Discovery of 6-(Aminomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo[1,2- α]pyrimidine-2-carboxamides as potent, selective dipeptidyl peptidase-4 (DPP4) inhibitors

Article abstract of DOI:10.1021/jm100634a

Continued structure-activity relationship (SAR) exploration within our previously disclosed azolopyrimidine containing dipeptidyl peptidase-4 (DPP4) inhibitors led us to focus on an imidazolopyrimidine series in particular. Further study revealed that by replacing the aryl substitution on the imidazole ring with a more polar carboxylic ester or amide, these compounds displayed not only increased DPP4 binding activity but also significantly reduced human ether-à-go-go related gene (hERG) and sodium channel inhibitory activities. Additional incremental adjustment of polarity led to permeable molecules which exhibited favorable pharmacokinetic (PK) profiles in preclinical animal species. The active site binding mode of these compounds was determined by X-ray crystallography as exemplified by amide 24c. A subsequent lead molecule from this series, (+)-6-(aminomethyl)-5-(2,4-dichlorophenyl)-N-(1-ethyl-1H- pyrazol-5-yl)-7-methylimidazo[1,2-a]pyrimidine-2-carboxamide (24s), emerged as a potent, selective DPP4 inhibitor that displayed excellent PK profiles and in vivo efficacy in ob/ob mice.

Full text of DOI:10.1021/jm100634a

5620 J. Med. Chem. 2010, 53, 5620–5628
DOI: 10.1021/jm100634a
Discovery of 6-(Aminomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo[1,2-a]pyrimidine-
2-carboxamides as Potent, Selective Dipeptidyl Peptidase-4 (DPP4) Inhibitors
Wei Meng,*,† Robert P. Brigance,† Hannguang J. Chao,† Aberra Fura,§ Thomas Harrity,‡ Jovita Marcinkeviciene,
Stephen P. O’Connor,† James K. Tamura,^ Dianlin Xie,^ Yaqun Zhang,^ Herbert E. Klei,^ Kevin Kish,^
Carolyn A. Weigelt,# Huji Turdi,† Aiying Wang,‡ Robert Zahler,† Mark S. Kirby,‡ and Lawrence G. Hamann†
†Departments of Discovery Chemistry, ‡Metabolic Diseases, §Pharmaceutical Candidate Optimization, Chemical Enzymology,
^Gene Expression & Protein Biochemistry, and #Computer-Assisted Drug Design, Bristol-Myers Squibb, Research and Development,
P.O. Box 5400, Princeton, New Jersey 08543-5400
Received May 25, 2010
Continued structure-activity relationship (SAR) exploration within our previously disclosed azolo-
pyrimidine containing dipeptidyl peptidase-4 (DPP4) inhibitors led us to focus on an imidazolopyrim-
idine series in particular. Further study revealed that by replacing the aryl substitution on the imidazole
ring with a more polar carboxylic ester or amide, these compounds displayed not only increased DPP4
ꢀ
binding activity but also significantly reduced human ether-a-go-go related gene (hERG) and sodium
channel inhibitory activities. Additional incremental adjustment of polarity led to permeable molecules
which exhibited favorable pharmacokinetic (PK) profiles in preclinical animal species. The active site
binding mode of these compounds was determined by X-ray crystallography as exemplified by amide 24c.
A subsequent lead molecule from this series, (þ)-6-(aminomethyl)-5-(2,4-dichlorophenyl)-N-(1-ethyl-
1H-pyrazol-5-yl)-7-methylimidazo[1,2-a]pyrimidine-2-carboxamide (24s), emerged as a potent, selective
DPP4 inhibitor that displayed excellent PK profiles and in vivo efficacy in ob/ob mice.
Introduction
GLP-1 (9-36). This quick inactivation process leads to an
apparent half-life of 60-90 s for GLP-1 (7-36), and there is
evidence that less than 50% of released active GLP-1 (7-36)
can reach circulation because of this natural degradation
mechanism.6 It is therefore apparent that a DPP4 inhibitor
can prevent degradationofand leadtopotentiation of GLP-1,
which will further improve glucose and insulin homeostasis.7
DPP4 is a nonclassical, sequence-specific serine protease that
is ubiquitously expressed throughout the body. Membrane-
bound DPP4 is especially highly expressed in the endothelium
of the capillary bed in close proximity to intestinal L-cells
where GLP-1 is secreted. There is also a soluble form of DPP4
circulating in plasma, although this form plays little role in the
cleavage of GLP-1.8 A number of small molecule DPP4 inhi-
bitors, including vildagliptin 1 (approved in Europe, Brazil,
and Mexico),9 sitagliptin 2,10 and saxagliptin 3,11 have been
granted regulatory approval, and alogliptin 4 (Figure 1) has
been advanced to late stage human clinical trials.12 Each of
these agents has demonstrated an ability to lower blood glu-
cose and HbA1c levels and to improve glucose tolerance in
type 2 diabetic patients.13
Glucogon-like peptide-1 (GLP-1a) based therapeutics have
emerged in the recent years as novel and promising treatments
for type 2 diabetes.1 Upon ingestion of meals, the incretin
hormone GLP-1 (7-36), the active and natural form of GLP-1,
is secreted from intestinal L-cells where it then stimulates insulin
secretion, inhibits glucogon release, delays gastric emptying,
and promotes β-cell trophism, all benefiting glucose homeo-
stasis in both animal models and human.2 GLP-1 levels in type 2
diabetics are markedly reduced, but exogenous GLP-1 infusion
can lead to normal insulin response to glucose,3 thereby form-
ing the basis for GLP-1 and its analogues as novel treatments
of type 2 diabetes. Exenatide, a GLP-1 analogue, was appro-
ved by the FDA in 2005 for treating type 2 diabetes.4
GLP-1 (7-36) is the most potent incretin hormone with a
half-maximal effective concentration of 10 pM for its effects
on pancreatic β-cells.5 For GLP-1 (7-36) to exert its biological
functions, it must circulate and bind to the GLP-1 receptor,
which is highly expressed in pancreatic β-cells. Following secre-
tion under normal physiological conditions, GLP-1 (7-36) is
rapidly degraded by DPP4 (EC 3.4.14.5) to afford inactive
By building on a successful DPP4 program that discovered
saxagliptin, at BMS, we continued to expand our medicinal
chemistry efforts at discovering novel scaffolds that can act
as DPP4 inhibitors. From a previous report from our labo-
ratories,14 we discovered several novel series of azolopyrimidine
amines as DPP4 inhibitors which were intrinsically potent at
DPP4 and selective over the other dipeptidyl peptidase (DPP)
family of enzymes (see compound 5, Figure 1 as an example).
Most of the structures in these series contain an aromatic or
heteroaromatic group on the azolo ring, and this substitution
*To whom correspondence should be addressed. Phone: 609-818-
4976. Fax: 609-818-3550. E-mail: wei.meng@bms.com.
a Abbreviations: DPP4, dipeptidyl peptidase-4; GLP-1, glucogon-like
ꢀ
peptide-1; DPP, dipeptidyl peptidase; hERG, human ether-a-go-go related
gene; SAR, structure-activity relationship; PK, pharmacokinetic; KMnO4,
potassium permanganate; TFA, trifluoroacetic acid; MsCl, methanesulfonyl
chloride; NaN3, sodium azide; PPh3, triphenyl phosphine; EDAC, 1-ethyl-
3-(3-dimethylaminopropyl)carbodiimide hydrochloride; HOAt, 1-hydroxy-
7-azabenzotriazole; PD, pharmacodynamic; PAMPA, parallel membrane
permeability assays; oGTT, oral glucose tolerance test; FAP, fibroblast
activation protein; IA, intraarterial.
r
pubs.acs.org/jmc
Published on Web 07/15/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5621
Scheme 1a
Figure 1. Structures of selected DPP4 inhibitors.
appeared to enhance the binding affinity to DPP4. However,
all of these compounds also displayed high levels of the human
ꢀ
ether-a-go-go related gene (hERG) and sodium channel inhi-
bition. To minimize undesired channel activities by reducing
lipophilicity, we sought to replace the aromatic or hetero-
aromatic group on the azolo ring with more polar functional
groups. This structure-activity relationship (SAR) investigation
led us to discover a series of imidazolopyrimidine amides as a
highly potent and selective class of DPP4 inhibitors as repre-
sented by 24s. Free of hERG and sodium channel activities,
24s displayed an excellent balance of DPP4 binding activity,
selectivity over other DPPs, in vivo efficacy, and pharmaco-
kinetic (PK) profiles in several preclinical species that repre-
sents an advanced lead with drug-like properties as a potential
treatment for type 2 diabetes.
a Reagents: (a) AcOH, Me2NH, IPA, 67%; (b) EtOH, 75 ꢀC, 30% 11a,
20% 11b; (c) NH2NH2 H2O, EtOH, 75 ꢀC, 100%; (d) EtOH, 70 ꢀC, 59%;
3
(e) KMnO4, acetone, 57%; (f) TFA, CH2Cl2, 100%; (g) ClCOOEt, Et3N,
THF, then NaBH4, THF, H2O, 0 ꢀC, 45%; (h) MsCl, Et3N, CH2Cl2;
(i) NaN3, DMF, 50 ꢀC; (j) PPh3, THF, H2O, 50 ꢀC, 50% for 3 steps.
