cis-2,5-dicyanopyrrolidine inhibitors of dipeptidyl peptidase IV: Synthesis and in vitro, in vivo, and x-ray crystallographic characterization

Article abstract of DOI:10.1021/jm0600085

Inhibitors of the glucagon-like peptide-1 (GLP-1) degrading enzyme dipeptidyl peptidase IV (DPP-IV) have been shown to be effective treatments for type 2 diabetes in animal models and in human subjects. A novel series of cis-2,5-dicyanopyrrolidine α-amino amides were synthesized and evaluated as inhibitors of dipeptidyl peptidase IV (DPP-IV) for the treatment of type 2 diabetes. 1-({[1-(Hydroxymethyl)cyclopentyl]amino}-acetyl)pyrrolidine-2,5-cis- dicarbonitrile (1c) is an achiral, slow-binding (time-dependent) inhibitor of DPP-IV that is selective for DPP-IV over other DPP isozymes and proline specific serine proteases, and which has oral bioavailability in preclinical species and in vivo efficacy in animal models. The mode of binding of the cis-2,5-dicyanopyrrolidine moiety was determined by X-ray crystallography. The hydrochloride salt of 1c was further profiled for development as a potential new treatment for type 2 diabetes.

Full text of DOI:10.1021/jm0600085

3068
J. Med. Chem. 2006, 49, 3068-3076
Articles
cis-2,5-Dicyanopyrrolidine Inhibitors of Dipeptidyl Peptidase IV: Synthesis and in Vitro,
in Vivo, and X-ray Crystallographic Characterization
Stephen W. Wright,* Mark J. Ammirati, Kim M. Andrews, Anne M. Brodeur, Dennis E. Danley, Shawn D. Doran,
Jay S. Lillquist, Lester D. McClure, R. Kirk McPherson, Stephen J. Orena, Janice C. Parker, Jana Polivkova, Xiayang Qiu,
Walter C. Soeller, Carolyn B. Soglia, Judith L. Treadway, Maria A. VanVolkenburg, Hong Wang, Donald C. Wilder, and
Thanh V. Olson
Pfizer Global Research and DeVelopment, Eastern Point Road, Groton, Connecticut 06340
ReceiVed January 4, 2006
Inhibitors of the glucagon-like peptide-1 (GLP-1) degrading enzyme dipeptidyl peptidase IV (DPP-IV) have
been shown to be effective treatments for type 2 diabetes in animal models and in human subjects. A novel
series of cis-2,5-dicyanopyrrolidine R-amino amides were synthesized and evaluated as inhibitors of dipeptidyl
peptidase IV (DPP-IV) for the treatment of type 2 diabetes. 1-({[1-(Hydroxymethyl)cyclopentyl]amino}-
acetyl)pyrrolidine-2,5-cis-dicarbonitrile (1c) is an achiral, slow-binding (time-dependent) inhibitor of DPP-
IV that is selective for DPP-IV over other DPP isozymes and proline specific serine proteases, and which
has oral bioavailability in preclinical species and in vivo efficacy in animal models. The mode of binding
of the cis-2,5-dicyanopyrrolidine moiety was determined by X-ray crystallography. The hydrochloride salt
of 1c was further profiled for development as a potential new treatment for type 2 diabetes.
Scheme 1. Preparation of cis-2,5-Dicyanopyrrolidine Inhibitorsa
Type 2 diabetes (formerly non-insulin-dependent diabetes
mellitus) is a severe and increasingly prevalent disease.1
Diabetics may suffer debilitating cardiovascular, eye, kidney,
and nerve damage and are at risk of premature handicap and
death due to these and other diabetic complications, which are
the result of glucose toxicity caused by their hyperglycemia. A
progressive reduction in insulin sensitivity and insulin secretion
are hallmarks of the disease, which eventually result in failure
of the pancreatic islet cells and dependence on exogenous
insulin. The incretin hormone glucagon-like peptide 1 (GLP-1)
is a potent stimulator of endogenous insulin release. GLP-1 has
beneficial effects on islet â-cell function and insulin sensitivity
without induction of hypoglycemia.2 Unfortunately, GLP-1 is
rapidly degraded in vivo. Inhibition of GLP-1 degradation by
dipeptidyl peptidase IV (DPP-IV) has emerged as a promising
approach for the treatment of Type 2 diabetes3 and has been
aggressively pursued by numerous laboratories.4 We sought to
identify novel inhibitors of DPP-IV that were structurally distinct
from the wide variety of inhibitors previously reported by other
groups. In particular, we wanted to explore an achiral cis-2,5-
dicyanopyrrolidine template, to learn if such compounds were
active as inhibitors of DPP-IV. Molecular modeling based on
published DPP-IV enzyme-inhibitor crystal structures5 sug-
gested that the cis-2,5-dicyanopyrrolidine template was not a
good starting point for compound design. However, it is well-
known that binding site residues can move to accommodate a
ligand, and 4,5-methano-bridged inhibitors of DPP-IV have been
reported.6 Therefore, we opted to test the idea that a cis-2,5-
dicyanopyrrolidine template might inhibit DPP-IV. This paper
reports the results of our investigation.
a Conditions: (a) 1-(Hydroxymethyl)cyclopentylamine (3 equiv), MeCN,
20 °C, 18 h.
Scheme 2. Preparation of Bromoacetamide 2a
a Conditions: (a) 2,5-Dimethoxytetrahydrofuran, KCN, 0.1 M aqueous
citric acid, 20 °C, 72 h; (b) BrCH2COBr, MeCN, 70 °C, 16 h.
the preparation of 1c is illustrated for example. Coupling of
the bromoacetamide 2 with an excess (g3 equiv) of the
appropriate primary amine was carried out in acetonitrile at 20
°C. On the basis of previously reported work detailing the
stability of nitrile-containing inhibitors of DPP-IV, we elected
to employ only amines bearing a quaternary R-carbon center.7
The products were converted to their hydrochloride salts by
treatment with 1 equiv of hydrogen chloride in methanol
following purification by silica gel chromatography.
The requisite bromoacetamide 2 was prepared by the route
shown in Scheme 2. Strecker reaction of 2,5-dimethoxytetrahy-
drofuran and aminodiphenylmethane with KCN in 0.1 M
aqueous citric acid solution afforded a precipitate containing
Synthesis. The compounds were prepared by parallel syn-
thesis according to the procedure outlined in Scheme 1, in which
* To whom correspondence should be addressed. Telephone: 860-441-
5831; fax: 860-715-4483; email: stephen.w.wright@pfizer.com.
10.1021/jm0600085 CCC: $33.50 © 2006 American Chemical Society
Published on Web 05/02/2006

Dicyanopyrrolidine Inhibitors of Dipeptidyl Peptidase
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11 3069
Scheme 3. Alternative Preparation of Bromoacetamide 2a
Table 1. Effect of Amine Substituent on DPP-IV Enzyme Inhibitory
Activity
entry
R
DPP-IV IC50, nMa
1a
1b
1c
1d
1e
1f
3-noradamantyl
72
74
104
106
164
1-(3-HO)adamantylb
1-(HOCH2)cyclopentyl
1-adamantyl
1-(C6H5)-1-cyclobutyl
t-BuCH2(CH3)2C
208
1g
1h
1i
2-exo-norbornyl
C6H5(CH3)2C
(4-FC6H4)CH2(CH3)2C
tert-butyl
237
261
651
663
1j
1k
1l
1-(HOCH2)cyclohexyl
1-methylcyclohexylc
HOCH2(CH3)2C
687
689
1m
1n
1o
1p
1q
1r
1s
1110
1160
1220
>3000
>3000
>3000
> 3000
> 3000
> 3000
1-(C6H5CH2)cyclobutyl
1-(HOCH2)-4-THPd
1-(ethynyl)cyclohexyl
1-(ethyl)cyclohexyle
3,5,7-trifluoro-1-adamantylf
(CH3)2CH(CH3)2Cc
CH3(C2H5)2Cc
a Conditions: (a) Diethyl meso-dibromoadipate, toluene, 80 °C, 16 h;
(b) H2 (50 psi), Pd/C, MeOH, 20 °C, 18 h; (c) NH3, MeOH, 20 °C, 7 d; (d)
BOC2O, H2O, 1,4-dioxane, 20 °C, 18 h; (e) POCl3, imidazole, C5H5N,
CH2Cl2, 0 °C f 20 °C, 90 m; (f) BrCH2COBr, MeCN, 20 °C, 1 h.
