Synthesis and biological evaluation of glucagon-like peptide-1 receptor agonists

In this study, a series of fused-heterocyclic derivatives were systematically designed and synthesized using an efficient route, and evaluated in terms of GLP-1R agonist activity. We employed short synthetic steps and reactions that are tolerant of the presence of various functional groups and suitable for parallel operations to enable the rapid generation of libraries of diverse and structurally complex small molecules. Of the compounds synthesized, 3-(8-chloro-6-(trifluoromethyl)imidazo[1,2-a] pyridin-2-yl)phenyl methanesulfonate (8e) was the most potent agonist with an EC50 of 7.89 μM, and thus is the compound with the greatest potential for application. These findings represent a valuable starting point for the design and discovery of small-molecule GLP-1R agonists that can be administered orally. Electronic supplementary material The online version of this article (doi:10.1007/s12272-013-0253-9) contains supplementary material, which is available to authorized users.


Introduction
Type 2 diabetes mellitus (DM2), a state of hormonal disruption and incretin deficiency, is increasingly becoming a worldwide epidemic (Kwak and Ha 2013). Current drugs utilized in the treatment of DM2 have well-established shortcomings: (1) increasing body weight and (2) increasing loss of b-cell function (Whitehouse 1997;Giugliano et al. 2009). However, the recent emergence of incretinbased therapies, which focus on glucagon-like peptide-1 (GLP-1), has attracted much interest.
GLP-1 is a peptide hormone of 30 amino acid residues. As a peptide, it has a very short half-life (\2 min) (Deacon et al. 1995). Such a short half-life has limited the utility of native GLP-1 in the treatment of DM2. The effort to identify GLP-1 analogues has resulted in the development of the drugs exenatide (Sennik et al. 2011;Buse et al. 2004) and liraglutide (Sjöholm 2010;Hribal and Sesti 2010). However, the requirement for injection limits the clinical utility of these peptide drugs. Therefore, orally active, small-molecule agonists of the GLP-1 receptor (GLP-1R) are highly sought after (Murphy and Bloom 2007). Figure 1 shows synthetic small molecule agonists reported by several groups (Teng et al. 2000;Wang et al. 2009;Teng et al. 2007;Kopin 2004;Gong et al. 2010). Compound 6b, characterized by a novel imidazopyridine hit core, was identified from a library of 10,000 heterocyclic small molecules (Gong et al. 2010). As a small and drug-like active molecule, it represents an interesting starting point for the development of novel drugs. Therefore, we selected this compound as a model. In an effort to move away from the labile ester group of the phenol, we planned a synthetic pathway of new derivatives of imidazo[1,2-a]pyridinebased molecules (Fig. 2). To evaluate the structure-activity relationship, we designed and synthesized a series of heterocyclic derivatives containing a ring-junction nitrogen using a three-dimensional (3D) pharmacophore model reported previously (Gong et al. 2010) (Fig. 2). For the first stage, only combinations of five-and six-membered rings are considered, including imidazo[1,5-a]pyridine, imidazo[1,2-a]pyrimidine and imidazo[1,2-a]pyrazine. We employed short synthetic steps and reactions that are tolerant of the presence of various functional groups and suitable for parallel operations to enable the rapid generation of libraries of diverse, structurally complex, small molecules.

Chemistry
All the chemicals used in synthesis were supplied by Aldrichand TCI, and were used without further purification. All solvents were purified and stored in a dry condition. Reaction progress was determined by thin-layer chromatography (TLC) on Merck TLC Silica gel 60 F245 plates. Column chromatography was carried out using a silica gel 60 Merck). NMR spectra were recorded on Agilent 400 instruments operating at 400 MHz for 1 H and 100 MHz for 13 C, and Agilent 500 instruments operating at 500 MHz for 1 H and 125 MHz for 13 C. Chemical shifts are expressed as parts per million (ppm) with tetramethylsilane as the internal standard. MS spectra were recorded on an Agilent G6530A Q-TOF.

