1-[[(3-Hydroxy-1-adamantyl)amino]acetyl]-2-cyano-(S)-pyrrolidine: A potent, selective, and orally bioavailable dipeptidyl peptidase IV inhibitor with antihyperglycemic properties

Article abstract of DOI:10.1021/jm030091l

Dipeptidyl peptidase IV (DPP-IV) inhibition has the potential to become a valuable therapy for type 2 diabetes. The synthesis and structure - activity relationship of a new DPP-IV inhibitor class, N-substituted-glycyl-2-cyanopyrrolidines, are described as well as the path that led from clinical development compound 1-[2-[5-cyanopyridin-2-yl)amino]ethylamino]acetyl-2-cyano-(S)pyrrolidine (NVP-DPP728, 8c) to its follow-up, 1-[[(3-hydroxy-1-adamantyl)amino]acetyl]-2-cyano-(S)pyrrolidine (NVP-LAF237, 12j). The pharmacological profile of 12j in obese Zucker fa/fa rats along with pharmacokinetic profile comparison of 8c and 12j in normal cynomolgus monkeys is discussed. The results suggest that 12j is a potent, stable, selective DPP-IV inhibitor possessing excellent oral bioavailability and potent antihyperglycemic activity with potential for once-a-day administration.

Full text of DOI:10.1021/jm030091l

2774
J . Med. Chem. 2003, 46, 2774-2789
1-[[(3-Hyd r oxy-1-a d a m a n tyl)a m in o]a cetyl]-2-cya n o-(S)-p yr r olid in e: A P oten t,
Selective, a n d Or a lly Bioa va ila ble Dip ep tid yl P ep tid a se IV In h ibitor w ith
An tih yp er glycem ic P r op er ties
Edwin B. Villhauer,*,† J ohn A. Brinkman,† Goli B. Naderi,† Bryan F. Burkey,‡ Beth E. Dunning,‡ Kapa Prasad,§
Bonnie L. Mangold,| Mary E. Russell,‡ and Thomas E. Hughes‡
Novartis Institute for Biomedical Research, One Health Plaza, East Hanover, New J ersey 07936
Received February 21, 2003
Dipeptidyl peptidase IV (DPP-IV) inhibition has the potential to become a valuable therapy
for type 2 diabetes. The synthesis and structure-activity relationship of a new DPP-IV inhibitor
class, N-substituted-glycyl-2-cyanopyrrolidines, are described as well as the path that led from
clinical development compound 1-[2-[5-cyanopyridin-2-yl)amino]ethylamino]acetyl-2-cyano-(S)-
pyrrolidine (NVP-DPP728, 8c) to its follow-up, 1-[[(3-hydroxy-1-adamantyl) amino]acetyl]-2-
cyano-(S)-pyrrolidine (NVP-LAF237, 12j). The pharmacological profile of 12j in obese Zucker
fa/fa rats along with pharmacokinetic profile comparison of 8c and 12j in normal cynomolgus
monkeys is discussed. The results suggest that 12j is a potent, stable, selective DPP-IV inhibitor
possessing excellent oral bioavailability and potent antihyperglycemic activity with potential
for once-a-day administration.
Ch a r t 1
In tr od u ction
Dipeptidyl peptidase IV (DPP-IV, EC 3.4.14.5) is a
ubiquitous yet highly specific serine protease that
cleaves N-terminal dipeptides from polypeptides with
L-proline or L-alanine at the penultimate position.1 The
biological activities of many circulating regulatory pep-
tides are altered or abolished by the action of DPP-IV
in vitro.2 However, in part because of the multiplicity
of enzymes exhibiting DPP-IV-like activity,3 the in vivo
role of DPP-IV in mediating the cleavage of endogenous
peptides and the consequences of its inhibition has yet
to be established. One exception is with the incretin
known as glucagon-like peptide-1 (GLP-1), the most
potent insulinotropic hormone known.4 Numerous stud-
ies with DPP-IV5 and DPP-IV inhibitors6-10 support a
principal role of DPP-IV in the inactivation of GLP-1
in vivo. More importantly, the contribution of DPP-IV
catalytic activity to blood glucose control through GLP-1
inactivation has recently been confirmed.11 Because of
multiple benefits of GLP-1 augmentation, DPP-IV in-
hibition has been recognized as a mechanistic approach
of potential value in the treatment of type 2 diabetes.12
Extending the duration of action of GLP-1 would
stimulate insulin secretion, inhibit glucagon release,13
and slow gastric emptying,14 each a benefit in the
control of glucose homeostasis in patients with type 2
diabetes or its predecessor syndromes. DPP-IV inhibi-
tion, through the preservation of active GLP-1 levels,
has the potential to slow or even prevent the progression
of type 2 diabetes by stimulating insulin gene expression
and biosynthesis, increasing the expression of the
â-cell’s glucose-sensing mechanism and promoting genes
involved in the differentiation (neogenesis) of â-cells.15
GLP-1 may play a role in acutely suppressing appetite
in humans,16 in mediating peripheral glucose uptake,17
and may confer maximum glucose sensitivity to the
hepatoportal glucose sensor.18 As the glucose lowering
effects of GLP-1 are dependent on elevated blood glucose
and subside as glucose levels return to normal, the
probability of hypoglycemia during treatment with a
DPP-IV inhibitor is expected to be very low.19 With few
exceptions,20-22 DPP-IV inhibitors resemble the P2-P1
dipeptidyl substrate cleavage product, where the P-1
site contains a proline mimic.23,24 A straightforward
replacement of the normally cleaved P-1 substrate
amide (R in Chart 1) with an electrophile provides both
irreversible (R ) P(O)(OPh)2, CONHOCOR′) and re-
versible (R ) B(OH)2, H, CN) inhibitors.25 The nitrile
is especially interesting and provides nanomolar inhibi-
tion and chemical stability adequate for oral adminis-
tration (Xaa-(2S)-cyanopyrrolidines26-28 and Xaa-(4R)-
cyanothiazolidines29). Cyclohexylglycine-(2S)-cyanopyr-
rolidine (9a ) is one of the more potent, selective, and
stable representatives of this nitrile class (Ki of 1.4 nM,
>1000-fold selectivity over closely related peptidases
and t1/2 stability of > 48 h at pH 7.4).27
* To whom correspondence should be addressed. Phone: 862-778-
8481. Fax 973-781-2188. E-mail: edwin.villhauer@pharma.novartis.com.
† Chemistry Department, Metabolic and Cardiovascular Diseases
Research.
‡ Metabolic Diseases Pharmacology Department, Metabolic and
Cardiovascular Diseases Research.
§ Chemical Development.
| Drug Metabolism & Pharmacokinetics Department.
10.1021/jm030091l CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/24/2003

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2775
Sch em e 1a
a
Reagents: (i) 1a f 2a and 2f, 1b f 2b respectively, ClCH2COCl for 2a and 2b, and BrCH2CH2COCl for 2f, THF or CH2Cl2, base, RT,
18 h; (ii) 2a , 2b, and 2f for 3a , 3b, and 3f respectively, TFAA, THF or CH2Cl2, RT; (iii) 3b, 2-(2-aminoethylamino)-5-cyanopyridine,
K2CO3, THF, RT; (iv) 3a or 3f (3f only for 8p ), THF, K2CO3, various amines, RT; then TFA/CH2Cl2 only for 11q; (v) 8c, 0.5 equiv of
tributylamine, EtOAc, reflux, 96 h.
Until recently, a constant in DPP-IV inhibitor design
had been an L-amino acid with a protonatable, N-
terminal primary amine in the P-2 site (i.e., 9a ). The
observation that substrates containing N-methylglycine
in the P-2 site were cleaved23 raised the possibility that
structurally more complicated N-substituted glycines
may be tolerated at the P-2 site. Through the previous
while the addition of 3-bromopropionyl chloride to 1a
led to 2f. Dehydration of amides 2a , 2b, and 2f with
trifluoroacetic anhydride (TFAA) provided nitriles 3a ,
3b, and 3f, respectively, which were then coupled with
various amines to provide the desired 2-cyanopyrrolid-
ides 8a -k , 8o, 8p , 9b, 9e, 9h , 10a -h , 11a -a k , 12 a -b,
12 h -j either as the free base, monohydrochloride,
dihydrochloride, or TFA salt. Amines used were either
commercially available or were known and prepared
from literature. These 2-cyanopyrrolidides are stable for
months to years if kept as dry solids but will convert to
the cyclic amidine with t1/2 between 48 h to >70 days
in buffered aqueous medium (pH 7.4). They appear as
mixtures of cis and trans amide rotomers in solution
according to NMR. Compounds 8c and 12j are trans
amide rotomers in crystalline form as evidenced by the
X-ray crystallographic analysis,34 possessing a solubility
of >100 mg/mL in distilled water. With minor modifica-
tions, the solution synthesis (general method A in
Scheme 1) has provided 8c and 12j on the 100-kg scale.
Intramolecular cyclization of 8c with 0.5 equiv of
tributylamine produced cyclic amidine 8q after refluxing
in EtOAc for 96 h. As shown in Scheme 2, the conversion
of 1a to nitriles 9a , 9d , 9f, and 9g was carried out
through DIC coupling with N-t-boc-L-R-cyclohexylgly-
cine, N-t-boc-L-valine, N-t-boc-3-methyl-2-methylami-
nobutyric acid, and cyclohexanepropionic acid, respec-
tively, followed by amide dehydration using POCl3 for
use of a tandem resin-solution parallel synthesis,30
a
number of diverse P-2 site N-substituted glycines were
prepared and found to provide potent inhibition when
combined with a (2S)-cyanopyrrolidide in the P-1 site.31-33
Herein is described the structure activity relationship
of this class of DPP-IV inhibitors which has led to the
selection of slow binding inhibitors NVP-DPP728 (8c)
and NVP-LAF237 (12j) (Chart 1) as clinical develop-
ment candidates for type 2 diabetes. In addition, a
solution-based method for multigram synthesis of this
P-2 site N-substituted glycine DPP-IV inhibitor class is
described as well as the pharmacological profile of a
selective DPP-IV inhibitor, 12j, which exhibits excellent
potency and oral bioavailability.
Ch em istr y
Pyrrolidine and thiazolidine analogues 8a -q, 9a -h ,
10a -h , 11a -a k , and 12a ,b,h -l were prepared as
described in Schemes 1 and 2 and listed in Tables 1-3.
As shown in Scheme 1, the addition of chloroacetyl
chloride to 1a and 1b led to 2a and 2b, respectively,

2776 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Villhauer et al.
Sch em e 2a
a
Reagents: (i) 1a , BocHNCHR′COOH, HOBt, CH2Cl2, NMM, DIC, RT, 18 h; then POCl3, pyridine, imidazole, -30 °C, 1 h; then HCl(g),
Et2O, RT, 18 h; (ii) 9a f 9c, cyclohexanone, NaB(O2CCH3)3H, ClCH2CH2Cl, RT, 18 h; (iii) 1a f 9f, N-t-boc-3-methyl-2-methylaminobutyric
acid, HOBt, CH2Cl2, NMM, DIC, RT, 18 h; then POCl3, pyridine, imidazole, -30 °C, 1 h; then HCl(g), Et20, RT, 18 h; (iv) 1a ,
3-cyclohexanepropionic acid, CH2Cl2, DMAP, DIC, RT, 6 h, then TFAA, CH2Cl2, 0 °C, 2 h; (v) 1a , 1c, 1d , and 1e f 2a , 2c, 2d , and 2e
respectively, ClCH2COCl, THF or CH2Cl2, K2CO3, base, RT, 18 h; (vi) K2CO3, THF or CH2Cl2, RT, 3-hydroxy-1-aminoadamantane, 2a for
12k and 2e for 12l; then TFA, CH2Cl2, RT for 12l; (vii) 2c, 2d , and 2a for 8l, 8m , and 8n , respectively, K2CO3, THF, 2-(2-aminoethylamino)-
5-nitropyridine for 8l and 8m and 2-(2-aminoethylamino)-5-cyanopyridine for 8n , RT.
9a , 9d , 9f or TFAA for 9g. Coupling of amine 9a with
cyclohexanone and reduction of the resulting imine
using sodium triacetoxyborohydride provided amine 9c.
Coupling of 2a and 2e with 3-hydroxy-1-aminoadaman-
tane33 provided 12k and the t-boc ester precursor of 12l,
respectively. Acid-catalyzed removal of the t-boc-ester
using TFA in CH2Cl2 led to 12l. Coupling of 2-(2-
aminoethylamino)-5-nitropyridine with 2c and 2d pro-
vided 8l and 8m , respectively, while coupling of 2-(2-
aminoethylamino)-5-cyanopyridine with 2f provides 8n .
by coupling of the resulting adamantylamines with 3a
provided 12m -r , respectively.
Resu lts a n d Discu ssion
The pyrrolidide, thiazolidide, and N-substituted gly-
cine amide derivatives described above were tested for
inhibition of DPP-IV derived from human colonic car-
cinoma cells (Caco-2),35 and the results are included in
Tables 1-3. An earlier manuscript30 described the solid-
phase synthesis of a library of ∼200 N-substituted
glycine-2-cyano-(S)-pyrrolidides which provided 8a as
one of a few low nanomolar DPP-IV inhibitors. Starting
with 8a , a focused structure activity profiling effort was
initiated by first introducing changes to the P-2 site
amine, 2-(2-aminoethyamino)-5-nitropyridine. Replace-
ment of the p-nitro group of 8a with electrophilic groups,
such as chloride (8b), cyano (8c), and trifluoromethyl
(8e), provided modest decreases in inhibitory potency
(4-, 3-, and 10-fold, respectively) while the o-trifluoro-
methyl analogue 8f significantly reduced potency (>100-
fold reduction). Electron-withdrawing group substitu-
tions on the 2-aminopyridine ring do not appear to be
essential since only a 4-fold reduction in inhibitory
potency relative to the p-nitro analogue 8a is obtained
with the p-hydrogen substitute (8g, IC50 ) 33 nM).
Adamantyl analogues 12c-g and 12m -r were pre-
pared as described in Scheme 3 and are listed in Table
3. The N-substituted glycine ester 4, prepared by
coupling 1-adamantaneamine and ethyl bromoacetate,
was reacted with benzyl chloroformate followed by ester
hydrolysis with LiOH to provide glycine derivative 5.
The free acid in 5 was DIC-coupled with various amines
followed by CBZ-deprotection via palladium-catalyzed
hydrogenation to provide N-adamantyl glycine amide
derivatives 12c-g. The hydroxyl group in carbamate
633 was coupled with isocyanates or carbamoyl chlorides
to provide bis-carbamates 7a -e while coupling 6 with
acetyl chloride provided ester 7f. CBZ-deprotection of
7a -e via palladium-catalyzed hydrogenation followed

