(3R,4S)-4-(2,4,5-Trifluorophenyl)-pyrrolidin-3-ylamine inhibitors of dipeptidyl peptidase IV: Synthesis, in vitro, in vivo, and X-ray crystallographic characterization

Article abstract of DOI:10.1016/j.bmcl.2007.07.081

A series of pyrrolidine based inhibitors of dipeptidyl peptidase IV were developed from a high throughput screening hit for the treatment of type 2 diabetes. Potency, selectivity, and pharmacokinetic properties were optimized resulting in the identification of a pre-clinical candidate for further profiling.

Full text of DOI:10.1016/j.bmcl.2007.07.081

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5638–5642
(3R,4S)-4-(2,4,5-Trifluorophenyl)-pyrrolidin-3-ylamine inhibitors
of dipeptidyl peptidase IV: Synthesis, in vitro, in vivo,
and X-ray crystallographic characterization
Stephen W. Wright,* Mark J. Ammirati, Kim M. Andrews, Anne M. Brodeur,
Dennis E. Danley, Shawn D. Doran, Jay S. Lillquist, Shenping Liu, Lester D. McClure,
R. Kirk McPherson, Thanh V. Olson, Stephen J. Orena, Janice C. Parker,
Benjamin N. Rocke, Walter C. Soeller, Carolyn B. Soglia, Judith L. Treadway,
Maria A. VanVolkenburg, Zhengrong Zhao and Eric D. Cox
Pfizer Global Research and Development, Cardiovascular and Metabolic Diseases, MS 8220-3141,
Eastern Point Road, Groton, CT 06340, USA
Received 22 May 2007; revised 20 July 2007; accepted 23 July 2007
Available online 22 August 2007
Abstract—A series of pyrrolidine based inhibitors of dipeptidyl peptidase IV were developed from a high throughput screening hit
for the treatment of type 2 diabetes. Potency, selectivity, and pharmacokinetic properties were optimized resulting in the identifica-
tion of a pre-clinical candidate for further profiling.
Ó 2007 Elsevier Ltd. All rights reserved.
Type 2 diabetes is a severe and increasingly prevalent dis-
ease.1 Diabetics may suffer debilitating cardiovascular,
eye, kidney, and nerve damage and are at risk of prema-
ture handicap and death due to these and other diabetic
complications, which are the result of glucose toxicity
caused by their hyperglycemia. A progressive reduction
in insulin sensitivity and insulin secretion are hallmarks
of the disease, which eventually resultin failure of the pan-
creatic islet cells and dependence on exogenous insulin.
The incretin hormone glucagon-like peptide 1 (GLP-1)
is a potent stimulator of endogenous insulin release.
GLP-1 has beneficial effects on islet b-cell function and
insulin sensitivity without induction of hypoglycemia.2
Studies in rodents have indicated that GLP-1 may stop
or reverse the loss of b-cell function.3 Unfortunately,
GLP-1 is rapidly degraded in vivo by the serine protease
dipeptidyl peptidase IV (DPP4); therefore inhibition of
DPP4 has emerged as a promising approach for the treat-
ment of Type 2 diabetes.4 This has been substantiated by
the results of clinical trials of several inhibitors,5 including
vildagliptin 1 (LAF-237),6 sitagliptin 2 (MK-0431),7 and
saxagliptin 3 (BMS-477118).8
We have previously reported our studies with the
inhibitor 4 based on a cis-2,5-dicyanopyrrolidine tem-
plate.9 The quinolone 5a, previously prepared as part
of an antibacterial program, was identified in a high
throughput screen of our corporate compound collec-
tion as an inhibitor of DPP4 (IC50 = 140 nM). We were
intrigued by the unusual structure of 5a and sought to
determine whether it could be parlayed into a novel
series of DPP4 inhibitors. This paper reports our dis-
covery of a series of 3-amino-4-phenyl pyrrolidine
inhibitors of DPP4.10
O
CN
O
CN
HO
H
HO
N
N
N
NH2
3
1
F
N
N
O
CN
HO
H
N
F
N
N
N
F3C
NH2
F
O
NC
4
2
H2N
F
N
N
N
O
Keywords: DPPIV inhibitors; Diabetes; Structure-based design.