3-(6-(Trifluoromethyl
General synthetic procedure for (8b and 8e) To a solution of 7a (98 mg, 0.35 mmol) or 7b (121 mg, 0.45 mmol) in pyridine (5 mL) was added methanesulfonyl chloride (66 mg, 0.60 mmol) dropwise with stirring overnight in an ice bath. The reaction mixture was quenched with water in an ice bath and extracted with ethyl acetate (3 9 30 mL). The combined organic phases were washed with water, 1 N HCl, and brine, dried, and filtered and concentrated in vacuo. The residue was purified by silica gel column chromatography (hexane/ethyl acetate = 4/1, Rf = 0.25).
After stirring for 2 h at room temperature, the reaction mixture was quenched with water in an ice bath and then extracted with ethyl acetate (3 9 30 mL). The combined organic phases were washed with water, 1 N HCl, and brine, dried, and filtered and concentrated. The residue was purified by silica gel column chromatography (hexane/ ethyl acetate = 8/1, Rf = 0.23). General synthetic procedure for (12a-c) Pyridium bromide perbromide (1.79 g, 5.60 mmol) was added to a solution of 10a-c (0.9 g, 5.08 mmol) in AcOH (100 mL) with stirring for 3 h at room temperature. The reaction mixture was poured into ice-cold water and then extracted with ethyl acetate (3 9 50 mL). The combined organic phases were washed with saturated aqueous NaHCO 3 , water, and brine, dried, and filtered and concentrated in vacuo to give crude 11a-c as a yellow oil (1.29 g, 98 %). The resulting crude 11a-c could be used without further purification. To a stirred solution of bromomethylketone 11a-c (1.30 g, 5.1 mmol) and aminopyridine 4 (0.82 g, 5.1 mmol) in EtOH (80 mL) was added NaHCO 3 (0.43 g, 5.1 mmol) at room temperature. The reaction mixture was heated to reflux for 8 h. After removing EtOH, the residue was extracted with ethyl acetate and water. The combined organic phases were washed with water, 1 N HCl, and brine, dried, and filtered and concentrated in vacuo. The residue was purified by silica gel column chromatography (12a-b, hexane/ethyl acetate = 1/1, Rf = 0.24; 12c, hexane/ethyl acetate = 4/1,  (21a) To a solution of 19 (257 mg, 0.72 mmol) in THF (20 mL) was added a solution of NaOH (50 mg, 1.25 mmol) in water (10 mL) with stirring for 3 h at room temperature. After removing THF, the resulting mixture was extracted with ethyl acetate. The combined organic phases were washed with water and brine, dried, and filtered and concentrated in vacuo. The resulting crude 20 could be used without further purification. Cyclohexanecarboxylic chloride (28 mg, 0.19 mmol) was added to a solution of 20 (50 mg, 0.16 mmol), TEA (19 mg, 0.19 mmol), and DMAP (4 mg, 0.03 mmol) in anhydrous CH 2 Cl 2 (20 mL) slowly in an ice bath. After stirring for 3 h at room temperature, the reaction mixture was poured into ice water and then extracted with CH 2 Cl 2 (3 9 20 mL). The combined organic phases were washed with 1 N HCl, water, and brine, dried, and filtered and concentrated in vacuo. General synthetic procedure for (27a-d) POCl 3 (0.3 mL, 3.40 mmol) was added to a mixture of 24 or 26a-c (0.17 mmol) and pyridine (0.93 mL, 11.60 mmol) in anhydrous dichloroethane (14 mL) at room temperature. The reaction mixture was heated to reflux for 7 h. After cooling to room temperature, the reaction mixture was concentrated, filtered, and extracted with ethyl acetate. The combined organic phases were washed with 1 N HCl, water, and brine, dried, and filtered and concentrated in vacuo. The residue was purified by silica gel column chromatography (hexane/ethyl acetate = 2/1, Rf = 0.24).

Chemistry
The general synthetic pathway yielding the novel derivatives 6a, 6b, and 8a-8f is outlined in Scheme 1. 3-Hydroxy acetophenone 1 was converted into 3-acetoxy acetophenone 2 by acetylation. Treatment of substituted acetophenone 2 with bromine in the presence of AlCl 3 in Et 2 O (Bunders et al. 2010) afforded á-bromomethylketone 3. Subsequently, 3 and the substituted 2-aminopyridines 4, 5 were allowed to react in the presence of sodium bicarbonate in refluxing ethanol (Fookes et al. 2008), resulting in the generation of compounds 6a and 6b. The deacetylation of compounds 6a and 6b with sodium hydroxide afforded compounds 7a and 7b, respectively. The akylation or acylation of 7a and 7b furnished the target compounds 8a-8f.
The syntheses of compounds 37a-37e are detailed in