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2777
Ta ble 1. DPP-IV Inhibitor Data for Templates 8-10a
compd
R
U
V
W
X
Y
Z
*
DPP-IVb
8a
8b
8c
8d
8e
8f
8g
8h
8i
8j
8k
8l
8m
8n
8o
8p
8q
9a
9b
9c
9d
9e
NO2
Cl
CN
H
CF3
H
H
H
H
H
CN
NO2
NO2
CN
CN
NO2
NH
NH
NH
NH
NH
NH
NH
NH
O
CH2
NH
NH
NH
NH
NCH2
NH
CN
CH
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
S
N
N
N
N
N
N
N
CH
CH
CH
N
N
N
N
N
N
1
S
S
S
S
S
S
S
S
S
S
R
8 ( 3
36 ( 5
CN
CN
CN
CN
CN
CN
CN
CN
CN
CN
H
CH
CH
N
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
22 ( 4
163 ( 39
83 ( 4
989 ( 218
33 ( 4
298 ( 113
635 ( 112
491 ( 40
CH
CCF3
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
1170 ( 21
7400 ( 1,500
3000 ( 600
434 000 ( 20 000
102 ( 10
18 700 ( 1600
1540 ( 520
2 ( 0.5
H
CONH2
CN
CN
CH2
CH2
CH2
S
S
S
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
N
N
N
N
N
N
CH
N
H
c-C6H11
H
c-C6H11
CH(CH3)2
c-C6H11
c-C6H11
H
CH(CH3)2
CH3
64 (33
18 600 ( 3900
2 ( 1
H
192 ( 112
9610 ( 3590
>500 000, n ) 2
103 000 ( 8000
612 ( 148
80 ( 27
9f
CH(CH3)2
H
H
H
H
H
CH2OH
CH2OH
H
H
H
9g
9h
10a
10b
10c
10d
10e
10f
10g
10h
c-C6H11
c-C6H11
1
2
3
3
4
5
6
10
54 ( 30
26 ( 17
104 ( 14
63 ( 13
53 ( 32
338 ( 7
a
b
Values are IC50 (nM) expressed as the mean ( SD of three independent determinations unless otherwise noted. Primary DPP-IV
assay in human Caco-2 cells; procedure described in Experimental Section.
Replacement of the 2-aminopyridine ring in 8g, with
pyrimidine (8d ), aniline (8h ), phenyl (8j), and phenoxy
(8i) provided increasingly less potent inhibition (5-, 9-,
15-, and 19-fold, respectively, relative to 8g). Increasing
from the ethyl diamine chain found in 8c to the propyl
diamine analogue 8o provided a modest (4-fold) decrease
in inhibitory potency while a dramatic reduction (2300-
fold) was observed when the R-amino acid in 8a was
substituted with a â-amino acid (8p ).
As expected, replacement of the P-1 site 2(S)-cyano
group in 8c, which plays a critical role as a transition-
state mimetic36 with a 2(R)-cyano group (8k ), signifi-
cantly reduced potency (53-fold). (Most of the inhibitory
potency seen with 8k can be attributed to ∼1% con-
tamination with 8c which is due to ∼1% L-prolinamide
contamination in the commercial source of the D-
prolinamide starting material. This contamination may
also explain the previously reported identical dissocia-
tion kinetics between 8c and 8k (koff of 1.5 × 10-3 s-1).36)
Replacement of the 2(S)-cyano group with either a
hydrogen (8l) or a 2(S)-amide (8n ) dramatically de-
creased inhibitory potency (900-fold compared to 8a and
20 000-fold compared to 8c, respectively). Both des-
cyanopyrrolidide (8l) and des-cyanothiazolidide (8m )
provided low µM IC50s, similar to what had been
reported for such P-1 site carbon- and sulfur-containing
rings in DPP-IV inhibitor classes possessing amino acid
groups in the P-2 site.37 Further evaluation of p-cyano
analogue (8c) in various in vitro DPP-IV and selectivity
assays (Table 4) led to its nomination as our first
development compound. When compared with the p-
nitro analogue 8a and cyclohexylglycine-(2S)-cyanopyr-
rolidine (9a ),27 8c was of comparable activity against
human plasma DPP-IV and was significantly more
selective for DPP-IV over closely related peptidases,
such as post-proline-cleaving enzyme (PPCE) and/or
DPP-II.38 Selectivity of 8c for human plasma DPP-IV
over PPCE and DPP-II was 27 000- and 16 000-fold,
respectively. In addition, the in vitro specificity of 8c
was profiled in over 100 receptor and enzyme assays,
and no significant binding was observed (10 µM). Under
neutral and basic aqueous conditions, the P-2 site amine
can nucleophilically attack the carbon of the pyrrolidide-
nitrile to form the 70-fold less active (attempts to
convert 8c to 8q resulted in a 1% impurity of 8c, which
may account for most of the observed inhibitory potency
seen for 8q. Attempts to further purify cyclic amidine
8q via HPLC were not pursued due to facile conversion
to the ketopiperazine) cyclic amidine 8q. Under in vitro
assay conditions employed, this intramolecular cycliza-

2778 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Ta ble 2. DPP-IV Inhibitor Data for Template 11a
Villhauer et al.
compd
RN
DPP-IVb
444 ( 35
compd
RN
DPP-IVb
11a
11b
11c
11d
11e
11f
11g
11h
11i
11jh
11k
11l
11m
11n
11o
ethylamine
propylamine
butylamine
pentylamine
hexylamine
heptylamine
11p
11q
2-amino-2-methyl-1,3-propanediol
2-amino-2-methylproprionic acid
11r to 11z shown below
372 ( 92
10 900 ( 6900
475 ( 45
261 ( 132
266 ( 107
282 ( 15
291 ( 12
346 ( 18
466 ( 18
398 ( 142
11a a
benzylamine
988 ( 146
436 ( 37
11a b 1-naphthalene-methylamine
11a c phenethylamine
11a d 1,1-dimethylphenethylamine
11a e
11a f
5-hydroxypentylamine
3,3-dimethylbutylamine
1-ethylpropylamine
(1S)-1-(hydroxymethyl)propylamine 256 ( 132
1-isopropyl-2-methylpropylamine
tert-butylamine
1,1-dimethylpropylamine
1,1,3,3-tetramethylbutylamine
2-amino-2-methylpropan-1-ol
305 ( 51
24 ( 5
3,3-diphenylpropylamine
2-aminoindane
377 ( 76
133 ( 47
10 900 ( 5100 11a g 2-aminobenzimidazole
35 ( 7
105 ( 16
123 ( 30
218 ( 62
>1 000 000, n ) 2
15 ( 8
11a h exo-2-aminonorbornane
11a i
11a j
(1R,2R,3R,5S)-(-)-isopinocampheylamine
1-adamantanemethylamine
16 ( 5
598 ( 204
242 ( 23
11a k (-)-cis-myrtanylamine
compd
RN
X for AV, AW, or AY
DPP-IVb
11r
11s
11t
11u
11v
11w
11x
11y
11z
AW
AV
AV
AY
AV
AY
AV
AY
AV
5-cyano-2-aminopyridine
5-cyano-2-aminopyridine
5-trifluoromethyl-2-aminopyridine
5-trifluoromethyl-2-aminopyridine
2-aminobenzothiazole
2-aminobenzothiazole
c-C6H11CONH
29 ( 2
5 ( 1
7 ( 0
296 ( 110
10 ( 1
228 ( 49
18 ( 6
23 ( 6
14 ( 2
c-C6H11CONH
p-Cl-C6H4O
a
b
Values are IC50 (nM) expressed as the mean ( SD of three independent determinations. Primary DPP-IV assay in human Caco-2
cells; procedure described in Experimental Section.
Ta ble 3. DPP-IV Inhibitor Data for Adamantyl Template 12a
compd
R
Y
V
W
Z
1
DPP-IVb
12a
12b
12c
12d
12e
12f
12g
12h
12i
12j
12k
12l
12m
12n
12o
12p
12q
12r
H
H
CN
CH2
3 ( 2
13 ( 6
H
H
H
H
H
H
H
H
H
H
H
H
CH2
CH2
CH2
CHOH
1
0
2
1
7000 ( 1300
24 200 ( 5300
52 600 ( 8500
>1 000 000, n ) 2
26 100 ( 4900
7 ( 1
CH2CH3
CH3
OH
OH
OH
H
CH3
H
H
H
H
H
H
H
CN
CN
CN
CONH2
COOH
CN
CN
CN
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
1
1
1
1
1
1
1
1
1
1
1
17 ( 0
3.5 ( 1.5
931 000 ( 105 000
477 000 ( 89 000
22 ( 7
OCONHC(CH3)3
OCONH((C6H5) p-CH3O)
OCONHC6H5
OCON(CH(CH3)2)2
OCONH(c-C6H11
OCOCH3
149 ( 15
62 ( 5
CN
CN
CN
183 ( 15
)
H
H
64 ( 16
70 ( 16
a
b
Values are IC50 (nM) expressed as the mean ( SD of three independent determinations. Primary DPP-IV assay in human Caco-2
cells; procedure described in Experimental Section.
tion was slow (t1/2 > 2 days), resulting in less than 1%
of 8c converting during the time frame of the experi-
ments.
According to a recent crystal structure of human DPP-
IV complexed with the known inhibitor (valine pyrro-
lidide),39 the P-2 site valine points into a large cavity

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2779
Sch em e 3a
a
Reagents: (i) benzyl chloroformate, CH2Cl2, K2CO3, RT, 20 h; (ii) LiOH, THF/H2O, 20 h; (iii) NMM, HOBt, DIC, CH2Cl2, various
amines; (iv) 10% Pd/C, H2 (1 atm), EtOAc, RT; (v) 6, appropriate isocyante, TMSCl and either CH2Cl2, RT, 18 h or ClCH2CH2Cl, 50 °C,
18 h for 7a -c,e; diisopropylcarbamoyl chloride for ClCH2CH2Cl, 50 °C, 18 h for 7d ; and acetyl chloride, pyridine, DMAP, ClCH2CH2Cl,
RT, 24 h for 7f; (vi) 10% Pd/C, H2 (1 atm), EtOH, RT; (vii) 3a , THF, K2CO3, RT, amines derived from 7a -f to provide 12m -r , respectively.
Ta ble 4. DPP-IV Inhibition and Selectivity Assaysa
Caco-2b
rat plasmab
human plasmab
PPCEc
DPP-IId
8a
8c
9a
12j
8.0 ( 3.0
22.0 ( 2.0
2.0 ( 0.3
3.5 ( 1.5
17.3 ( 0.34
6.0 ( 1.0
2.8 ( 0.2
2.3 ( 0.1
8.73 ( 0.79
7.0 ( 1.7
3.15 ( 0.19
2.7 ( 0.1
16 000 ( 1200
190 000 ( 46 000
41 000 ( 14 000
210 000 ( 40 000
12 000 ( 580
110 000 ( 5800
102 000 ( 20 000
>500 000
a
Values are IC50 (nM) expressed as the mean ( SEM of three independent determinations; procedures described in Experimental
Section. Primary DPP-IV assays. c Extract from human erythrocytes. Extract from bovine kidney homogenate.
b
d
in the catalytic binding site and does not make specific
contacts with DPP-IV. Due to the importance of the 2(S)-
cyano group and the use of the P-1 site pyrrolidide ring
in our N-substituted glycine series, one can assume that
our inhibitors bind DPP-IV in a fashion similar to that
observed for valine pyrrolidide. During our SAR effort
around 8a , we explored the steric limitations of DPP-
IV’s S-2 site cavity first with inhibitors that possess both
P-2 site N- and C-substitutions. Whereas proline is a
well-known N- and C-substituted cyclic amino acid for
the P-2 site in DPP-IV inhibitors,37 little is known about
the effect of acyclic N- and C-substituted amino acids.
Initial studies suggested that there are significant steric
restrictions in this cavity when inhibitors possess both
acyclic C- and N-substitutions. Whereas either a C-
substituted (9a ) or N-substituted (9b) cyclohexyl group
provides low nM inhibition (Table 1), positioning cyclo-
hexyl groups at both the C-and N-terminus (9c) signifi-
cantly reduces potency (9300-fold less compared to 9a
and 290-fold less compared to 9b). The steric restriction
was also observed with significantly smaller lipophilic
groups. Whereas a single isopropyl substitution is
tolerated on either the P-2 site carbon (9d ) or nitrogen
(9e), isopropyl substitution at carbon and methyl at
nitrogen (9f) was not well tolerated providing signifi-
cantly reduced potency compared to C-substituted 9d
(4800-fold less). It was found that simple lipophilic and
branched C-substitutions provided significantly greater
inhibitory effect than N-substitution analogues (9a /9d
vs 9b/9e). One can speculate that the N- and C-substi-
tution pattern plays a role in alignment of the basic
terminal amine of inhibitors to form the required salt
bridges with DPP-IV’s glutamic acids, Glu 205 and Glu
206.39 Inhibitor alignment between DPP-IV’s salt bridge
formed by Glu 205 and Glu 206, the oxyanion hole, and
the S1-pocket may also explain the lack of potency
observed with the P-2 site â-amino acid (8p ) discussed
above. As with DPP-IV substrates,1,2 a basic, primary,
or secondary amine is essential for inhibitory potency
as replacement of the N-terminal amine in 9b with
carbon (9g) or N-methylation of 9b to provide a tertiary
amine (9h ) dropped potency by >7800-fold and 1600-
fold, respectively.
The study of P-2 site N-substituted glycines continued
with the evaluation of various lipophilic ring-sizes
(Table 1). The S-2 site pocket of DPP-IV appears to
tolerate a large number of lipophilic rings, providing
IC50s ranging from 53 nM to 80 nM for four to eight
members (10b, 10c, 9b, 10f, 10g) with only a modest
drop in potency (5- to 10-fold) for the very small three-
membered (10a ) or rather large 12-membered (10h )
rings (Table 1). Unlike the simpler ring series where
the cyclopentyl (10c) and cyclohexyl (9b) rings are of
similar potency, the hydroxymethylcyclohexyl analogue
10e is 4-fold less active than the hydroxymethylcyclo-
pentyl analogue 10d .
As with the above ring systems, inhibitory potency
did not vary dramatically (IC50 ) 261 to 475 nM, Table
2) between the simple N-terminal linear chain exten-
sions (ethyl to heptyl, 11a -f), the hydrophilic 5-hy-
droxypentyl 11g, or with chain-end sterics found in 3,3-
dimethylbutylamine 11h . As linear chain extensions

2780 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Villhauer et al.
were observed to be 4- to 8-fold less potent than ring
systems branched near the terminal amine, acyclic
chains branching R to the terminal amine were exam-
ined further (Table 2). Starting with the simplest
branched analogue, isopropyl derivative 9e, it was found
that methylene extensions (1-ethylpropylamine, 11i) or
methylene and hydroxyl extensions (1-hydroxymethyl-
propylamine, 11j) only slightly decreased potency. How-
ever, steric limits were quickly exceeded with the bis-
gem-dimethyl extension (11k ) which provided a 60-fold
reduction in potency (IC50 ) 10.9 µM) compared to 9e.
Fully branched carbons adjacent to the P-2 site amine
proved more interesting as demonstrated with the
hydroxymethylcyclopentylamine (10d , IC50 ) 26 nM)
and the tert-butylamine groups (11l, IC50 ) 35 nM).
Simple lipophilic and hydrophilic chain extensions on
the tert-butyl group of 11l were well tolerated as shown
with methyl (11m ), tert-butyl (11n ), monohydroxyl
(11o), or bis-hydroxyl (11p ) groups that only slightly
decreased inhibitory potency relative to 11l (3- to 10-
fold). However, replacement of a methyl group from 11l
with a carboxylic acid (11q) dramatically reduced
potency (300-fold) which may reflect hydrophobic and/
or steric limitations.
analogue 11a g is >7500-fold less potent than the
saturated 2-aminoindane analogue 11a f supports the
importance of a basic, protonatable P-2 site amine in
the N-substituted glycine template for DPP-IV inhibi-
tion.
Further explorations of sterics adjacent to the P-2 site
amine concentrated on multiple cyclic rings and pro-
vided favorable results (Tables 2 and 3). Bicyclic ring
systems containing 2-aminonorbornane (11a h ) and iso-
pinocampheylamine (11a i) or tricyclic ring systems
containing 1- and 2-adamantylamine (12a and 12b,
respectively) all provided very potent inhibitory effect.
With an IC50 of 3 nM, 12a was one of the most potent
P-2, N-substituted glycine derivatives found. However,
lipophilic extensions on 12a as subtle as the 3-ethyl-
(12h ) and 3,5-dimethyl-1-adamantylamine (12i) ana-
logues somewhat reduced inhibitory potency compared
to 12a (2- and 6-fold). The importance of steric ring bulk
adjacent to the P-2 site amine is further supported by
the dramatic potency decrease seen with methylene
insertion between the steric group and the amine in the
form of the 1-adamantanemethylamine group in 11a j
and (-)-cis-myrtanylamine group in 11a k (Table 2).
With a 1-adamantyl group as the optimal P-2 site
substitution, P-1 site requirements were further evalu-
ated starting with the des-cyano analogue of 12a (12c).
Although 2300-fold less potent as compared to its 2(S)-
cyano counterpart, pyrrolidine (12c) was noticeably
more potent than both azetidine (12d ) and piperidine
(12e) P-1 site ring analogues (4- and 8-fold, respec-
tively). As expected, the narrow lipophilic pocket formed
by the stacking of Tyr662 and Tyr666 in DPP-IV’s
catalytic S-1 site39 did not tolerate the hydrophilic
nature of the P-1 site 3-pyrrolinol in 12f which was
>330 000-fold less potent than its pyrrolidide counter-
part (12c). Interestingly, the tetrahydroisoquinoline P-1
site analogue 12g was as potent as the piperidine P-1
site counterpart 12e, suggesting either the narrow
pocket of DPP-IV’s catalytic S-1 site can tolerate a
phenyl-extended P-1 site analogue of 12e or that 12g
binds DPP-IV differently than 12e.
Cyclic and acyclic branching on the P-2 site pharma-
cophore of 8c and 8e was well tolerated. Substitution
of a gem-dimethyl group into the 1,2-diaminoethyl chain
of 8c provided the equipotent 11r , while replacement
with a trans-1,4-diaminocyclohexyl group provided in-
creased potency as shown with 11s (4-fold) and 11t (12-
fold), respectively. Substitutions on the trans-1,4-di-
aminocyclohexyl template were well tolerated as shown
by the modest potency reductions through replacing the
5-cyano-2-aminopyridine of 11s with a 2-aminoben-
zothiazole (11v) or a with a cyclohexylamide (11x) (2-
and 3-fold reduction, respectively). However, replace-
ments with 4-aminopiperidine provided a different
profile than observed with the trans-1,4-diaminocyclo-
hexyl group. Replacement with 4-aminopiperidine for
the 1,2-diaminoethyl chain of 8e (11u ) and the trans-
1,4-diaminocyclohexyl chain in 11t and 11v (11w )
provided >20-fold decreased in inhibitory potency,
whereas similar low nM potency was seen for 1,4-di-
aminocyclohexyl (11x) and 4-aminopiperidine (11y)
analogues containing the cyclohexylamide groups. The
advantage of geometrical restriction also carried over
into the amino alcohol side chain 11z, which provided
a 45-fold increase in inhibitory potency over the 2-di-
aminoethane analogue 8i.
Examination of primary metabolites of 1-and 2-ada-
mantylamine analogues 12a and 12b suggest that
monohydroxylation on the adamantyl ring may be well-
tolerated in DPP-IV’s catalytic S-2 site. To study the
effect of monohydroxylation and avoid incorporation of
an additional chiral center in monohydroxylation, the
3-hydroxylated-1-aminoadamantane analogue 12j was
prepared. 12j was noticeably more potent in the Caco-
2, rat-, and human plasma DPP-IV assays (6-, 3-, and
3-fold) as well as showing superior efficacy in various
in vitro selectivity assays (Table 4) when compared with
both 8c and literature’s cyclohexylglycine-(2S)-cyano-
pyrrolidine (9a ). 12j was profiled in over 100 receptor
and enzyme assays, and no significant binding was
observed (10µM). Because of the sterics imposed by the
adamamtyl group, the ability for 12j to cyclize to an
analogue of imidine 8q was significantly reduced (>30-
fold) compared to that of 8c and 9a . To date, 12j is the
most stable DPP-IV inhibitor, possessing a P-1 site
transition-state mimetic. The dramatic reduction in
inhibitory potency of 2(S)-amide (12k ) and 2(S)-acid
(12l) compared to 12j (266 000- and 14 000-fold, respec-
tively) provides further support that 12j binds in the
As with the simple aliphatic chains described above,
the inhibitory potency of phenyl-substituted aliphatic
P-2 site amine extensions, such as benzylamine (11a a ),
1-naphthalenylmethylamine (11a b), phenethylamine
(11a c), 3-phenylpropylamine (8j), and 3,3-diphenylpro-
pylamine (11a e), did not vary dramatically (IC50 ) 305
nM to 988 nM, Table 2). Numerous substitutions on
benzylamine (11a a ) and phenethylamine (11a c) were
explored and showed little effect on potency compared
to these unsubstituted analogues (data not shown).
Compared to the simpler phenethyl derivative 11a c,
sterics in the form of a gem-dimethyl group adjacent to
the P-2 site amine (11a d ) and a conformational con-
strained ring analogue (11a f) proved more potent (13-
and 2-fold). The fact that the 2-aminobenzimidazole