1t
1u
CH3CH2(CH3)2Cc
the desired cis-dinitrile 3, along with N-benzhydrylpyrrole (5)
and lesser amounts of the corresponding trans-dinitrile isomer
(4).8 Purification of the desired cis-dinitrile 3 was accomplished
by recrystallization from acetonitrile. Direct conversion of 3 to
the bromoacetamide 2 was accomplished by reaction of 3 with
bromoacetyl bromide in acetonitrile at 70 °C, followed by
chromatographic purification to remove the byproduct benzhy-
dryl bromide.9
a Recombinant wild-type human enzyme. Values are means of at least
three experiments; standard deviations are (15%. An IC50 of >3000
indicates that no curve was noted in the dose-response up to 3 µM. b See
ref 14. c Prepared from the corresponding tertiary alcohol with TMSN3 and
BF3 etherate; see Experimental Section. d 4-Tetrahydropyranyl. e Prepared
by hydrogenation of 1-ethynylcyclohexylamine. f See ref 15.
Table 2. Effect of Amine Substituent on DPP-IV Enzyme Inhibitory
Activity
Alternatively, bromoacetamide 2 could be prepared from
diethyl meso-dibromoadipate10 as shown in Scheme 3.11 Reac-
tion with excess benzylamine afforded predominantly the cis-
isomer of the diethyl ester 6. Crystallization of the hydrochloride
salt afforded pure 6, which was then converted to the free base
and subjected to hydrogenolysis to afford 7. Treatment with
excess ammonia afforded the crystalline diamide 8, which was
converted to the N-BOC derivative 9. Dehydration with phos-
phoryl chloride gave the BOC-protected dinitrile 10, which upon
exposure to bromoacetyl bromide gave the bromoacetamide 2.
The amines that were coupled with bromoacetamide 2 were,
in general, commercially available amines (Table 1). In certain
cases, the desired amine fragment was not commercially
available (1l, 1z). These amines were prepared from the
corresponding tertiary alcohols via the intermediate azide by
treatment with azidotrimethylsilane and boron trifluoride ether-
ate12 followed by hydrogenation over palladium on carbon.
1-(Alkoxymethyl)-1-cyclopentylamines (Table 2) were prepared
from commercially available 1-(hydroxymethyl)-1-cyclopentyl-
amine by a three-step sequence (Scheme 4).
Evaluation of Enzyme Inhibition. Recombinant human
DPP-IV activity was measured by the release of 4-methoxy-2-
aminonaphthylamine from the artificial substrate Gly-Pro-4-
methoxy-â-naphthylamide.13 We were gratified to find that the
cis-dinitrile pyrrolidine scaffold did indeed afford compounds
with DPP-IV inhibitory activity. Screening a variety of primary
amines bearing a tertiary carbon center at the R position revealed
that the best potency was generally attained with those amines
in which the R-carbon was part of a cyclic or polycyclic
framework (Table 1). In particular, the adamantane (1b, 1d),
noradamantane (1a), norbornane (1g), and cyclopentane (1c)
entry
R
DPP-IV IC50, nMa
1c
1v
1w
1x
1y
HOCH2
104
64
84
91
123
201
454
702
955
b
CH3CH2OCH2
b
CH3OCH2
b
CH3CH2CH2CH2OCH2
CH3CH2CH2OCH2
b
c
1z
CH3
b
1aa
1bb
1cc
1dd
1ee
1ff
1gg
(CH3)2CHCH2OCH2
d,b
C6H11OCH2
b
CH3OCH2CH2
(CH3)3COCH2
b
1860
>3000
>3000
>3000
1-CONHCH3
1-CO2CH3
b
HOCH2CH2
a Recombinant wild-type human enzyme. Values are means of at least
three experiments; standard deviations are (15%. An IC50 of >3000
indicates that no curve was noted in the dose response up to 3 µM.
b Prepared from N-Cbz1-(hydroxymethyl)-1-cyclopentylamine; see Ex-
perimental Section. c Prepared from the corresponding tertiary alcohol with
TMSN3 and BF3 etherate; see Experimental Section. d (Cyclcohexyloxy)-
methyl.
fragments afforded compounds with acceptable enzyme inhibi-
tory potency. Acyclic hydrocarbon fragments were generally
less potent (for example, 1s, 1t, 1u). Also noted was an apparent
preference for a cyclopentane ring as opposed to a cyclohexane
ring (1c vs 1k).

3070 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11
Wright et al.
Scheme 4. Preparation of
1-(Alkoxymethyl)-1-cyclopentylaminesa
Figure 3. Slow binding kinetics of DPP-IV enzyme inhibition by 1c.
a Conditions: (a) Cbz-Cl, CH2Cl2, 0 °C, 1 h; (b) R-Br, Ag2O, DMF, 20
°C, 18 h; (c) H2, Pd-C, MeOH, 20 °C, 18 h.
Figure 1. Inactive trans analogues.
Figure 4. Slow binding kinetics of DPP-IV enzyme inhibition by 1c.
by incubation of enzyme for 100 min in the presence or absence
of 300 nM 1c, followed by 100-fold dilution into a reaction
mixture containing 1 mM H-Ala-Pro-4-nitroanilide.
Compounds exhibiting DPP-IV IC50 e 120 nM were subse-
quently screened for their ability to inhibit other enzymes with
DPP-IV-like activity: DPP-II, DPP-III, DPP-VIII, DPP-IX,
FAP, and APP.18 No compounds had any significant inhibitory
activity against any of these enzymes (all IC50 values were .
30 µM).
Figure 2. Inactive cis analogues.
Based on the synthetic flexibility of the cyclopentane
framework, and upon the results of preliminary in vivo studies
with 1b and 1c (vide infra), further analogue generation centered
upon targeted analogues incorporating a cyclopentyl fragment
in the amine residue (Table 2).
Compounds with a substituent containing the CH2O subunit
(1v, 1w, 1x) were clearly preferred over other functional groups
(for example, 1ee and 1ff). Steric encumbrance of this subunit
decreased potency (1aa, 1dd), suggesting that perhaps this
subunit was also involved in a binding interaction. The
remarkable loss of potency observed when the CH2O subunit
was extended to a CH2CH2O subunit as in 1cc and 1gg further
supported this conclusion. In addition, the potency preference
for a cyclopentane ring over a cyclohexane ring was again
observed (for example, 1z vs 1l).
The trans-dinitrile isomers of 1c and 1d (14a, 14b) were
prepared from 4 in the same manner as 1c and 1d and found to
have DPP-IV IC50 . 30 µM (Figure 1). The cis-5-methyl
analogue 15,16a the cis-5-phenyl analogue 16,16b and the cis-4-
hydroxy analogue 1716c were also inactive (IC50 . 30 µM,
Figure 2). The kinetics of DPP-IV inhibition by 1c were studied
at various concentrations and revealed that the binding of 1c to
DPP-IV was time dependent (Figure 3) and reversible (Figure
4). No evidence was found to indicate that the covalently bound
imidate (vide infra) underwent hydrolysis to the corresponding
carboxylic acid. Activity was assayed by spectrophotometric
measurement (405 nm) of the liberation of 4-nitroaniline from
H-Ala-Pro-4-nitroanilide.17 Time dependent binding of 1c to
DPP-IV was assessed by the addition of substrate (100 µM)
and inhibitor (at the concentrations indicated) to the enzyme at
t ) 0. Dissociation of the DPP-IV-1c complex was assayed
In Vivo Hypoglycemic Evaluation. Compounds exhibiting
DPP-IV IC50 e 100 nM were subsequently dosed orally at an
initial dose of 10 mg kg-1 in fasted, diabetic KK/H1J mice,
and the response of the mice to a subsequent oral glucose
challenge was measured. [1-(S)-1-cyclohexyl-2-oxo-(3,3,4,4,-
tetrafluoropyrrolidin-1-yl)ethyl]amine (18, DPP-IV IC50 ) 110
nM) was used as a positive control in these experiments.19
This was performed to demonstrate that the compounds could
effect glucose lowering in vivo by DPP-IV inhibition, as
subsequent pharmacokinetic evaluation of the compounds did
not examine in vivo hypoglycemic activity per se. The initial
dose of 10 mg kg-1 was selected in an attempt to compensate
for the unknown pharmacokinetics of the test compound in the
KK/H1J mice. Selected compounds were subsequently tested
at lower doses. Due to interassay variability, the compounds
were scored simply as exhibiting significant or nonsignificant
inhibition of glucose excursion in the oral glucose tolerance
test (Table 3).