Chemistry
Imidazolopyrimidine ester 19 was synthesized as shown in
Scheme 1. Knoevenagel condensation of t-butylacetoacetate6
and 2,4-dichlorobenzaldehyde7 in 2-propanol with a catalytic
amount of dimethylamine and acetic acid provided a 1:1 E/Z
mixture of acrylate 8 in 67% yield. Condensation of 2-amino-
pyrimidine 9 with ethyl bromopyruvate 10 in ethanol at 75 ꢀC
gave imidazolopyrimidine 11 as a mixture of two regio-
isomers, which underwent hydrazine hydrate-mediated clea-
vage to afford substituted imidazole 12 in good overall yield.15
Condensation of acrylate 8 and imidazole 12 in ethanol at 70 ꢀC
afforded substituted imidazolodihydropyrimidine 13, which
was immediately subjected to oxidation with KMnO4 to give
diester 14. The t-butyl ester was next converted to an amino-
methyl moiety in a five-step sequence. First, hydrolysis of the
t-butylester of 14 with TFA afforded acid 15 quantitatively.
Mixed anhydride formation of acid 15 with ethyl chlorofor-
mate, followed by reductionwithNaBH4, provided alcohol 16
in 45% yield. The alcohol was then treated with MsCl to form
benzyl chloride 17, which was converted to azide 18 using
NaN3. Final reduction of azide 18 with PPh3 afforded the
amine ester 19 in 50% yields for the three steps.
During our characterization of benzylamine 19, as well as
intermediates 16, 17, and 18, it was noticed that the two ben-
zylic protons next to the amino group existed as a pair of doublet
of doublets in the proton NMR spectrum, indicating that
these two protons are nonequivalent. This further suggested
the possibility of restricted rotation along the biaryl bond and
the existence of atropisomers. Variable temperature NMR up
to 120 ꢀC did not lead to coalescence of these doublets. When
ester 19 was subjected to chiral HPLC methods, it became clear
that two distinct compounds existed. Additional chemical and
instrumental analyses led to the conclusion that 19 is comprised
of a pair of stable, noninterconverting enantiomers in which
the chirality arises from atropisomerism. To more accurately
assess their enzymatic activity, 19 was separated on a chiral
HPLC column to give each atropisomer 20 and 21 (Figure 2).
The absolute stereochemistry was assigned based on the X-ray
crystal structure determination for an earlier intermediate 15
and later confirmed by X-ray crystallography in the active site
of DPP4.
To improve upon the potentially poor metabolic stability of
ethyl ester 19, we set out to replace the ester moiety with
metabolically more stable amides. To expedite SAR studies in
this series, a combinatorial approach was taken to prepare
these amides using a wide variety of sterically and electro-
nically differentiated amines. In anticipation of different acti-
vity for the two enantiomers, most of the amides were synthe-
sized as single atropisomers and only atropisomerically pure
molecules were carried onto PK and pharmacodynamic (PD)
studies. Because racemic azide 18 displayed superior separa-
tion behavior over 19 on chiral HPLC, a large amount of this
intermediate 18 was processed to provide chrial azide 22. Mild
chemical conditions were employed for the subsequent steps
to minimize any possible racemization. Therefore, the ethyl
ester of 22 was saponified by LiOH to form acid 23, which was
then coupled with different amines using standard EDAC/
HOAt conditions to afford the desired amides. A PPh3-
mediated reduction of the azide, followed by purification by
preparatory HPLC, provided final products 24a-ee (Scheme 2)
as TFA salts in moderate to high yields. As analyzed by chiral
HPLC, no interconversion of the atropisomers occurred during

5622 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Meng et al.
the above chemical sequence. In addition, no racemization was
observed when amides such as 24c and 24s were heated in
water at 100 ꢀC for up to 10 days.
Results and Discussion
The DPP4 inhibitory activity for compounds 19-21 and
24a-ee is summarizedinTable 1. Compoundswere incubated
with human recombinant DPP4 under steady state conditions
to detect the cleavage of the pseudosubstrate, Gly-Pro-pNA.
We were encouraged by the observation of subnanomolar
DPP4 activity (Ki = 0.7 nM) for the first analogue that was
prepared in this series, prototype ester 19. Furthermore, upon
separation of the two atropisomers, we observed that eutomer
21 was about 2-fold more potent than 19 with DPP4 Ki of
0.4 nM, whereas 20 was more than 10-fold less active, indicat-
ing most of the activityfor 19resided inone of its enantiomers.
On the basis of this finding, all subsequent amides were pre-
pared in the more active atropisomer series. All of the amide
analogues exhibited excellent DPP4 inhibitory activity, with
Ki’s ranging from sub- to single-digit nanomolar. Steric and
electronic effects did not appear to affect the binding activity
to any appreciable extent, as small alkyl amides (24a), cyclic
amides (24b, 24c, 24d, 24e), and heteroaromatic amides (24l,
24p, 24q, 24r) represented in this group all displayed compar-
able activity, suggesting orientation into a solvent exposed
region of the active site. Introduction of strong electron with-
drawing groups on the amide further increased DPP4 binding
affinity, as showninsulfonamides24i and 24j and sulfone 24k.
Figure 2. Structures of atropisomers 20 and 21.
Scheme 2a
a Reagents: (a) LiOH H2O, THF, H2O, 94%; (b) EDAC, HOAt,
3
amine, THF, iPr2NEt; (c) PPh3, THF, H2O, 50-90%.
Table 1. DPP4/DPP8/DPP9 Inhibition for Compounds 19, 20, 21, and 24a-eea
a Values represent the mean of multiple determinations.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5623
In particular, methyl sulfonamide amide 24i is the most potent
compound prepared in this series, with DPP4 Ki of 0.2 nM.
We initially focused on the very potent sulfonamide 24i for
additional biological evaluation. As a TFA salt, 24i showed
excellent aqueous solubility with calculated cLogP of 1.15.
While the strong hydrophilicity provided an advantage in
terms of formulation for preclinical animal studies, we ques-
tioned if this molecule might be too polar to cross biological
membranes. When 24i was tested in caco-2 cell monolayer and
parallel membrane permeability assays (PAMPA), very low
permeability values of <15 and 0 nm/s, respectively, were
obtained, suggestive of no membrane permeability for this
compound.16 Additionally, after a 3 mg/kg oral dose in ob/ob
mice for 24i, Cmax at 1 h was 45 nM, likely insufficient for any
PD effect despite its high potency. It was further confirmed
that when24i was administeredorally inob/obmice for anoral
glucose tolerance test (oGTT), there was no reduction in
glucose excursion. Similar lack of in vivo efficacy was obser-
ved for other potent and polar compounds, including sulfon-
amide 24j and sulfone 24k. While these analogues were the
most potent DPP4 inhibitors from the series, they were also
the most polar molecules, with cLogP’s in the range of 1-2,
which likely contributed to their low permeability. It was
reasoned that an increase of lipophilicity by removing some
polar functional groups on these molecules may lead to higher
permeability across the membranes. Thus, less polar amides
were synthesizedandevaluated. Itwas discovered thatmost of
the amides lacking additional polar sulfonamide, amine, or
sulfone functionality displayed high PAMPA permeability.
For example, comparing 24b with 24i, removal of methyl sulfon-
amide from 24i decreased DPP4 binding activity by 10-fold
but increased PAMPA permeability significantly from un-
measurable to 518 nm/s. Heterocyclic amides such as 24l and
24s also showed excellent PAMPA permeabilities of 587 and
669 nm/s, respectively, with morpholino amide 24c measuring
464 nm/s in the same assay.
is the most potent analogue from this series against DPP8
and DPP9, with selectivity for DPP4 of only 16- and 52-fold,
respectively. Selectivity for five-membered ring hetero-
aromatic amides varied from 500- to 2000-fold, mostly due to
their varying degrees of activity at DPP4. By selecting com-
pounds with appropriate DPP4 potency from this subgroup,
1000-fold selectivity can be achieved as shown in amides
such as 24s.
As reported earlier,14 when the azolo ring of the azolopyr-
imidine core was substituted with an aromatic or hetero-
aromatic group (5, Figure 1), the analogues displayed signifi-
cant hERG and sodium channel activities. Various predictive
binding models for hERG have been developed, and based on
in silico modeling predictions for hERG channel blockers,
molecular features that may contribute to increased hERG
inhibitory activity include hydrophobicity, high molecular
weight, the presence of a basic group such as an amine, and
a V-shape topology arising from ring-linker arrangements.22
Compound 5 contains several of these structural attributes
which may explain its high hERG activity. To disrupt some of
these features that may contribute to the hERG activity, we
explored the SAR off of the five-membered ring of the hetero-
bicyclic core by replacing the pendant aromatic or hetero-
aromatic system with more polar functionality. It was reasoned
that modifications of this type should increase the overall
polarity of the analogues and reduce the number of aromatic
systems, thereby destroying the resulting V-shaped arrange-
ment. Indeed, when the phenyl ring on 5 was replaced with a
more polar ester or amide group, the compounds in general
showed reduced ion channel activities. Notably, ester 19 had
23% inhibition for hERG at 10 μM, compared with 74% for 5.