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2781
F igu r e 1. Obese male Zucker rats were orally dosed with
CMC or 12j (10 µmol/kg) immediately following the -15 min
blood sample collection. Plasma DPP-IV activity (left) and
levels of GLP-1 (7-36 amide) (right) were measured during
the OGTT. Data were the mean ( SEM of nine rats/group;
procedures are described in Experimental Section. For both
DPP-IV activity and GLP-1 (7-36 amide) levels, values for
the 12j-treated rats were significantly different from control
at all time points (Student’s t-test, p < 0.05, except for time
-15 min GLP-1 (7-36 amide) levels that were not significant).
Plasma DPP-IV activity of -5 min control samples was 3.4 (
0.9 mU/mL.
F igu r e 2. Obese male Zucker rats were orally dosed with
CMC or 10 µmol/kg 12j immediately following the -15 min
blood sample collection. Glucose was administered orally
(1 g/kg) immediately following the time 0 min blood collection.
Data were the mean ( SEM of eight rats/group; procedure is
described in Experimental Section. Analysis of the glucose and
insulin AUC reveal statistically significant differences between
the CMC and 12j groups.
Student’s t-test) in the 12j-treated rats 15 min after
drug administration and before the glucose challenge
(0 min). Following the glucose challenge, GLP-1 (7-36
amide) in the drug-treated rats peaked at 5 min at levels
that were close to 5 times those of CMC-treated rats
(13.5 ( 3.0 vs 2.8 ( 0.2 pM, p ) 0.001), and returned
back to baseline by 15 min. This return to baseline of
GLP-1 (7-36 amide) levels by 15 min while DPP-IV is
completely inhibited is fully in line with the glucose-
dependency of GLP-1 (7-36 amide) secretion from the
intestinal L-cell. The results in Figure 2 support the
ability of 12j to both significantly decrease glucose
excursions and stimulate insulin secretion. Obese ani-
mals receiving CMC were markedly glucose intolerant,
with peak glucose levels of 294 ( 19 mg/dL at 20 min
while 12j-treatment decreased the glucose excursion
with noticeable differences starting after 5 min and
provided peak glucose levels that reached only 241 ( 8
mg/dL at 15 min. Interestingly, while insulin levels were
no different between the two groups at the start of the
experiment (-15 min), insulin levels were elevated 40%
before the glucose challenge in the 12j-treated group.
This modest increase most likely resulted from 12j-
induced preservation of active GLP-1 (7-36 amide)
described above. Upon glucose challenge, plasma insulin
levels of 12j-treated rats peaked at levels twice those
of vehicle-treated rats and returned to baseline levels
by 45 min. The 0 to 45 min insulin AUC was 275%
higher in 12j-treated rats with a corresponding 35%
reduction of the 0-45 min glucose AUC. The glucose-
stimulated insulin release index (insulin AUC 0-45/
glucose AUC 0-45) was ∼5-fold greater in the 12j-
treated rats.
The pharmacokinetic profile of 12j was compared to
8c in normal cynomolgus monkeys. The absolute oral
bioavailability of both 8c and 12j were both excellent
(>90%) following a 1 µmol/kg oral dose while clearance
from plasma was moderate ∼1.3 l/h/kg and ∼1.5 l/h/kg,
respectively. The steady-state volumes of distribution
are also moderate for both 8c (0.8 L/kg) and 12j (0.7
L/kg), suggesting that both are distributed principally
in the body fluids. The Cmax for 12j (293 nM at 72 min)
is 2.7-fold lower than for 8c (805 nM at 36 min) while
the terminal elimination half-life of 12j (90 min) is 2.6-
fold longer than for 8c (35 min). Differences seen in the
pharmacokinetic profiles of 8c and 12j correlated well
with observed DPP-IV inhibition pharmacodynamic
differences. When comparing the effect on DPP-IV
S1- and S2-catalytic sites of DPP-IV. Substitutions on
the 3-hydroxyl group of 12j in the form of carbamates
(12m -q) and ester (12r ) resulted in significantly re-
duced potency compared to 12j. In the carbamate series,
potency correlated with size as the diisopropyl analogue
12p was 8-fold less potent than the tert-butyl analogue
12m while the 4-methoxyphenyl analogue 12n was
2-fold less potent than both the similarly sized phenyl
and cyclohexyl analogues 12o and 12q, respectively.
Evaluation of our first development compound (8c)
in rat,9 monkey,30 and human40 had previously sup-
ported the connection between DPP-IV inhibition and
improvement in oral glucose tolerance through an
increase in active GLP-1 levels. In a recent chronic
study, 8c was shown to improve glucose tolerance in
both normal and glucose-intolerant mice through im-
proved islet function as judged by increased GLUT-2
expression, increased insulin secretion, and protection
from increased islet size in insulin resistance.41 In a
continuing effort to explore the pharmacological poten-
tial of the N-substituted glycine-2-(S)-cyanopyrrolidide
class of inhibitor, we evaluated the antidiabetic poten-
tial of 12j in an oral glucose tolerance test (OGTT) with
obese Zucker (fa/fa) rats which are profoundly insulin-
resistant, are markedly glucose intolerant, and repre-
sent a rather severe model of Type II diabetes. The
results shown in Figure 1 demonstrate that 12j is a
potent, orally active inhibitor of plasma DPP-IV activity
that provides increased levels of GLP-1 (7-36 amide).
Administration of 12j (10 µmol/kg, po) 15 min prior to
a glucose challenge inhibited plasma DPP-IV activity
>90% within 10 min and throughout the 90 min study.
The activity in CMC-treated plasma obtained at -5 min
was 3.4 ( 0.9 mU/mL and was set as 100%. DPP-IV
activity of rats dosed with CMC vehicle declined ap-
proximately 35% at 20 and 45 min postglucose chal-
lenge. This decline in the control group is attributed to
both plasma volume expansion caused by increased
osmolarity following the glucose load and to possible
plasma dilution caused by frequent blood sampling and
donor blood replacement. Prior to 12j administration
(-15 min) plasma levels of GLP-1 (7-36 amide) were
no different (CMC vs 12j; 2.4 ( 0.2 vs 2.9 ( 0.2 pM,
(not statistically significant), Student’s t-test). However,
GLP-1 (7-36 amide) levels were 60% higher (p < 0.001,

2782 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Villhauer et al.
mercially available unless otherwise indicated. Experimentals
for 8b, 8d -j, 8o, 9b , 9e, 9h , 10a -h , 11a -a k , 12a -b , and
12h -i follow General Method A and are found in the Sup-
porting Information.
1-Ch lor oa cetyl-2-(S)-p yr r olid in eca r boxa m id e (2a ). A
solution of L-prolinamide (1a ) (5.0 g, 43.9 mmol) in THF (100
mL) was added dropwise over 0.5 h to a stirred mixture of
chloroacetyl chloride (3.50 mL, 43.9 mmol) and K2CO3 (24.0
g, 175 mmol) in THF (50 mL) at ambient temperature. The
reaction mixture was then stirred at room temperature for 18
h and filtered and the filtrate concentrated to provide the
desired amide 2a as a clear, taffy (7.46 g, 89%); 1H NMR
(CDCl3) 1.94-2.45 (m, 4H), 3.54-3.65 (m, 1H), 3.65-3.75 (m,
1H), 4.08-4.18 (m, 2H), 4.57-4.67 (m, 1H), the two amide
hydrogens spread from 6.1 to 7.2 with small broad peaks at
6.18 and 7.04; MS m/z (rel intensity, %) 191 (M+, 100), 193
(M+, 33).
F igu r e 3. Effects of 12j (black squares) and 8c (open
triangles) following oral administration of a single 1 µmol/kg
dose. Values are means ( SEM of three monkeys 12j or
means ( range of two monkeys 8c; procedure is described in
Experimental Section.
1-Ch lor oa cetyl-2-cya n o-(S)-p yr r olid in e (3a ). A solution
of L-prolinamide (1a ) (10.0 g, 87.7 mmol) in THF (300 mL) was
added dropwise over a period of 0.75 h to a mechanically
stirred solution of chloroacetyl chloride (6.99 mL, 87.7 mmol)
and K2CO3 (48.5 g, 351 mmol) in THF (100 mL). After another
2 h of being stirred, the reaction was filtered to remove
potassium salts, and the filtrate was dried over Na2S04 (30
g). This solution of 2a was again filtered, and to the colorless
filtrate was added trifluoroacetic anhydride (19.8 mL, 0.140
mmol) in one portion. The resulting clear, light amber reaction
was stirred for 1 h at RT. EtOAc (30 mL) was added prior to
rotovap concentration (water bath <45 °C) with two more
EtOAc chasings (2×) once the concentrate was ∼25 mL. The
resulting clear, light-orange solution was then partitioned
between EtOAc and saturated NaHCO3, and the aqueous layer
was washed twice with EtOAc. The combined organic layers
were then successively washed with water and brine, dried
over Na2SO4, filtered, and concentrated to obtain 1-chloro-
acetyl-2-(S)-pyrrolidinenitrile as clear, amber oil. This oil
solidified quickly to provide the target product 3a (7.82 g, 52%)
inhibition in normal Cynomolgus monkeys (Figure 3),
maximum inhibition of plasma DPP-IV activity (∼95%)
was observed approximately 2 h postdose of 12j (1 µmol/
kg, po) while g50% inhibition of DPP-IV was observed
within 30 min postdose and persisted for g10h (not
shown). Maximum inhibition of plasma DPP-IV activity
(∼95%) occurred within 30 min postdose with 8c (1
µmol/kg, po) but inhibition of g50% persisted only for 4
to 5 h postdose. As expected with this glucose-dependent
mechanism, no hypoglycemia was observed. As a re-
versible DPP-IV inhibitor with a relatively short half-
life, development candidate 8c might most effectively
be taken with a meal when GLP-1 secretion is at
maximal rate, while development candidate 12j, with
its longer half-life, provides a better profile for once-a-
day treatment if desired. The results of these pharma-
cological studies suggest that 12j is a potent, selective,
and orally active inhibitor of DPP-IV that improves
insulin secretion and glucose homeostasis with a profile
wholly consistent with increased action of glucagon-like
peptide-1.
1
as a yellow-white solid; mp 53-57 °C; H NMR (CDCl3) (4 to
1 mixture of trans/cis amide rotomers): δ 2.10-2.40 (m, 4H),
3.55-3.66 (m, 1H), 3.66-3.79 (m, 1H), 4.03-4.21 (m, 0.4H,
CH2Cl), 4.09 (s, 1.6H, CH2Cl), 4.76 (m, 0.8H, CHCN), 4.87 (dd,
0.2H, J ) 7.4 and 2.2 Hz, CHCN); 13C NMR (CDCl3): δ 165.15,
117.72, 46.95, 46.82, 46.64, 46.37, 41.41, 32.32, 29.81, 25.05,
22.61; MS m/z (rel intensity, %) 173 (M+, 100), 175 (M+, 33).
Gen er a l Meth od A. A 0.5 M solution of 3a , 3b, or 3f (1
equiv) in THF was added dropwise over 0.5 h into an ice-
water-cooled 0.3 M mixture of the appropriate amine (2 to 3
equiv) and K2CO3 (3 equiv) in THF. The resulting reaction was
stirred at ice-water temperature for 2 h and then at room
temperature for 1 to 3 days. The resulting mixture was then
filtered to remove K2CO3, concentrated, and then partitioned
between CH2Cl2 and saturated aqueous NaHCO3. The aqueous
layer was then washed with CH2Cl2 (2×), and the combined
organic layers were washed with water, dried over Na2SO4,
concentrated in vacuo, and chromotographed on silica gel with
5% CH3OH in CH2Cl2 or 20% EtOAc in hexane to provide the
free base of the target product. Alternatively, the reaction may
be carried out using CH2Cl2 as solvent in the reaction while
EtOAc may be substituted for CH2Cl2 in the partitioning with
water. The free base may be converted to the mono- or
dihydrochloride salt by slowly passing HCl gas into a 0.1 M
solution (THF or Et2O) of the free base. The resulting solids
are then filtered and dried in vacuo. The amine coupling step
with 3a or 3b provided chromatographed yields that ranged
between 30 to 50%.
Exp er im en ta l Section
Ch em istr y. All melting points (mp) were obtained on a
Thomas-Hoover capillary melting point apparatus and are
uncorrected. Proton NMR (1H NMR) and carbon NMR (13C
NMR) spectra were obtained on
a Bruker AC 300-MHz
spectrometer. Chemical shifts were recorded in ppm (δ) and
were reported relative to the solvent peak or TMS. Chemical
shifts recorded for carbon NMR spectra comprised both trans
and cis amide rotomers without further delineation. Mass
spectra were run on a Finnigan Mat 4600 spectrometer.
Elemental analyses, performed by Robertson labs, are within
0.4% of theoretical values. Analytical results were within
(0.4% of the theoretical value. Fractional moles of water and
solvent frequently found in some analytical samples were not
removed despite 24-48 h of drying in a vacuum and were
confirmed, where possible, by their presence in the 1H NMR
spectrum. Optical rotation measurements were performed on
a J ASCO P-1030 polarimeter, 589 nm, cell length 1 dm, at 20
°C. Column chromatography separations were carried out
using Merck silica gel 60 (mesh 230-400) employing a SIMS/
Biotage apparatus. Reagents and solvents were purchased
from common suppliers and were utilized as received. All
reactions were conducted under a nitrogen atmosphere at room
temperature with a calcium sulfate drying tube unless noted
otherwise. Yields were of purified product and were not
optimized. All starting materials and amines were com-
1-[[[2-[(5-Nitr o-2-pyr idin yl)-am in o]eth yl]am in o]acetyl]-
2-cya n o-(S)-p yr r olid in e, Dih yd r och lor id e (8a ). Method A:
2-(2-aminoethylamino)-5-nitropyridine31 as amine and use of
3a . Product was a light vanilla-colored solid: mp ) 205-208
1
°C decomp; H NMR (D2O) (trans and cis rotomers in a ratio
of 9:1): δ 2.01-2.17 (m, 2H), 2.20-2.30 (q, 2H, J ) 5 Hz),
3.36-3.49 (m,3H), 3.55-3.64 (m, 1H), 3.82 (t, 2H, J ) 5 Hz),
4.10 (s,2H), 4.72 (t, 9/10H, J ) 5H, HCN), 4.90 (bd, 1/10H,

