Design and synthesis of DPP-4 inhibitor for the treatment of type 2 diabetes

Article abstract of DOI:10.1016/j.cclet.2009.11.014

An efficient stereoselective synthesis of the rigid aza-bicyclo[3.2.0]heptane scaffold has been developed to provide 2-cyano pyrrolidine alpha-amino amide 1 as DPP-4 inhibitor.

Full text of DOI:10.1016/j.cclet.2009.11.014

Available online at www.sciencedirect.com
Chinese Chemical Letters 21 (2010) 253–256
www.elsevier.com/locate/cclet
Design and synthesis of DPP-4 inhibitor for the treatment
of type 2 diabetes
*
Peng Cho Tang , Zhi Gang Lin, Yang Wang, Fang Long Yang, Qian Wang,
Jian Hong Fu, Lei Zhang, Ai Shen Gong, Jing Jing Luo, Jun Dai, Gao Hong She,
Dan Dan Si, Jun Feng
Shanghai Hengrui Pharmaceuticals Co. Ltd., 279 Wenjing Road, Shanghai 200245, China
Received 31 August 2009
Abstract
An efficient stereoselective synthesis of the rigid aza-bicyclo[3.2.0]heptane scaffold has been developed to provide 2-cyano
pyrrolidine alpha-amino amide 1 as DPP-4 inhibitor.
# 2009 Peng Cho Tang. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
Keywords: DPP-4 inhibitor; Aza-bicyclo[3.2.0]heptane scaffold; 2-Cyano pyrrolidine alpha-amino amide derivative; Synthesis; Type 2 diabetes
Type 2 diabetes mellitus may be effectively treated by agents that induce the biosynthesis and secretion of insulin
during periods of hyperglycemia. Two endogenous peptides that stimulate glucose-dependent insulin secretion are the
incretin hormones glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) [1].
Continuous infusion of GLP-1 in patients with type 2 diabetes has resulted in significant decreases in plasma glucose
and hemoglobin A1c levels [2]. However, active GLP-1 is rapidly converted to inactive GLP-1 by the serine protease
dipeptidyl peptidase IV (DPP-4), thus limiting its therapeutic practicality. Inhibition of DPP-4 increases the level of
endogenous intact GLP-1. Consequently, inhibition of DPP-4 is rapidly emerging as a novel therapeutic approach for
the treatment of type 2 diabetes [3,4]. Saxagliptin [5] (Onglyza1, BMS-477118) was recently approved by US FDA to
be marketed for the treatment of type 2 diabetes (Fig. 1). The 1,6 spatial relationship of the NH2 and CN group is
required for inhibition for this class of compound, but also favors the intramolecular cyclization to form the rearranged
compound 3 which is devoid of inhibitory activity [6]. The presence of the cyclopropyl group on the pyrrolidine ring,
as a steric barrier, has been shown to discourage the cyclization and therefore increases the stability and half life of
BMS-477118 [7]. We reasoned the cyclobutyl group would be sterically as effective as the cyclopropyl group and the
introduction of the gem-difluoro group, which otherwise was much more difficult to be introduced in cyclopropyl
group, would further provide electronic repulsion to the incoming nitrogen nucleophile which may result in longer
acting. Herein, we described the synthesis and biological activities of the novel difluorocyclobutyl pyrrolidine
derivative 1 (Fig. 1).
* Corresponding author.
E-mail address: tangpc@shhrp.com (P.C. Tang).
1001-8417/$ – see front matter # 2009 Peng Cho Tang. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
doi:10.1016/j.cclet.2009.11.014

254
P.C. Tang et al. / Chinese Chemical Letters 21 (2010) 253–256
1. Experimental
Compound 1 was prepared from difluorocyclobutyl pyrrolidine amide 4 and adamantylglycine 5 [7] (Fig. 1).
As shown in Scheme 1, anti-[2+2] cycloaddition of the enantiomerically pure enecarbamate 6, which derived from
ethyl Boc-D-pyroglutamate [8], with dichloroketene generated in situ occurred to provide the desired product 7 in 65%
yield to set up the stereochemistry at the ring junction of the bicyclic core. Reductive dechlorination of 7 proceeded
cleanly with Zn dust in the presence of ammonium chloride to provide the corresponding azabicyclic cyclobutanone 8
in 70% yield. Treatment of 8 with DAST gave the difluoromethylene derivative 9. As anticipated, conformational
inversion of the carboxylic acid ethyl ester in 9 employing LDA occurred cleanly in good yield (10:9 = 9:1). After
hydrolysis and amidation, compound 4 was obtained, which coupled with adamantylglycine 5 [7] under HATU
conditions. Dehydration of amide 11 with TFAA, followed by deprotection of the N-terminus, gave the target
compound 1 as its TFA salt [9].
2. Results and discussion
Inhibitory selectivity for DPP-4 over DPP-8 and DPP-9 was particularly emphasized. The inhibition of the latter
two enzymes may be associated with profound toxicity [10,11]. The data for DPP-4, DPP-8 and DPP-9 inhibitions
were summarized in Table 1. Compound 1 exhibited excellent potency against DPP-4 and high selectivity over DPP-8
and DPP-9. The IC50 of compound 1 was 27 nmol/L. The DPP-4 IC50/DPP-8 IC50 ratio was 99-fold compared with
BMS-477118 which ratio was 28-fold. At the same time, selectivity against DPP-9 was also observed.
In conclusion, an efficient stereoselective synthesis of the rigid aza-bicyclo[3.2.0]heptane scaffold was developed
starting from compound 6 which led to the synthesis of new DPP-4 inhibitor 1. Despite the potency and favorable
selectivity, development of compound 1 has to await more in-depth evaluation of related analogs because of its
moderate pharmacokinetic profile. With an efficient synthesis in hand, further medicinal chemistry efforts focusing on
chemical diversification of intermediates such as 8, 9 and 10 that possess multiple diversifiable functional groups to
identify new DPP-4 inhibitors meeting the target profiles concerning the stability, selectivity, pharmacokinetics and
efficacy are in progress (Table 2).
Acknowledgments
The authors thank Mr. Sun Piao Yang for his vision and support in new drug discovery in China, Dr. Wang Wei Bo
for ideas and suggestions on the initial phase of the synthesis, and Mr. Wu Qian of Analytical Chemistry Group of
1
Shanghai Hengrui Pharmaceuticals Co., Ltd. for H NMR and mass spectroscopy services.
Fig. 1. Rational design of compound 1.

