinear regression analysis using Prism
software (GraphPad Software, Inc., San Diego, CA).
20. In the course of the preparation of this manuscript we became aware of
simultaneous efforts by other researchers around some of the compounds
described in this manuscript. See: Wagner, H.; Schoenafinger, K.; Jaehne, G.;
Gaul, H.; Buning, C.; Tschank, G.; Werner, U. PCT Patent Application WO 05/
012312A1.
18. Sedo, A.; Malik, R. Biochim. Biophys. Acta 2001, 1550, 107.