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2783
J ) 7 Hz, HCN), 6.81 (d,1H, J ) 9 Hz), 8.27 (bd, 1H, J ) 9
Hz), 8.90 (d, 1H, J ) 1 Hz); (isobutane/DCI) MH/e 318 (MH+)
of free base; Anal. (C14H18N6O3‚2HCl‚0.2H2O‚0.1THF) C, H,
N, Cl.
sulfate drying tube for 2h. The potassium salts were removed
via filtration to yield a clear colorless solution of 1-chloro-
acetylthiazolidine (2d ) which was immediately taken onto the
next step and assumed quantitative (3.7 g). To a 500 flask was
dissolved 12.1 g (67.0 mmol) of 2-(2-aminoethylamino)-5-
nitropyridine into 50 mL of tetrahydrofuran, and the mixture
was then cooled in an ice bath. To this cooled mixture was
then added a solution of 3.7 g (22.0 mmol) of 1-chloroacetylthi-
azolidine in the tetrahydrofuran solution (from previous step)
dropwise over 30 min. Following the procedure outlined in 8l,
the dihydrochloride salt of 8m (1.80 g, 25%) was provided as
an off-white solid; mp ) 203 °C-205 °C; 1H NMR (D2O): δ 3.01
(t, 1H, J ) 6.2 Hz), 3.09 (t, 1H, J ) 6.2 Hz), 3.37(t, 2H, J )
5.2 Hz), 3.67(t, 1H, J ) 6.2 Hz), 3.71(t, 1H, J ) 6.2 Hz), 3.79
(t, 2H, J ) 5.9 Hz), 4.10 (d, 2H, 9.6 Hz), 4.43 (d, 2H, J ) 11.8
Hz), 6.76(d, 1H, J ) 9.2 Hz), 8.30 (dd, 1H, J ) 9.6, 2.6 Hz),
9.03 (d, 1H, J ) 2.9 Hz); (ESI-pos) MH/e 312 (MH+ of free
base); Anal. (C12H17N5O3S1‚2HCl), C, H, N.
1-[[[2-[(5-Cyan o-2-pyr idin yl)am in o]eth yl]am in o]acetyl]-
2-cya n o-(S)-p yr r olid in e, Dih yd r och lor id e (8c). Method A:
2-(2-aminoethylamino)-5-cyanopyridine31 used as amine and
use of 3a . Product was an off-white solid; OR: [d] ) -77.152,
12.25 mg/mL in MeOH; white solid; mp: 155-157 °C; 1H NMR
(D2O) (trans and cis amide rotomers in a ratio of 5:1): δ 2.15-
2.30 (m, 2H), 2.35-2.43 (m, 2H), 3.50-3.60 (m, 1H), 3.63-
3.75 (m, 1H), 3.80 (t, 2H, J ) 6 Hz), 3.90 (t, 2H, J ) 6 Hz),
4.19 (s, 2H), 4.85 (t, 5/6H, J ) 6 Hz, HCN), 5.00 (dd, 1/6H,
J ) 8 and 1 Hz, HCN), 6.99 (d, 1H, J ) 9 Hz), 7.97 (dd, 1H,
J ) 11 and 1 Hz), 8.50 (s, 1H). (isobutane/DCI) MH/e 299
(MH+) of free base; Anal. (C15H18N6O‚2.0HCl‚0.19H2O‚0.30THF)
C, H, N.
1-[[[2-[(5-Cyan o-2-pyr idin yl)am in o]eth yl]am in o]acetyl]-
2-cya n o-(S)-p yr r olid in e, Dih yd r och lor id e (8k ). D-Pro-
linamide (1b) (commercial source contaminated with ∼1%
L-prolinamide) was converted to 1-chloroacetyl-2-cyano-(R)-
pyrrolidine (3b) following the procedure of 3a . The synthesis
for dihydrochloride of 8c was then followed. Product was a
N-[2-[(5-Cya n o-2-p yr id in yl)a m in o]et h yl]glycyl-L-p r o-
lin a m id e (8n ). A solution of 2a (1.00 g, 5.24 mmol) in CH2-
Cl2 (20 mL) was added dropwise over 10 min to an ice-water-
cooled mixture of 2-(2-aminoethylamino)-5-cyanopyridine (1.70
g, 10.5 mmol) and K2CO3 (2.90 g, 21.0 mmol) in CH2Cl2 (40
mL). The reaction was then stirred at ice-water temperature
for 2 h and then room temperature for 18 h. The potassium
salts were removed via filtration, and the filtrate was concen-
trated. The residue was chromatographed with 10% CH3OH
in CH2Cl2 and then with a 90/10/1 mixture of CH2Cl2/CH3OH/
NH4OH which provided the desired product (558 mg, 34%) as
1
white solid; mp ) 152-154 °C; H NMR (D2O) (trans and cis
amide rotomers in a ratio of 5:1): δ 2.17-2.27 (m, 2H), 2.35-
2.45 (m, 2H), 3.48-3.60 (m, 2H), 3.67-3.75 (m, 1H), 3.79-
3.87 (m, 1H), 3.93 (t, 2H, J ) 6 Hz), 4.22 (s, 2H), 4.87 (t, 5/6H,
J ) 6 Hz, CHCN), 5.00 (dd, 1/6H, J ) 8 and 2 Hz, CHCN),
7.01 (d, 1H, J ) 9 Hz), 7.99 (dd, 1H, J ) 10 and 1 Hz), 8.48 (d,
1H, J ) 1 Hz); (isobutane/DCI) MH/e 299 (MH+ of free base);
Anal. (C15H18N6O1‚1.95HCl‚0.50H2O‚0.50THF) C, H, N.
1
a white solid; mp ) 77-80 °C; H NMR (CDCl3) (2 to 1 ratio
of trans/cis amide rotomers): δ 1.86-2.10 (m, 2H), 2.10-2.23
(m, 1H), 2.23-2.32 (m, 1H), 2.82-3.03 (m, 2H), 3.31-3.61 (m,
4H), 3.48 (s, 1.33H, trans conformer), 3.48 (s, 0.66H, cis amide
conformer), 3.65-3.57 (m, 1H), 4.47 (dd, 1/3 H, J ) 4.9 and
2.3 Hz, cis amide conformers), 4.52 (d,t 2/3H, J ) 8.3 and 2.3
Hz, trans amide comformer), 6.09 (bs, 1/3H, cis amide con-
former), 6.34 (bs, 2/3H, trans amide conformer), 6.41-6.52 (m,
1H), 6.45 (d, 2/3H, J ) 9.0 Hz, trans amide conformer), 6.48
(s, 1/3H, J ) 7.9 Hz, cis amide conformer), 6.82 (bs, 1/3H, cis
amide conformer), 6.93 (bs, 2/3H, trans amide conformer), 6.93
(bs, 2/3H, trans amide conformer), 7.43-7.56 (m, 1H), 8.33 (d,
1H, J ) 2.3 Hz); (ESI) MH/e 317 (MH+ of free base); Anal.
(C17H20N6O2‚0.05H2O) C, H, N.
1-[2-(5-Nitr op yr id in -2-yl)a m in o]eth yla m in o]a cetylp yr -
r olid in e, Dih yd r och lor id e (8l). A solution of pyrrolidine 1c
(2.50 g, 35.0 mmol), triethylamine (3.90 mL, 38.7 mmol), and
20.0 mg of (dimethylamino)pyridine (DMAP) in CH2Cl2 (20.0
mL) was added, dropwise to an ice-cold solution of 3.08 mL
(38.7 mmol) of chloroacetyl chloride in 20.0 mL of CH2Cl2, over
a period of 60 min. The resulting solution was then stirred at
ice-water temperature under a calcium sulfate drying tube
for 2 h and then at room temperature for 16 h. The solution
was poured into 150 mL of EtOAc and the resulting white
precipitate filtered. The filtrate was washed with ethyl acetate
and concentrated in vacuo to obtain 3.80 g of crude 1-chloro-
acetylpyrrolidine (2c). In a 200 mL flask was dissolved 14.3 g
(79.0 mmol) of 2-[(5-nitropyridin-2-yl)amino]ethylamine into
50 mL of tetrahydrofuran, and the mixture was then ice-water
cooled. To this cooled mixture was added a solution of 3.80 g
(26.0 mmol) of 1-chloroacetylpyrrolidine (2c) in 25.0 mL of
tetrahydrofuran, via addition funnel, over 40 min. The reaction
was then stirred at ice-water temperature for 2 h under a
calcium sulfate drying tube and then allowed to stir at room
temperature for 18 h. The solvent was then removed in vacuo
and the resulting oily paste partitioned between CH2Cl2 and
water. The aqueous layer was then washed twice with meth-
ylene chloride, and the combined organic layers were washed
successively with water and brine, dried over sodium sulfate,
concentrated, and then purified on silica gel, employing a
mixture of 5% methanol in methylene chloride as the eluent
to yield the free base as a golden solid. The dihydrochloride
salt was obtained following Method A to provide 5.42 g (57%
yield) of 8l a tan solid; mp 220-222 °C ; 1H NMR (D2O): δ
1.92-2.12 (m, 4H), 3.51 (t, 2H, J ) 5.9 Hz), 3.58(t, 4H, J ) 6
Hz), 4.05 (t, 2H, J ) 6 Hz), 4.20 (s, 2H), 7.08 (d, 1H, J ) 10
Hz), 8.48 (d, 1H, J ) 10 Hz), 9.05 (d, 1H, J ) 1 Hz); (isobutane/
DCI) MH/e 294 (MH+ of free base); Anal. (C13H19N5O1‚2HCl‚
0.2H2O‚0.2THF) C, H, N.
1-[3-[[2-[(5-Nit r o-2-p yr id in yl)a m in o]et h yl]a m in o]-1-
oxopr opyl]-2-cyan o-(S)-pyr r olidin e, Dih ydr och lor ide (8p).
Method A: 2-(2-aminoethylamino)-5-nitropyridine31 as amine
and 3f (generated by replacing chloroacetyl chloride with
3-bromopropionyl chloride in the 3a procedure). Product was
a light yellow solid, mp: softens at 80 °C and foams 82-84
°C; 1H NMR (CD3OD) (trans and cis amide rotomers in a ratio
of 9:1): δ 2.10-2.22 (m, 2H), 2.22-2.32 (m, 2H), 2.90 (t, 2H,
J ) 5.6 Hz), 3.32-3.43 (m, 4H), 3.48-3.60 (m, 1H), 3.65-3.78
(m, 1H), 3.87 (t, 2H, J ) 5.6 Hz), 4.80 (t, 9/10H, J ) 5.3 Hz,
CHCN), 5.02 (t, 1/10H, J ) 6 Hz, CHCN), 6 0.83 (d, 1H, J )
9.4 Hz), 8.30 (dd, 1H, J ) 9.4, 2.3 Hz), 9.04 (dd, 1H, J ) 7.9,
2.6 Hz); (isobutane/DCI) MH/e 299 (MH+ of free base); Anal.
(C15H20N6O3‚2HCl‚0.15H2O‚0.35THF) C, H, N.
6-[[2-[(8a S)-h exa h yd r o-1-im in o-4-oxop yr r olo[1,2-a ]p y-
r azin -2(1H)-yl]eth yl]am in o]-3-pyr idin ecar bon itr ile, Mon o-
h yd r och lor id e (8q). A cloudy, white mixture of 8c as the
monohydrochloride30 (0.400 g, 1.20 mmol) and tributylamine
(110 mg, 0.60 mmol) in EtOAc (5.00 mL) was heated at reflux
for 4 days. The resulting cloudy, white reaction was then
filtered from the reaction, washed with EtOAc (10 mL), and
dried under reduced pressure to provide a 99/1 mixture of the
target product (by HPLC) and starting material, respectively,
as a white solid (0.150 g, 45%); mp ) 263-265 °C decomp; 1H
NMR (D2O): δ 1.60-1.76 (m, 1H), 1.82-1.99 (m, 1H), 1.99-
2.15 (m, 1H), 2.38-2.50 (m, 1H), 3.12-3.25 (m, 1H), 3.42-
3.62 (m, 2H), 3.62-3.90 (m, 3H), 4.05 (d, 1H, J ab ) 16.9 Hz),
4.37 (d,d 1H, J ab ) 16.9, 2.2 Hz), 4.97-4.69 (m, 1H), 6.53 (d,
1H, J ) 9.2 Hz), 7.60 (d,d 1H, J ) 9.2,1.8 Hz), 8.21 (d, J ) 1.8
Hz, 1H); MS m/z 299 (MH+ of free base); Anal. (C15H18N6O1‚
1.0HCl) C, H, N.
1-[2-(5-Nitr op yr id in -2-yl)a m in o]eth yla m in o]a cetylth i-
a zolid in e, Dih yd r och lor id e (8m ). A solution of chloroacetyl
chloride (2.5 g, 22.0 mmol) in 25 mL of tetrahydrofuran was
added dropwise (over 20 min) into an ice-water cold solution
containing 2.0 g (22.0 mmol) of thiazolidine (1d ), 12.4 g (90.0
mmol) of K2CO3, and 50 mL of tetrahydrofuran. The reaction
was then stirred at ice-water temperature under a calcium