Pharmacokinetic Evaluation. Certain compounds that were
active in the KK/H1J mouse oral glucose tolerance test were
advanced to study of their pharmacokinetic properties in the
Sprague-Dawley rat. Oral pharmacokinetic evaluations were
conducted using the compound delivered in 0.5% methyl-
cellulose formulation at 5 mg kg-1 (Table 4). Intravenous
pharmacokinetic evaluations were conducted using the com-
pound in sterile saline solutions at 1 mg kg-1
.
On the basis of its pharmacokinetic properties in the rat,
compound 1c was advanced to further pharmacokinetic profiling
in the beagle dog and cynomolgus monkey (Table 5). For these

Dicyanopyrrolidine Inhibitors of Dipeptidyl Peptidase
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11 3071
Table 3. In Vivo Hypoglycemic Activity of DPP-IV Inhibitors
Figure 5. [1-(S)-1-Cyclohexyl-2-oxo-(3,3,4,4,-tetrafluoropyrrolidin-
1-yl)ethyl]amine.
DPP-IV
in vivo glucose
lowering
%
entry
R
IC50, nMa
loweredb
1v
1a
1b
1w
1x
1c
1c
1c
1d
18
CH3CH2OCH2
3-noradamantyl
1-(3-HO)adamantyl
CH3OCH2
CH3CH2CH2CH2OCH2
1-(HOCH2)cyclopentyl
1-(HOCH2)cyclopentyl
1-(HOCH2)cyclopentyl
1-adamantyl
64
72
74
84
91
104
104
104
106
110
significant
significantc
significantd
significant
not significant
significant
significante
significantf
significantd
significant
79
62
45
74
16
67
59
46
55
57
a Recombinant wild-type human enzyme. Values are means of at least
three experiments; standard deviations are (15%. b Data are means of 10
animals per group; standard deviations are e15%. See Experimental
Sectionfor details. c Inhibitor dosed at 5 mg kg-1
.
d Inhibitor dosed at 3
mg kg-1
.
e Inhibitor dosed at 2.5 mg kg-1
.
f Inhibitor dosed at 0.3 mg kg-1
.
Figure 6. Plasma DPP-IV ativity and GLP-1 levels in KK mice dosed
with 1c.
Table 4. In Vivo Pharmacokinetic Parameters of DPP-IV Inhibitors in
the Rat
DPP-IV
CLh,
mL/min/kg
Vss,
L/kg
t1/2
h (iv)
,
t1/2
h (po)
,
F,
%
compd
IC50, nMa
1b
1w
1c
74
84
104
106
130
98
70
2.7
1.6
1.4
3.1
0.4
0.2
0.3
0.6
1.4
42
48
77
nd
0.64
2.7
1d
80
ndb
a Recombinant wild-type human enzyme. Values are means of at least
three experiments; standard deviations are (15%. b Not determined.
Table 5. In Vivo Pharmacokinetic Parameters of 1c
DPP-IV
unbound
fraction
CLh,
Vss,
t1/2
,
t1/2
,
F,
%
species IC50, nMa
mL/min/kg L/kg h (iv) h (po)
rat
dog
monkey
human
38
109
92
100
100
84
70
22
42
15b
1.4
1.3
1.7
1.2b
0.3
0.8
0.6
2.7
5.3
2.4
2.0c
77
93
42
70b
69
77
a Plasma enzyme assay. Values are means of at least three experiments;
standard deviations are (20%. b Projected values based on allometric
scaling. c Projected values based on combination of clearance and volume
predictions.
Figure 7. Plasma DPP-IV activity in beagles dosed with 1c.
administration of 1c resulted in increased plasma levels of GLP-
1. An initial experiment was conducted in fed KK mice, which
were dosed with 5 mg kg-1 1c po at t ) 0 min. Blood samples
were collected at intervals over 2 h for the measurement of
GLP-1 and DPP-IV activity (Figure 6). Following dosing of
1c, total plasma DPP-IV activity was reduced by approximately
50%. Plasma GLP-1 increased to levels approximately three
times those recorded at t ) 0 min.
A second experiment was conducted in beagle dogs, using a
procedure adapted from one previously reported for demonstra-
tion of the effects of DPP-IV inhibition on plasma GLP-1
levels.20 The dogs were fasted for 18 h, after which the t ) 0
blood sample was taken and they were dosed with water or 0.5
mg kg-1 1c in water. At t ) 30 min, a meal comprising 9 mL
kg-1 of a 1:1 v/v mixture of dairy sour cream and 0.5 M sucrose
was administered.21 Additional blood samples were taken for
analysis at the times indicated (Figures 7 and 8). As anticipated,
plasma DPP-IV enzyme activity was reduced in the treatment
group to less than 50% of plasma DPP-IV activity in the vehicle
control group. Likewise, plasma GLP-1 levels were increased
by two to three times in the treatment group relative to the
vehicle control group. Analysis of the area under the curve for
the GLP-1 response was analyzed and indicated a significant
elevation of GLP-1 in the group treated with 1c (Figure 9).
experiments, oral pharmacokinetic evaluations were conducted
using the compound delivered in sterile water at 1 mg kg-1
.
Intravenous pharmacokinetic evaluations were conducted at 1
mg kg-1 using sterile saline as the vehicle. Simultaneously, 1c
was evaluated for its ability to inhibit rat, dog, monkey, and
human DPP-IV using enzyme derived from plasma. Similar
potency against all four isoforms was observed. Since in vitro
metabolism of 1c was not detected, human clearance was
predicted using allometric scaling of rat, dog, and monkey
clearance data with and without protein binding consideration.
The compound exhibited good bioavailability in all species,
a low volume of distribution, and moderate clearance. Rapid
absorption was also noted in all species as indicated by Tmax
times that were 1 h or less. To assess whether nonlinear exposure
might be observed with 1c, this compound was dosed in rats at
higher amounts. We were pleased to note dose proportional
increases in exposure (Cmax and AUC0-∞) following oral doses
of 5, 15, and 50 mg kg-1 in a 0.5% methylcellulose vehicle.
Effects on Endogenous GLP-1 Levels. Having established
that 1c resulted in improved glucose tolerance in vivo following
an oral glucose challenge, we sought to demonstrate that

3072 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11
Wright et al.
Figure 10. Electron density map of 1c bound at rHDPP-IV active site
at 1.6 σ. Yellow is the binding mode of 1c; alternative binding mode
in green would also fit the electron density, but makes less sense in
light of in vitro SAR findings. The CH2OH group of S630 is shown in
red.
Figure 8. Plasma GLP-1 levels in beagles dosed with 1c.
Figure 9. GLP-1 AUC in beagles dosed with 1c.
Further Profiling. On the basis of the results obtained with
1c, this compound was further profiled to determine its
suitability for clinical development. The mono-hydrochloride
salt of 1c was found to be an anhydrous crystal form that melted
at ∼173 °C that was nonhygroscopic under kinetic measure-
ments. In the biological model system of phosphate buffered
saline (PBS, pH ) 6.5) the solubility of the mono-hydrochloride
salt form was determined to be >76 mg mL-1. The solubility
of the mono-hydrochloride salt form in unbuffered water was
difficult to determine precisely, but was in the range of 200 to
300 mg mL-1, with final solution pH ) 2.0.
