1-((3S,4S)-4-Amino-1-(4-substituted-1,3,5-triazin-2-yl) pyrrolidin-3-yl)-5,5-difluoropiperidin-2-one inhibitors of DPP-4 for the treatment of type 2 diabetes

Article abstract of DOI:10.1016/j.bmcl.2011.01.055

A 3-amino-4-substituted pyrrolidine series of dipeptidyl peptidase IV (DPP-4) inhibitors was rapidly developed into a candidate series by identification of a polar valerolactam replacement for the lipophilic 2,4,5-trifluorophenyl pharmacophore. The additi

Full text of DOI:10.1016/j.bmcl.2011.01.055

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1810–1814
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
1-((3S,4S)-4-Amino-1-(4-substituted-1,3,5-triazin-2-yl) pyrrolidin-3-yl)-5,
5-difluoropiperidin-2-one inhibitors of DPP-4 for the treatment
of type 2 diabetes
⇑
Kim M. Andrews, David A. Beebe, John W. Benbow , David A. Boyer, Shawn D. Doran, Yu Hui, Shenping Liu,
⇑
R. Kirk McPherson, Constantin Neagu, Janice C. Parker, David W. Piotrowski , Steven R. Schneider,
Judith L. Treadway, Maria A. VanVolkenberg, William J. Zembrowski
Pfizer Global Research and Development, Groton Laboratories, 558 Eastern Point Road, Groton, CT 06340, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A 3-amino-4-substituted pyrrolidine series of dipeptidyl peptidase IV (DPP-4) inhibitors was rapidly
developed into a candidate series by identification of a polar valerolactam replacement for the lipophilic
2,4,5-trifluorophenyl pharmacophore. The addition of a gem-difluoro substituent to the lactam improved
overall DPP-4 inhibition and an efficient asymmetric route to 3,4-diaminopyrrolidines was developed.
Advanced profiling of a subset of analogs identified 5o with an acceptable human DPP-4 inhibition profile
based on a rat PK/PD model and a projected human dose that was suitable for clinical development.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 17 December 2010
Revised 13 January 2011
Accepted 14 January 2011
Available online 31 January 2011
Keywords:
Type 2 diabetes
Dipeptidyl peptidase
DPP-4
Type 2 diabetes mellitus (TTDM) is a chronic disorder character-
ized by hyperglycemia coupled with a gradual decline in insulin
sensitivity and insulin secretion. An increasing incidence of obesity
has, in part, fueled the rise in new TTDM diagnoses and continues
to make the treatment of TTDM a critical global health care issue.1
Present pharmaceutical therapy often fails to provide adequate
long-term glycemic control and many patients are unable to
achieve their targeted plasma glucose levels through available reg-
imens. The incretin hormones glucagon-like peptide-1 (GLP-1) and
gastrointestinal inhibitory peptide (GIP) are released post-pran-
dially from the L-cells of the intestine and stimulate the release
of insulin from pancreatic b-cells.2 However, these incretins are
rapidly degraded in vivo by peptidases. Dipeptidyl peptidase IV
(DPP-4) is a widely distributed serine protease that specifically
cleaves N-terminal dipeptides from polypeptides with proline or
alanine at the penultimate position. In vivo administration of
DPP-4 inhibitors to patients results in higher circulating concentra-
tions of endogenous GLP-1 and subsequent decrease in plasma
glucose. Long term treatment with a DPP-4 inhibitor leads to a
reduction in circulating HbA1c (glycosylated hemoglobin). It has
been suggested that DPP-4 inhibition may also offer the poten-
tial to improve the insulin producing function of the pancreas
through either b-cell preservation or regeneration.3 From a safety
standpoint, DPP-4 inhibition offers a significantly reduced risk of
hypoglycemia because the endogenous insulin secretagogue
(GLP-1) is generated only in response to a glucose stimulus. There-
fore, DPP-4 inhibition has emerged as a promising new treatment
of type 2 diabetes.4
Several DPP-4 inhibitors are now approved drugs, including
first in class inhibitor sitagliptin 1, as well as vildagliptin 2
and saxagliptin 3. Several other candidates such as gosogliptin
45 and others6 have been reported in advanced clinical trials
(Fig. 1). We were interested in finding additional novel small
molecular weight inhibitors of DPP-4 that lacked the 2-cyano-
pyrrolidide as a back-up to our clinical candidate 4. This account
describes the discovery of a pyrrolidine series of DPP-4 inhibitors
5 and the tactics used to manage pre-clinical safety risks to
identify a clinical candidate.