*
O
F
Corresponding author. Tel. +1 860 441 5831; fax: +1 860 715
4483; e-mail: stephen.w.wright@pfizer.com
5a
OH
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.07.081

S. W. Wright et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5638–5642
5639
Bn
N
Substructure searching and screening quickly revealed
key SAR information (Table 1): the trans-3,4-disub-
stituted pyrrolidine was more potent than the cis iso-
mer, and enzyme inhibitory potency was critically
dependent upon the substitution pattern of the phe-
nyl ring.
a
b, c
NO2
O2N
6
7
Bn
N
H
N
d
O
O
We hypothesized that the trans-3-amino-4-phenyl pyr-
rolidine was the key pharmacophore, and docking
experiments provided further support for this hypothe-
sis. Recognizing that the quinolone fragment of 5 was
undesirable,11 we screened a variety of heterocycles as
potential replacements for the quinolone fragment using
the trans-3-amino-4-phenyl pyrrolidine scaffold. These
compounds were prepared as racemates using the route
shown in Scheme 1.
NH
NH
O
O
8
9
Het
N
e, f
H2N
10
Scheme 1. Reagents and conditions: (a) N-benzyl-N-(methoxy)methyl-
N-trimethylsilylmethylamine, TFA, CH2Cl2, 0 °C; (b) H2, RaNi,
MeOH–NH3, 25 °C; (c) Boc2O, Et3N, THF, 25 °C; (d) H2, Pd/C,
EtOH, AcOH, 50 °C; (e) chloroheterocycle, i-Pr2NEt, t-BuOH, 110 °C
(microwave); (f) HCl, 1,4-dioxane, 25 °C.
Nitrostyrene 6 was treated with the azomethine ylide
generated from N-benzyl-N-(methoxy)methyl-N-trim-
ethylsilylmethylamine to provide pyrrolidine ( )7.12
Reduction of ( )7 with hydrogen in the presence of
Raney nickel was followed by protection with BOC
anhydride to give ( )8. Subsequent debenzylation
with hydrogen and Pd/C in acetic acid afforded
( )9, which was coupled with heteroaryl chlorides
using parallel synthesis to provide ( )10a–10m. This
effort identified the 6-phenylpyrimidine ( )10a as the
best replacement for the quinolone, albeit with some
loss of potency (Table 2).13 Pharmacokinetic evalua-
tion of ( )10a in the rat showed a correlation be-
tween clearance predicted from in vitro microsomal
screens and in vivo clearance, and acceptable bio-
availability (32%), suggesting that derivatives of
( )10a could be found that had acceptable pharmaco-
kinetic properties.
Further improvement in potency was realized by the
introduction of a 2,4,5-trifluorophenyl substitution
pattern on the phenyl ring attached to the pyrrolidine
4-position.14 The necessary 2,4,5-trifluorophenyl tem-
plate (( )12, Scheme 2) was prepared from 11 by the
same sequence of reactions presented in Scheme 1.
Further SAR development was focused on the phenyl
ring attached to the pyrimidine ring as shown in Scheme
3. Suzuki coupling of an appropriately substituted
boronic acid with 4,6-dichloropyrimidine afforded the
Table 2. Heterocycle and DPP4 enzyme inhibition assay results for
compounds 10a–10m
Table 1. Aryl substitution, pyrrolidine stereochemistry, and DPP4
enzyme inhibition assay results for compounds 5a–5k
Het
N
H2N
Ar
H2N
N
F
N
N
O
O
Compound
Heterocycle
DPP4
IC50, nMa
OH
Compound
Ar
3-NH2-4-Ar
stereochemistry
DPP4
IC50, nMa
10a
10b
10c
10d
10e
10f
10g
10h
10i
6-Phenyl-pyrimidin-4-yl
790
1400
6-(3-Cyano)phenyl-pyrimidin-4-yl
6-(4-Methoxy)phenyl-pyrimidin-4-yl
6-(3-Chloro)phenyl-pyrimidin-4-yl
(3-Cyano)pyridin-2-yl
5a
5b
5c
5d
5e
5f
4-FC6H4
trans
cis
140
2000
2300
2900
2-CH3OC6H4
2-CH3OC6H4
C6H5
trans
trans
trans
trans
trans
trans
trans
trans
trans
200
150
>3000
>3000
>3000
>3000
>3000
>3000
>3000
>3000
>3000
6-(2-Chloro)phenyl-pyrimidin-4-yl
6-Chloro-2-phenyl-pyrimidin-4-yl
9-Methyl-9H-purin-6-yl
4-H2NC6H4
4-iPrC6H4
>3000
>3000
>3000
>3000
1000
5g
5h
5i
4-MeO2CC6H4
4-CH3OC6H4
3,4-Cl2C6H4
4-Me2NC6H4
4-HOC6H4
6-(4-Hydroxy)phenyl-pyrimidin-4-yl
6-(4-Chloro)phenyl-pyrimidin-4-yl
(5-Acetyl)pyridin-2-yl
10j
10k
10l
5j
5k
>3000
>3000
1-Quinoxalin-2-yl
1-Quinolin-2-yl
10m
a Recombinant wild type human enzyme. Means of at least three
experiments; standard deviations are 15%. An IC50 of >3000 indi-
cates that no curve was noted in the dose–response up to 3 lM; see
Ref. 9.