Within the amideseries, whena heterocyclesuch as pyrazole is
present in the amide portion of the molecule, the analogues
also showed improvement in channel activities compared with
the corresponding nonamides. For example, pyrazole amide
24s had 31% and 2% inhibition for hERG and sodium channel
at 10 μM, respectively.
Having identified membrane permeability as one prerequi-
site for efficacy, these compounds were tested for other related
peptidase activities. DPP4 belongs to a larger family of pep-
tidases including DPP8,17 DPP9,18 DPP2,19 and fibroblast
activation protein (FAP).20 While all members of the DPP
familypreferentiallycleave Xaa-Pro-and Xaa-Ala-dipeptides
(where Xaa is any amino acid except proline) from the
N-terminus of proteins, however, the physiological substrates
and clinicalrelevance of theserelated peptidases have not been
definitively established21 except for DPP4. To date, com-
pounds with varying degrees of selectivity for DPP4 versus
DPP8 and DPP9 have been studied in a human clinical setting
and are well tolerated. As we continue to explore novel
chemical scaffolds directed against this promising target for
chronic treatment for type 2 diabetes, selectivityover the other
DPPswas nonetheless incorporated intothesenext generation
inhibitors as a medicinal chemistry goal. All of the amides
24a-ee were highly selective for DPP4 over DPP2 and FAP,
however, varying degrees of selectivity over DPP8 and DPP9
were observed in this series (Table 1). Esters and alkyl amides
were most selective over DPP8 and DPP9, with selectivity
ratios typically greater than 1000-fold. For the very potent
sulfonamide amide 24i, the selectivity for DPP4 over DPP8
and DPP9 is approaching 10000-fold. On the other hand,
pyridine containing amides 24r, 24w, and 24bb-24ee are often
less than 500-fold selective for DPP4, suggesting the pyridine
ring may have an additional interaction with DPP8 and DPP9
at the corresponding region of the active site. Compound 24bb
To achieve better understanding of the binding mode, we
were able to obtain X-ray crystal structures of both the DPP4
apo enzyme and DPP4 bound as a cocrystal with inhibitor 24c
in the active site (Figure 3).23 Consistent with earlier reports in
the literature,24 the DPP4 binding pocket is located at the
interface of the R/β-hydrolase and β-propeller domains. The
catalytic triad (S630, H740, and D708), which is highly
conserved in the DPP4 gene family (including DPP8 and
DPP9), resides in the R/β-hydrolase domain. On the other
hand, E205 and E206, which typically interact with the
N-terminus of DPP substrates, are located in the β-propeller
domain, which is less conserved across the DPP4 gene family.
From analysis of the X-ray crystal structure of 24c bound to
DPP4, it is clearly demonstrated that this compound binds in
the active site with the 2,4-dichlorophenyl group fitting tightly
into the hydrophobic S1 pocket, whereas the primary benzylic
amino group forms salt bridges to E205 and E206. The amine
also exhibited hydrogen binding interactions with Y662. These
key interactions anchor the molecule into the DPP4 binding
site and also serve to explain why the alternate orientation of
amine and dichlorophenyl groups in the opposite atropisomer
would not fit well in this binding mode, leading to weakened
binding affinity. Consistent with our observed SAR for this
series, the amide portion of the molecule projects to a large,
solvent-exposed area of the enzyme pocket, which can tolerate
a variety of different groups, including those with polar
functionality. Unlike dipeptide mimetic DPP4 inhibitors such

5624 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Meng et al.
monitored at different time intervals from 0 to 2 h. At 1 and
3 μmol/kg, 24c significantly reduced glucose excursion at 15 min
by 66% and 63%, respectively. In the same experiment, 24s
reduced glucose excursion up to 40% and 58% after 1 and
3 μmol/kg oral dosing with concurrent increase of insulin
AUC 0-2 h by 93% and 129%. In the latter experiment, the
plasma DPP4 activity was inhibited by 62% and 52.8%,
respectively, after 1 and 3 μmol/kg dosing. The reduced glu-
cose excursion, increase in insulin secretion during an oral
glucose challenge, combined with reduced plasma DPP4 acti-
vity, supports thatDPP4inhibition was themechanismfor the
observed PD effects.
Conclusions
In summary, we have developed a novel series of imidazolo-
pyrimidine amides as potent DPP4 inhibitors. By replacing
the aryl substituent on an azolopyrimidine core scaffold from
a previously disclosed series with a more polar amide group,
the resultant molecules displayed enhanced DPP4 inhibitory
activity, as well as markedly reduced hERG and sodium channel
activities. Polarity of the compounds was further modified to
improve the balance of off-target activity, DPP4 inhibition
and plasma exposure levels, leading to amide-containing inhi-
bitors possessing drug-like properties as represented by 24c
and 24s. X-ray crystallography of 24c from this class bound to
DPP4 confirmed that these compounds bind in the DPP4
binding pocket in a mode consistent with our observed SAR.
Lead molecule 24s from this series demonstrated potent DPP4
activity, selectivity over DPP2, DPP8, and DPP9, low hERG,
sodium and calcium ion channel activities, an excellent PK
profile in preclinicalspecies, andfavorablePDeffects. Further
evaluation of these compounds as potential treatments for
type 2 diabetes is ongoing and will be reported in due course.
Figure 3. X-ray crystal structure of 24c and DPP4.
Table 2. PK Profiles for 24c and 24sa
CL
(iv) mL/min/kg
% of parent
in urine
compd/species Vss L/kg t1/2
h
F %
24c/rat
24s/rat
24c/dog
1.3
3.9
1.5
1.6
25.4
43.2
7.9
56-79
4.4-5.9
41
54
64
76
2.1
4.6
a Dose: IA: 5 μmol/kg, po: 2 μmol/kg.
as saxagliptin and vildagliptin, where the hydroxy group of
S630 from the catalytic triad plays a crucial role in binding,
there was no obvious interaction of this catalytic residue with
24c, and combined with the fact that most of the key contacts
withandsubsequentbinding energy from24c withinthe active
site of DPP4 derives from residues in the β-propeller domain,
this may help to explain why this chemotype provides high degree
of selectivity for DPP4 over DPP8 and DPP9.25
Experimental Section
All reactions were carried out under a static atmosphere of argon
or nitrogen and stirred magnetically unless otherwise noted. All
reagents used were of commercial quality and were obtained from
Aldrich Chemical Co., Sigma Chemical Co., Lancaster Chemical
Co., or Acros Chemical Co. 1H (400 MHz) and 13C (100 MHz)
NMR spectra were recorded on a JEOL GSX400 spectrometer
using Me4Si as an internal standard unless otherwise noted.
1H (500 MHz) and 13C (125 MHz) NMR spectra were recorded
on a JEOL JNM-ECP500spectrometer. Chemical shiftsare given
in parts per million (ppm) downfield from internal reference
Me4Si in δ units, and coupling constants (J values) are given in
hertz (Hz). Selected data are reported in the following manner:
chemicalshift, multiplicity, coupling constants. All reactionswere
carried out using commercially available anhydrous solvents
from Aldrich Chemical Co. or EM Science Chemical Co. unless
otherwise noted. All flash chromatographic separations were per-
formed using E. Merck silica gel (particle size, 0.040-0.063 mm).
Reactions were monitored by TLC using 0.25 mm E. Merck silica
gel plates (60 F254) and were visualized with UV light, with 5%
phosphomolybdic acid in 95% EtOH, or by a sequential treat-
ment with 1 N HCl/MeOH followed by ninhydrin staining.
LC/MS data were recorded on a Shimadzu LC-10AT equipped
with a SIL-10A injector, a SPD-10AV detector, normally operat-
ing at 220 nm, and interfaced with a Micromass ZMD mass spectro-
meter. LC/MS or HPLC retention times, unless otherwise noted,
are reported using a Phenomenex Luna C-18 4.6 mm ꢀ50 mm
columneluted with a 4 min gradientfrom 0 to 100% B, where A=
10% MeOH-90% H2O-0.1% TFA and B=90% MeOH-10%
H2O-0.1% TFA. All solvents were removed by rotary evapora-
tion under vacuum using a standard rotovap equipped with a
dry ice condenser. All filtrations were performed with a vacuum
On the basis of their highly potent and selective DPP4 inhi-
bitory activity, as well as favorable permeability and low off-
target activity, 24c and 24s were carried on for additional PK
evaluationin preclinicalanimalmodels. Keyparameters obta-
ined from rat PK studies for 24c and 24s after a 5 μmol/kg
intraarterial (IA) and 2 μmol/kg oral dose are summarized in
Table 2. The vehicle for these studies is water, in which these
compounds are freely soluble, another favorable drugability
property shared by the present class of compounds. The Cmax
for 24c at 0.6 h is 324 nM with an AUC0-¥ of 714 nM h.