Based on emerging clinical data and PK/PD modeling, the labora-
tory objectives for the next generation effort was to provide P80%
inhibition of DPP-4 for at least 8 h in humans with a once-a-day oral
therapy. Candidate compounds would demonstrate >500-fold
selectivity against DPP-2 and DPP-8.7 The successful candidate
would also demonstrate a wide safety margin in in vitro safety as-
says: a therapeutic index versus the hERG patch-clamp assay for
QTc prolongation (hERG IC50/DPP-4 Ki P 300), a clean profile in
the in vitro genetic toxicity assays (Ames and micronucleus), and
minimal activity in a broad panel of receptors and enzymes. In addi-
⇑
Corresponding authors.
E-mail address: david.w.piotrowski@pfizer.com (D.W. Piotrowski).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.055

K. M. Andrews et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1810–1814
1811
N
N
F3C
O
N
F
N
N
N
H
N
CN
HO
N
HO
H2N
O
O
1
NH2
NC
F
2
3
F
N
N
N
N
N
N
F
O
( )n
R1
X
N
N
N
F
F
N
F
N
O
O
N
R3
R3
NH2
NH2
F
N
H
4
5
6
Figure 1. Approved DPP-4 inhibitors and advanced clinical candidates.
tion, physicochemical property space consistent with reducing
in vivo toxicological outcomes (c log P <3, TPSA >75) was targeted.8
The 2,4,5-trifluorophenyl moiety has been established as a
proline amide replacement in the S1 pocket of DPP-4 inhibitors.9
Earlier efforts from Pfizer laboratories identified potent, isoform-
selective DPP-4 inhibitors containing the 3-aminopyrrolidine core
(e.g., 6) as a suitable scaffold to explore substituents in the S2
showed that the more polar lactams were effective pharmacophores
for this template (Table 1). The butyrolactams (5b, 5e, 5h) were less
potent inhibitors of DPP-4 than the aryl congeners (5a, 5d, 5g) but
were significantly less lipophilic (
D
c log PÀ2.0). Addition of a meth-
ylene unit provided valerolactams (5c, 5f, 5i) that showed improved
DPP-4 inhibition. In this case, the 10-fold gain in potency in going
from butyrolactam to valerolactam corresponded to the gain in
binding energy from a small, optimally-buried lipophilic group.13
The reduction in c log P offered by the lactam moieties brought a sig-
nificant decrease in intrinsic clearance in human liver microsomes.
The co-crystal structure of compound 5c with human DPP-4
confirmed the expected binding mode in the active site. The
primary amine makes hydrogen bonds to Tyr662, Glu205, and
Glu206. The valerolactam occupies the S1 pocket (Tyr662,
Tyr666, Val656, Val711, Tyr631, Trp659) and the lactam carbonyl
make a hydrogen bond with Asn710. The phenyl extends into a
pocket comprised of Phe357, Arg358, Glu206, and Ser209. The
pocket.10 However, compounds like
c log P >1.2) than typical DPP-4 inhibitors (Fig. 1), which led to
6 were more lipophilic
(D
moderate to high intrinsic clearance in human microsomes. There-
fore, we sought a suitable polar replacement for the 2,4,5-trifluor-
ophenyl moiety that would result in a more efficient use of the
lipophilicity contained in the pharmacophore. Superposition of
crystal structures 4 and 6 showed that the carbonyl of pyrrolidine
amide of 4 and the 2-fluoro of 6 occupied the same space (Fig. 2).