2784 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Villhauer et al.
r-(S)-Cycloh exylglycin e-2-cyan o-(S)-pyr r olidin e, Mon o-
h yd r och lor id e (9a ). This compound was synthesized accord-
ing to a reported general procedure26 with slight modifications.
To a stirring, opaque-white solution of N-t-boc-L-R-cyclohex-
ylglycine (2.00 g, 7.77 mmol) and 1-hydroxybenzotriazole
hydrate (1.03 g, 7.77 mmol) in CH2Cl2 (30 mL) was added
N-methylmorpholine (2.36 g, 23.3 mmol) and then diisopro-
pylcarbodiimide (2.45 g, 19.4 mmol). After DIC addition, the
reaction became clear and then cloudy again after 10 min. At
this time, the L-prolinamide (1a ) (0.890 g, 7.77 mol) was added.
The cloudy reaction was then stirred at room temperature for
18 h. The solids were then removed via filtration, and the
filtrate was concentrated to provide a golden oil. Purification
via chromatography using 3% MeOH/CH2Cl2 as eluent pro-
vided 1.90 g (69%) of R-(S)-N-t-boc-cyclohexylglycine-2-(S)-
pyrrolidinecarboxamide. POCl3 (1.31 mL, 14.0 mmol) was
added over 3 min to a CH3CN/dry ice bath cooled (-35 °C)
solution of the above t-boc-protected amide (1.90 g, 5.38 mmol)
and imidazole (0.475 g, 7.0 mmol) in dry pyridine (30.0 mL).
The reaction was stirred at between -30 °C and -20 °C for 1
h, warmed to room temperature, and then concentrated via
rotovap. Purification via chromatography using 3% MeOH in
CH2Cl2 as eluent provided R-(S)-N-t-boc-cyclohexylglycine-2-
(S)-pyrrolidinecarbonitrile (0.775 g, 43%) as a white solid; TLC
Rf: 0.6 (10% MeOH/CH2Cl2). A saturated HCl/diethyl ether
solution (2.00 mL) was added to a clear, colorless solution of
the above t-boc protected nitrile (0.772 g, 2.31 mmol) in diethyl
ether (20 mL). The resulting clear, colorless solution was
stirred at room temperature for 18 h and then concentrated
in vacuo to yield the target compound (0.600 g, 96%) as a white
1.16 (d, 3H, J ) 7.0 Hz), 2.10 (m, 5H), 2.69(s, 2.7 H), 2.71 (s,
0.3H), 3.65-3.82 (m, 2H), 4.18 (d, 1H, J ) 5.15 Hz), 4.89 (d,d,
0.9H, J ) 8.1, 4.4 Hz, CHCN), 5.08(d,d, 0.1H, J ) 7.0, 2.6 Hz,
CHCN). 13CNMR (CD3OD): δ 17.73, 18.59, 26.22, 30.73, 31.26,
33.29, 66.08, 119.14, 119.70, 167.53. (ESI) MH/e 210 (MH+ of
free base). Anal.(C11H19N3O1‚1HCl), C, H, N.
1-(3-C yc lo h e x yl-1-oxo p r o p y l)-2-c y a n o-(S )-p y r r o li-
d in e (9g). To a cloudy solution of 3-cyclohexanepropionic acid
(1.0 g, 6.40 mmol) and L-prolinamide (1a ) (730 mg, 6.40 mmol)
in CH2Cl2 (25 mL) were added DMAP (20 mg) and DIC (890
mg, 7.04 mmol). The initially clear reaction was stirred for a
total of 6 h to provide a cloudy white mixture. The solids were
removed via filtration, and the filtrate was concentrated to
provide a white solid. Purification via chromotography using
3% CH3OH in CH2Cl2 as eluent provided 1-(3-cyclohexyl-1-
oxopropyl)-(S)-prolinamide (1.37 g, 5.44 mmol) as a white solid.
To an ice-water cooled solution of this amide (1.37 g, 5.44
mmol) in CH2Cl2 (25 mL) was slowly added TFAA (1.54 mL,
10.9 mmol) over 10 min and then stirred at ice-water
temperature for 2 h. The reaction was then partitioned
between methylene chloride and saturated aqueous sodium
bicarbonate, and the aqueous layer was washed twice with
methylene chloride. The combined organic layers were then
washed successively with water and brine and then dried over
sodium sulfate. Purification via chromatography using 40%
EtOAc in hexane as eluent provided the desired product (871
1
mg, 69%) as a white solid; mp ) 40-42 °C; H NMR (CDCl3)
(7 to 1 mixture of trans/cis amide rotomers): δ 0.82-1.02 (m,
2H), 1.11-1.33 (m, 4H), 1.51-1.63 (m, 2H), 1.63-1.77 (m, 5H),
2.07-2.25 (m, 2H), 2.25-2.40 (m, 4H), 3.41-3.52 (m, 1H),
3.57-3.70 (m, 1H), 4.59 (dd, 1/8 H, J ) 8.1 and 1.8 Hz, CHCN),
4.75 (d, 7/8H, J ) 7.7, 2.2 Hz, CHCN); (ESI) MH/e 235 (MH+
of free base). Anal. (C14H22N2O1) C, H, N.
1
solid: mp ) 188-190 °C; H NMR (CD3OD) (10 to 1 mixture
of trans to cis amide rotomers); δ 1.06-1.46 (m, 4H), 1.65-
1.99 (m, 7H), 2.04-2.42 (m, 4H), 3.70 (t, 2H, J ) 6.3 Hz), 4.04
(d, 10/11 H, J ) 6.3 Hz), 4.26 (d, 1/11H, J ) 5.1 Hz), 4.84 (dd,
10/11 H, J ) 7.7 and 5.2 Hz, CHCN), 5.06 (dd, 1/11H, J ) 6.3
and 3.7 Hz, CHCN); MS m/z 236 (M+); Anal. (C13H21N3O1‚
1.0HCl‚0.81H2O‚0.09 Et2O) C, H, N.
1-[[(1-Ad a m a n tyl)a m in o]a cetyl]p yr r olid in e, Mon oh y-
d r och lor id e (12c). To an ice-water cooled mixture of 1-ami-
noadamantane (15.9 g, 105.0 mmol), K2CO3 (19.4 g, 140 mmol)
and CH2Cl2 (168 mL) was added a solution of ethyl bromoac-
etate (5.17 mL, 140 mmol) and CH2Cl2 (50 mL) dropwise over
20 min. The resulting cloudy, yellow-white mixture was then
stirred at ice-water temperature for 1.5 h, then at RT for 18
h and then filtered through a plug of Celite (50 g). The solids
and Celite plug were then washed with 100 mL of CH2Cl2, and
the combined organic fractions were concentrated in vacuo to
provide a cloudy, yellow-white oil. Purification via chroma-
tography (4/1 mixture of hexane/ethyl acetate as eluent)
followed by concentration in vacuo provided 6.9 g (83% yield)
of ester 4 as a white solid. To an ice-water cooled mixture of
ester 4 (6.87 g, 28.9 mmol), K2CO3 (15.9 g, 115 mmol), and
CH2Cl2 (80 mL) was added a solution of benzyl chloroformate
(4.34 mL, 30.4 mmol) in CH2Cl2 (26 mL) over 15 min. The
resulting cloudy, white mixture was then stirred at ice-water
temperature for 1.5 h, then at RT for 18 h, filtered, and
concentrated in vacuo to provide a white crunchy solid.
Purification via chromatography (9/1 mixture of hexane/ethyl
acetate as eluent) followed by concentration in vacuo provided
7.89 g (73.6% yield) of CBZ-protected 4 as a white solid. A
solution of this CBZ-protected 4 (7.89 g, 21.3 mmol), THF/
H2O: 1/1 (85 mL), and LiOH‚H2O (7.14 g, 170 mmol) was
stirred at 50 °C for 22h, cooled to RT, and poured into a 303
mL mixture of EtOAc/H2O/AcOH (200/100/3 ratio). The organic
layer is then isolated, washed with 2 × 50 mL of H2O, dried
over MgSO4, filtered, and concentrated in vacuo to provide (6.7
g, 92%) of acid 5 as an off-white solid. To a solution of acid 5
(0.500 g, 1.46 mmol) in CH2Cl2 (20 mL) were added HOBT
(0.195 g, 1.46 mmol), NMM (0.440 g, 4.37 mmol), and finally
DIC (0.460 g, 3.64 mmol). After stirring at room temperature
for 10 min, pyrrolidine (0.195 g, 1.46 mmol) was added to the
cloudy-white solution via syringe, and the reaction was stirred
at RT for 18 h. A CH2Cl2/NaHCO3(aq) workup and purification
via chromatography (40% EtOAc in hexane as eluent) provided
0.428 g (74% yield) of the CBZ-protected 12c as a white solid.
Deprotection of this intermediate was accomplished by shaking
0.428 g (1.08 mmol) of the CBZ-protected 12c in a solution of
EtOAc (50 mL), 10% Pd/C (0.100 g) and H2 (1atm) at RT for
1-[(S)-Cycloh exyl(cycloh exyla m in o)a cet yl]-2-cya n o-
(S)-p yr r olid in e (9c). Cyclohexanone (1.60 g, 15.0 mmol) was
added to a partially dissolved mixture of 9a (1.40 g, 5.1 mmol)
in 1,2-dichloroethane (56.0 mL). To the ice-water cooled
mixture was then added the sodium triacetoxyborohydride
(2.18 g, 10.0 mmol) in portions over 10 min. The ice-water
bath was then removed and the reaction stirred at room
temperature for 18 h. The reaction was partitioned between
ethyl acetate and water, and the ethyl acetate layer was dried
over Na2SO4, filtered, and concentrated. Purification via chro-
motography using 2% CH3OH in CH2Cl2 as eluent provided
(0.100 g, 6.3% yield) of the target as a white solid: mp ) 78-
80 °C; 1H NMR (CD3OD) (5 to 1 mixture of trans/cis amide
rotomers): δ 1.00-1.35 (m, 10 H), 1.55-1.85 (m, 10H), 1.88-
1.95 (m, 1H), 2.09-2.22 (m, 2H), 2.22-2.38 (m, 2H), 2.38-
2.52 (m, 1H), 3.56-3.67 (m, 2H), 3.69-3.80 (m, 1H), 4.82 (dd,
4/5H, J ) 8.1 and 4.4 Hz, CHCN), 5.17 (dd, 1/5H, J ) 7.4 and
1.4 Hz, CHCN); (ESI) MH/e 318 (MH+ of free base); Anal.
(C19H31N3O1‚1.2H2O‚0.2 CH3OH) C,H; N: calcd, 12.16; found,
11.53.
1-[(2S)-2-Am in o-3-m eth yl-1-oxobu tyl]-2-cya n o-(S)-p yr -
r olid in e, Mon oh yd r och lor id e (9d ). Same procedure as in
9a with N-t-boc-L-valine as a replacement for N-t-boc-L-R-
cyclohexylglycine. Product was a white solid; mp 162-164 °C;
1H NMR (CD3OD) (trans and cis amide rotomers in a ratio of
9:1) δ 0.98 (d, 1/3H, J ) 7.0 Hz), 1.08 (d, 3H, J ) 7.0 Hz), 1.14
(d, 2.67H, J ) 7.0 Hz), 2.07-2.42 (m, 5H), 3.72 (t, 2H, J ) 7.0
Hz), 4.10 (d, 1H, J ) 5.9 Hz), 4.84 (d,d, 9/10 H, J ) 8.1, 4.8
Hz, CHCN), 5.08 (d,d, 1/10H, J ) 6.2, 3.7 Hz, CHCN); (ESI)
MH/e 196 (MH+ of free base); Anal. (C10H17N3O1‚1HCl‚0.04H2O‚
0.14 EtOAc), C, H, N.
1-[(2S)-3-Meth yl-2-(m eth yla m in o)-1-oxobu tyl]-2-cya n o-
(S)-p yr r olid in e, Mon oh yd r och lor id e (9f). Same procedure
as in 9a with N-t-boc-3-methyl-2-methylamino-butyric acid as
a replacement for N-t-boc-L-R-cyclohexylglycine. Product was
a white solid; mp ) 223-225 °C; 1H NMR (CD3OD) (trans and
cis amide rotomers in a ratio of 9:1) δ 1.10 (d, 3H, J ) 7.0 Hz),

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2785
18 h. The solution was filtered through Celite and concentrated
in vacuo to yield the free base of 12c as a clear oil. HCl gas
was gently bubbled into a solution of this free base in THF
(25 mL) for ∼20 s stirred at RT for 15 min and then
concentrated in vacuo to provide the monohydrochloride salt
of 12c as a white solid (63% yield from 5); mp 242 °C, sublimes;
1H NMR (D2O) δ 1.67 (bq, 6H, J ab) 16.5 Hz), 1.89 (s, 6H),
1.80-2.00 (m, 4H), 2.16 (bs, 3H), 3.40 (q, 4H, J ) 6.3 Hz), 3.87
(s, 2H); (ESI) MH/e 263 (MH+ of free base); Anal. (C16H26N2O1‚
1HCl‚0.5H2O‚0.1THF) C, H, N.
procedure for 8n . Purification with chromatography (4%
MeOH in CH2Cl2 as eluent) provided a 42% yield over two
steps of the t-boc ester of 12l as a light yellow solid. Cleavage
of the t-boc ester was obtained by stirring at room temperature
with 2 equiv of TFA in 0.5 M solution of CH2Cl2 overnight.
The resulting TFA salt of 12l was obtained as a white,
hygroscopic solid in a 92% yield upon concentration in vacuo;
mp ) 80-82 °C; 1H NMR (CD3OD) (4 to 1 mixture of trans to
cis amide rotomers): δ 1.62 (s, 2H), 1.73 (s, 4H), 1.81-1.91-
(m,2H),1.89 (s, 4H), 2.01-2.14 (m, 2H), 2.24-2.40 (m, 2H),
2.40(s, 2H), 3.54-3.73 (m,2H), 4.01(s,2H), 4.49(d,d, 4/5H, J )
8.8, 3.3 Hz, CHCN), 4.69 (d,d,1/5H, J ) 8.5, 2.6 Hz, CHCN);
(ESI) MH/e 435 (MH+ of free base); Anal. (C17H26N2O4‚1.3TFA‚
0.28H2O) C, H, N. Note: An excess 0.3 equiv of TFA could not
be removed even with several toluene chases, triturations with
Et2O, and high vacuum (∼1 mm) over several days.
1-[(1-Ad a m a n tyl)a m in o]a cetyl]a za tid in e (12d ). Follow-
ing procedure for 12c but replacing pyrrolidine with azetidine
as the amine. Product was a white solid (67% yield from 5):
1
mp 65-67 °C; H NMR (CDCl3): δ 1.52-1.72 (m, 10H), 1.85
(bs, 2H), 2.05 (bs, 3H), 2.22-2.35 (m, 2H), 3.18 (s, 2H), 4.06
(t, 2H, J ) 7.7 Hz), 4.12 (t, 2H, J ) 7.7 Hz); (ESI) MH/e 249
(MH+ of free base); Anal. free base (C15H24N2O1‚0.06H2O) C,
H, N.
1-[[[3-[[(ter t-Bu tyla m in o)ca r bon yl]oxy]-1-a d a m a n tyl]-
a m in o]a cetyl]-2-cya n o-(S)-p yr r olid in e; Mon oh yd r och lo-
r id e (12m ). To a mixture of 1-aminoadamantan-3-ol32 (5.00
g, 30 mmol) and potassium carbonate (6.20 g; 45 mmol) in 150
mL of THF was added benzyl chloroformate (4.70 g, 33.0 mmol)
in dropwise fashion over a 10 min period. The mixture was
then stirred at RT for 2 h and then partitioned between ethyl
acetate and water. The aqueous layer was washed twice with
ethyl acetate (100 mL), and the combined organic layers were
then washed successively with 100 mL of aqueous 2 N sodium
hydroxide, water, and brine, dried over sodium sulfate, filtered,
and concentrated in vacuo to provide (8.03 g, 85% yield) of
1-benzylcarbamoyladamantane-3-ol (6) as a white solid. To a
clear solution of 1-benzylcarbamoyladamantane-3-ol (6) (1.00
g, 3.32 mmol) and tert-butyl isocyanate (380 µL, 3.32 mmol)
in 30 mL of CH2Cl2 was then syringe-added TMSCl (20.0 µL,
0.17 mmol). This reaction was then stirred at RT for 18 h,
concentrated in vacuo, and purified on silica gel with 20% ethyl
acetate in hexane as eluent to yield 3-[[(tert-butlyamino)-
carbonyl]oxy]-1-benzylcarbamoyladamantane (7a ) as a white
solid in quantitative yield. A mixture of 3-[[(tert-butylamino)-
carbonyl]oxy]-1-benzylcarbamoyladamantane (1.50 g, 3.75 mmol)
and 10% palladium on carbon (400 mg) in ethanol (150 mL)
was then shaken under hydrogen (50 psi) for 24 h, filtered
through Celite, and concentrated in vacuo to provide 3-[[(tert-
butylamino)carbonyl]oxy]-1-aminoadamantane as a clear oil
in 99% yield. Method A was then followed using 3-[[(tert-
butylamino)carbonyl]oxy]-1-aminoadamantane as the amine
1-[[(1-Ad a m a n tyl)a m in o]a cetyl]p ip er id in e (12e). Fol-
lowing procedure for 12c but replacing pyrrolidine with
piperidine as amine. Product was a white solid (68% yield from
5); mp 80-82 °C; 1H NMR (CDCl3): δ 1.48-1.70 (m, 18H), 2.03
(bs, 3H), 3.10(t, 2H, J ) 5.1 Hz), 3.40 (s, 2H), 3.52 (t, 2H, J )
5.1 Hz); (ESI) MH/e 277 (MH+ of free base); Anal. (C17H28N2-
O1‚0.1EtOAc); C, H, N.
1-[[(1-Ad a m a n tyl)a m in o]a cetyl]-3-p yr r olid in ol, Mon o-
h yd r och lor id e (12f). Following procedure for 12c but replac-
ing pyrrolidine with 3-pyrrolidinol as amine. Product was a
white solid (51% yield from 5); mp 230-232 °C; 1H NMR
(D2O): (3 to 2 ratio of amide rotomers: δ 1.65 (q, 6H, J ) 16.5
Hz), 1.87 (s, 6H), 1.91-2.10 (m, 2H), 2.15 (s, 3H), 3.37-3.63
(m, 3H), 3.83-3.97 (m, 2H), 4.47 (bs, 0.6H, OH of one amide
rotomer), 4.54 (bs, 0.4H, OH of other amide rotomer); (ESI)
MH/e 279 (MH+ of free base); Anal. (C16H26N2O2‚1.0HCl‚
0.95H2O‚0.35THF); C, H, N.
1,2,3,4-Tetr a h yd r o-2-[[(1-a d a m a n tyl)a m in o]a cetyl]-iso-
qu in olin e, m on oh yd r och lor id e (12g). Following procedure
for 12c but replacing pyrrolidine with 1,2,3,4-tetrahydroiso-
quinoline as amine. Product was a white solid (53% yield from
5); mp ) 260-262 °C; 1H NMR (D2O) (1 to 1 trans to cis amide
rotomers) δ 1.65 (bq, 6H, J ) 15.3 Hz), 1.86 (s, 6H), 2.13 (bs,
3H), 2.84 (t, 1H, J ) 6 Hz), 2.91 (t, 1H, J ) 6 Hz), 3.62-3.75
(m, 2H), 4.05 (s, 2H, one rotomer), 4.06 (s, 2H, other rotomer),
4.59 (s, 1H, one amide rotomer), 4.63 (s, 1H, other amide
rotomer), 7.15-7.25 (m, 4H); (ESI) MH/e 325 (MH+ of free
base); Anal. (C21H28N2O1‚1HCl‚0.896H2O‚0.14THF) C, H, N.
1-[[(3-H yd r oxy-1-a d a m a n t yl)a m in o]a cet yl]-2-cya n o-
(S)-p yr r olid in e (12j). Method A: 3-hydroxy-1-aminoadaman-
tane32 as amine and use of 3a . Product was a white solid; mp:
138-140 °C (when recrystallized from EtOAc and 2-pro-
panol: mp: 148-150 °C); OR: [d] ) -78.3° (c ) 9.73, MeOH);
1H NMR (CDCl3) (5 to 1 mixture of trans to cis amide
rotomers): δ 1.47-1.79 (m, 12H), 1.67 (s, 3H), 2.05-2.25 (m,
2H), 2.25-2.38 (m, 4H), 3.40-3.54 (m, 1H), 3.45 (d, 2H, J )
1.9 Hz), 3.57-3.67 (m, 1H), 4.77 (dd, 5/6H, J ) 7.5, 2.3 Hz,
CHCN), 4.86 (dd, 1/6H, J ) 7.9, 1.9 H, CHCN); (ESI) MH/e
304 (MH+ of free base), Anal. (C17H25N3O2) C, H, N.
N-(3-Hyd r oxy-1-a d a m a n tyl)glycyl-L-p r olin a m id e (12k ).
Following procedure for 8n and using 3-hydroxy-1-aminoada-
mantane32 as amine. Product was a white, hygroscopic solid;
mp ) 94-96 °C; 1H NMR (DMSO): (trans and cis amide
rotomers in a ratio of 2:1) δ 1.24-1.54 (m, 10H), 1.54-1.63
(m, 3H), 1.70-2.01 (m, 4H), 2.11 (bs, 2H), 2.17 (bs, 1H), 3.17
(s, 4/3H), 3.32 (s, 2/3H), 3.36-3.47 (m,1H), 3.47-3.59 (m,1H),
4.18 (dd, 2/3H, J ) 9.2 and 2.9 Hz), 4.30 (dd, 1/3H, J ) 8.4
and 2.9), 6.92 (bs, 2/3H, amide), 7.18 (bs, 1/3H, amide), 7.28
(bs, 2/3H, amide), 7.57 (bs,1/3H, amide); IR (KBr): 1640.4,
1677.3 (cm-1); (ESI) MH/e 322 (MH+ of free base); Anal.
(C17H27N3O3‚1.5H2O‚0.22 CH3OH) C, H, N.
1
to provide 12m as a yellow solid; mp: 210-212 °C; H NMR
(CD3OD) (9 to 1 mixture of trans to cis amide rotomers): δ
1.26 (s, 9H), 1.62-1.71 (m, 2H), 1.92 (s, 4H), 1.98-2.44 (m,
12H), 3.51-3.62 (m, 1H), 3.70-3.78 (m, 1H), 4.04 (d, 2H, J )
6.6 Hz), 4.82 (t, 9/10H, J ) 5.5 Hz, CHCN), 5.07 (d,d, 1/10H,
J ) 6.6, 2.9, CHCN); (ESI) MH/e 403 (MH+ of free base); Anal.
(C22H34N4O3‚1.0HCl‚0.67H2O‚0.40THF) C, H, N, Cl.
1-[[[3-[[[(4-Meth oxyp h en yl)a m in o]ca r bon yl]oxy]-1-a d -
a m a n tyl]a m in o]a cetyl]-2-cya n o-(S)-p yr r olid in e; Mon o-
h yd r och lor id e (12n ). The procedure for 12m was followed
except in the second step to form 7b where an equivalent of
4-methoxyphenyl isocyanate replaces tert-butyl isocyanate, 1,2-
dichloroethane was used as solvent instead of methylene
chloride, and the reaction is stirred at 50 °C for 18 h. Product
was provided as a yellow solid (55% yield from 7b); mp: 212-
214 °C; 1H NMR (CD3OD) (9 to 1 mixture of trans to cis amide
rotomers): δ 1.63-1.79 (m, 2H), 1.94 (s, 4H), 2.08-2.34 (m,
8H), 2.46 (s, 4H), 3.51-3.62 (m, 2H), 3.75 (s, 3H), 4.07 (d, 2H,
J ) 7.0 Hz), 4.82 (t, 9/10H, J ) 5.8 Hz, CHCN), 5.09 (d,d,
1/10H, J ) 7.0, 2.2, CHCN), 6.83 (d,d, 2H, J ) 8.8, 1.1 Hz),
7.27 (d, 2H, J ) 8.4 Hz); (ESI) MH/e 453 (MH+ of free base);
Anal. (C25H32N4O4‚1.0HCl‚1.15H2O‚0.55THF) C, H, N, Cl.
1-[[[3-[[(P h en ylam in o)car bon yl]oxy]-1-adam an tyl]am i-
n o]a cetyl]-2-cya n o-(S)-p yr r olid in e; Mon oh yd r och lor id e
(12o). The procedure for 12m was followed except in the
second step to form 7c where an equivalent of phenyl isocy-
anate replaces tert-butyl isocyanate, 1,2-dichloroethane was
used as solvent instead of methylene chloride, and the reaction
was stirred at 50 °C for 18 h. Product was a light yellow solid
N-(3-Hyd r oxy-1-a d a m a n tyl)glycyl-L-p r olin e, Mon otr i-
flou r oa cetic Acid (12l). Following the procedure for 2a ,
L-proline-tert-butyl ester (1e) was coupled with chloroacetyl
chloride to provide 2e (golden oil) which was then coupled with
3 equiv of 3-hydroxy-1-aminoadamantane32 following the
1
(48% yield from 7c); mp ) 205-207 °C; H NMR (CD3OD) (9