3
The systemic plasma clearance of 24c and 24s were 25.4 and
43.2 mL/min/kg, respectively, indicating high clearance in
rats. The fraction of dose excreted in urine collected from
the rats over a 24 h period, as unchanged parent compound
was 56-79% for 24c and 4.4-5.9% for 24s, indicating renal
clearance to be significant for 24c. The steady state volume
of distribution was 1.3 and 3.9 L/kg for 24c and 24s. The
estimated terminal elimination half-life after IA administra-
tion was 1.6 and 2.1 h for 24c and 24s, respectively. Despite the
relatively high clearance, the absolute oral bioavailability for
24c and 24s were excellent at 54 and 64%. Compound 24c was
also evaluated in dog PK profiling with 76% bioavailability
and a t1/2 of 4.6 h that may be suitable for once-a-day dosing.
An ob/ob mouse model was utilized to evaluate the in vivo
PD effects in an oGTT. In this model, compounds were
administered orally in water to the mice 1 h prior to an oral
glucose (2 g/kg) bolus. Blood glucose and insulin levels were

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5625
unless otherwise noted. Purity of all intermediates and final com-
pounds was determined to be >95% by HPLC (Phenomenex
Luna C-18 4.6 mm ꢀ50 mm column eluted with a 4 min gradient
from 0 to 100% B, where A=10% MeOH-90% H2O-0.1% TFA
and B = 90% MeOH-10% H2O-0.1% TFA and detection at
220 nm) and/or elemental analyses.
(980 mg, 6.32 mmol) and 8 (2.03 g, 6.44 mmol) were dissolved in
ethanol (15 mL) and heated to 75 ꢀC for 16 h. The reaction was
concentrated under reduced pressure to obtain a 10:1 mixture
(by HPLC) of 13 and isomeric 6-tert-butyl 3-ethyl 5-(2,4-dichloro-
phenyl)-7-methyl-5,8-dihydroimidazo[1,2-a]pyrimidine-3,6-di-
carboxylate 3-isomer (3.07 g) as a yellow foam. A small amount
of this mixture (350 mg) was purified by silica gel chromato-
graphy to obtain 13 (191 mg, 59%) and 3-isomer (19.2 mg, 5.9%)
as white solids. For 2-isomer 13: 1H NMR (400 MHz, CDCl3)
δ 10.19 (br s, 1H), 7.43 (d, J=2.2 Hz, 1H), 7.38 (s, 1H), 7.34 (d,
J=7.2 Hz, 1H), 7.21 (dd, J=2.2, 7.2 Hz, 1H), 6.67 (s, 1H), 4.29
E/Z-tert-Butyl 2-(2,4-dichlorobenzylidene)-3-oxobutanoate (8).
To a stirred solution of tert-butylacetoacetate 6 (1.50 g, 9.48 mmol)
and 2,4-dichlorobenzaldehyde 7 (1.66 g, 9.48 mmol) in 2-propanol
(10 mL) was added acetic acid (23.9 μL, 0.400 mmol) and
dimethyl amine (2 M solution in THF, 200 μL, 0.400 mmol).
The reaction was kept at RT for 3 days and was concentrated
under reduced pressure. The resulting yellow oil was purified by
silica gel chromatography (0-5% EtOAc in hexanes) to obtain 8
(2.03 g, 67%) as a 1:1 mixture of E/Z isomers. On the basis of
NOE experiments, the Z isomer is the slower moving compound
on thin layer chromatography (100% hexanes, Rf = 0.60) and
the E isomer is the faster moving isomer (100% hexanes, Rf =
0.70). For Z isomer: 1H NMR (400 MHz, CDCl3) δ 7.68 (s, 1H),
7.49 (d, J=8.4 Hz, 1H), 7.47 (d, J=2.2 Hz, 1H), 7.24 (dd, J=
2.2, 7.3 Hz, 1H), 2.44 (s, 3H), 1.45 (s, 9H). 13C NMR (125 MHz,
CDCl3) δ 194.5, 165.8, 138.0, 136.4, 135.4, 135.3, 130.8, 130.2,
129.7, 127.0, 83.2, 27.8, 26.9. Anal. Calcd for C15H16Cl2O3: C,
57.16; H, 5.11; Found: C, 57.26; H, 4.93. For E isomer: 1H NMR
(400 MHz, CDCl3) δ 7.75 (s, 1H), 7.45 (d, J=2.2 Hz, 1H), 7.28
(d, J=8.3 Hz, 1H), 7.21 (dd, J=2.2, 8.3 Hz, 1H), 2.26 (s, 3H),
1.55 (s, 9H). 13C NMR (125 MHz, CDCl3) δ 202.0, 162.8, 138.4,
136.2, 135.1, 135.0, 130.8, 130.4, 129.7, 127.3, 82.7, 31.1, 27.9.
Anal. Calcd for C15H16Cl2O3: C, 57.16; H, 5.11; Found: C,
57.34; H, 5.04.
(m, 2H), 2.57 (s, 3H), 1.34 (t, J=7.0 Hz, 3H), 1.26 (s, 9H). 13
C
NMR (125 MHz, CDCl3) δ 164.6, 162.2, 148.1, 141.6, 138.6,
134.6, 132.9, 130.1, 130.0, 129.2, 128.0, 119.7, 80.2, 60.4, 54.5,
28.1, 19.0, 14.3. Anal. Calcd for C21H23Cl2N3O4: C, 55.76; H,
5.12; N, 9.29; Found: C, 55.49; H, 5.04; N, 8.99. HRMS: Anal.
Calcd. for C21H24Cl2N3O4 452.1144; Found: 452.1128 (M þ H)þ.
6-tert-Butyl 2-Ethyl 5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidine-2,6-dicarboxylate (14). To a stirred solution of
13 (191 mg, 0.422 mmol) in acetone (5 mL) was added KMnO4
(66.7 mg, 0.422 mmol). After 1 h, the reaction was filtered through
Celite and concentrated under reduced pressure. The resulting
residue was diluted with CH2Cl2 and extracted with H2O (2ꢀ).
The organic layer was dried over MgSO4, concentrated under
reduced pressure, and recrystallized from EtOAc to obtain 14
(108 mg, 57%) as white crystals. 1H NMR (400 MHz, CDCl3) δ
7.68 (d, J=1.8 Hz, 1H), 7.53 (s, 1H), 7.50 (dd, J=1.8, 7.6 Hz,
1H), 7.35 (d, J=7.6 Hz, 1H), 4.43 (m, 2H), 2.77 (s, 3H), 1.41 (t,
J = 7.2 Hz, 3H), 1.29 (s, 9H). 13C NMR (100 MHz, CDCl3) δ
163.7, 162.7, 159.8, 147.0, 141.1, 138.5, 138.2, 134.5, 131.4, 130.3,
128.1, 127.8, 119.0, 114.1, 83.8, 61.4, 27.7, 24.3, 14.3. Anal.
Calcd for C21H21Cl2N3O4: C, 56.01; H, 4.70; N, 9.33; Found: C,
55.79; H, 4.43; N, 9.14.
5-(2,4-Dichlorophenyl)-2-(ethoxycarbonyl)-7-methylimidazo-
[1,2-a]pyrimidine-6-carboxylic Acid (15). To a stirred solution of
14 (1.14 g, 2.53 mmol) in CH2Cl2 (5 mL) was added TFA (5 mL).
The reaction was heated to 60 ꢀC for 30 h and concentrated
under reduced pressure to obtain 15 (0.998 g, 100%) as yellow
oil. 1H NMR (400 MHz, CDCl3) δ 7.80 (d, J=1.3 Hz, 1H), 7.65
(s, 1H), 7.59 (d, J=1.3 Hz, 2H), 4.36 (q, J=7.0 Hz, 2H), 2.78 (s,
3H), 1.36 (t, J = 7.0 Hz, 3H). 13C NMR (100 MHz, CDCl3) δ
161.4, 157.9, 154.9, 142.7, 138.0, 134.6, 133.2, 129.6, 127.8, 125.9,
124.8, 124.0, 113.7, 109.9, 56.0, 25.2, 9.6. Anal. Calcd for
C17H13Cl2N3O4: C, 51.79; H, 3.32; N, 10.66; Found: C, 51.57;
H, 3.05; N, 10.42.