a Recombinant wild type human enzyme. Means of at least three
experiments; standard deviations are 15%. An IC50 of >3000 indi-
cates that no curve was noted in the dose–response up to 3 lM; see
Ref. 9.

5640
S. W. Wright et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5638–5642
Table 3. Aryl substitution and DPP4 enzyme inhibition assay results
H
N
for compounds 14a–14w, 15a, and 15b
F
F
a - d
NO2
O
F
R
NH
R
N
O
N
N
N
F
N
N
F
11
12
F
N
F
Scheme 2. Reagents and conditions: (a) N-benzyl-N-(methoxy)methyl-
N-trimethylsilylmethylamine, TFA, CH2Cl2, 0 °C; (b) H2, RaNi,
MeOH–NH3, 25 °C; (c) Boc2O, Et3N, THF, 25 °C; (d) H2, Pd/C,
EtOH, AcOH, 50 °C.
F
H2N
H2N
15
F
14
F
F
F
Compound
R
DPP4 IC50, nMa
14a
14b
14c
14d
14e
14f
14g
14h
14i
Hb
2-Cl
21
260
93
R
N
2-CN
2-COCH3
2-F
N
N
Cl
R
230
57
N
a
b, c
N
N
N
F
2-OMe
3-Cl
3-CN
180
190
37
Cl
H2N
Cl
3-COCH3
3-F
19
51
13
14
F
F
14j
14k
14l
3-OMe
4-Cl
4-CN
47
Scheme 3. Reagents and conditions: (a) Boronic acid, Pd(PPh3)4,
Na2CO3, DME, H2O, 90 °C; (b) ( )12, i-Pr2Net, t-BuOH, 110 °C; (c)
HCl, 1,4-dioxane, 25 °C.
470
130
160
51
14m
14n
14o
14p
14q
14r
14s
14t
14u
14v
14w
15a
15b
4-COCH3
4-F
4-OMe
3-CF3
3-OCF3
3-CH2OH
3-CO2Hb
120
71
4-chloro-6-(substituted)phenyl pyrimidines 13, which
were aminated with ( )12 and deprotected with HCl/
dioxane.
1300
32
5.0
14
This effort revealed that potency was optimal with a sub-
stituent meta- to the biaryl bond, and further effort was
directed toward increasing the diversity of meta- substit-
uents and replacement of the meta- substituted benzene
ring by a 3-pyridyl group (Table 3).
3-NHCOCH3
3-NO2
3-OH
120
70
H
OMe
31
37
a Recombinant wild type human enzyme. Means of at least three
experiments; standard deviations are 15%. An IC50 of >3000 indi-
cates that no curve was noted in the dose–response up to 3 lM; see
Ref. 9.
Compounds 14a, 14j, and 15b emerged as compounds of
interest based on potency and in vitro microsomal sta-
bility. These compounds were prepared as single enanti-
omers from the Boc-protected pyrrolidine enantiomer
(+)12, which was obtained by classical resolution of
the racemate ( )12 using (S)-naproxen (See Table 4).15
b See also Ref. 10.
Table 4. DPP4, DPP2, DPP8 enzyme inhibition assay results and
predicted human extraction ratio for compounds 14a, 14j, and 15b
Compound (+)15b was selected for pharmacokinetic
16
study in the rat and dog (Table 5)
based on its po-
Compound DPP4
DPP2
DPP8
Predicted
tency, selectivity,17 in vivo activity,18 and lack of CYP
inhibition.