Experiments were also conducted in order to address the
potential concern that metabolism of the dicyanopyrrolidine ring
moiety in 1c could liberate cyanide ion by a reverse Strecker
reaction. First, metabolite identification studies were completed
for both in vitro (microsomes and hepatocytes; rat, dog, monkey,
human) and in vivo (plasma and urine; rat, dog, monkey)
samples. Products corresponding to the loss of the cyano group
were not detected in any of the samples. Next, to further confirm
this finding, in vivo plasma samples (rat, dog, monkey) were
analyzed for thiocyanate ion, as cyanide ion is detoxified in
vivo by conversion to thiocyanate.22 Thiocyanate is found in
plasma at basal levels ranging from 10 to 100 µM. The basal
level of thiocyanate ion in human plasma has been reported to
be around 30 µM.23 Previously collected plasma samples were
therefore analyzed for thiocyanate ion. In rats (15 mg kg-1, po),
dogs (3 mg kg-1 po), and monkeys (3 mg kg-1 po), no increases
in thiocyanate were detected through 24 h when compared to
the T ) 0 sample, providing further support for the conclusion
cyanide ion was not liberated from 1c following in vivo
administration. Further confidence in the projected chronic safety
of 1c was gained by a preclinical 5-day toxicity study in rats.
Male and female Sprague - Dawley rats (3 of each sex) were
dosed with 500 mg/kg of 1c for 5 days. Clinical signs, clinical
chemistry, hematology, and postmortem findings (macroscopic
and microscopic) were all unremarkable. A single-dose safety
Figure 11. X-ray crystal structure of 1c bound to rHDPP-IV showing
(clockwise from top) side chains of Y547, Y631, E205, E206, N710,
H740, S630, and Y666. R125, ordered water molecules, and the
hydrogen bond network present between the ordered water molecules,
compound 1c, and protein atoms have been omitted for clarity.
study in the dog in which 1c was administerd at 20 mg kg-1 iv
likewise produced no remarkable findings.24
X-ray Crystallography. To better understand the binding
mode of our compounds, we cocrystallized recombinant human
DPP-IV with 1c (Figures 10 and 11). The DPP-IV active site is
defined by a catalytic triad (S630, H740, D708), an oxyanion
hole (Y547 OH, Y631 N), and specificity residues (E205, E206,
Y662, Y666, N710). One nitrile group is attacked by S630 to
form a covalent link in the crystal structure. The other nitrile
fits into the active site by forcing the Y547 side chain to move
by ∼2.8 Å. This nitrile nitrogen is likely hydrogen bonded to
the Y666 OH (3.1 Å) and is 3.5 Å from Y547 OH. The π cloud
of the triple bond stacks with the edge of Y666. The pyrrolidine
ring occupies the S1 pocket (Y662, Y666, V656, V711, Y631,
W659) and compares well with that in previous structures
(shifted 0.4 Å, coordinates error 0.28 Å). The P2 backbone is
recognized by E205 and E206 (3.0 and 2.9 Å to the secondary
amine, respectively), N710 (2.7 Å to carbonyl), and Y662 (2.9
Å to carbonyl). The cyclopentane ring is in the large S2 site,
but surprisingly does not appear make any hydrophobic interac-
tions with the protein (all protein atoms < 4.0 Å to the
cyclopentane ring are oxygen). The alcohol makes a hydrogen
bond with the R125 side chain directly (3.1 Å) and via H2-
O128 (OH to H2O 2.6 Å, H2O to R125 2.7 Å). This hydrogen
bonding would appear to be the basis for the preference observed
in vitro for the CH2O subunit (1c, 1v, 1w, 1x) over the CH2-
CH2O subunit (1cc and 1gg) or other functional groups (1ee,
1ff). This hydrogen bond would also be consistent with the

Dicyanopyrrolidine Inhibitors of Dipeptidyl Peptidase
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11 3073
1-Bromoacetyl-pyrrolidine-2,5-cis-dicarbonitrile (2 from 3).
To a solution of 3.00 g (10.4 mmol) of 3 in 40 mL of MeCN was
added 1.1 mL (12 mmol) of bromoacetyl bromide with stirring at
70 °C. The mixture was stirred at 70 °C for 16 h. The mixture was
concentrated, and the residue was flash chromatographed on silica
gel eluting with 1:1 hexanes-EtOAc to afford 2.47 g (99%) of 2
as an oil that crystallized upon storage below 0 °C. 1H NMR
(CDCl3): 4.94 (m, 1 H); 4.68 (m, 1 H); 4.01 (d, J ) 13.2 Hz, 1
H); 3.96 (d, J ) 13.2 Hz, 1 H); 2.52 (m, 1 H); 2.40 (m, 3 H). 13C
NMR (CDCl3): 165.1, 116.8, 47. 6, 47.5, 31.6, 28.9, 25.9. APCI
MS: 242, 244 (MH+, Br isotope pattern). mp 49-50 °C. Anal.
(C8H8BrN3O) C, H, N.
previous observation that steric encumbrance of this subunit
resulted in decreased potency (1aa, 1dd).
Comparison of this crystal structure with published data
shows that with one nitrile group on the pyrrolidine ring, the
covalent modification of S630 would position the intermediate
imidate nitrogen perfectly in the oxyanion hole, forming
hydrogen bonds with Y547 OH and Y631 N. Binding of the
cis-dinitrile 1c likewise leads to imidate ester formation;
however, the interaction of the imidate nitrogen with the
oxyanion hole is no longer possible. Movement of Y547,
however, results in favorable binding of the remaining nitrile
group with Y666 via π-stacking and hydrogen bonds.
Further structural work was conducted by NMR using 1c
bearing 13C labels (>98%) on both nitrile carbons.25 The 13C
NMR of (13CN)2 1c in 25 mM Tris buffer at pH 7.5 showed
two signals at approximately 118 and 119 ppm. Upon addition
of recombinant human DPP-IV to the sample, the resonance at
118 ppm was replaced by a resonance at approximately 177
ppm, indicating that binding to the enzyme was accompanied
by imidate formation.
1-Benzylpyrrolidine-2,5-cis-dicarboxylic Acid Diethyl Ester
Hydrochloride (6). A solution of 253 g (700 mmol) of diethyl
meso-2,5-dibromoadipate in 1000 mL of toluene was stirred at 80
°C while 253 mL (2300 mmol) of benzylamine was added over
the course of 1 h. Upon completion of the addition of benzylamine,
the solution was kept at 80 °C with stirring overnight. The mixture
was cooled to room temperature and the precipitate was filtered
and washed twice with toluene. The washings were combined with
the filtrate and washed water, then twice with pH ) 7 5% potassium
phosphate solution, brine, and dried (MgSO4). Filtration and
evaporation afforded an amber oil that was dissolved in 900 mL
of ethyl acetate and treated with 150 mL of 4 M HCl in 1,4-dioxane
at 20 °C. The mixture was cooled with stirring in ice for 20 m.
The resulting white precipitate was filtered, washed with ethyl
acetate, and dried under dynamic vacuum to afford 164 g (70%)
Summary
1-({[1-(Hydroxymethyl)cyclopentyl]amino}acetyl)pyrrolidine-
2,5-cis-dicarbonitrile (1c) is an achiral inhibitor of DPP-IV that
selectively inhibits DPP-IV over DPP-II, DPP-III, DPP-VIII,
DPP-IX, APP, and FAP. It exhibited oral bioavailability in the
rat, dog, and monkey and displayed in vivo efficacy in a mouse
OGTT model. It has a high free fraction, low volume of
distribution, moderate clearance, and moderate half-life. The
mode of binding of the cis-2,5-dicyanopyrrolidine moiety was
determined by X-ray crystallography and determined to involve
a covalent interaction of S630 with one nitrile group, while the
other nitrile forces the Y547 side chain to move and subse-
quently makes a π-stacking interaction and H-bond with Y666.
The secondary amine is recognized by E205, E206, N710, and
Y662. The covalent nature of the interaction with S630 was
indicated by 13C NMR. The hydrochloride salt of 1c was
subjected to further profiling for development as a potential new
treatment for type 2 diabetes. Preliminary safety assessments
with 1c indicated no safety liability with this new class of
compounds.