We reasoned that incorporation of an N-linked lactam would offer
H-bonding potential through the carbonyl (to Asn710) and the
ability to lower log P. Recent literature examples have also shown
that a lactam moiety can function as a proline amide surrogate.11
Thus, we focused on optimization of more polar S1 pocket binding
groups.
pyrimidine ring is involved in a
p-stacking interaction with
Phe357 (Fig. 3).14
Using the valerolactam template 12, a range of substituents
(heterocycles, amide, sulfonamides; data not shown) was rapidly
probed using parallel chemistry. The 2,4-pyrimidine or triazine
heterocycles offered the best balance of ease of synthesis and
flexibility to explore a range of substituents. In general, analogs
made with these heterocycles resulted in high selectivity over
An initial synthetic route to the requisite (3S,4S)-diaminopyrr-
olidine core was fashioned from work reported by Still.12 The
commercially available unnatural tartaric acid 7 was readily
converted into the bis-azide 8. Mono-reduction under Staüdinger
conditions and reaction of the amine with an appropriate
c- or
the DPP-2 isoform (IC50 >30 lM) regardless of the substitution
d-chloro acid followed by cyclization generated the azido butyro-
or valerolactam, 9 or 10, respectively. Reduction of the azide and
protecting group manipulation afforded the pyrrolidine templates
11 or 12 for analog preparation.
A comparison of a small series of five- and six-membered lactam
analogs to the corresponding 2,4,5-trifluorophenyl compounds
pattern. However, for pyridinyl 5j or phenyl 5k analogs, the
DPP-4 inhibition and selectivity over DPP-8 remained sub-optimal.
As demonstrated by analogs in Table 1, amine substituents off of
the pyrimidine are suitable replacements for aryl moieties and
make more efficient use of the lipophilicity. Incorporation of 3,3-
difluoropyrrolidine as an aryl isostere provided analogs 5l and
5m that achieved modest potency gains and higher selectivity over
DPP-8 (Table 2).
The rigidity of the S1 pocket appears to leave little room for
additional substitution on the lactam moiety but subtle changes
of substituents in the S1 binding pocket can have a profound im-
pact on potency. Superposition of the crystal structure of 5c with
multiple crystal structures of DPP-4 inhibitors with fluorine atoms
occupying the S1 pocket (sitagliptin, gosogliptin, 6) showed some
unoccupied space, suggesting that additional interactions could
be achieved. Thus, incorporation of a gem-difluoro substitution at
C5 of the valerolactam moiety gave 5n, a compound that retained
selectivity over DPP-8 and improved inhibition of DPP-4 by about
threefold (Table 2). Further refinement of the heterocycle 5o or
substituent 5p led to additional analogs meeting many of our
objectives. Advanced screening of selected compounds in routine
in vitro safety assays (hERG patch-clamp, micronucleus and Ames
assays) did not uncover any issues with this series.
Figure 2. Superposition of compounds 4 (magenta) and 6 (cyan).

1812
K. M. Andrews et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1810–1814
Table 1
DPP-4 inhibition and ADME properties of N1-substituted 3-aminopyrrolidines 5
R1
N
N
N
H2N
R2
a
b
Compd
R1
R2
DPP-4 IC50 (nM)
c log P
TPSA
55
HLM CLint
(lL/min/mg)
A =
5a
13
3.47
85
F
O
N
B =
F
5b
488
1.43
75
<7.0
F
O
C =
5c
5d
5e
47
1.99
1.36
88
84
<7.2
56
N
A
B
1.6
153
O
N
À0.68
105
<7.0
N
5f
5g
C
A
12
19
À0.12
105
67
<7.5
38
2.76
5h
5i
B
C
1400
74
0.72
1.28
87
87
<7.0
<7.0
N
H
a
Values are means of three experiments.
Values show the intrinsic clearance of a compound as measured by the disappearance of the parent compound in cultured human liver microsomes.
b
triazine analogs 5m and 5o have a pKa2 that is three log units lower
and
D
c log PÀ0.8 than the corresponding pyrimidines.16 In order to
test whether these properties would have an impact on bioavail-
ability, 5l–5o were selected for pharmacokinetic (PK) profiling
(Table 3). Compounds 5m and 5o exhibited moderate clearance,
moderate volume of distribution, moderate half-life and good bio-
availability, while 5l and 5n were characterized by higher clear-
ance and shorter half-life. Based on the desirable PK properties in
rat, 5o was advanced to PK profiling in dog, where moderate clear-
ance, moderate volume of distribution, moderate half-life and good
bioavailability were again noted. Of significance, a twofold increase
in bioavailability for triazines versus pyrimidine (cf. 5l vs 5m, 5n vs
5o), which was consistent with log P mediated clearance reduction,
was observed. An ex vivo measurement of DPP-4 inhibition from
these rat PK studies (5 mg/kg, po) indicated that 5l–5m were capa-
ble of inhibiting DPP-4 at the IC50 level for at least 16 h.17
Compound 5o was assessed in an advanced battery of in vitro
assays (Table 4). Of note was the high selectivity over closely re-
lated serine proteases, P450 enzymes and a wide panel of enzymes,
receptors and ion channels. On the basis of desirable drug-like
properties, excellent PK properties and expected PD effect in ro-
dents, a human dose projection from modeling was conducted
for 5o. The human dose for 5o was predicted by PK–PD simulation
utilizing the projected human pharmacokinetic parameters and
the associated DPP-4 inhibition.18 Modeling suggested that a
100 mg dose of 5o would provide >80% inhibition of DPP-4 for
20 h in humans.