2786 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Villhauer et al.
to 1 mixture of trans to cis amide rotomers): δ 1.62-1.81 (m,
2H), 1.97 (s, 4H), 2.10-2.32 (m, 8H), 2.47 (s, 4H), 3.51-3.62
(m, 2H), 4.06 (d, 2H, J ) 7.0 Hz), 4.82 (t, 9/10H, J ) 5.9 Hz,
CHCN), 5.08 (d,d, 1/10H, J ) 7.0 and 2.2 Hz, CHCN), 6.99
(d,d,d, 1H, J ) 7.0, 7.0, 1.1 Hz), 7.24 (d,d,d, 2H, J ) 7.4, 7.4,
2.2 Hz), 7.37 (d, 2H, J ) 8.8 Hz); (ESI) MH/e 423 (MH+ of free
base); Anal. (C24H30N4O3‚1.0HCl‚1.06H2O‚0.64THF) C, H, N.
DP P -IV In h ibition Mea su r em en t ex Vivo. Ra t, Hu m a n ,
Mon k ey P la sm a Assa ys. Human, rat, or monkey plasma was
used as the source of DPP-IV in the assay. The standard assay
was modified from a previously published method.43 Five µL
of plasma was added to 96-well flat-bottom microtiter plates,
followed by the addition of 5 µL of 80 mM MgC12 in assay
buffer (25 mM HEPES, 140 mM NaC1, 1% RIA-grade BSA,
pH 7.8). After a 5-min preincubation at room temperature, the
reaction was initiated by the addition of 10 µL of assay buffer
containing 0.1 mM substrate (H-Gly-Pro-AMC; AMC is 7-amino-
4-methylcoumarin). The plates were covered with aluminum
foil (or kept in the dark) and incubated at room temperature
for 20 min. After incubation, fluorescence was measured using
a CytoFluor II fluorometer (excitation 380 nm/ emission 460
nm). Test compounds and solvent controls were added as 2
µL additions, and the assay buffer volume was reduced to 13
µL. A standard curve of free AMC was generated using 0-50
µM solutions of AMC. The curve generated, which was linear,
was used for interpolation of substrate consumption (catalytic
activity in nmoles substrate cleaved /min).
1-[[[3-[[[(Diisop r op yl)a m in o]ca r bon yl]-oxy]-1-a d a m a n -
tyl]a m in o]a cetyl]-2-cya n o-(S)-p yr r olid in e; Mon oh yd r o-
ch lor id e (12p ). The procedure for 12m was followed except
in the second step to form 7d where an equivalent of diiso-
propylcarbamoyl chloride replaces tert-butyl isocyanate, 1,2-
dichloroethane was used as solvent instead of methylene
chloride, and the reaction was stirred at 50 °C for 18 h. Product
1
was a white solid (63% yield from 7d ); mp: 148-150 °C; H
NMR (CD3OD) (8 to 1 mixture of trans to cis amide ro-
tomers): δ 1.21 (bs, 12H), 1.61-1.78 (m, 2H), 1.95 (s,4H),
2.04-2.36 (11H, m), 2.36-2.48 (m, 4H), 3.50-3.61 (m, 1H),
4.06 (d, 2H, J ) 7.4 Hz), 4.82(t, 8/9H, J ) 5.9 Hz, CHCN),
5.09 (d,d, 1/9H, J ) 7.0, 2.6 Hz, CHCN); (ESI) MH/e 431 (MH+
of free base); Anal. (C24H36N4O3‚1.0HCl‚1.2H2O‚0.40THF) C,
H, N.
DP P -II In h ibition Mea su r em en t in Vitr o. An extract of
bovine kidney homogenate, partially purified by ion-exchange
and adenosine deaminase chromatography, was used as the
source of DPP-II in the assay.44-46 The standard assay was
modified from a previously published method.47 Twenty mi-
crograms of DPP-II-containing fraction diluted to a final
volume of 60 µL in assay buffer (0.2 M Borate, 0.05 M Citrate,
pH 5.3) was added to 96-well flat-bottom microtiter plates,
followed by the addition of 10 µL of 10 mM o-phenanthroline
(to inhibit aminopeptidase activity) and 20 µL of 5 mM
substrate (H-Lys-Ala-AMC; AMC is 7-amino-4-methylcou-
marin). The plates were incubated at 37 °C for 30 min. After
incubation, fluorescence was measured using a CytoFluor II
fluorometer (excitation 380 nm/ emission 460 nm). Test com-
pounds and solvent controls were added as 20 µL additions,
and assay buffer volume is reduced to 50 µL. A standard curve
of AMC was generated using 0 to 100 µM of AMC. The curve
generated, which was linear, was used for interpolation of
catalytic activity (in nmoles substrate cleaved/min).
P ost -P r olin e Clea vin g E n zym e (P P CE ) In h ib it ion
Mea su r em en t in Vitr o. A cytosolic extract of human eryth-
rocytes, partially purified by ion-exchange chromatography,
was used as the source of PPCE in the assay. The standard
assay is modified from a previously published method.48 PPCE-
containing fraction (350 ng protein) diluted to a final volume
of 90 µL in assay buffer (20 mM NaPO4, 0.5 mM EDTA, 0.5
mM DTT, 1% BSA, pH 7.4) was added to 96-well flat-bottom
microtiter plates, followed by the addition of 10 µL of 0.5 mM
substrate (Z-Gly-Pro-AMC; AMC is 7-amino-4-methylcou-
marin). The plates were incubated at room temperature for
30 min. After incubation, fluorescence was measured using a
CytoFluor II fluorometer (excitation 380 nm/ emission 460 nm).
Test compounds and solvent controls were added as 20 µL
additions, and the assay buffer volume was reduced to 70 µL.
A standard curve of free AMC was generated using 0 to 5 µM
solutions of AMC. The curve generated, which was linear, was
used for interpolation of catalytic activity (in nmoles substrate
cleaved/min).
In Vivo Obese Ma le (fa /fa ) Zu ck er Ra t Stu d ies. Effect
of 12j on DP P -IV Activity, Active GLP -1 Levels, a n d
Glu cose a n d In su lin Excu r sion s. Studies were performed
on obese male Zucker (fa/fa) rats (Charles River Labs, Cam-
bridge, MA); controls (n ) 9) and 12j-treated (n ) 9). These
rats were purchased at 7 weeks of age, cannulated at 7.5
weeks, and studied beginning at around 11 weeks of age. In
the morning of the oral glucose tolerance test (OGTT), the rats
were “fasted” by removing food before the lights were turned
on, after which they were transferred to the experiment room
at 8:00 a.m.. 12j was dissolved in vehicle solution (0.5%
carboxymethylcellulose (CMC) and 0.2% Tween 80). The
cannulas were connected to sampling tubing (PE-100, 0.034
in. i.d. × 0.06 in. o.d.), which were filled with saline. After
30-40 min cage acclimation, a 0.5 mL baseline blood sample
was taken at t ) -15 min, and the rats were then orally dosed
1-[[[3-[[[(Cycloh exyl)a m in o]ca r bon yl]-oxy]-1-a d a m a n -
tyl]a m in o]a cetyl]-2-cya n o-(S)-p yr r olid in e; Mon oh yd r o-
ch lor id e (12q). The procedure for 12m was followed except
in the second step to form 7e where an equivalent of cyclohexyl
isocyanate replaces tert-butyl isocyanate, 1,2-dichloroethane
was used as solvent instead of methylene chloride, and the
reaction was stirred at 50 °C for 18 h. Product was provided
as a light brown solid (41% yield from 7e); mp: 155-157 °C;
1H NMR (CD3OD) (9 to 1 mixture of trans to cis amide
rotomers): δ 1.08-1.40 (m, 6H), 1.56-1.78 (m, 5H), 1.78-1.98
(m, 6H), 1.98-2.47 (m, 12H), 3.43-3.61 (m, 2H), 4.04 (d, 2H,
J ) 7.0 Hz), 4.83 (t, 9/10 H, J ) 5.9, Hz, CHCN), 5.09 (dd,
1/10H, J ) 7.5, 1.8, CHCN); (ESI) MH/e 429 (MH+ of free base);
Anal. (C24H36N4O3‚1.0HCl‚0.77H2O‚0.32THF) C, H, N.
1-[[3-Acet yloxy-1-a d a m a n t yl)a m in o]-a cet yl]-2-cya n o-
(S)-p yr r olid in e; Mon oh yd r och lor id e (12r ). The procedure
for 12m was followed except in the second step where a
standard acylation of 6 is performed using 1.2 equiv of acetyl
chloride, 3.0 equiv of pyridine, 0.1 equiv of DMAP, and 1,2-
dichloroethane which are all stirred at room temperature for
24 h. Product obtained was a light yellow taffy (45% yield from
7f); 1H NMR (CD3OD) (6 to 1 mixture of trans to cis amide
rotomers): δ 1.55-1.72 (6H, m), 1.94 (s, 3H), 1.98-2.34 (m,
13H), 3.42-3.58 (m, 1H), 3.44 (d, 1H, J ) 6.6 Hz), 3.62-3.75
(m, 1H), 4.77 (t, 6/7H, J ) 5.5 Hz, CHCN), 5.09 (d,d,1/7H, J )
6.6 and 3.3 Hz, CHCN); (ESI) MH/e 346 (MH+ of free base);
Anal. (C19H27N3O3‚0.38H2O) C, H, N.
In Vitr o Stu d ies. DP P -IV In h ibition Mea su r em en t in
Vitr o: Ca co-2 Assa y. An extract from human colonic carci-
noma cells (Caco-2; American Type Culture Collection; ATCC
HTB 37) was used as the source of DPP-IV in the assay. The
cells were differentiated to induce DPP-IV expression as
described by previously.42 Cell extract was prepared from cells
solubilized in lysis buffer (10 mM Tris-HC1, 0.15 M NaC1, 0.04
T.I.U. (trypsin inhibitor unit) aprotinin, 0.5% nonidet-P40, pH
8.0) then centrifuged at 35 000g for 30 min at 4 °C to remove
cell debris. The assay was conducted by adding 20 µg of
solubilized Caco-2 protein, diluted to a final volume of 125 µL
in assay buffer (25 mM Tris-HC1 pH 7.4, 140 mM NaC1, 10
mM KC1, 1% bovine serum albumin) to 96-well flat-bottom
microtiter plates. The reaction was initiated by adding 25 µL
of 1 mM substrate (H-Ala-Pro-pNA; pNA is p-nitroaniline). The
reaction was run at room temperature for 10 min, and then
19 µL of 25% glacial acetic acid was added to stop the reaction.
Fluorescence was measured using a CytoFluor II fluorometer
(excitation 380 nm/ emission 460 nm). Test compounds and
solvent controls were added as 30 µL additions, and the assay
buffer volume was reduced to 95 µL. A standard curve of free
p-nitroaniline was generated using 0-100 µM pNA in assay
buffer. The curve generated, which was linear, was used for
interpolation of substrate consumption (catalytic activity in
nmoles substrate cleaved /min).