Ethyl 5-(2,4-Dichlorophenyl)-6-(hydroxymethyl)-7-methylimidazo-
[1,2-a]pyrimidine-2-carboxylate (16). To a stirred solution of 15
(0.998 g, 2.53 mmol) in THF (10 mL) was added ClCOOEt
(0.289 mL, 3.04 mmol) and Et3N (0.529 mL, 3.80 mmol). After
2 h, the reaction was filtered and concentrated under reduced
pressure to obtain the mixed anhydride (1.18 g) as brown oil. To
a stirred solution of crude mixed anhydride (1.18 g, 2.53 mmol)
in THF (10 mL) at 0 ꢀC was added NaBH4 (144 mg, 3.80 mmol)
in H2O (0.5 mL). After 1 h, the reaction was quenched by 1N
HCl and diluted with EtOAc. The organic layer was extracted
with 1N HCl, satd aq NH4Cl, and brine before drying over
MgSO4. Filtration, concentration under reduced pressure and
purification by silica gel chromatography gave 16 (430 mg, 45%)
as a light-yellow solid. 1H NMR (400 MHz, CDCl3) δ 7.70 (d,
J=7.9 Hz, 1H), 7.63 (d, J=1.8 Hz, 1H), 7.49 (dd, J=1.8, 7.9 Hz,
1H), 7.36 (s, 1H), 4.68 (d, J=12.3 Hz, 1H), 4.36 (m, 3H), 2.75 (s,
3H), 1.35 (t, J = 7.0 Hz, 3H). 13C NMR (100 MHz, CDCl3) δ
164.5, 162.6, 147.0, 140.8, 137.8, 136.3, 134.1, 132.7, 130.1, 128.4,
127.1, 121.4, 113.8, 61.1, 57.9, 23.5, 14.2. Anal. Calcd for
C17H15Cl2N3O3: C, 53.70; H, 3.97; N, 11.05; Found: C, 53.48;
H, 3.65; N, 10.80.
Ethyl Imidazo[1,2-a]pyrimidine-2-carboxylate (11a) and Ethyl
Imidazo[1,2-a]pyrimidine-3-carboxylate (11b). 2-Aminopyrim-
idine 9 (5.00 g, 52.6 mmol) and ethyl bromopyruvate 10 (90%,
7.35 mL, 52.6 mmol) were dissolved in ethanol (80 mL) and
heated to 75 ꢀC for 16 h. The reaction was concentrated under
reduced pressure and diluted with CH2Cl2 and satd aq NaH-
CO3. The organic layer was washed with satd aq NaHCO3 (2ꢀ),
and the aq layers were extracted with CH2Cl2 (3ꢀ). The com-
bined organic layers were dried over MgSO4 and concentrated
under reduced pressure. The resulting brown oil was suspended
in cold CH2Cl2 and filtered. The filter cake was washed with cold
CH2Cl2 to obtain 11a (3.02 g, 30%) as a light-yellow solid. The
mother liquor contains a mixture of 11a and 11b (6.03 g, 60%),
which was first purified by silica gel chromatography followed
by recrystallization from EtOAc to obtain 11b (2.04 g, 20%).
For 2-isomer 11a: 1H NMR (500 MHz, CDCl3) δ 8.69 (dd, J=
2.2, 6.6 Hz, 1H), 8.67 (dd, J=2.2, 4.4 Hz, 1H), 8.22 (s, 1H), 7.01
(dd, J = 3.9, 6.6 Hz, 1H), 4.46 (q, J = 7.2 Hz, 2H), 1.43 (t, J =
7.2 Hz, 3H). 13C NMR (125 MHz, CDCl3) δ 162.8, 152.2, 147.8,
137.7, 134.4, 115.3, 110.0, 61.2, 14.2. Anal. Calcd for C9H9-
N3O2: C, 56.54; H, 4.74; N, 21.97; Found: C, 56.25; H, 4.61; N,
1
21.96. For 3-isomer 11b: H NMR (500 MHz, CDCl3) δ 9.57
(dd, J=2.2, 7.2 Hz, 1H), 8.74 (dd, J=2.2, 4.4 Hz, 1H), 8.45 (s,
1H), 7.17 (dd, J=4.4, 7.2 Hz, 1H), 4.44 (q, J=7.2 Hz, 2H), 1.44
(t, J=7.2 Hz, 3H). 13C NMR (125 MHz, CDCl3) δ 160.1, 152.0,
150.7, 142.4, 135.3, 114.3, 110.3, 60.7, 14.2. Anal. Calcd for
C9H9N3O2: H, 4.74; N, 21.97; Found: H, 4.55; N, 22.20.
Ethyl 2-Amino-1H-imidazole-4-carboxylate (12). To a stirred
solution of 11a or 11b (1.00 g, 5.23 mmol) in ethanol (40 mL) was
added hydrazine monohydrate (0.284 mL, 5.75 mmol). The reac-
tion was heated to 75 ꢀC for 16 h and was concentrated under
reduced pressure. The resulting light-yellow solid was suspended
in diethyl ether and filtered to collect 12 (810 mg, 100%) as a
white solid. 1H NMR (400 MHz, CD3OD) δ 7.27 (s, 1H), 4.24
(q, J = 7.0 Hz, 2H), 1.31 (t, J = 7.0 Hz, 3H). 13C NMR (125
MHz, CD3OD) δ 165.5, 156.1, 131.8, 125.7, 63.8, 17.3. Anal.
Calcd for C6H9N3O2: C, 46.44; H, 5.84; N, 27.08; Found: C,
46.17; H, 5.65; N, 27.28.
Ethyl 6-(Chloromethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidine-2-carboxylate (17). To a stirred solution of 16
(430 mg, 1.13 mmol) in CH2Cl2 (10mL) wasaddedMsCl(0.105mL,
1.36 mmol) and Et3N (0.236 mL, 1.70 mmol). After 10 h,
6-tert-Butyl 2-Ethyl 5-(2,4-dichlorophenyl)-7-methyl-5,8-dihydro-
imidazo[1,2-a]pyrimidine-2,6-dicarboxylate (13). Compounds 12

5626 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Meng et al.
1
the reaction was concentrated under reduced pressure and diluted
with EtOAc. The mixture was extracted with H2O and brine before
drying over MgSO4. Filtration and concentration under redu-
ced pressure gave 17 (448 mg) as yellow oil. 1H NMR (400 MHz,
CDCl3) δ 7.72 (d, J=2.2 Hz, 1H), 7.57 (dd, J=2.2, 8.4 Hz, 1H),
7.46 (d, J=8.4 Hz, 1H), 7.45 (s, 1H), 4.53 (d, J=12.3 Hz, 1H),
4.43 (m, 2H), 4.23 (d, J=12.3 Hz, 1H), 2.86 (s, 3H), 1.40 (t, J=
7.5 Hz, 3H). 13C NMR (100 MHz, CDCl3) δ 162.8, 162.1, 147.1,
141.4, 138.6, 137.9, 134.2, 131.6, 130.8, 128.7, 126.4, 118.4, 114.0,
61.3, 39.5, 23.3, 14.2. Anal. Calcd for C17H14Cl3N3O2: C, 51.21;
H, 3.53; N, 10.54; Found: C, 50.92; H, 3.45; N, 10.25.
homogeneity index, >99.9% ee. H NMR (400 MHz, CDCl3)
δ 7.72 (d, J=2.2 Hz, 1H), 7.60 (dd, J=2.2, 8.3 Hz, 1H), 7.45 (d,
J=8.3 Hz, 1H), 7.43 (s, 1H), 4.40 (q, J=7.2 Hz, 2H), 4.32 (d, J=
13.8 Hz, 1H), 4.19 (d, J=13.8 Hz, 1H), 2.78 (s, 3H), 1.39 (t, J=
7.2 Hz, 3H). 13C NMR (100 MHz, CDCl3) δ 162.7, 162.3, 147.1,
141.5, 138.6, 137.8, 134.3, 132.0, 130.8, 128.7, 126.5, 116.5, 113.9,
61.2, 48.1, 23.6, 14.2. Anal. Calcd for C17H14Cl2N6O2: C, 50.38;
H, 3.48; N, 20.73; Cl, 17.49; Found: C, 50.49; H, 3.58; N, 20.47;
Cl, 17.71. HRMS: Anal. Calcd for C17H15Cl2N6O2 405.0628;
Found: 405.0640 (M þ H)þ.
(þ)-6-(Azidomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidine-2-carboxylic Acid (23). To a stirred solution of
22 (1.00 g, 2.47 mmol) in THF (20 mL) and H2O (2 mL) was
Ethyl 6-(Azidomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidine-2-carboxylate (18). To a stirred solution of crude
17 (448 mg, 1.12 mmol) in DMF (5 mL) was added NaN3 (110 mg,
1.69 mmol). The reaction was heated to 50 ꢀC for 1 h. After cool-
ing, the reaction was diluted with EtOAc and extracted with H2O
and brine before drying over MgSO4. Filtration and concentration
under reduced pressure gave 18 (450 mg) as yellow oil. 1H NMR
(400 MHz, CDCl3) δ 7.72 (d, J=2.2 Hz, 1H), 7.59 (dd, J=2.2, 7.2
Hz, 1H), 7.43 (s, 1H), 7.43 (d, J=7.0 Hz, 1H), 4.41 (m, 2H), 4.31
(d, J=14.0 Hz, 1H), 4.18 (d, J=14.0 Hz, 1H), 2.79 (s, 3H), 1.40
(t, J=7.0 Hz, 3H). 13C NMR (125 MHz, CDCl3) δ 162.7, 162.3,
147.1, 141.5, 138.6, 137.7, 134.3, 132.0, 130.7, 128.7, 126.5, 116.5,
113.9, 61.2, 48.1, 23.6, 14.2. Anal. Calcd for C17H14Cl2N6O2: C,
50.38; H, 3.48; N, 20.73; Found: C, 50.33; H, 3.47; N, 20.66.