With a suitable compound in hand, improvements to the syn-
thetic route for the diaminopyrrolidine core were undertaken
(Scheme 2). Specifically, the azide intermediates in the initial
synthetic route are generally not acceptable in a development pro-
cess route.19 Compound 13, synthesized as described in Scheme 1,
was treated with Burgess reagent to affect cyclization to sulfamate
Figure 3. Binding interactions in the active site. Compound 5c co-crystallized with
human DPP-4.
Operating in higher probability safety space (c log P <3, TPSA >
75) not only brings in vivo safety benefits but also increases prob-
ability of obtaining compounds with higher solubility and lower
intrinsic clearance.15 We found that a careful balance of intrinsic
permeability and ionization state is needed to maximize the frac-
tion of the dose absorbed. In general, compounds in Table 2 had
moderate permeability when assessed in an RRCK (low efflux
MDCK) assay. However, 3,3-difluoropyrrolidinopyrimidine analogs
5l, 5n, and 5p have a second basic center (pKa2 ꢀ7.6) that effec-
tively renders them partially dibasic at physiological pH. The

K. M. Andrews et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1810–1814
1813
Table 2
DPP-4 inhibition and membrane permeability of N-substituted (3S,4S)-3-amino-4-(valerolactam-N-yl)pyrrolidines
R1
N
R2
X
N
N
O
H2N
N
R3
R3
a
b
Compd
5j
R1
R2
H
R3
H
X
DPP-4 IC50 (nM)
DPP-8 IC50 (nM)
c log P
1.43
TPSA
98
pKa2
3.03
2.6
RRCKc Papp AB (10À6 cm/s)
N
CH
39
11,400
2.7
MeO
5k
H
H
N
59
26,000
1.11
88
35.7
5l
5m
H
H
H
H
CH
N
26
37
>30,000
>30,000
1.20
0.39
79
92
7.58
4.26
1.8
3.8
F
F
5n
H
F
CH
8.4
>30,000
1.48
79
7.58
3.4
N
5o
5p
H
CH3
F
F
N
CH
23
13
9
>30,000
>30,000
0.67
1.98
92
79
4.26
7.59
5.0
3.9
a
b
c
Values are means of three experiments.
Calculated pKa from ACD version 9.0.
RRCK permeability assessment is conducted with low efflux Madin-Darby canine kidney (MDCK) cells, where the cells are double-FACS sorted to isolate the cells that
showed no P-gp/MDR functional expression.
14.20 Opening of sulfamate 14 with potassium phthalimide and
subsequent loss of sulfinic acid provided a differentiated diamine
substrate that was treated with hydrazine hydrate to afford amin-
ocarbamate 15. Acylation of the amine with 5-chloro-4,4-difluoro-
pentanoic acid21 followed by base-induced cyclization and benzyl
deprotection provided key pyrrolidine intermediate 16. Displace-
ment of the appropriately substituted chloro-heterocycle and
acidic deprotection of the methyl carbamate afforded the desired
target 5o.
In conclusion, a 3-amino-4-substituted pyrrolidine series of
DPP-4 inhibitors was rapidly developed into a candidate series. A
more polar valerolactam replacement for the lipophilic 2,4,5-triflu-
orophenyl pharmacophore was identified and the addition of a
gem-difluoro substituent to the lactam improved overall DPP-4
Table 3
Pharmacokinetic profiles of selected DPP-4 inhibitorsa
b
c
d
Compd Species CLp (ml/min/kg) Vdss (L/kg) t1/2 (h) F (%) fu
5l
Rat
Rat
Rat
Rat
Dog
60.1
36.8
84
46
7.2
2.3
4.7
5.7
8.4
1.9
0.78
5.0
2.4
3.6
5.9
21
52
50
90
150
0.75
0.86
0.82
0.83e
5m
5n
5o
a
Rats (n = 2) were dosed with 1 mg/kg (iv) and 5 mg/kg (po) in saline and 5%
methyl cellulose, respectively.
b
Clearance of compound from plasma.