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2787
with CMC or 12j (10 µmol/kg), after which additional baseline
blood samples were taken at t ) -5, -2.5, and 0 min. The
animals were then administered an oral glucose solution (10%
glucose, 1 g/kg) immediately after t ) 0′. The rest of the
samples were taken at 1, 3, 5, 10, 15, 20, 30, 45, 60, 75, and
90 min. Throughout the OGTT, an equal volume of donor blood
was used to replace the blood withdrawn during sampling.
Donor blood was obtained from donor rats through cardiac
puncture. The collected blood samples (0.5 mL) were im-
mediately transferred into chilled Eppendorf tubes containing
50 µL of EDTA: trasylol (25 mg/mL of 10 000 trasylol (FBA
Pharmaceuticals)) and used for the measurement of glucose
and insulin levels and DPP-IV activity. Larger blood samples
(0.75 mL) were collected at t ) -15, 0, 5, 10, 15, and 30 min
for GLP-1 (7-36 amide) measurements. To these tubes, the
DPP-IV inhibitor valine pyrrolidide49 was added to yield a final
concentration in the blood of 1 µM. Technical difficulties with
obtaining blood samples after minute 20 for one rat in both
the CMC and 12j groups resulted in the inability to calculate
glucose and insulin AUC data for those rats, leading to AUC
data with an n ) 8/group. Measurement of plasma glucose was
made using a modification of a Sigma Diagnostics glucose
oxidase kit. DPP-IV activity was measured in plasma samples
obtained at -5, 0, 20, 45, and 90 min DPP-IV activity as
previously described in the above ex vivo rat plasma experi-
mental. Plasma levels of GLP-1 (7-36 amide) were measured
using the GLP-1 (active) Elisa Kit (Linco Research Cat# EGLP-
35K, St. Charles, MO). A double antibody RIA method using
a rat-specific antiinsulin antibody from Linco Research (#1013,
lot # ARI-02 (16T), St. Charles, MO) was used to measure
plasma insulin. All data were expressed as the mean ( SEM
for each experimental group of rats. Statistical analysis of data
was performed using a two-tailed unpaired Student’s t-test
(Microsoft Excel 5.0), or a two-way repeated measures ANOVA
with a Tukey post-ANOVA test (SPSS Inc. SigmaStat for
windows, version 2.03).
In Vivo Cyn om olgu s Mon k ey P K/P D Stu d ies Usin g 8c
a n d 12j. Ketamine-anesthetized male healthy cynomolgus
monkeys received either 8c (n ) 2) or 12j (n ) 3) (dissolved in
CMC/Tween-80) by oral gavage (1.007 µmol/kg), and by
intravenous administration (0.399 µmol/kg) (dissolved in
saline). For iv study, compound was administered (0.4 mL/kg
over 1 min) in 0.9% saline as vehicle. Different monkeys were
used for each dosage regimen. Basal blood samples were
collected at -10 min and immediately prior to administration
of compound. Blood samples were collected at 0.03, 0.08, 0.17,
0.25, 0.33, 0.42, 0.5, 0.75, 1, 1.25, 1.5, 2, 2.5, 3, 3.5, 4, 7, 12,
and 25 h postdose for both routes of administration. Blood was
obtained into heparin-coated syringes, transferred to micro-
centrifuge tubes, and centrifuged to separate the plasma. The
plasma was stored at -80 °C in fresh microcentrifuge tubes
until assay. DPP-IV activity was measured in a similar
manner was as previously described in the above ex vivo rat
and human plasma experimentals. Plasma DPP-IV activities
were calculated and expressed as ‘percent of baseline’ to reduce
variability due to individual differences in plasma enzyme
activity. Area-under-curve (AUC) values for DPP-IV activity
were calculated from time (hours after dose) vs effect (percent
inhibition) curves from individual animals using the trapezoi-
dal method. The ratio of dose-normalized effect AUC for oral/
intravenous administration routes was taken as an estimate
of effect bioavailability. Parent drug concentrations were
determined using an HPLC/MS/MS method with a limit of
mediate, Elina Dunn, Huiping H. Gu, Lori A. Bolognese,
Xue Li, Wieslawa M. Maniara, Michele Valentin, and
Stephen C. Weldon for their assistance in the in vivo
analysis of 12j, and Robert C. Anderson and Philip A.
Bell for helpful discussions. We also thank Dr. Gregory
Bebernitz for helpful comments while reviewing this
manuscript.
Su p p or tin g In for m a tion Ava ila ble: ORTEP drawing
and atomic coordinate information for 12j and experimental
details for 8b, 8d -j, 8o, 9b, 9e, 9h , 10a -h , 11a -a k , 12a ,b,
and 12h -i. This material is available free of charge via the
Internet at http://pubs.acs.org.
Refer en ces
(1) (a) Yaron, A.; Nadier, F. Proline-Dependent Structural and
Biological Properties of Peptides and Proteins. Crit. Rev. Bio-
chem. Mol. Biol. 1993, 28, 31-81. (b) Demuth, H.; Heins J .
Catalytic mechanism of Dipeptidyl Peptidase-IV. In Dipeptidyl
Peptidase-IV (CD26) in Metabolism and the Immune Response;
Fleischer, B., Ed.; Landes, R. G.: Austin, TX, 1995; Vol. 3, pp
1-35.
(2) Mentlein, R. Dipeptidyl-peptidase IV (CD26)-role in the inactiva-
tion of regulatory peptides. Regul. Pept. 1999, 85, 9-24.
(3) (a) J acotot, E.; Callebaut, C.; Blanco, J .; Krust, B.; Neubert, K.;
Barth, A.; Hovanessian, A. G. Dipeptidyl-peptidase IV-B, a novel
form of cell-surface-expressed protein with dipeptidyl-peptidase
IV activity. Eur. J . Biochem. 1996, 239, 248-258. (b) Schneider,
B. L.; Thevananther, S.; Moyer, M. S.; Walters, H. C.; Rinaldo,
P.; Devarajan, P.; Sun, A. Q.; Dawson, P. A.; Ananthanarayanan,
M. Cloning and Characterization of a Novel Peptidase from Rat
and Human Ileum. J . Biol. Chem. 1997, 272, 31006-31015. (c)
Duke-Cohan, J . S.; Gu, J .; McLaughlin, D. F.; Xu, Y.; Freeman,
G. J .; Schlossman, S. F. Attractin (DPPT-L), a member of the
CUB family of cell adhesion and guidance proteins, is secreted
by activate human T lymphocytes and modulate immune cell
interactions. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 11336-
11341. (d) Niedermeyer, J .; Enenkel, B.; Park, J . E.; Lenter, M.;
Rettig, W. J .; Damm, K.; Schnapp, A. Mouse fibroblast-activation
protein-conserved Fap gene oranization and biochemical function
as a serine protease. Eur. J . Biochem. 1998, 254, 650-654. (e)
Chiravuri, M.; Schmitz, T.; Yardley, K.; Underwood, R.; Dayal,
Y.; Huber, B. T. A Novel Apoptotic Pathway in Quiescent
Lymphocytes Identified by Inhibition of a Post-Proline Cleaving
Aminodipeptidase: A Candidate Target Protease, Quiescent Cell
Proline Dipeptidase. J . Immunol. 1999, 163, 3092-3099.
(4) (a) Holst, J . J . Glucagon-like peptide-1 (GLP-1) a newly discov-
ered GI hormone. Gastroenterology 1994, 107, 1048-1055. (b)
Orsakov C. Glucagon-like peptide-1, a new hormone of the
enteroinsular axis. Diabetologia 1992, 35, 701-711. (c) Drucker,
D. J . Glucagon-Like Peptides. Diabetes 1998, 47, 159-169.
(5) (a) Mentlein, R.; Gallwitz, B.; Schmidt, W. E. Dipeptidyl-
peptidase IV hydrolyzes gastric inhibitory polypeptide, glucagon-
like peptide-1 (7-36), peptide histidine methionine and is
responsible for their degradation in human serum. Eur. J .
Biochem. 1993, 214, 829-835. (b) Deacon, C. F.; J ohnsen, A.
H.; Holst, J . J . Degradation of glucagon-like peptide-1 by human
plasma in vitro yields an N-terminally truncated peptide that
is a major endogenous metabolite in vivo. J . Clin. Endocrinol.
Metab. 1995, 80, 952-957. (c) Kieffer, T. J .; McIntosh, C. H. S.;
Pederson, R. A. Degradation of glucose-dependent insulinotropic
polypeptide and truncated glucagon-like peptide-1 in vitro and
in vivo by dipeptidyl peptidase IV. Endocrinology 1995, 136,
3585-3596. (d) Lene, H.; Deacon, C. F.; Orskov, C.; Holst, J . J .
Glucagon-like peptide-1-(7-36) amide is transformed to gluca-
gon-like peptide-1-(9-36) amide by dipeptidyl peptidase IV in
the capillaries supplying the L cells of the porcine intestine.
Endocrinology 1999, 140, 5356-5363.
(6) Ahre´n, B.; Holst, J . J .; Martensson, H.; Balkan, B. Improved
glucose tolerance and insulin secretion by inhibition of dipeptidyl
peptidase IV Eur. J . Pharmacol. 2000, 404 (1/2), 239-245.
(7) Deacon, C. F.; Hughes, T. E.; Holst, J . J . Dipeptidyl peptidase
IV inhibition potentiates the insulinotropic effect of glucagon-
like peptide 1 in the anesthetized pig. Diabetes 1998, 47, (5),
764-769.
(8) (a) Pederson, R. P.; White, H. A.; Schlenzig, D.; Pauly, R. P.;
McIntosh, C. R. P.; Demuth, H. U. Improved glucose tolerance
in Zucker fatty rats by oral administration of the dipeptidyl
peptidase IV inhibitor isoleucine thiazolidide. Diabetes 1998, 47,
(8) 1253-1258. (b) Pauly, R. P.; Demuth, H.-U.; Rosche, R.;
Schmidt, J .; White H. A.; Lynn F.; McIntosh, C. H. S.; Pederson,
R. A. Improved glucose tolerance in rats treated with the
dipeptidyl peptidase IV (CD26) inhibitor Ile-thiazolidide Me-
tabolism 1999, 3, 385-389. (c) Pospisilik, J . A.; Stafford, S. G.;
quantification of
1 ng/mL. Pharmacokinetic parameters
were calculated using noncompartment modeling, and the
AUC was calculated using the linear trapezoidal method.
Absolute oral bioavailability was calculated by (AUC0-∞po ×
399)/(AUC0-∞iv × 1007).
Ack n ow led gm en t. The authors thank Dr. Phillip
E. Fanwick from Purdue University for conducting the
X-ray crystallographic analysis of 12j. The authors
thank Eric M. Loeser for monohydrochloride of 8q,
George T. Lee for a generous quantity of a key inter-