Ethyl 6-(Aminomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidine-2-carboxylate (19). To a stirred solution of 18
(450 mg, 1.11 mmol) in THF (10 mL) and H2O (0.5 mL) was added
PPh3 polymer bound (3 mmol/g, 750 mg, 2.25 mmol). The reac-
tion was heated to 50 ꢀC for 24 h. The reaction was filtered and
concentrated to give crude 19 (315 mg, 74% crude yield) as
yellow oil. A part (115 mg) was purified by reverse phase pre-
paratory HPLC to give 19, TFA salt (102 mg, 50% for 3 steps)
added LiOH H2O (155 mg, 3.70 mmol). After heating to 50 ꢀC
3
for 16 h, the reaction was concentrated under reduced pressure
and diluted with EtOAc. The aqueous layer was acidified by 1N
HCl to pH=1. The organic layer was washed with satd aq NH4Cl
and brine before drying over MgSO4. Filtration and concentra-
tion under reduced pressure provided 23 (874 mg, 94%, >99%
ee) as a white solid; [R]25D þ77.8ꢀ (c 2.38, MeOH). 1H NMR (400
MHz, CDCl3) δ 8.71 (br s, 1H), 7.67 (s, 1H), 7.59 (d, J=8.2 Hz,
1H), 7.55 (d, J=8.2 Hz, 1H), 7.51 (s, 1H), 4.33 (d, J=13.8 Hz,
1H), 4.24 (d, J = 13.8 Hz, 1H), 2.79 (s, 3H). 13C NMR (100
MHz, CDCl3) δ 164.1, 163.6, 146.5, 142.0, 138.8, 136.9, 134.2,
132.3, 130.7, 128.9, 126.2, 117.3, 114.4, 48.1, 23.6. Anal. Calcd
for C15H10Cl2N6O2 0.1EtOAc 0.7H2O: C, 46.40; H, 3.09; N,
3
3
21.08; Found: C, 46.21; H, 2.99; N, 20.93. HRMS: Anal. Calcd
for C15H11Cl2N6O2 377.0321; Found: 377.0313 (M þ H)þ.
(þ)-(6-(Azidomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidin-2-yl)(morpholino)methanone. To a stirred solu-
tion of 23 (150 mg, 0.400 mmol) in THF (10 mL) was added
morpholine (52.2 μL, 0.597 mmol), HOAt (81.3 mg, 0.597 mmol),
EDAC (114 mg, 0.597 mmol), and iPr2NEt (138 μL, 0.795 mmol).
The reaction was kept at RT for 2 h and was concentrated under
reduced pressure. The residue was diluted with EtOAc, and the
organic layer was washed with 1N HCl, 1N NaOH, and brine prior
to drying over anhydrous MgSO4. Filtration and concentration
under reduced pressure afforded the amide (177 mg, 99% crude
yield) as a clear glass. A small amount was purified by silica gel
chromatography (100% EtOAc) to afford the product (99% purity,
>99% ee); [R]25D þ82.7ꢀ (c 2.26, MeOH). 1H NMR (400 MHz,
CDCl3) δ 7.70 (d, J=2.2 Hz, 1H), 7.56 (dd, J=2.2, 8.2 Hz, 1H),
7.50 (s, 1H), 7.35 (d, J=8.2 Hz, 1H), 4.56 (m, 2H), 4.31 (d, J=
14.3 Hz, 1H), 4.18 (d, J=14.3 Hz, 1H), 3.78 (s, 6H), 2.81 (s, 3H).
13C NMR (100 MHz, CDCl3) δ 161.7, 161.7, 146.0, 141.9, 141.6,
138.7, 134.4, 131.9, 130.8, 128.7, 126.6, 116.2, 114.6, 67.4, 66.9,
48.1, 47.3, 43.2, 23.5. Anal. Calcd for C19H17Cl2N7O2: C, 51.13;
H, 3.83; N, 21.96; Found: C, 50.97; H, 3.57; N, 21.75.
(þ)-(6-(Aminomethyl)-5-(2,4-dichlorophenyl)-7-methylimidazo-
[1,2-a]pyrimidin-2-yl)(morpholino)methanone (24c). To a stirred
solution of the above amide (177 mg, 0.397 mmol) in THF (8 mL)
and H2O (0.8 mL) was added PPh3 polymer bound (3 mmol/g,
199 mg, 0.600 mmol) and heated to 50 ꢀC for 24 h. The reaction
was filtered, concentrated under reduced pressure, and purified
by reverse phase HPLC to give 24c, TFA salt (121 mg, 56% yield,
>99% ee) as a white solid; [R]24.3D þ54.12ꢀ (c 2.62, MeOH). 1H
NMR (400 MHz, CD3OD) δ 7.92 (d, J=2.2 Hz, 1H), 7.73 (dd,
J=2.2, 8.4 Hz, 1H), 7.62 (d, J=8.4 Hz, 1H), 7.55 (s, 1H), 4.19
(d, J=15.0 Hz, 1H), 4.01 (d, J=15.0 Hz, 1H), 4.01 (br s, 2H),
3.73 (br s, 6H), 2.86 (s, 3H). 13C NMR (100 MHz, CD3OD) δ
164.6, 164.5, 162 (TFA), 148.1, 145.3, 141.8, 140.5, 135.7, 133.8,
132.6, 130.8, 128.0, 120 (TFA), 117.1, 114.9, 68.2, 44.7, 37.8,
1
as a white solid. H NMR (400 MHz, CD3OD) δ 7.94 (d, J =
1.8 Hz, 1H), 7.75 (dd, J=1.8, 8.3 Hz, 1H), 7.64 (s, 1H), 7.63 (d,
J=8.3 Hz, 1H), 4.36 (q, J=7.0 Hz, 2H), 4.18 (d, J=14.9 Hz,
1H), 4.02 (d, J=14.9 Hz, 1H), 2.83 (s, 3H), 1.35 (t, J=7.0 Hz,
3H). HRMS: Anal. Calcd for C17H17Cl2N4O2 379.0729; Found:
379.0736 (M þ H)þ.
(-)- and (þ)-Ethyl 6-(aminomethyl)-5-(2,4-dichlorophenyl)-
7-methylimidazo[1,2-a]pyrimidine-2-carboxylate (20) and (21).
For 20 and 21, 50 mg of 19 was separated on a chiral OJ column
(15% B isocratic, 20 min, A=heptane, B=MeOH-EtOH (1:1)
with 0.1% DEA) to obtain 20 (15 mg, 86.5% ee) [R]24.6D -25.8ꢀ
(c 3.15, MeOH) and 21 (17 mg, 88.5% ee) [R]24.9D þ24.9ꢀ (c 3.11,
1
MeOH). For (-) isomer 20, H NMR (400 MHz, CD3OD) δ
7.93 (d, J=2.2 Hz, 1H), 7.74 (dd, J=2.2, 8.3 Hz, 1H), 7.64 (d,
J= 8.3 Hz, 1H), 7.62 (s, 1H), 4.36 (q, J=7.0 Hz, 2H), 4.18 (d,
J=14.9 Hz, 1H), 4.02 (d, J=14.9 Hz, 1H), 2.82 (s, 3H), 1.35 (t, J=
7.0 Hz, 3H). 13C NMR (100 MHz, CD3OD) δ 164.1, 163.3, 162
(TFA), 147.9, 144.2, 139.3, 137.8, 134.7, 133.0, 131.5, 129.9,
127.0, 120 (TFA), 116.7, 115.3, 61.8, 36.8, 23.6, 14.1. For (þ)
isomer 21, 1H NMR (400 MHz, CD3OD) δ 7.94 (d, J=2.2 Hz,
1H), 7.74 (dd, J=2.2, 8.4 Hz, 1H), 7.63 (d, J=8.4 Hz, 1H), 7.62
(s, 1H), 4.36 (q, J=7.5 Hz, 2H), 4.17 (d, J=14.5 Hz, 1H), 4.02
(d, J= 14.5 Hz, 1H), 2.82 (s, 3H), 1.35 (t, J= 7.5 Hz, 3H). 13
C
NMR (100 MHz, CD3OD) δ 163.3, 162.3, 147.5, 143.6, 138.7,
137.5, 134.0, 132.2, 130.9, 129.2, 126.2, 115.7, 114.6, 61.1, 36.2,
22.8, 13.3.