IC50, nMa IC50, lMa IC50, lMa human CLb
(+)14a
(+)14j
(+)15b
13
14
18
15
19
3.0
1.6
3.2
0.81
<0.31
<0.31
The co-crystal structure of (+)15b confirmed the ex-
pected binding mode (Fig. 1). The primary amine makes
hydrogen bonds to Y662, E205, and E206. The trifluor-
ophenyl ring occupies the S1 pocket (Y662, Y666, V656,
V711, Y631, and W659). The pyridine extends into a
pocket comprised of F357, R358, E206, and S209, and
the methoxy group forms a hydrogen bond to R358.
The pyrimidine ring is involved in a pi-stacking interac-
tion with F357.
6.4
a Recombinant wild type human enzyme. Means of at least three
experiments; standard deviations are 15%; see Ref. 9.
b Predicted hepatic clearance (CLh) from human liver microsomal
lability assay divided by human liver blood flow (20 mL/min/kg).
properties, and selectivity versus other DPP isoforms.
Compound (+)15b was identified as a compound for fur-
ther pre-clinical profiling due to its PK/PD profile in the
rat.
In conclusion, a high throughput screening hit identified
a trans-3-amino-4-phenyl pyrrolidine as a template for a
novel series of DPP4 inhibitors. By use of structure-
based design and parallel synthesis, the series was
advanced with improvements in potency, ADME
X-Ray crystal structures have been deposited in the
RCSB protein data bank with code 2QJR.

S. W. Wright et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5638–5642
5641
Table 5. Pharmacokinetic parameters for compound 15b in rat and
dog
Pacini, G.; Foley, J. E.; Schweizer, A. Diabetes Care 2005,
28, 1936.
6. 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.
7. Kim, D.; Wang, L.; Beconi, M.; Eiermann, G. J.;
Fisher, M. H.; He, H.; Hickey, G. J.; Kowalchick, J. E.;
Leiting, B.; Lyons, K.; Marsilio, F.; McCann, M. E.;
Patel, R. A.; Petrov, A.; Scapin, G.; Patel, S. B.; Sinha
Roy, R.; Wu, J. K.; Wyvratt, M. J.; Zhang, B. B.; Zhu,
L.; Thornberry, N. A.; Weber, A. E. J. Med. Chem.
2005, 48, 141.
Species Unbound
CL,
fraction (%) mL/min/kg L/kg h (i.v.) h (p.o.)
Vss, t1/2
,
t1/2
,
F%
Rat
Dog
6
17
6
4.4
4.8
0.67 2.8
1.9
3.3
nda
57
5.1
nda
50b
Human
1.3b
2.0b
a Not determined.
b Projected values based on rat single-species allometric scaling (see
Ref. 19).
8. 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.
9. Wright, S. W.; Ammirati, M. J.; Andrews, K. M.;
Brodeur, A. M.; Danley, D. E.; Doran, S. D.; Soglia, C.
B.; Lillquist, J. S.; McClure, L. D.; McPherson, R. K.;
Orena, S. J.; Parker, J. C.; Polivkova, J.; Qiu, X.; Soeller,
W. C.; Treadway, J. L.; VanVolkenburg, M. A.; Wilder,
D. C.; Olson, T. V. J. Med. Chem. 2006, 49, 3068.
10. This template was independently discovered at Abbott
Laboratories from a high throughput screening hit, see:
Backes, B. J.; Longenecker, K.; Hamilton, G. L.; Stewart,
K.; Lai, C.; Kopecka, H.; von Geldern, T. W.; Madar, D.
J.; Pei, Z.; Lubben, T. H.; Zinker, B. A.; Tian, Z.;
Ballaron, S. J.; Stashko, M. A.; Mika, A. K.; Beno, D. W.
A.; Kempf-Grote, A. J.; Black-Schaefer, C.; Sham, H. L.;
Trevillyan, J. M. Bioorg. Med. Chem. Lett. 2007, 17, 2005.
11. Miolo, G.; Violaa, G.; Vedaldia, D.; Dall’Acquaa, F.;
Fravolinib, A.; Tabarrinib, O.; Cecchetti, V. Toxicol. In
Vitro 2002, 16, 683.
Figure 1. Compound (+)15b co-crystallized with human DPP4 show-
ing hydrogen bond from Arg 358 to methoxypyridine.