1
of the hydrochloride salt of 6. H NMR (D2O): 7.34 (m, 5 H);
4.50 (s, 2 H); 4.46 (m, 2 H); 3.99 (q, 4 H); 2.43 (m, 2 H); 2.09 (m,
2 H); 1.04 (t, 6 H). 13C NMR (D2O): 168.9, 131.2, 131.0, 130.8,
129.7, 67.9, 64.7, 60.2, 27.5, 13.3. APCI MS: 306 (MH+). mp
120-122 °C. Anal. (C17H23NO4 HCl H2O) C, H, N.
Pyrrolidine-2,5-cis-dicarboxylic Acid Diethyl Ester (7). A
solution of 164 g (480 mmol) of 6 in 250 mL of CH2Cl2 was stirred
vigorously with 250 mL of 5 M NH4OH for 15 m at 20 °C. The
CH2Cl2 solution was separated, washed with brine, dried (Na2SO4),
and concentrated to give a colorless oil. The oil was dissolved in
500 mL of MeOH, 4.8 g of 10% Pd/C was added, and the mixture
was shaken under 50 psi of H2 at room-temperature overnight. The
mixture was filtered through Celite, and the solvent was evaporated
1
to provide 98 g (95%) of 7 as an oil. H NMR (D2O): 4.44 (m,
2H); 4.13 (q, 4H); 2.31 (m, 2H); 2.09 (m, 2H); 1.12 (t, 6H). 13C
NMR (D2O): 169.3, 64.3, 60.4, 27.4, 13.3. APCI MS: 216 (MH+).
Anal. (C10H17NO4 HCl) C, H, N.
Pyrrolidine-2,5-cis-dicarboxylic Acid Diamide (8). 7 (96.4 g,
448 mmol) was dissolved in 1300 mL of 7 M NH3 in MeOH. The
reaction vessel was closed tightly, and the mixture was allowed to
stand at 20 °C for about 7 days until the reaction was complete.
The crystals that formed were filtered, washed with ethanol, and
dried to afford 65.4 g (93%) of 8 as a white solid. 1H NMR (D2O):
3.69 (m, 2H); 2.03 (m, 2H); 1.67 (m, 2H). 13C NMR (D2O): 181.3,
60.8, 30.2. APCI MS: 158 (MH+). mp: 229-230 °C. Anal.
(C6H11N3O2) C, H, N.
Experimental Section
All procedures involving animals were reviewed and approved
by the Pfizer Institutional Animal Care and Use Committee.
1-Benzhydryl-pyrrolidine-2,5-cis-dicarbonitrile (3). To 900
mL of 0.1 M aqueous citric acid solution was added 8.24 g (45.0
mmol) of aminodiphenylmethane, followed by 4.00 g (61.4 mmol)
of KCN. The mixture was stirred at 20 °C until the reactants had
dissolved, after which 11.00 g (83.2 mmol) of 2,5-dimethoxytet-
rahydrofuran was added in one portion. The mixture was stirred at
20 °C for about 72 h, after which 14.9 g of precipitate was collected
by filtration, washed with water, and dried. The precipitate was
dissolved in a minimum of acetonitrile with heating and allowed
to cool, whereupon 3.90 g (30%) of 3 precipitated and was collected
2,5-cis-dicarbamoylpyrrolidine-1-carboxylic Acid tert-Butyl
Ester (9). A solution of 3.73 g (23.7 mmol) of 8 in 30 mL of H2O
was treated with a solution of 5.70 g (26.1 mmol) of di-tert-butyl
dicarbonate in 60 mL of 1,4-dioxane. The mixture was stirred at
20 °C overnight, after which the mixture was concentrated to
1
provide 5.98 g (100%) of 9 as a white solid. H NMR (CD3OD):
4.26 (m, 2 H); 2.30 (m, 2 H); 1.99 (m, 2 H); 1.42 (s, 9 H). 13C
NMR (CD3OD): 182.0, 181.8, 157.9, 85.1, 65.5, 65.2, 34.0, 33.6,
31.3. APCI MS: 258 (MH+). mp: 196-197 °C. Anal. (C11H19N3O4)
C, H, N.
1
by filtration. H NMR (CDCl3): 7.53 (m, 4 H); 7.30 (m, 4 H);
7.24 (m, 2 H); 5.11 (s, 1 H); 3.98 (m, 2 H); 2.35 (m, 4 H). 13C
NMR (CDCl3): 140.4, 129.4, 128.4, 127.9, 117.0, 69.1, 50.9, 28.9.
APCI MS: 167 (Ph2CH+ fragment). mp: 194-197 °C. Anal.
(C19H17N3) C, H, N.
2,5-cis-Dicyanopyrrolidine-1-carboxylic Acid tert-butyl Ester
(10). Imidazole (7.08 g, 104 mmol) and 9 (6.70 g, 26.0 mmol)
were dissolved in 65 mL of dry pyridine and 260 mL of CH2Cl2..
The solution was cooled to 0 °C, and 19 mL (200 mmol) of POCl3
was added dropwise. The mixture was stirred at 0 °C for 30 m and
at 20 °C for 1 h. Water (4 mL) in pyridine (12 mL) was added
dropwise with stirring and ice cooling to quench excess POCl3.
The mixture was then evaporated, and the residue was triturated
Evaporation of the mother liquor gave a residue that was flash
chromatographed on silica gel eluting with 4:1 hexanes-EtOAc
to provide 0.80 g (6%) of 1-benzhydryl-pyrrolidine-2,5-trans-
dicarbonitrile 4: 1H NMR (CDCl3): 7.53 (m, 4 H); 7.34 (m, 4 H);
7.25 (m, 2 H); 4.89 (s, 1 H); 3.86 (m, 2 H); 2.48 (m, 2 H); 2.28
(m, 2 H). APCI MS: 167 (Ph2CH+ fragment). mp: 187-188 °C.

3074 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11
Wright et al.
with dry THF and filtered. The filtrate was stirred for 10 min with
50 g of Amberlite IR-120 H+ ion-exchange resin and filtered
through a pad of anhydrous MgSO4. Evaporation and recrystalli-
zation of the residue from i-Pr2O afforded 4.24 g (74%) of 10 as
(t, 3 H). 13C NMR (CDCl3): 79.0, 73.4, 61.9, 37.8, 24.7, 23.0,
10.8. APCI MS: 158 (MH+). Anal. (C9H19NO): C, H, N.
DPP-IV Enzyme Inhibition Screening Assay. An amount of
150 µL of an enzyme-substrate solution containing 50 µM Gly-
Pro-4-methoxy-â-naphthylamide hydrochloride in 50 mM pH 7.3
Tris buffer containing 0.1 M sodium chloride, 0.1% (v/v) Triton,
and 50 µU/mL DPP-IV29 was pipetted into microtiter wells of a
polystyrene 96-well plate maintained at 4 °C. Five microliters per
well of test compound was added, bringing the final test compound
concentrations to 3 µM to 10 nM per well. Diprotin A28 (100 µM)
was used as a positive control. The entire assay is incubated 18 h
at 37 °C and then stopped by adding 10 µL of a solution containing
0.5 mg/mL Fast Blue B in a buffer comprising 0.1 M pH 4.2 sodium
acetate and 10% (v/v) Triton X-100 to each well, followed by
shaking for approximately 5 min at room temperature, after which
the absorption maximum at 525 nm was determined. IC50 deter-
minations were made from dose-response curves determined in
triplicate. The reported IC50 values are the mean of at least three
such determinations.
Plasma Enzyme Inhibition Assay. Aliquots (12.5 µL) of plasma
were loaded onto a 96-well plate, and 10 µL of buffer pH 7.5 100
mM HEPES containing 0.2 mg mL-1 of BSA was added to each
aliquot. The plate was briefly mixed on a titer plate shaker, sealed
with plate sealing tape, and then placed in an incubator at 37 °C
for 5 min. The reaction was initiated by the addition of 2.5 µL of
substrate solution (500 µM H-Gly-Pro-AMC in buffer) to the wells,
and the plate was shaken briefly. The plate was then read over a 5
min period in a Wallac Victor2 1420 Multilabel Counter, using an
excitation wavelength of 360 nm and recording the emission at
460 nm. IC50 determinations were made from dose-response curves
determined in triplicate. The reported IC50 values are the mean of
at least three such determinations.