(þ)-Ethyl 6-(Azidomethyl)-5-(2,4-dichlorophenyl)-7-methyl-
imidazo[1,2-a]pyrimidine-2-carboxylate (22). 10.0 g of racemate
18 was separated by supercritical fluid chromatography on
Chiralpak OF, 250 mm ꢀ 20 mm, 10 μm, at 35 ꢀC, 50 mL/min;
mobile phase, CO2/MeOH/DEA, 65/35/0.1; injection volume,
3.5 mL of 33 mg/mL sample solution; detector wavelength,
220 nm to obtain (-)-ethyl 6-(azidomethyl)-5-(2,4-dichlorophenyl)-
7-methylimidazo[1,2-a]pyrimidine-2-carboxylate (5.00 g, 50%)
and 22 (5.00 g, 50%) as white solids. For (þ)-isomer 22: HPLC
Chiralpak OF, 4.6 mm ꢀ 250 mm, 10 μm; isocratic CO2/MeOH/
DEA, 60/40/0.1 over 25 min, 2.0 mL/min; tR=17.94 min; 100%
24.2. Anal. Calcd for C19H19Cl2N7O2 1.5TFA: C, 44.24; H,
3
3.45; N, 11.62; Found: C, 44.40; H, 3.72; N, 11.90.
In Vitro DPP4 Inhibition Assays. Inhibition of human DPP4
activity was measured under steady-state conditions by follow-
ing the absorbance increase at 405 nm upon the cleavage of
the pseudosubstrate, Gly-Pro-pNA. Assays were performed in
96-well plates using a Thermomax plate reader. Typical reac-
tions contained 100 μL of ATE buffer (100 mM Aces, 52 mM

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5627
Tris, 52 mM ethanolamine, pH 7.4), 0.45 nM enzyme, either 120
or 1000 μM of substrate (S < Km and S > Km, Km = 180 μM)
and 10 or 11 concentrations of the inhibitor. To ensure steady-
state conditions for slow-binding inhibitors, enzyme was pre-
incubated with the compound for 40 min prior to substrate
addition. All serial inhibitor dilutions were in DMSO and final
solvent concentration did not exceed 1%. Inhibitor potency was
evaluated by calculating IC50’s at each substrate concentration
and then converting them to Ki’s by assuming competitive
inhibition according to the equation Ki = IC50/[1 þ (S/Km)].
All inhibitors were competitive as judged by close agreement of
Ki values obtained from assays at high and low substrate con-
centrations. In cases where IC50 at the low substrate concentra-
tion was close to the enzyme concentration used in the assay, the
data were fit to the Morrison equation to account for the deple-
tion of the free inhibitor.
oGTT in ob/ob Mice. Male 13-14 week-old ob/ob mice
(Jackson Laboratories) were maintained under constant tempe-
rature and humidity conditions, a 12 h/12 h light-dark cycle,
and had free access to a 10% fat rodent diet (D1245B Research
Diets) and tap water. After an overnight fasting period, animals
were dosed orally with vehicle (water) or DPP4 inhibitor
(1, 3 μmol/kg) at -60 min. Two blood samples were collected
at -60 and 0 min by tail bleed for glucose and insulin determina-
tions. Glucose (2 g/kg) was then administered orally (at 0 min).
Additional blood samples were collected at 15, 30, 60, 90, and
120 min for glucose and insulin determinations. Blood samples
were collected into EDTA containing tubes (Sarstedt). Plasma
glucose was determined with an Accu-Chek Advantage (Roche)
glucometer. Plasma insulin was assayed using a mouse insulin
ELISA kit (ALPCO Diagnostics). Data represent the mean of
12-24 mice/group. Data analysis was performed using one way
ANOVA followed by Dunnett’s test. All procedures were per-
formed according to BMS-IACUC guidelines.
(4) (a) Gentilella, R.; Bianchi, C.; Rossi, A.; Rotella, C. M. Exenatide:
a review from pharmacology to clinical practice. Diabetes, Obes.
Metab. 2009, 11, 544–556. (b) Gonzalez, C.; Beruto, V.; Keller, G.;
Santoro, S.; Di Girolamo, G. Investigational treatments for type 2
diabetes mellitus: Exenatide and Liraglutide. Expert Opin. Invest.
Drugs 2006, 15, 887–895.
(5) Holst, J. J. Therapy of type 2 diabetes mellitus based on the actions
of glucagon-like peptide-1. Diabetes Metab. Res. Rev. 2002, 18,
430–441.
(6) Hansen, L.; Deacon, C. F.; Orskov, C.; Holst, J. J. Glucagon-like
peptide-1-(7-36)amide is transformed to glucagon-like peptide-
1-(9-36)amide by dipeptidyl peptidase IV in the capillaries supply-
ing the L cells of the porcine intestine. Endocrinology 1999, 140,
5356–5363.
(7) (a) Nagakura, T.; Yasuda, N.; Yamazaki, K.; Ikuta, H.; Yoshikawa,
S.; Asano, O.; Tanaka, I. Improved glucose tolerance via enhanced
glucose-dependent insulin secretion in dipeptidyl peptidase IV-deficient
Fischer rats. Biochem. Biophys. Res. Commun. 2001, 284, 501–506.
(b) Marguet, D.; Baggio, L.; Kobayashi, T.; Bernard, A.; Pierres, M.;
Nielsen, P. F.; Ribel, U.; Watanabe, 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, 6874–6879.
(8) (a) Durinx, C.; Lambeir, A.; Bosmans, E.; Falmagne, J.; Bergh-
mans, R.; Haemers, A.; Scharpe, S.; De Meester, I. Molecular
characterization of dipeptidyl peptidase activity in serum. Soluble
CD26/dipeptidyl peptidase IV is responsible for the release of X-Pro
dipeptides. Eur. J. Biochem. 2000, 267, 5608–5613. (b) Iwaki-Egawa, S.;
Watanabe, Y.; Kikuya, Y.; Fujimoto, Y. Dipeptidyl peptidase IV from human
serum: purification, characterization, and N-terminal amino acid sequence.
J. Biochem. 1998, 124, 428–433.
(9) Villhauer, E. B.; Brinkman, J. A.; Naderi, G. B.; Burkey, B. F.;
Dunning, B. E.; Prasad, K.; Mangold, B. L.; Russell, M. E.;
Hughes, T. E. 1-[[(3-Hydroxy-1-adamantyl)amino]acetyl]-2-cyano-
(S)-pyrrolidine: A potent, selective, and orally bioavailable dipep-
tidyl peptidase IV inhibitor with antihyperglycemic properties.
J. Med. Chem. 2003, 46, 2774–2789.
(10) 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-(trifluoro-
methyl)-5,6-dihydro[1,2,4]triazolo[4,3-a]pyrazin-7(8H)-yl]-1-(2,4,5-tri-
fluorophenyl)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.
(11) Augeri, D. J.; Robl, J. A.; Betebenner, D. A.; Magnin, D. R.;
Khanna, A.; Robertson, J. G.; Wang, A.; Simpkins, L. M.; Taunk,
P.; Huang, Q.; Han, S. P.; Abboa-Offei, B.; Cap, M.; Xin, L.; Tao,
L.; Tozzo, E.; Welzel, G. E.; Egan, D. M.; Marcinkeviciene, J.;
Chang, S. Y.; Biller, S. A.; Kirby, M. S.; Parker, R. A.; Hamann,
L. G. Discovery and preclinical profile of saxagliptin (BMS-477118):
a highly potent, long-acting, orally active dipeptidyl peptidase IV
inhibitor for the treatment of Type 2 diabetes. J. Med. Chem. 2005,
48, 5025–5037.
Acknowledgment. We thank the Department of Discovery
Analytical Sciences at BMS for their assistance in the char-
acterization of the compounds reported herein and for main-
tenance of various analytical instruments and the Department
of Pharmaceutical Candidate Optimization for performing
in vitro and in vivo assays to aid in the profiling of these
compounds.
Supporting Information Available: 1H and 13C NMR data,
and elemental analyses for compounds 24a, 24b, 24d-24i, and
24l-24ee. This material is available free of charge via the Internet
at http://pubs.acs.org.
(12) Feng, J.; Zhang, Z.; Wallace, M. B.; Stafford, J. A.; Kaldor, S. W.;
Kassel, D. B.; Navre, M.; Shi, L.; Skene, R. J.; Asakawa, T.;
Takeuchi, K.; Xu, R.; Webb, D. R.; Gwaltney, S. L. Discovery of
alogliptin: a potent, selective, bioavailable, and efficacious inhibi-
tor of dipeptidyl peptidase IV. J. Med. Chem. 2007, 50, 2297–2300.
(13) For other small molecular DPP4 inhibitors, see Havale, S. H.; Pal,
M. Medicinal chemistry approaches to the inhibition of dipeptidyl
peptidase-4 for the treatment of type 2 diabetes. Bioorg. Med.
Chem. 2009, 17, 1783–1802 and references cited therein.