P.C. Tang et al. / Chinese Chemical Letters 21 (2010) 253–256
255
Scheme 1. Reagents and conditions: (a) 1.5 equiv. Cl2CHCOCl, 2.0 equiv. Et3N, 40 8C, 65%; (b) 11.0 equiv. Zn dust, MeOH (sat. NH4Cl aqueous),
rt, 70%; (c) 2.5 equiv. DAST, CH2Cl2, 0 8C ! 40 8C, 57%; (d) 1.5 equiv. LDA, 2.0 equiv. i-PrOH, THF, ꢀ78 8C, 70%; (e) 3.0 equiv. LiOH, EtOH/
H2O, rt, 100%; (f) 1.2 equiv. EDCI, 1.2 equiv. HOBt, 1.5 equiv. DIPEA, 3.0 equiv. (NH4)2CO3, THF, rt, 85%; (g) HCl, ethyl acetate, rt; (h) 1.0 equiv.
HATU, 5, 2.0 equiv. DIPEA, DMF, rt, 72%; (i) 2.5 equiv. TFAA, 5.0 equiv. pyridine, THF, 0 8C ! rt, 50%; (j) TFA, CH2Cl2, rt, 98%.
Table 1
Potency and selectivity of compound 1.
Compound
IC50 (mmol/L)
DPP-4
DPP-8
DPP-9
1
BMS-477118
0.027
0.019
2.66
0.54
1.66
0.307
Table 2
Pharmacokinetic parameters of 1 (mean ꢁ SD)a.
Parameters
tmax (h)
Cmax (ng/mL)
t1/2 (h)
AUC 0-t (ng/mL h)
Vz/F (l/kg)
CLz/F (l/h/kg)
1
0.32 ꢁ 0.17
Oral administrations of 3.0 mg/kg to Sprague–Dawley rats.
270 ꢁ 144
0.92 ꢁ 0.21
347 ꢁ 200
14.4 ꢁ 7.46
11.7 ꢁ 6.74
a
References
[1] J.F. Gautier, S. Fetita, E. Sobngwi, et al. Diabetes Metab. 31 (2005) 233.
[2] M. Zander, S. Madsbad, J.L. Madsen, Lancet 359 (2002) 824.
[3] S.L. Gwaltney, J.A. Stafford, Annu. Rep. Med. Chem. 40 (2005) 149.
[4] E.B. Villhauer, J.A. Brinkman, G.B. Naderi, et al. J. Med. Chem. 46 (2003) 2774.
[5] J.U. Peters, Curr. Top. Med. Chem. 7 (2007) 579.

256
P.C. Tang et al. / Chinese Chemical Letters 21 (2010) 253–256
[6] D.R. Magnin, J.A. Robl, R.B. Sulsky, et al. J. Med. Chem. 47 (2004) 2587.
[7] D.J. Augeri, J.A. Robl, D.A. Betebenner, et al. J. Med. Chem. 48 (2005) 5025.
[8] L.A. Joa˜o Carlos, A.S. Regina Helena de, R.D.C. Carlos, J. Braz. Chem. Soc. 14 (2003) 27.
[9] All compounds were characterized by 1H NMR and MS etc. Data for compound 1 (TFA salt): pale yellow solid, 1H NMR (DMSO-d6,
400 MHz): d 1.06–1.55 (m, 14H), 2.16 (s, 2H), 2.42–2.49 (m, 1H), 2.56–2.78 (m, 1H), 3.19–3.23 (m, 1H), 3.89 (m, 1H), 3.99 (s, 1H),
4.94 (s, 1H), 5.40–5.43 (m, 1H), 8.26 (s, 3H). MS (ESI) m/z: 539 [M+1]+ 366.4.
[10] G.R. Lankas, B. Leiting, R.S. Roy, et al. Diabetes 54 (2005) 2988.
[11] J.S. Rosenblum, J.W. Kozarich, Curr. Opin. Chem. Biol. 7 (2003) 496.