KK/H1J Mouse Oral Glucose Tolerance Test. KK/H1J mice
were fasted overnight with free access to water. After fasting, (time
“t” ) 0), 25 mL of blood was drawn from the retro-orbital sinus
and added to 0.025% heparinized saline (100 mL) on ice. The mice
(10 per group) were then orally dosed with a solution of a test
compound in 0.5% methylcellulose (0.2 mL per mouse). Two
controls groups received only 0.5% methylcellulose. At t ) 15 min,
the mice were bled, as described above, and then dosed with 1 mg
kg-1 glucose in distilled water (0.2 mL per mouse). The first control
group was dosed with glucose. The second control group was dosed
with water. At t ) 45 min, the mice were again bled as described
above. The blood samples were centrifuged, and the plasma was
collected and analyzed for glucose content. Data were expressed
as percent inhibition of glucose excursion relative to the two control
groups.30
Rat PK Determination. Male Sprague-Dawley rats (200-250
g) implanted with jugular vein cannulas (JVC) were obtained from
Charles River Laboratories. Test compounds were administered as
either an intravenous bolus via the JVC or by oral gavage. The
intravenous dose was administered as a solution in sterile saline at
1.0 mg/kg and a dose volume of 1.0 mL/kg. The oral dose was
administered as a solution in 0.5% methycellulose at 5.0 mg/kg
and a dose volume of 10 mL/kg. Blood samples (0.25 mL) were
collected at multiple time points from 0 to 24 h and placed into
tubes containing lithium heparin. The blood samples were then
centrifuged at 12000 rpm for 10 min. The plasma was removed
and divided into two aliquots, one for determination of test
compound plasma concentrations (pharmacokinetic analysis) and
the second for determination the percent DPP-IV inhibition (phar-
macodynamic effect). The plasma samples were frozen at -70 °C
until analysis. The rat plasma samples were analyzed for test
compound concentrations by LC/MS/MS using an Applied Bio-
systems API 4000 mass spectrometer. Briefly, test compound
standard curves were prepared in control rat plasma with a dynamic
range of 1.0-2000 ng/mL. Aliquots (0.02 mL) of both standards
and samples were placed into Marsh tubes in a 96-well block.
Proteins were precipitated by addition of 0.1 mL acetonitrile
containing 0.1 mg/mL of internal standard. The 96-well blocks were
vortexed and then centrifuged at 3000 rpm for 5 min. The resulting
1
a white solid. H NMR (CD3OD): 4.68 (m, 2 H); 2.49-2.32 (m,
4 H); 1.52 (s, 9 H). 13C NMR (CD3OD): 152.8, 118.7, 118.3, 83.0,
30.4, 29.5, 27.2. APCI MS: 222 (MH+). mp: 113-114 °C. Anal.
(C11H15N3O2) C, H, N.
General Procedure for the Preparation of Inhibitors 1: 1-
({[1-(Hydroxymethyl)cyclopentyl]amino}acetyl)pyrrolidine-2,5-
dicarbonitrile (1c). A solution of 0.149 g (0.615 mmol) of 2 in 2
mL of MeCN was added dropwise to a solution of 0.213 g (1.85
mmol) of 1-(hydroxymethyl)-1-cyclopentylamine in 4 mL of
MeCN. The mixture was stirred for 48 h at room temperature, after
which the mixture was evaporated. The residue was flash chro-
matographed on silica gel (eluting with 95:5 CH2Cl2-MeOH) to
afford 0.120 g (70%) of 1c as the free base, which was then taken
up in 4 mL of MeOH, treated with 1 equiv of HCl in MeOH, and
concentrated. The residue was taken up in MeCN, concentrated,
and then crystallized from MeCN to afford 0.128 g (95%) of 1c as
the hydrochloride salt. 1H NMR (D2O): 5.04 (m, 1 H); 4.78 (m, 1
H); 4.14 (d, J ) 16.6 Hz, 1 H); 4.04 (d, J ) 16.6 Hz, 1 H); 2.54-
2.35 (m, 4 H); 1.82-1.68 (m, 8 H); 1.62-15.4 (m, 2H). 13C NMR
(D2O): 165.7, 118.0, 117.5, 70.9, 62.9, 47.6, 47.4, 44.4, 32.0, 30.9,
28.3, 24.0. APCI MS: 277 (MH+). mp: 173-175 °C. Anal.
(C14H20N4O2 HCl): C, H, N.
General Procedure for the Preparation of Tertiary Carbi-
namines from Tertiary Alcohols Using TMSN3 and BF3:
1-Methylcyclopentylamine Hydrochloride (19, amine for 1z).
A solution of 2.00 g (20.0 mmol) of 1-methylcyclopentanol26 in
25 mL of benzene was treated with 3.18 mL (23.9 mmol) of
azidotrimethylsilane and 3.04 mL (24.0 mmol) of boron trifluoride
etherate. After 24 h, the solution was poured into 50 mL of 1 M
sodium bicarbonate solution and stirred for 30 min. Solid sodium
bicarbonate was added as needed to maintain pH > 7. The benzene
layer was separated and dried over anhydrous CaCl2. The benzene
solution was diluted with 25 mL of methanol and hydrogenated
over 0.45 g of 10% Pd/C catalyst for 90 min at 45 psi and 20 °C.
The reaction mixture was then filtered through diatomaceous earth,
1.7 mL of 12 M hydrochloric acid was added to the filtrate, and
the solvent was evaporated to provide 1.92 g (71%) of 1-methyl-
cyclopentylamine hydrochloride 19 as a white solid. 1H NMR
(D2O): 1.59 (m, 8 H); 1.22 (s, 3 H). 13C NMR (D2O): 61.4, 37.7,
24.6, 23.7. APCI MS: 100 (MH+). mp: 261-262 °C. Anal.
(C6H13N HCl): C, H, N.
General Procedure for the Preparation of Tertiary Carbi-
namines from N-Cbz-1-(hydroxymethyl)-1-cyclopentylamine:
[1-(Propoxymethyl)cyclopentyl]amine (13y, amine for 1y). A
solution of 11.5 g (99.8 mmol) of 1-(hydroxymethyl)-1-cyclopen-
tylamine in 50 mL of CH2Cl2 was added over 30 min to a solution
of 7.1 mL (50 mmol) of Cbz-Cl in 100 mL of CH2Cl2 with cooling
in ice.27 Stirring at 0 °C was continued for 30 min, and the mixture
was filtered and concentrated to provide 12.48 g (100%) of 11 as
1
a white solid. H NMR (CDCl3): 7.41 (m, 5 H); 5.11 (s, 2 H);
5.04 (br s, 1 H); 3.73 (s, 2 H); 1.87-1.66 (m, 8 H). APCI MS:
250 (MH+). mp: 56-57 °C (from hexane). A mixture of 11 (4.98
g, 20 mmol) and freshly prepared Ag2O (9.27 g, 40.0 mmol) in 15
mL of DMF was treated with allyl bromide (12 mL, excess) and
stirred overnight at 20 °C. The reaction mixture was filtered and
evaporated, and the residue was chromatographed on silica gel
(eluting with 9:1 hexanes-EtOAc) to provide 5.31 g (91%) of 12
(R ) allyl) as an oil. 1H NMR (CDCl3): 7.36-7.27 (m, 5 H); 5.84
(m, 1 H); 5.23 (d of d, J ) 17.0 Hz, 1.6 Hz, 1 H); 5.14 (d of d, J
) 10.4 Hz, 1.5 Hz, 1 H); 5.04 (br d, 2 H); 3.95 (m, 2 H); 1.92-
1.88 (m, 2 H); 1.75-1.66 (m, 4 H); 1.58-1.52 (m, 2 H). APCI
MS: 290 (MH+). A mixture of 2.80 g (9.67 mmol) of 12 (R )
allyl) and 0.80 g of 10% Pd on C was shaken under 50 psi of
hydrogen overnight at 20 °C. The mixture was filtered and
concentrated to provide 1.52 g (98%) of [1-(propoxymethyl)-
1
cyclopentyl]amine (13y, R ) propyl) as a clear liquid. H NMR
(CD3OD): 3.42 (t, 2 H); 3.27 (s, 2 H); 1.78-1.45 (m, 8 H); 0.93

Dicyanopyrrolidine Inhibitors of Dipeptidyl Peptidase
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11 3075
Dipeptidyl Peptidase IV Inhibitors with Enhanced Chemical Stabil-
ity: Interplay between the N-Terminal Amino Acid Alkyl Side
Chain and the Cyclopropyl Group of R-Aminoacyl-L-cis-4,5-metha-
noprolinenitrile-Based Inhibitors. J. Med. Chem. 2004, 47, 2587-
2598.
supernatant was removed and placed into a new 96-well block and
taken to dryness at 50 °C under a nitrogen stream. Residues were
reconstituted in mobile phase (60% 5 mM ammonium acetate and
40% acetonitrile). Aliquots (0.01 mL) were then injected onto the
LC/MS/MS for analysis.