2788 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13
Villhauer et al.
Demuth, H.-U.; Brownsey, R.; Parkhouse, W.; Finegood, D. T.;
McIntosh, C. H. S. Pederson, R. A. Long-Term Treatment with
the Didpeptidyl Peptidase IV Inhibitor P32/98 causes sustained
improvements in glucose tolerance, insulin sensitivity, hyper-
insulinemia, and â-cell glucose responsiveness in VDF (fa/fa)
Zucker Rats. Diabetes 2002, 51, 943-950.
ferentiation of Human Pancreatic Islet-Derived Progenitor Cells
into Insulin-Producing Cells. Endocrinology 2002, 143 (8), 3152-
3161. (h) Tourrel, C.; Bailbe, D.; Lacorne, M.; Meile, M. J .;
Kergoat, M.; Portha, B. Persistent Improvement of Type 2
Diabetes in the Goto-Kakizaki Rat Model by Expansion of the
B-Cell Mass During the Prediabetic Period with Glucagon-Like
Peptide-1 or Exendin-4. Diabetes 2002, 51, 1443-1452. (i)
Perfetti, R.; Zhou, J .; Doyle, M. E.; Egan, J . M. Glucagon-Like
Peptide -1 Induces Cell Proliferation and Pancreatic-Duodenum
Homeobox-1 Expressoin and Increases Endocrine Cell Mass in
the Pancreas of Old, Glucose-Intolerant Rats. Endocrinology
2000, 141 (12), 4600-4605. (j) Li, Y.; Hansotia, T.; Yusta, B.;
Ris, F.; Halban, P. A.; Drucker, D. J . Glucagon-like Peptide-1
Receptor Signaling Modulates beta Cell Apoptosis. J . Biol. Chem.
2003, 278 (1), 471-478.
(9) (a) Balkan, B.; Kwasnik, L.; Miserendino, R.; Holst, J . J .; Li, X.
Inhibition of dipeptidyl peptidase IV with NVP-DPP728 in-
creases plasma GLP-1 (7-36 amide) concentrations and im-
proves oral glucose tolerance in obese Zucker rats. Diabetologia
1999, 42 (11), 1324-1331. (b) Mitani, H.; Takimoto, M.; Hughes
T. E.; Kimur, M. Dipeptidyl peptidase IV inhibition improves
impaired glucose tolerance in high-fat diet-fed rats: study using
a Fisher 344 rat substrain deficient in its enzyme activity.
J apanese J ournal of Pharmacology 2002, 88, 442-450. (c)
Mitani, H.; Takimoto, M and Kimura, M. Dipeptidyl peptidase
IV inhibitor, NVP-DPP728, ameliorates early insulin response
and glucose tolerance in aged rats but not in aged Fisher 344
rats lacking its enzyme activity. J pn J . Pharmacol. 2002, 88,
451-458.
(10) Sudre, B.; Broqua, P.; White, R. B.; Ashworth, D.; Evans, E. M.;
Haigh, R.; J unien, J .-L.; Aubert, M. L. Chronic Inhibition of
Circulating Dipeptidyl Peptidase-IV by FE 999011 Delays the
Occurrence of Diabets in Male Zucker Diabetic Fatty Rats.
Diabetes 2002, 51, 1461-1469.
(11) Marguet, D.; Baggio, L.; Kobayashi, T.; Bernard, A.-M.; Pierres,
M.; Nielsen, P. F.; Ribel, U.; Wantanabe, T.; Drucker, D. J .;
Wagtmann, N. Enhanced insulin secretion and improved glucose
tolerance in mice lacking CD26. Proc. Natl. Acad. Sci. U.S.A.
2000, 97, 6864-6879.
(12) Holst, J . J .; Deacon, C. F. Inhibition of the activity of Dipeptidyl-
Peptidase IV as a treatment for Type 2 Diabetes. Diabetes 1998,
47, 1663-1670. (b) Zander, M.; Madsbad, S.; Madsen, J . L.;
Holst, J . J . Effect of 6-week course of glucagon-like peptide 1 on
glycaemic control, insulin sensitivity, and â-cell function in type
2 diabetes; a parallel-group study. Lancet 2002, 359, 824-830.
(13) (a) Ahren, B.; Larsson, J .; Holst, J . J . Effects of glucagon-like
peptide-1 on islet function and insulin sensitivity in noninsulin
dependent diabetes mellitus. J . Clin. Endocrinol. Metab. 1997,
82, 473-478. (b) Rachman, J .; Borrow, B. A.; Levy, J . C.; Turner
R. C. Near normalizatoin of diurnal glucose concentrations by
continuous administration of glucagon-like peptide 1 (GLP-1)
in subjects with NIDDM. Diabetologia 1997, 40, 205-211. (c)
Nauck, M. A.; Wollschlager, D.; Werner, J .; Holst, J . J .; Orskov,
C.; Creutzfeldt, W.; Willms, B. Effects of subcutaneous glucagon-
like peptide 1 (GLP-1 [7-36 amide]) in patients with NIDDM.
Diabetologia 1996, 39, 1546-1553. (d) Drucker, D. J . Biological
actions and therapeutic potential of the glucagon-like peptides.
Gastroenterology 2002, 122 (2), 531-544. (e) Egan, J .; Clocquet,
A. R.; Elahi, D. The Insulinotropic Effect of Acute Exendin-4-
administered to Humans: Comparison of Nondiabetic State to
Type 2 Diabetes. J . Clin. Endocrinol. Metab. 2002, 87 (3), 1282-
1290.
(16) (a) Flint, A.; Raben, A.; Ersboll, A. K.; Holst, J . J .; Astrup, A.
The effect of physiological levels of glucagon-like peptide-1 on
appetite, gastric emptying, energy and substrate metabolism in
obesity. Int. J . Obes. 2001, 25, 781-792.
(17) (a) Doyle, M. E.; Egan, J . M. Glucagon-like peptide-1. Recent
Prog. Horm. Res. 2001, 56, 377-399. (b) Villanueva-Pen˜acarrillo,
M. L.; Puente, J .; Redondo, A.; Clemente, F.; Valverde, I. Effect
of GLP-1 treatment on GLUT2 and GLUT4 expression in Type
1 and Type 2 rat diabetic models. Endocrine 2001, 15 (2), 241-
248. (c) Sandhu, H.; Wiesenthal, S. R.; MacDonald, P. E.; McCall,
R. H.; Tchipashvili, V.; Rashid, S.; Satkunarajah, M.; Irwin, D.
M.; Shi, Z. Q.; Brubaker, P. L.; Wheeler, M. B.; Vranic, M.;
Efendic, S.; and Giacca, A. Glucagon-Like Peptide 1 increases
inslin sensitivity in depancreatized dogs. Diabetes 1999, 48,
1045-1053. (d) Pospisilik, J . A.; Stafford, S. G.; Demuth, H.-U.;
McIntosh, H. S.; Pederson, R. A. Long-Term Treatment with
Dipeptidyl Peptidase IV Inhibitor Improves Hepatic and Pe-
ripheral Insulin Sensitivity in the VDF Zucker Rat - A Eugly-
cemic-Hyperinsulinemic Clamp Study. Diabetes 2002, 51, 2677-
2683.
(18) (a) Burcelin, R.; Da Costa, A.; Drucker, D.; Thorens, B. Glucose
competence of the hepatoprotal vein sensor requires the presence
of an activated glucagon-like peptide-1 receptor. Diabetes 2001,
50, 1720-1728. (b) Balkan, B.; Li, X. Portal GLP-1 administra-
tion in rats augments the insulin response to glucose via
neuronal mechanisms. Am. J . Physiol. Regul. Integ. Comput.
Physiol. 2000, 279: R1449-R1454.
(19) (a) J uana, F.; Valdeolmillos, M. Glucose-dependent stimulatory
effect of glucagon-like peptide 1(7-36) amide on the electrical
activity of pancreatic â-cells recorded in vivo. Diabetes 1999, 48
(4), 754-757. (b) Nauck, M. A.; Heimessat, M. M.; Behle, K.;
Holst, J . J .; Nauck, M. S.; Ritzel, R.; Hufner, M.; Schmiegel,
Wolff, H. Effects of glucagon-like peptide 1 on counterregulatory
hormone responses, cognitive functions, and insulin secretion
during hyperinsulinemic, stepped hypoglycemic clamp experi-
ments in healthy volunteers. J . Clin. Endocrinol. Metab. 2002,
87 (3), 1239-1246.
(20) These nonsubstrate DPP-IV inhibitors provide low micromolar
IC50’s, and, except for refs 20(a) and 20(b), contain a protona-
table amine: (a) Shimazawa, R.; Takayama, H.; Kato, F.; Kato,
M.; Hashimoto, Y. Nonpeptide small-molecular inhibitors of
Dipeptidyl Peptidase IV; N-phenylphthalimide analogues. Bioorg.
Med. Chem. Lett. 1999, 9, 559-562. (b) Miyachi, H.; Kato, M.;
Kato, F.; Hashimoto, Y. Novel potent nonpeptide aminopeptidase
N inhibitors with a cyclic imide skeleton. J . Med. Chem. 1998,
41, 3, 263-265. (c) Yamada, M.; Okagaki, C.; Higashijima, T.;
Tanaka, S.; Ohnuki, T.; Sugita, T. A potent dipeptide inhibitor
of dipeptidyl peptidase IV. Bioorg. Med. Chem. Lett. 1998, 8,
1537-1540. (d) Coppola, G. M.; Zhang, Y. L.; Schuster, H. F.;
Russell, M. E.; Hughes T. E. 1-Aminomethylisoquinoline-4-
carboxylates as novel Dipeptidyl Peptidase IV inhibitors. Bioorg.
Med. Chem. Lett. 2000, 10, 1555-1558. (e) Kanstrup, A. B.;
Christiansen, L. B.; Lundbeck, J . M.; Kristiansen; M. Prepara-
tion of piperazinylpurinediones as inhibitors of dipeptidylpep-
tidase IV. Patent WO 02002560, J uly 4th, 2002.
(21) Irreversible oligopeptides with the N-terminal X-Pro-X sequence
(low micromolar IC50) - Rahfeld, J .; Schierhorn, M.; Hartrodt,
B.; Neubert, K.; Heins, J . Are Diprotin A and (Ile-Pro-Ile) and
Diprotin B (Val-Pro-Leu) Inhibitors or Substrates of Dipeptidyl
Peptidase IV? Biochim. Biophys. Acta 1991, 1076, 314-316.
(22) Suicide substrate (latent quinoniminium methide electrophile
- large cyclopeptide). Nguyen, C.; Blanco, J .; Mazaleyrat, J . P.;
Krust, B.; Callebaut, C.; J acotot, E.; Hovanessian, A. G.;
Wakselman, M. Specific and Irreversible Cyclopeptide Inhibitors
of Dipeptidyl Peptidase IV Activity of the T-Cell Activation
Antigen CD26. J . Med. Chem. 1998, 41, 2100-2110.
(14) (a) Wettergren, A.; Schjoldager, B.; Martensen, P. E.; Myhre,
J .; Christiansen, J .; Holst, J . J . Truncated GLP-1 (proglucagon
72-107 amide) inhibits gastric and pancreatic functions in
man. Dig. Dis. Sci. 1993, 38, 665-673. (b) Nauck, M. A.;
Niedereichholz, U.; Ettler, R.; Holst, J . J .; Orskov, C.; Ritzel,
R.; Schmigel, W. H. Glucagon-like peptide-1 inhibition of gastric-
emptying outweighs its insulinotropic effects in healthy humans.
Am. J . Physiol. 1997, 273, E981-E988. (c) Nauck, M. A. Is GLP-1
an incretin hormone? Diabetologia 1999, 42, 373-379. (d) Egan,
J .; Clocquet, A. R.; Elahi, D. The Insulinotropic Effect of Acute
Exendin-4-Administered to Humans: Comparison of Nondia-
betic State to Type 2 Diabetes. J . Clin. Endocrinol. Metab. 2002,
87 (3), 1282-1290.
(15) (a) Habener, J . F. Glucagonlike peptide-1 agonist stimulation
of â-cell growth and differentiation. Curr. Opin. Endocrinol.
Diabetes 2001, 8 (2), 74-81. (b) Ling, Z.; Wu, D.; Zambre, Y.;
Flamez, D.; Drucker, D. J .; Pipeleers, D. G.; Schuit, F. C.
Glucagon-like peptide 1 receptor signaling influences topography
of islet cells in mice. Virchows Arch. 2001, 438 (4), 382-387. (c)
Aspelun, F.; Egan, J . M.; Slezak, L. A.; Sritharan, K. C.; Elahi,
D.; Andersen, D. K. Glucagon-like peptide-1 and exendin-4
improve glucose tolerance and induce islet cell growth in
pancreatitis in rats. Surg. Forum 2000, 51, 40-42. (d) Kemp,
D. M.; Habener, J . F. Insulinotropic hormone glucagon-like
peptide 1 (GLP-1) activation of insulin gene promoter inhibited
by p38 mitogen-activated protein kinase. Endocrinology 2001,
142 (3), 1179-1187. (e) Parkes, D. G.; Pittner, R.; J odka, C.;
Smith, P.; Young, A. Insulinotropic actions of exendin-4 and
glucagon-like peptide-1 in vivo and in vitro. Metab. Clin. Exp.
2001, 50 (5), 583-589. (f) Farilla, L.; Hui, H.; Berolotto, C.; Kang,
E.; Bulotta, A.; Di Mario, U.; Perfetti, R.. Glucagon-like Peptide-1
promotes islet cell growth and inhibits apoptosis in Zucker
diabetic rats. Endocrinology 2002, 143 (11), 4397-4408. (g)
Abraham, E. J .; Leech, C. A.; Lin, J . C.; Zuelwski, H.; Habener,
J . F. Insulinotropic Hormone Glucagon-Like Peptide-1 dif-
(23) Heins, J .; Welker, P.; Scho¨nlein, C.; Born, I.; Hartrodt, B.;
Neubert, K.; Tsuru, D.; Barth, A. Mechanism of Proline-Specific
Proteinases; (I) Substrate Specificity of Dipeptidyl Peptidase IV
from Pig Kidney and Protein-Specific Endopeptidase from Fla-
vobacterium meningosepticum. Biochim. Biophys. Acta 1988,
954, 161-169.

Antihyperglycemic Dipeptidyl Peptidase IV Inhibitor
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 13 2789
(24) (a) Barth, A.; Heins, J .; Fischer, G.; Neubert, K.; Schneeweiss,
B. Dipeptidyl-peptidase IV: mechanism and specificity of the
substrate cleavage. Mol. Cell. Regul. Enzyme Act., Part 1-3
1984, 46, 297-339. (b) Scho¨n, E.; Born, I.; Demuth, H.-U.; Faust,
J .; Neubert, K.; Steinmetzer, T.; Barth, A.; Ansorge, S. Dipeptidyl
Peptidase IV in the Immune System., Biol. Chem. Hoppe-Seyler
1991, 372, 305-311.
(37) Augustyns, K. J . L.; Lambeir, A. M.; Borloo, M.; De Meester, I.;
Vedernikova, I.; Vanhoof, G.; Handriks, D.; Scharpe´, S.; Hae-
maers, A. Pyrrolidides; synthesis and structure-activity rela-
tionship as inhibitors of dipeptidyl peptidase IV. Eur. J . Med.
Chem. 1997, 32, 301-309.
(38) PPCE and DPP-II are the standard enzymes in DPP-IV selectiv-
ity studies. (a) Belyaev, A.; Zhang, X.; Augustyns, K.; Lambeir,
A. M.; De Meester, I.; Verdernikova, I.; Scharpe´, S.; Haemers,
A. Structure-Activity Relationship of Diaryl Phosphonate Esters
a Potent Irreversible Dipeptidyl Peptidase IV Inhibitors. J . Med.
Chem. 1999, 42, 1041-1052. (b) Sedo, A.; Malik, R. Dipeptidyl
peptidase IV-like molecules: homologous proteins or homologous
activities. Biochim. Biophys. Acta 2001, 1550, 107-116.
(39) Rasmussen, H. P.; Branner, S.; Wiberg, F. C.; Wagtmann, N.
Crystal structure of human dipeptidyl peptidase IV/CD26 in
complex with a substrate analogue. Nature Struct. Biol. 2003,
10 (1), 19-25.
(25) For comprehensive reviews on DPP-IV inhibitors, please see the
following. (a) Villhauer, E. B.; Coppola, G. M.; Hughes, T. E.
DPP-IV inhibition and Therapeutic Potential. Annu. Rep. Med.
Chem. 2001, 36, 191-200. (b) Augustyns, K.; Bal, G.; Thonus,
G.; Belyaev, A.; Zhang, X. M.; Bollaert, W.; Lambeir, A. M.;
Durinx, C.; Goossens, F.; Haemers A. The Unique Properties of
Dipeptidyl-Peptidase IV (DPP-IV/CD26) and the Therapeutic
Potential of DPP-IV Inhibitors Curr. Med. Chem. 1999, 6, 311-
327. (c) Snow, R. J .; Bachovchin, W. W. Boronic acid inhibitors
of dipeptidyl peptidase IV; a new class of immunosuppressive
agents. Adv. Med. Chem. 1995, 3, 149-177.
(40) (a) Rothenberg, P.; Kalbag, J .; Smith, H. T.; Gingerich, R.;
Nedelman, J .; Villhauer, E.; McLeod, J .; Hughes, T. Treatment
(26) J enkins, P. D.; J ones, D. M. DP-IV-serine protease inhibitors,
Patent WO 95/15309, J une 8th, 1995.
with
a DPP-IV Inhibitor, NVP-DPP728, Increases Prandial
(27) Ashworth, D. M.; Atrash, B.; Baker, G. R.; Baxter, A. J .; J enkins,
P. D. 2-Cyanopyrrolidides as potent, stable inhibitors of Dipep-
tidyl Peptidase IV. Bioorg. Med. Chem. Lett. 1996, 6, (10), 1163-
1166.
(28) Li, J .; Wilk, E.; Wilk, S. Aminoacylpyrrolidine-2-nitriles: Potent
and stable inhibitors of Dipeptidyl-Peptidase IV (CD26). Arch.
Biochem. Biophys. 1995, 323 (1), 148-154.
(29) Ashworth, D. M.; Atrash, B.; Baker, G. R.; Baxter, A. J .; J enkins,
P. D. 4-Cyanothiazolidides as very potent, stable inhibitors of
Dipeptidyl Peptidase IV. Bioorg. Med. Chem. Lett. 1996, 6, (22),
2745-2748.
(30) Villhauer, E. B.; Brinkman, J . A.; Naderi, G. B.; Dunning, B.
E.; Mangold, B. L.; Mone M. D.; Russell M. E.; Weldon, S. C.;
Hughes T. E. 1-[2-[(5-Cyanopyridin-2-yl)amino]ethylamino]-
Intact GLP-1 Levels and Reduces Glucose Exposure in Humans.
Diabetes 2000, 49, (1), A39. (b) Ahre´n, B.; Sinonsson, E.; Larsson,
H.; Landin-Oisson, M.; Torgeirsson, H.; Jansson, P. A.; Sandqvist,
M.; Båvenholm, P.; Efendic, S.; Eriksson, J . W.; Dickinson, S.;
Holmes, D. Inhibition of Dipeptidyl Peptidase IV improves
metabolic control over a 4 week study period in type 2 diabetes.
Diabetes Care 2002, 25 (5), 869-875.
(41) Reimer, M. K.; Holst, J . J .; Ahre´n, B. Long-term inhibition of
dipeptidyl peptidase IV improves glucose tolerance and preserves
islet function in mice. Eur. J . Endocrinol. 2002, 146, 717-727.
(42) Reisher, S. R.; Hughes, T. E.; Ordovas, J . M.; Schaefer, E. J .;
Feinstein, S. I. Increased expression of apolipoprotein genes
accompanies differentiation in the intestinal cell line Caco-2.
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5757-5761.
acetyl-2-(S)-pyrrolidine-carbonitrile:
a Potent, Selective and
(43) Kubota, T.; Flentke, G. R.; Bachovchin, W. W.; Stollar, B. D.
Involvement of dipeptidyl peptidase IV in an in vivo immune
response. Clin. Exp. Immunol. 1992, 89, 192-197.
Orally Bioavailable Dipeptidyl Peptidase IV Inhibitor with
Antihyperglycemic Properties, J . Med. Chem. 2002, 45, 2362-
2368.
(44) Macnair, R. D.; Kenny, A. J . Proteins of the Kidney Microvillar
Membrane: The amphipathic Form of Dipeptidyl Peptidase IV.
Biochem. J . 1979, 179, 379-395.
(31) Villhauer, E. B. N-(Substituted glycly)-2-cyanopyrrolidides,
Pharmaceutical compositions containing them and their use in
inhibiting Dipeptidyl Peptidase-IV. US 6,011,155, J anuary 4th,
2000.
(32) Villhauer, E. B. N-(Substituted glycly)-2-cyanopyrrolidides,
Pharmaceutical compositions containing them and their use in
inhibiting Dipeptidyl Peptidase-IV. US 6,166,063, December
26th, 2000.
(33) Villhauer, E. B. N-(Substituted glycly)-2-cyanopyrrolidides,
Pharmaceutical compositions containing them and their use in
inhibiting Dipeptidyl Peptidase-IV. US 6,432,969, August 13th,
2002.
(34) ORTEP drawings and atomic coordinate information of NVP-
DPP728 (8c) and NVP-LAF237 (12j) atomic coordinate informa-
tion can be obtained from Supporting Information of ref 30 and
this paper, respectively.
(45) Mantle, D. Characterization of dipeptidyl and tripeptidyl ami-
nopeptidases in human kidney soluble fraction. Clin. Chim. Acta
1991, 196, 135-142.
(46) DeMeester, I.; Vanhoof, G.; Lambeir, A. M.; Scharpe, S. Use of
immobilized adenosine deaminase for the rapid purification of
native human CD26/dipeptidyl peptidase IV. J . Immunol.
Methods 1996, 189, 99-105.
(47) Nagatsu, T.; Sakai, T.; Kojima, K.; Araki, E.; Sakakibara, S.;
Fukasawa, K.; Harada, M. A sensitive and specific assay for
dipeptidyl-aminopeptidase II in serum and tissues by liquid
chromatography-fluorometry. Anal. Biochem. 1985, 147, 80-85.
(48) Momand, J .; Clarke, S. Rapid degradation of D- and L-succin-
imide-containing peptides by a post-proline endopeptidase from
human erythrocytes. Biochemistry 1987, 26, 7798-7805.
(49) Deacon, C. F.; Hughes, T. E.; Holst, J . J . Dipeptidyl peptidase
IV inhibition potentiates the insulinotropic effect of glucagon-
like peptide 1 in the anesthetized pig. Diabetes 1998, 47, (5),
764-769.
(35) Described in Experimental Section.
(36) Hughes, T. E.; Mone, M. D.; Russell, M. E.; Weldon, S. C.;
Villhuaer, E. B. NVP-DPP728 (1-[[[2-[(5-cyanopyridin-2-yl)-
amino]ethyl]amino]acetyl]-2-cyano-(S)-pyrrolidine), a slow-bind-
ing inhibitor of Dipeptidyl Peptidase IV. Biochemistry 1999, 38,
11597-11603.
J M030091L