Volume of distribution calculated from the iv portion of the experiment.
The unbound fraction of drug was measured using fresh plasma.
The value measured in human plasma was 0.83.
c
d
e
Ph
N3
Ph
N
Table 4
Pharmacological and ADME properties of 5o
OH
N
N
e - g
a - d
CO2H
5oa
HO2C
Human Plasma DPP-4 Kib
14.4 1.6 nM (4)
2.1 0.3 nM (5)
>1000Â
O
OH
N3
N3
Rat Plasma DPP-4 Kib
Selectivityc
8
( )n
7
DPP-2
DPP-3
DPP-9
FAP
>30,000 nM
>30,000 nM
>30,000 nM
>30,000 nM
>30,000 nM
9 n = 0
10 n = 1
R1
N
N
N
H
N
POP
Projected Human CLd
2.9 mL/min/kg
1.5 L/kg
80
h - j
k, l
e
Projected Human Vd
O
Projected F%f
Boc
NH
N
CEREP broad receptor and enzyme profile
(68 assays)
CYP1A2, 2C9, 2D6, 3A4
IC50 >10
l
M
( )n
O
H2N
N
IC50 >15
l
M
11 n = 0
12
( )n
a
Values are means of at least three experiments.
Data are means SEM (n). Ki values were calculated using the Cheng–Prusoff
5
n = 1
b
equation.
Scheme 1. Chiral pool route to (3S,4S) 3-amino-4-(lactam-N-yl) pyrrolidine.
Reagent and conditions: (a) BnNH2, PhCH3; (b) Red-Al, THF, 0 °C; (c) MsCl, Et3N,
CH2Cl2; (d) NaN3, DMF, 100 °C; (e) PPh3, THF/H2O, heat; (f) Cl(CH2)3(CH2)n–CO2H.
propanephosphonic acid cyclic anhydride, Et3N, EtOAc, heat; (g) NaH, DMF;
(h) PPh3, THF/H2O, heat; (i) Boc2O, THF; (j) H2 (75 psi), 10% Pd/C, AcOH, EtOAc/
EtOH, 50 °C; (k) Het-Cl, t-BuOH, heat; (l) TFA, CH2Cl2; NaOH, H2O.
c
DPP-2, DPP-3, DPP-9, FAP, POP.9
d
Predicted by dog single species scaling and scaling of human in vitro microsome
data.
e
Projected by dog–human proportionality.
Projected value based on pre-clinical F%.
f

1814
K. M. Andrews et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1810–1814
5. Ammirati, M. J.; Andrews, K. M.; Boyer, D. A.; Brodeur, A. M.; Danley, D. E.;
Ph
O
Ph
Ph
Doran, S. D.; Hulin, B.; Liu, S.; McPherson, R. K.; Orena, S. J.; Parker, J. C.;
Polivkova, J.; Qiu, X.; Soglia, C. B.; Treadway, J. L.; VanVolkenburg, M. A.; Wilder,
D. C.; Piotrowski, D. W. Bioorg. Med. Chem. Lett. 2009, 19, 1991.
6. For alogliptin, see: Feng, J.; Zhang, Z.; Wallace, M. B., et al J. Med. Chem. 2007, 50,
2297; For linagliptin, see: Wang, Y.; Serradell, N.; Rosa, E.; Castaner, R. Drugs
Future 2008, 33, 473.
N
S
O
14
N
N
a
b - d
O
O
N
NH NH2
HO
OH
O
O
7. Some reports have suggested that inhibition of DPP-8/9 is associated with a
higher risk of toxicity in animal models, see: Lankas, G. R.; Leiting, B.; Roy, R. S.;
Eiermann, G. J.; Beconi, M. G.; Biftu, T., et al Diabetes 2005, 54, 2988; However,
more recently this assertion has been challenged, see: Wu, J.-J.; Tang, H.-K.;
Yeh, T.-K.; Chen, C.-M.; Shy, H.-S.; Chu, Y.-R.; Chien, C.-H.; Tsai, T.-Y.; Huang, Y.-
C.; Huang, Y.-L.; Huang, C.-H.; Tseng, H.-Y.; Jiaang, W.-T.; Chao, Y.-S.; Chen, X.