(7) Sitkoff, D.; Magnin, D.; Robl, J.; Sulsky, R. B.; Augeri, D. J.; Huang,
Y.; Taunk, P.; Betebenner, D. A.; Simpkins, L. M.; Robertson, J.
G.; Khanna, A.; Abboa-Offei, B.; Wang, A.; Cap, M.; Xing, L.; Tao,
L.; Malley, M.; Gougoutas, J. Z.; Huang, Q.; Han, S.-P.; Parker, R.
A.; Hamann, L. G. A Study of Chemical Stability of Novel, Potent
Inhibitors of DPP-IV. Abstracts of Papers, 226th American Chemical
Society National Meeting, New York, NY, Sept 7-11, 2003, ACS:
Washington, DC, 2003.
(8) (a) McIntosh, J. M. Robinson-Schopf Condensations with Succinal-
dehyde. J. Org. Chem. 1988, 53, 447-448. (b) Takahashi, K.; Saitoh,
H.; Ogura, K.; Iida, H. Efficient Synthesis of 1-Substituted 2,5-
Dicyanopyrrolidines by the Strecker Reaction Using Succindialdehyde
and Primary Amines and Their Stereochemistry. Heterocycles 1986,
24, 2905-2910.
(9) The corresponding N-benzyl analogue of 3 was prepared but failed
to undergo significant debenzylation upon treatment with bromoacetyl
bromide. Treatment of 3 with chloroacetyl chloride failed to afford
significant amounts of the desired chloroacetamide.
(10) Watson, H. A.; O’Neill, B. T. A Reinvestigation and Improvement
in the Synthesis of meso-2,5-Dibromoadipates by Application of Le
Chatelier’s Principle. J. Org. Chem. 1990, 55, 2950-2952.
(11) Fraenkel, G.; Duncan, J. H.; Wang, J. Restricted Stereochemistry of
Solvation of Allylic Lithium Compounds: Structural and Dynamic
Consequences. J. Am. Chem. Soc. 1999, 121, 432-443.
(12) Koziara, A.; Zwierzak, A. Iminophosphorane-Mediated Transforma-
tion of Tertiary Alcohols Into t-Alkylamines and Their N-Phospho-
rylated Derivatives. Tetrahedron Lett. 1987, 28, 6513-6516.
(13) Scharpe, S.; De Meester, I.; Vanhoof, G.; Hendriks, D.; van Sande,
M.; Van Camp, K.; Yaron, A. Assay of Dipeptidyl Peptidase IV in
Serum by Fluorometry of 4-Methoxy-2-naphthylamine. Clin. Chem.
1988, 34, 2299-2301.
(14) Lavrova, L. N.; Indulen, M. K.; Ryazantseva, G. M.; Korytnyi, V.
S.; Yashunskii, V. G. Pharm. Chem. J. (Engl. Transl.) 1990, 24,
29-34.
X-ray Crystallography. The protein construct contained residues
31-766 of DPP-IV and a hexa-his tag linked via a thrombin
cleavage site. The protein was secreted from insect cells and purified
through conventional chromatography. The purified protein was
then enzymatically deglycosylated and set up for crystallization.
The crystal diffracted 2.6 Å resolution at APS 17ID. The crystal
belonged to space group P43212 with a ) b ) 69.1 Å and c )
410.8 Å. The crystal mosaicity was 0.5°. The data were 96.3%
complete to 2.6 Å resolution with an Rmerge of 0.109. There were
149578 reflections and 30926 unique reflections, giving an average
redundancy of 4.8. The structure was solved using difference Fourier
methods with the program AMORE.31 The structure was refined
with 2 iterations of REFMAC refinement and manual model
building. The final R factor was 0.217 (Rfree 0.281). The final model
contained residues 31-766, and additionally contained eight
N-acetylglucosamine (NAG) residues on six sites (85 × 2, 150 ×
2, 219, 229, 281, 520), one inhibitor molecule 1c and 179 water
molecules.
Supporting Information Available: Elemental analyses for
compounds 1a-z, 1aa-gg, 2, 3, 6-10, 13y, 14a, 14b, 15-17,
and 19, observed APCI MS MH+ ions for compounds in Tables 1
and 2, and structures of additional inactive (IC50 > 500 nM) cis-
2,5-dicyanopyrrolidine inhibitors not presented in Tables 1 and 2.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(15) Jasys, V. J.; Lombardo, F.; Appleton, T. A.; Bordner, J.; Ziliox, M.;
Volkmann, R. A. Preparation of Fluoroadamantane Acids and
Amines: Impact of Bridgehead Fluorine Substitution on the Solution-
and Solid-State Properties of Functionalized Adamantanes. J. Am.
Chem. Soc. 2000, 122, 466-473.
(1) Fact Sheet No. 138. World Health Organization, Geneva, 2002.
(2) For a recent review, see: Knudsen, L. B. Glucagon-like Peptide-1:
The Basis of a New Class of Treatment for Type 2 Diabetes. J. Med.
Chem. 2004, 47, 4128-4134.
(3) For recent reviews, see: (a) Mest, H.-J.; Mentlein, R. Dipeptidyl
Peptidase Inhibitors as New Drugs for the Treatment of Type 2
Diabetes Diabetologia 2005, 48, 616-620. (b) Weber, A. E.
Dipeptidyl Peptidase IV Inhibitors for the Treatment of Diabetes. J.
Med. Chem. 2004, 47, 4135-4141. (c) Drucker, D. J. Therapeutic
Potential of Dipeptidyl Peptidase IV Inhibitors for the Treatment of
Type 2 Diabetes. Expert Opin. InVest. Drugs 2003, 12, 87-100. (d)
Augustyns, K.; Van der Veken, P.; Senten, K.; Haemers, A.
Dipeptidyl Peptidase IV Inhibitors as New Therapeutic Agents for
the Treatment of Type 2 Diabetes. Expert Opin. Ther. Pat. 2003,
13, 499-510.
(16) (a) Prepared from commercial (2S)-5-methylpyrrolidine-2-carboxylic
acid by (i) CH2N2, Et2O, THF; (ii) NH3, MeOH; (iii) (adamantan-
1-yl-tert-butoxycarbonyl-amino)-acetic acid,16d EDC, HOBT, DMF;
(iv) POCl3, imidazole, pyridine, CH2Cl2; (v) isomer separation by
chromatography; (vi) HCl, MeCN. (b) Prepared from commercial
(2S,5R)-Boc-5-phenyl-pyrrolidine-2-carboxylic acid by (i) 1,1′-car-
bonyldiimidazole, THF, then NH4OH.; (ii) POCl3, imidazole, pyri-
dine, CH2Cl2; (iii) BrCH2COBr, MeCN; (iv) 1-amino-1-cyclopen-
tanemethanol, MeCN. (c) Prepared from commercial N-Boc-L-(trans)-
4-hydroxyproline methyl ester by (i) NH3, MeOH; (ii) MsCl, Et3N,
THF; (iii) POCl3, imidazole, pyridine, CH2Cl2; (iv) NaOAc, DMSO;
(v) BrCH2COBr, MeCN; (vi) 1-adamantanamine, MeCN.; (vii) NH3,
MeOH. (d) See Van der Veken, P.; Senten, K.; Kertesz, I.; De
Meester, I.; Lambeir, A.-M.; Maes, M.-B.; Scharpe, S.; Haemers,
A.; Augustyns, K. Fluoro-Olefins as Peptidomimetic Inhibitors of
Dipeptidyl Peptidases. J. Med. Chem. 2005, 48, 1768-1780.