12. Terao, Y.; Kotaki, H.; Imai, N.; Achiwa, K. Chem.
Pharm. Bull. 1985, 33, 2762–2766.
13. Analogs incorporating amides, amines, sulfonamides, and
urea substituents placed on the pyrrolidine ring nitrogen
atom were also evaluated; in general these were less potent
as DPP4 inhibitors.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2007.07.081.
14. The application disclosing MK-0431 published concur-
rently with our efforts to increase potency by appropriate
substitution of this phenyl ring; for example, a 2,5-
difluorophenyl substitution improved potency to 61 nM
from 790 nM for 10a. We obtained internal co-crystal
structures using compounds similar to 2 and our pyrrol-
idine inhibitors and found a similar aryl binding mode in
these structures. See Edmondson, D. D.; Parmee, E.;
Weber, A. E.; Xu, J. Patent WO 03/000180, 2003; Chem.
Abstr. 2003, 138, 73536.
15. The absolute configuration of (+)12 ([a]D = 61° (c = 1.5,
EtOH) was based upon agreement with protein crystal
structures as well as identity to a sample of (+)12 prepared
by diastereoselective azomethine ylide cycloaddition
to (S)-4-phenyl-3-[(E)-3-(2,4,5-trifluoro-phenyl)-acryloyl]-
oxazolidin-2- one; see U.S. Patent 5,618,949, 1997; Chem.
Abstr. 1997, 126, 330549.
16. All procedures involving animals were reviewed and
approved by the Pfizer Institutional Animal Care and
Use Committee. Oral pharmacokinetic evaluations were
dosed in 0.5% methylcellulose formulation at 5 mg kgÀ1 in
the rat and in sterile saline solutions at 1 mg kgÀ1 in the
dog. Intravenous evaluations were conducted at 1 mg kgÀ1
in sterile saline. Pharmacodynamic determinations were
made by assay of DPP4 activity in plasma samples taken
from the PK study; see reference 7 for plasma DPP4
References and notes
1. Fact Sheet No. 138. World Health Organization, Geneva,
2002.
2. Knudsen, L. B. J. Med. Chem. 2004, 47, 4128.
3. Farilla, L.; Hui, H.; Bertolotto, C.; Kang, E.; Bulotta, A.;
Di Mario, U.; Perfetti, R. Endocrinology 2002, 143, 4397.
4. For recent reviews, see: (a) Mest, H.-J.; Mentlein, R.
Diabetologia 2005, 48, 616; (b) Weber, A. E. J. Med.
Chem. 2004, 47, 4135; (c) Drucker, D. J. Expert Opin.
Invest. Drugs 2003, 12, 87; (d) Augustyns, K.; Van der
Veken, P.; Senten, K.; Haemers, A. Expert Opin. Ther.
Patents 2003, 13, 499.
´
5. (a) Ahren, B.; Simonsson, E.; Larsson, H.; Landin-Olsson,
M.; Torgeirsson, H.; Jansson, P.-A.; Sandqvist, M.;
Bavenholm, P.; Efendic, S.; Eriksson, J. W.; Dickinson,
´
S.; Holmes, D. Diabetes Care 2002, 25, 869; (b) Ahren, B.;
Gomis, R.; Standl, E.; Mills, D.; Schweizer, A. Diabetes
´
Care 2004, 27, 2874; (c) Ahren, B.; Landing-Olsson, L.;
Jansson, P-A.; Sevensson, M.; Holmes, D.; Schweizer, A.
´
J. Clin. Endocrinol. Metab. 2004, 89, 2078; (d) Ahren, B.;

5642
S. W. Wright et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5638–5642
activity assay. PK and PD data for (+)15b are presented in
the Supporting information.
17. Compound (+)15b had IC50 > 30 lM against DPP3,
DPP9, FAP, and POP.
measured; see Ref. 9. See Supporting information for
in vivo glucose lowering data for (+)15b.
0.75
19. CLpredicted = (BWhuman/BWrat
Fuanimal
)
(Fuhuman/BWhuman), where BW, body weight,
(CLanimal BWanimal/
* *
)
*
18. Compound (+)15b was dosed in fasted, diabetic KK/H1J
mice and the response of the mice in an OGTT was
Fu, fraction unbound, CL, i.v. clearance, and 0.75,
allometric scaling factor.