Biochem. Pharmacol. 2009, 78, 203. We desired to maximize selectivity over
DPP-8..
8. (a) Price, D. A.; Blagg, J.; Jones, L.; Greene, N.; Wager, T. Expert Opin. Drug Metab.
Toxicol. 2009, 5, 921; (b) Hughes, J. D.; Blagg, J.; Price, D. A.; Bailey, S.;
DeCrescenzo, G. A.; Devraj, R. V.; Ellsworth, E.; Fobian, Y. M.; Gibbs, E. M.;
Gilles, R. W.; Greene, N.; Huang, E.; Krieger-Burke, T.; Loesel, J.; Wager, T.;
Whiteley, L.; Zhang, Y. Bioorg. Med. Chem. Lett. 2008, 18, 4872.
9. Benbow, J. W.; Andrews, K. A.; Aubrecht, J.; Beebe, D.; Boyer, D.; Doran, S.;
Homiski, M.; Hui, Y.; McPherson, K.; Parker, J. C.; Treadway, J.; VanVolkenberg,
M.; Zembrowski, W. J. Bioorg. Med. Chem. Lett. 2009, 19, 2220.
O
15
13
R1
N
N
R2
X
N
N
H
N
e - g
h, i
O
O
O
NH
H2N
N
O
F
F
F
F
16
5
10. (a) Wright, S. W.; Ammirati, M. J.; Andrews, K. M.; Brodeur, A. M.; Danley, D. E.;
Doran, S. D.; Lillquist, J. S.; Liu, S.; McClure, L. D.; McPherson, R. K.; Olson, T. V.;
Orena, S. J.; Parker, J. C.; Rocke, B. N.; Soeller, W. C.; Soglia, C. B.; Treadway, J. T.;
VanVolkenburg, M. A.; Zhao, Z.; Cox, E. D. Bioorg. Med. Chem. Lett. 2007, 17,
5638; (b) Corbett, J. W.; Dirico, K.; Song, W.; Boscoe, B. P.; Doran, S. D.; Boyer,
D.; Qiu, X.; Ammirati, M.; VanVolkenburg, M. A.; McPherson, R. K.; Parker, J. C.;
Cox, E. D. Bioorg. Med. Chem. Lett. 2007, 17, 6707.
11. (a) Liang, G.-B.; Qian, X.; Biftu, T.; Singh, S.; Gao, Y.-D.; Scapin, G.; Patel, S.;
Leiting, B.; Patel, R.; Wu, J.; Zhang, X.; Thornberry, N. A.; Weber, A. E. Bioorg.
Med. Chem. Lett. 2008, 18, 3706; (b) Benbow, J. W.; Piotrowski, D. W.; Hui, Y.
PCT Int. Appl., WO2007148185, 2007.; (c) Mattei, P.; Boehringer, M.; Di Giorgio,
P.; Fischer, H.; Hennig, M.; Huwyler, J.; Koçer, B.; Kuhn, B.; Loeffler, B. M.;
MacDonald, A.; Narquizian, R.; Rauber, E.; Sebokova, E.; Sprecher, U. Bioorg.
Med. Chem. Lett. 2010, 20, 1109.
12. (a) Weingarten, M. D.; Sekanina, K.; Still, W. C. J. Am. Chem. Soc., 1998, 120,
9112, Supplementary data.; (b) Rejman, D.; Kocalka, P.; Budesinsky, M.; Pohl,
R.; Rosenberg, I. Tetrahedron 2007, 63, 1243.
13. Pace, C. N. J. Mol. Biol. 1992, 226, 29.
14. The structure has been deposited in the Protein Data Bank with the deposition
code 3QBJ.