(17) Hughes, T. E.; Mone, M. D.; Russell, M. E.; Weldon, S. C.; Villhauer,
E. B. NVP-DPP728 (1-[[[2-[(5-Cyanopyridin-2-yl)amino]ethyl]ami-
no]acetyl]-2-cyano-(S)-pyrrolidine), a Slow-Binding Inhibitor of
Dipeptidyl Peptidase IV. Biochemistry 1999, 38, 11597-11603.
(18) (a) DPP-II: see McDonald, J. K.; Leibach, F. H.; Grindeland, R. E.;
Ellis, S. Purification of Dipeptidyl Aminopeptidase II (Dipeptidyl
Arylamidase II) of the Anterior Pituitary Gland. J. Biol. Chem. 1968,
243, 4143-4150. (b) Abramic M.; Zubanovic M.; Vitale L. Dipep-
tidyl Peptidase III from Human Erythrocytes. Biol. Chem. 1988, 369,
29-38. (c) DPP-VIII: see Abbot, C. A.; Yu, D. M.; Woollatt, E.;
Sutherland, G. R.; McCaughan, G. W.; Gorrell, M. D. Cloning,
Expression and Chromosomal Localization of a Novel Human
Dipeptidyl Peptidase (DPP) IV Homolog, DPP8. Eur. J. Biochem.
2000, 267, 6140-6150. (d) DPP-IX: see Olsen, C.; Wagtmann, N.
Identification and Characterization of Human DPP9, a Novel
Homologue of Dipeptidyl Peptidase IV. Gene 2002, 299, 185-193.
(e) FAP: see Scallan, M. J.; Raj, B. K. M.; Calvo, B.; Garin-Chesa,
P.; Sanz-Moncasi, M. P.; Healey, J. H.; Old, L. J.; Rettig, W. J.
Molecular Cloning of Fibroblast Activation Protein Alpha, a Member
of the Serine Protease Family Selectively Expressed in Stromal
Fibroblast of Epithelical Cancers. Proc. Natl. Acad. Sci. U.S.A. 1994,
91, 5657-5661. (f) APP: see Rusu I.; Yaron A. Eur. J. Biochem.
1992, 210, 93-100.
(4) Over 130 patent applications concerning DPP-IV inhibitors have been
published.
(5) Published structures of DPP-IV with bound inhibitors include: (a)
Kim, D.; Wang, L.; Beconi, M.; Eiermann, G. J.; Fisher, M. H.; He,
H.; Hickey, G. J.; Kowalchick, J. E.; Leiting, B.; Lyons, K.; Marsilio,
F.; McCann, M. E.; Patel, R. A.; Petrov, A.; Scapin, G.; Patel, S. B.;
Roy, R. S.; Wu, J. K.; Wyvratt, M. J.; Zhang, B. B.; Zhu, L.;
Thornberry, N. A.; Weber, A. E. (2R)-4-Oxo-4-[3-(Trifluoromethyl)-
5,6-dihydro[1,2,4]triazolo[4,3-a]pyrazin-7(8H)-yl]-1-(2,4,5-trifluo-
rophenyl)butan-2-amine: A Potent, Orally Active Dipeptidyl Pep-
tidase IV Inhibitor for the Treatment of Type 2 Diabetes. J. Med.
Chem. 2005, 48, 141-151. (b) Hiramatsu, H.; Yamamoto, A.; Kyono,
K.; Higashiyama, Y.; Fukushima, C.; Shima, H.; Sugiyama, S.; Inaka,
K.; Shimizu, R. The Crystal Structure of Human Dipeptidyl Peptidase
IV (DPPIV) Complex with Diprotin A. Biol. Chem. 2004, 385, 561-
564. (c) Oefner, C.; D′Arcy, A.; MacSweeney, A.; Pierau, S.;
Gardiner, R.; Dale, G. E. High-Resolution Structure of Human Apo
Dipeptidyl Peptidase IV/CD26 and Its Complex with 1-[({2-[(5-
Iodopyridin-2-yl)amino]-ethyl}amino)-acetyl]-2-cyano-(S)-pyrroli-
dine. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2003, D59, 1206-
1212. (d) Rasmussen, H. B.; Branner, S.; Wiberg, F. C.; Wagtmann,
N. Crystal Structure of Human Dipeptidyl Peptidase IV/CD26 in
Complex with a Substrate Analog. Nat. Struct. Mol. Biol. 2003, 10,
19-25.
(6) Magnin, D. R.; Robl, J. A.; Sulsky, R. B.; Augeri, D. J.; Huang, Y.;
Simpkins, L. M.; Taunk, P. C.; Betebenner, D. A.; Robertson, J. G.;
Abboa-Offei, B. E.; Wang, A.; Cap, M.; Xin, L.; Tao, L.; Sitkoff,
D. F.; Malley, M. F.; Gougoutas, J. Z.; Khanna, A.; Huang, Q.; Han,
S.-P.; Parker, R. A.; Hamann, L. G. Synthesis of Novel Potent

3076 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 11
Wright et al.
(19) Hulin, B. U.S. Patent 6812350 B1, Nov. 2, 2004 (Chem. Abstr. 140,
16964).
(20) Deacon, C. F.; Wamberg, S.; Bie, P.; Hughes, T. E.; Holst, J. J.
Preservation of Active Incretin Hormones by Inhibition of Dipeptidyl
Peptidase IV Suppresses Meal-Induced Incretin Secretion in Dogs.
J. Endocrinol. 2002, 172, 355-362.
(26) Walling, C.; Padwa, A. Positive Halogen Compounds. VI. Effects
of Structure and Medium on the â-Scission of Alkoxy Radicals. J.
Am. Chem. Soc. 1963, 85, 1593-1597.
(27) (a) Kaminski, J. J.; Bodor, N. S.; Higuchi, T. N-Halo Derivatives
IV: Synthesis of Low Chlorine Potential Soft N-Chloramine Systems.
J. Pharm. Sci. 1976, 65, 1733-1737. (b) Kaminski, J. J.; Bodor, N.
S. U.S. Patent 4036843, July 19, 1977 (Chem. Abstr. 87, 133914).
(28) Obtained from MP Biomedicals, Livermore, CA.; DPP-IV 5 mU/
mL stock.
(29) Obtained from Bachem Bioscience, Inc.; King of Prussia, PA.
(30) The glucose level in the animals receiving glucose but no test
compound represented 0% inhibition, and the glucose concentration
in the animals receiving only water represented 100% inhibition of
glucose excursion.
(31) Collaborative Computational Project, Number 4. The CCP4 suite:
programs for protein crystallography. Acta Crystallogr. 1994, D50,
760-763.
(21) The mixture was prepared from 500 mL of regular dairy sour cream
containing 17% fat blended with 500 mL of 1 M aqueous sucrose.
(22) See, for example: Breen, P. H.; Isserles, S. A.; Westley, J.; Roizen,
M. F.; Taitelman, U. Z. Effect of Oxygen and Sodium Thiosulfate
During Combined Carbon Monoxide and Cyanide Poisoning. Toxicol.
Appl. Pharmacol. 1995, 134, 229-234 and references therein.
(23) This value is for nonsmokers; see: Brocco, G.; Rossato, R.;
Fraccaroli, M.; Lippi, G. An Automated Method Based on Plasma
Thiocyanate Determination for the Identification of Exposure to
Tobacco Smoke. Eur. J. Lab. Med. 1999, 7, 106-110.
(24) The plasma Cmax attained by this dosing regimen was determined to
be 73 µM.
(25) 13C-Labeled 1c was prepared from 13C-labeled 3, which was prepared
by using K13CN in the Strecker reaction (Scheme 2).
JM0600085