(14) Brigance, R. P.; Meng, W.; Fura, A.; Harrity, T.; Wang, A.; Zahler,
R.; Kirby, M. S.; Hamman, L. G. Synthesis and SAR of azolopyr-
imidines as potent and selective dipeptidyl peptidase-4 (DPP4)
inhibitors for type 2 diabetes. Bioorg. Med. Chem. Lett. 2010, 20,
4395-4398.
(15) (a) Abignente, E.; Sacchi, A.; Laneri, S.; Rossi, F.; D’Amico, M.;
Berrino, L.; Calderaro, V.; Parrillo, C. Research on heterocyclic
compounds. XXXII. Synthesis and cyclooxygenase-independent
antiinflammatory and analgesic activity of imidazo[1,2-a]pyrimidine
derivatives. Eur. J. Med. Chem. 1994, 29, 279–286. (b) Commercon,
A. Preparation of bis(N-protected)-2-aminoimidazole-4-carboxalde-
hyde. Fr. Demande 1993, 12 pp. FR 2681323.
References
(1) (a) Wideman, R. D.; Kieffer, T. J. Mining incretin hormone path-
ways for novel therapies. Trends Endocrinol. Metab. 2009, 20, 280–
286. (b) Triplitt, C.; Wright, A.; Chiquette, E. Incretin mimetics and
dipeptidyl peptidase-IV inhibitors: potential new therapies for type 2
diabetes mellitus. Pharmacotherapy 2006, 26, 360–374. (c) Gallwitz,
B. Glucagon-like peptide-1-based therapies for the treatment of type 2
diabetes mellitus. Treat. Endocrinol. 2005, 4, 361–370.
(2) (a) Holst, J. J. Glucagon-like peptide-1 (GLP-1) a newly discovered
GI hormone. Gastroenterology 1994, 107, 1048–1055. (b) Drucker,
D. J. Glucagon-like peptides. Diabetes 1998, 47, 159–169.
(3) (a) Vilsboll, T.; Holst, J. J. Incretins, insulin secretion and type 2
diabetes mellitus. Diabetologia 2004, 47, 357–366. (b) Elahi, D.;
McAloon-Dyke, M.; Fukagawa, N. K.; Meneilly, G. S.; Sclater, A. L.;
Minaker, K. L.; Habener, J. F.; Andersen, D. K. The insulinotropic
actions of glucose-dependent insulinotropic polypeptide (GIP) and
glucagon-like peptide-1 (7-37) in normal and diabetic subjects. Regul.
Pept. 1994, 51, 63–75. (c) Nauck, M. A.; Bartels, E.; Oerskov, C.;
Ebert, R.; Creutzfeldt, W. Additive insulinotropic effects of exogenous
synthetic human gastric inhibitory polypeptide and glucagon-like pep-
tide-1-(7-36) amide infused at near-physiological insulinotropic hor-
mone and glucose concentrations. J. Clin. Endocrinol. Metab. 1993,
76, 912–917.
(16) Box, K.; Comer, J.; Huque, F. Correlations between PAMPA
permeability and log P. In Pharmacokinetic Profiling in Drug
Research: Biological, Physicochemical, And Computational Strate-
gies; Testa, B., Ed.; Verlag Helvetica Chimica Acta: Zurich, Switzerland,
2006; pp 243-257.
(17) (a) Abbott, C. A.; Yu, D. M. T.; Woollatt, E.; Sutherland, G. R.;
McCaughan, G. W.; Gorrell, M. D. Cloning, expression and

5628 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Meng et al.
chromosomal localization of a novel human dipeptidyl peptidase
(DPP) IV homolog, DPP8. Eur. J. Biochem. 2000, 267, 6140–6150.
(b) Bjelke, J., R.; Christensen, J.; Nielsen, P., F.; Branner, S.; Kanstrup,
A. B.; Wagtmann, N.; Rasmussen, H. B. Dipeptidyl peptidases 8 and 9:
specificity and molecular characterization compared with dipeptidyl
peptidase IV. Biochem. J. 2006, 396, 391–399.
and 9 inhibition in rodents revisited. Diabetes Obes. Metab. 2008,
10, 1057–1061. (b) Lankas, G. R.; Leiting, B.; Roy, R. S.; Eiermann,
G. J.; Beconi, M. G.; Biftu, T.; Chan, C. C.; Edmondson, S.; Feeney,
W. P.; He, H.; Ippolito, D. E.; Kim, D.; Lyons, K. A.; Ok, H. O.; Patel,
R. A.; Petrov, A. N.; Pryor, K. A.; Qian, X.; Reigle, L.; Woods, A.; Wu,
J. K.; Zaller, D.; Zhang, X.; Zhu, L.; Weber, A. E.; Thornberry, N. A.
Dipeptidyl peptidase IV inhibition for the treatment of type 2 diabetes.
Potential importance of selectivity over dipeptidyl peptidases 8 and 9.
Diabetes 2005, 54, 2988–2994.
(18) (a) Olsen, C.; Wagtmann, N. Identification and characterization of
human DPP9, a novel homologue of dipeptidyl peptidase IV. Gene
2002, 299, 185–193. (b) Ajami, K.; Abbott, C., A.; McCaughan, G., W.;
Gorrell, M., D. Dipeptidyl peptidase 9 has two forms, a broad tissue
distribution, cytoplasmic localization and DPIV-like peptidase activity.
Biochim. Biophys. Acta 2004, 1679, 18–28.
(22) (a) Aronov, A. M. Predictive in silico modeling for hERG channel
blockers. Drug Discovery Today 2005, 10, 149–155. (b) Song, M.;
Clark, M. Development and evaluation of an in silico model for hERG
binding. J. Chem. Inf. Model 2006, 46, 392–400. (c) Vaz, R. J.; Li, Y.;
Rampe, D. Human ether-a-go-go related gene (HERG): a chemist's
perspective. Prog. Med. Chem. 2005, 43, 1–18.
(19) (a) Maes, M. B.; Lambeir, A. M.; Gilany, K.; Senten, K.; Van der
veken, P.; Leiting, B.; Augustyns, K.; Scharpe, S.; De Meester, I.
Kinetic investigation of human dipeptidyl peptidase II (DPPII)-
mediated hydrolysis of dipeptide derivatives and its identification
as quiescent cell proline dipeptidase (QPP)/dipeptidyl peptidase
7 (DPP7). Biochem. J. 2005, 386, 315–324. (b) Araki, H.; Li, Y. H.;
Yamamoto, Y.; Haneda, M.; Nishi, K.; Kikkawa, R.; Ohkubo, I.
Purification, molecular cloning, and immunohistochemical localization
of dipeptidyl peptidase II from the rat kidney and its identity with
quiescent cell proline dipeptidase. J. Biochem. 2001, 129, 279–288.
(20) (a) Gorrell, M. D.; Yu, D. M. T. Diverse functions in a conserved
structure: the dipeptidyl peptidase IV gene family. In Trends in
Protein Research; Robinson, J. W., Ed.; Nova Science Publishers, Inc.:
Hauppauge, NY, 2005; pp 1-78. (b) Edosada, C. Y.; Quan, C.;
Wiesmann, C.; Tran, T.; Sutherlin, D.; Reynolds, M.; Elliott, J. M.;
Raab, H.; Fairbrother, W.; Wolf, B. B. Selective inhibition of fibroblast
activation protein protease based on dipeptide substrate specificity.
J. Biol. Chem. 2006, 281, 7437–7444.
(23) The X-ray crystal structure coordinates of 24c (BMS-762382) in
DPP4 are deposited in the Protein Data Bank (PDB code 3NOX).
(24) (a) Metzler, W. J.; Yanchunas, J.; Weigelt, C.; Kish, K.; Klei, H. E.;
Xie, D.; Zhang, Y.; Corbett, M.; Tamura, J. K.; He, B.; Hamann,
L. G.; Kirby, M. S.; Marcinkeviciene, J. Involvement of DPP-IV
catalytic residues in enzyme-saxagliptin complex formation. Pro-
tein Sci. 2008, 17, 240–250. (b) Thoma, R.; Loffler, B.; Stihle, M.;
Huber, W.; Ruf, A.; Hennig, M. Structural basis of proline-specific
exopeptidase activity as observed in human dipeptidyl peptidase-IV.
Structure 2003, 11, 947–959. (c) Engel, M.; Hoffmann, T.; Wagner, L.;
Wermann, M.; Heiser, U.; Kiefersauer, R.; Huber, R.; Bode, W.;
Demuth, H.-U.; Brandstetter, H. The crystal structure of dipeptidyl
peptidase IV (CD26) reveals its functional regulation and enzymatic
mechanism. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 5063–5068.
(25) Kirby, M.; Yu, D. M. T.; O’Connor, S. P.; Gorrell, M. D. Inhibitor
selectivity in the clinical application of dipeptidyl peptidase-4
inhibition. Clin. Sci. 2010, 118, 31–41.
(21) (a) Burkey, B. F.; Hoffmann, P. K.; Hassiepen, U.; Trappe, J.;
Juedes, M.; Foley, J. E. Adverse effects of dipeptidyl peptidases 8