Scheme 2. The second generation asymmetric route to (3S,4S) 3-amino-4-(lactam-
N-yl) pyrrolidine Reagent and conditions: (a) Burgess reagent, dioxane, heat; (b) K-
phthalimide, CH3CN, 80 °C; (c) 2 N HCl; (d) N2H4ÁH2O, EtOH, heat; (e) HO2C(CH2)3
CF2CH2Cl, propanephosphonic acid cyclic anhydride, Et3N, EtOAc, heat; (f) NaH,
THF/DMF; (g) H2 (75 psi), Pd(OH)2/C, EtOH/AcOH, 50 °C; h) Het-Cl, DIPEA, t-BuOH,
heat; (i) PPA, heat.
inhibition. The PK properties were enhanced by lowering the basi-
city and lipophilicity of the core heterocycle. Advanced profiling in
a rat PK/PD model and subsequent human projections identified 5o
as having an acceptable human DPP-4 inhibition profile with a pro-
jected dose of 100 mg/q.d. that was suitable for a drug develop-
ment candidate. A non-azide, asymmetric route to the candidate
compound was also discovered.
15. Compounds 5l, 5m, 5n and 5o all had HLM CLint <7.0 mL/min/kg and RLM CLint
<28.0 mL/min/kg. The crystalline free base of 5o had a solubility of 2.6 mgA/mL
in phosphate buffered saline (pHfinal 7.0) and a determined HLM CLint clearance
value of 3.1 mL/min/kg.
16. See: Mylari, B. L.; Withbroe, G. J.; Beebe, D. A.; Brackett, N. S.; Conn, E. L.;
Coutcher, J. B.; Oates, P. J.; Zembrowski, W. J. Bioorg. Med. Chem. 2003, 11, 4179.
for permeability/absorption benefits by replacement of a pyrimidine with a
triazine.
Acknowledgments
The authors thank Ben Rocke and Zhengrong Zhao for the
preparation of compounds 5d and 5g, respectively and Drs. David
Hepworth, Allyn Londregan, and Kentaro Futatsugi for reviewing
the manuscript.
17. The rat pharmacodynamic assay is described in: Kim, D.; Wang, L.; Beconi, M.;
Eiermann, G. J.; Fisher, M. H.; He, H.; Hickey, G. J.; Kowalchick, J. E.; Leiting, B.;
Lyons, K.; Marsilio, F.; McCann, M. E.; Patel, R. A.; Petrov, A.; Scapin, G.; Patel, S.
B.; Sinha Roy, R.; Wu, J. K.; Wyvratt, M. J.; Zhang, B. B.; Zhu, L.; Thornberry, N.
A.; Weber, A. E. J. Med. Chem. 2005, 48, 141.
18. Human PK parameters were used to generate predicted plasma concentration
versus time profiles at varying doses of 5o (WinNonlin version 5.2, Pharsight
Corp.). The projected human plasma concentration data was then used to
predict the associated DPP-4 inhibition using the equation [i = 1/(1 + Ki/I)],
where ‘i’ is DPP-4 inhibition, ‘Ki’ is the value calculated from the in vitro IC50
value, and ‘I’ is the unbound plasma concentration.
19. Differential scanning calorimetry run on diazide intermediate 8 indicated a
moderate to high thermal energy release that was within 100 °C of the
processing temperature.
20. Nicolaou, K. C.; Snyder, S. A.; Longbottom, D. A.; Nalbandian, A. Z.; Huang, X.
Chem. Eur. J. 2004, 10, 5581.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.01.055. These data
include MOL files and InChiKeys of the most important compounds
described in this article.
References and notes
1. (a) Flier, J. S. Cell 2004, 116, 337; (b) Hogan, P.; Dall, T.; Nikolov, P. Diabetes Care
2003, 26, 917.
2. Flatt, P. R.; Bailey, C. J.; Green, B. D. Front. Biosci. 2008, 13, 3648.
3. Drucker, D. J. Endocrinology 2003, 144, 5145.
4. For comprehensive reviews on DPP-4 inhibitors, see: (a) Idris, I.; Donnelly, R.
Diabetes Obes. Metab. 2007, 9, 153; (b) von Geldern, T. W.; Trevillyan, J. M. Drug
Dev. Res. 2006, 67, 627; (c) Weber, A. E. J. Med. Chem. 2004, 47, 4135.
21. Ethyl 4-chloro-4-oxobutyrate was treated with TMS-diazomethane to afford
the alpha-chloroketone. Conversion of the ketone into the gem-difluoro was
accomplished using the DAST reagent and saponification (LiOH/H2O) afforded
the desired acid.