Biology
The compounds prepared in this study were evaluated in terms of GLP-1R agonist activity using an in vitro activation efficacy assay in CHO-K1 cells (Chen et al. 2007), and the magnitude of the responses have been compared at two concentrations of compounds used. GLP-1 (7-37) was used as the positive control and DMSO (0.1 %) was used as the negative control. Induction values represent luciferase activities driven by CRE (cAMP response element). Compounds were grouped into three series according to fused-heterocyclic ring type.
In general, the first series of compounds, 6a-6b, 8a-8f, 12a-12c, and 33a-33c (Fig. 3), based on the imidazo[1,2a]pyridine structure and containing various substituted groups, generated higher responses than those of the second series. Compound 6b is the model compound, in which replacement of the acetyl group with propanyl-2-one 8d, mesyl 8e, or tosyl 8f resulted in a significant increase in magnitude of the response, suggesting that the hydrogenbond donor is preferred to be this region and that the length of linker affects binding to the ago-allosteric binding site of GLP-1R. To determine whether the chlorine in imidazo[1,2-a]pyridine is essential for its activity, compounds 6a and 8a-8c were synthesized. Compounds 8a and 8b showed good responses similar to that of compounds 8d and 8e at 10 lM. However, a loss of response was observed for compounds 6a and 8c. The bioisosteric replacement of the ester group (6a, 8b-8c) with an amide group (12a-12c) resulted in comparable activities; in particular, compound 12a exhibited a two fold enhanced response. On the contrary, compounds 33a and 33c exhibited decreased responses. However, compound 33b, which contains acyclohexanecarboxamide group, exhibited a moderate response. Further increases in concentration resulted in a significant drop in the responses of compounds 8a, 6a-6b, 8f, and 12a-12b to an about \1-fold increase, likely due to cytotoxicity at a high concentration (100 lM). Surprisingly, compound 8e, which generated the highest response at 10 lM, also exhibited the greatest response at a high concentration (100 lM).
Finally, a nitrogen atom was introduced into the sixmembered fused-heterocyclic ring to evaluate the effect of electron density on activity (Fig. 4). The majority of the compounds 37a-37e thus generated showed loss of responses compared to the first series. We speculated that the loss of responses might be attributable to a decreased interaction between the p-electron and the receptor.
Over the half of the synthesized compounds, the effects did not appear concentration-dependent. However the effects of the compounds on coupling the GLP-1R to the signaling way may well be concentration-dependent, but the responses measured did not appear concentrationdependent due to cytotoxicity.
In addition, selected compounds 8a, 8b and 8e which showed [2.5-fold increases at 10 lM were assayed further to determine concentration-response curves (Fig. 5) and calculate EC 50 values (Table 1). Compound 8e, bearing chlorine substitution imidazo[1,2-a]pyridine ring and mesyl group of benzene ring, was found to be a potent GLP-1R agonist exhibiting an EC 50 of 7.89 lM. Compounds 8a and 8b, without chlorine substitution of imidazo[1,2-a]pyridine ring, were about threefold less potent, with EC 50 values of 20 lM and 17 lM, respectively (Table 1). Concentration-response curves of selected compounds are shown in Fig. 5. The concentration are in a range from 1 lM to 100 lM. Compounds 8b and 8e showed above 50 % response (8b, 52 %; 8e, 58 %) at their EC 50 values, while compound 8a showed lower response 43 % at EC 50 value (Fig. 5). Thus, compound 8e may serve as a GLP-1R agonist with potential for application.
In conclusion, these new compounds, synthetic methodology developed and preliminary biological evaluation results could be helpful in further design and discovery of more potent GLP-1R agonists for the treatment of DM2. Fig. 3 In vitro responses of compounds 6a-6b, 8a-8f, 12a-12c and 33a-33c on CHO-K1 cells at 10 and 100 lM. The cells were transfected with the pCMV6-GLP-1R and pCRE-Luc plasmids. The transfected and cultured cells were incubated with different compounds for 8 h, and luciferase activity quantified using the Steady-Glo luciferase assay system. Vertical and horizontal axes show the fold increases compared to the control and the synthesized compounds, respectively. Values shown are mean ± SD of three independent experiments. Significant difference from 0.1 % DMSO treated group: *p B 0.05, **p B 0.01 and ***p \ 0.001 Fig. 4 In vitro responses of compounds 19, 21a-21b, 27a-27d, and 37a-37e on CHO-K1 cells at 10 and 100 lM. Experimental details are described in Fig. 3. Vertical and horizontal axes show the fold increases compared to the control and the synthesized compounds, respectively. Values shown are mean ± SD of three independent experiments. Significant difference from 0.1 % DMSO treated group: *p B 0.05, **p B 0.01, and ***p \ 0.001

Synthesis of GLP-1R agonists 597
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited. Values shown are mean ± SD of three independent experiments. For determined EC 50 values see Table 1 Table