br Results and discussion Compounds were tested for their bi
Results and discussion Compounds were tested for their binding affinity to human CRTH2 in a radioligand binding assay (3H-PGD2) using CHO cells stably transfected with human CRTH2. In addition, these compounds were assessed for their functional activity in PGD2 driven Ca2+ flux in KB8 cells expressing human CRTH2. As described above, we designed a novel isoquinoline scaffold and synthesized compound 15-1, based on the superimposition of the critical P1, P2, and P3 sites of known CRTH2 antagonists (Fig. 2). Inspection of the key interaction between the binding pocket of CRTH2 and the P1 and P2 sites of the CRTH2 antagonists identified a likely lipophilic side pocket on the P3 site. Considering that the lipophilic side pocket could influence CRTH2 potency and allow us to make a diverse library of derivatives, we decided therefore to first explore the acetyl moiety of the P3 site on 15-1 (Table 1). When the methyl group (R1) in the acyl moiety of 15-1 was replaced with bulkier cycloalkyl substituents such as a cyclopropyl (15-2), cyclohexyl (15-3), cyclohexylmethyl (15-4), and 1-adamantylmethyl (15-5), binding affinities were enhanced (15-2: IC50=50nM, 15-3: IC50=14nM, 15-4: IC50=20nM, 15-5: IC50=21nM) as compared with 15-1. However, these groups led to submicromolar antagonist potency in the functional assay. In addition, replacement of the methyl with a phenyl (15-7) resulted in further improvement in the binding affinity (15-7: IC50=7.9nM) and in the functional antagonist affinity (15-7: IC50=42nM). Other aromatic rings such as naphthalen-2-yl (15-8), 1H-indol-2-yl (15-10), and benzofuran-2-yl (15-11) led to greatly enhancing the binding affinity (15-8: IC50=4.0nM, 15-10: IC50=3.5nM, 15-11: IC50=3.8nM) and the functional activity as well (15-8: IC50=10nM, 15-10: IC50=4.2nM, 15-11: IC50=7.4nM). In contrast, monocyclic heteroaromatic rings including pyridin-2-yl (15-12), pyridin-4-yl (15-13), pyridazine-2-yl (15-14), pyrimidin-2-yl (15-15), and thiazole-4-yl (15-30) had moderate binding activity comparable to those containing methyl (15-1) and cyclopropyl (15-2) moieties. These results suggested that the lipophilicity of R1 group played a key role in increasing CRTH2 binding and functional potency. We decided to further explore the SAR of the potent compound 15-7 by preparing a set of compounds with a substitution on the terminal phenyl moiety (Table 2). At first, the compound 15-7 was modified by incorporating a chlorine AVE 0991 into the ortho, meta, and para positions of the phenyl ring. The meta-chlorophenyl (15-18) and para-chlorophenyl (15-19) groups enhanced the binding affinity (15-18: IC50=3.2nM, 15-19: IC50=3.4nM) and the functional antagonist activity (15-18: IC50=9.7nM, 15-19: IC50=9.1nM). In contrast, the ortho-chlorophenyl (15-17) led to a 4-fold reduction (IC50=45nM) in the binding affinity, and complete loss of the functional activity. Moreover, the meta- and para-dichlorophenyl derivative (15-20) was well tolerated in the binding and functional activity with CRTH2. For further SAR development we focused on the modification of the para-substitution of the phenyl part of 15-7. Replacement of a methyl (15-21), methoxy (15-22), trifluoromethyl (15-23), trifluoromethoxy (15-24), nitro (15-26) and dimethylamino (15-27) groups resulted in retention of the binding potency, simultaneously enhancing the functional activity. Introduction of a cyano (15-25) maintained the potent binding affinity (15-25: IC50=7.8nM), while the functional activity was greatly reduced (15-25: IC50=>1000nM). Additional lipophilic substituents such as a phenyl (15-28) and phenoxy (15-29) were tolerated with respect to the binding affinity (15-28: IC50=18nM, 15-29: IC50=14nM). Next, we examined the SAR of the amide linker moiety of 15-19 and 15-20 (Table 3). An N-methylamide 15-34, which was devoid of the hydrogen bond donor (NH) showed a 15-fold reduction in the binding affinity (15-34: IC50=210nM), suggesting that the hydrogen bond donor (NH) is likely to contribute to the greater binding affinity. The amide derivatives 15-30–15-32, which differ in the lengths of their linkers, displayed activity comparable to that of the parent compound 15-20. It is interesting to note that incorporation of an aminocarbonyloxymethylene linker (15-35) led to a 2-fold improvement in the binding and functional potency. Compounds (15-38 and 15-39) having the inverted amide linkers of 15-19 and 15-20, respectively, had acceptable binding affinity (15-38: IC50=8.2nM, 15-39: IC50=10nM), and functional activity (15-38: IC50=38nM, 15-39: IC50=13nM). In addition, an amidoethylene linker was also tolerable in the CRTH2 binding (15-40: IC50=7.8nM, 15-41: IC50=3.7nM), and the functional potency (15-40: IC50=12nM, 15-41: IC50=18nM). Next, we carried out additional SAR study of the linker moiety. When an aminomethylene (15-36) or urea (15-33) tether was incorporated in place of the amide linker, the resulting compounds maintained the potent CRTH2 binding affinity, but it led to a 2- to 4-fold reduction in functional activity compared to that of 15-19. In contrast, a sulfonamide linker (15-37) led to a significant reduction in the binding (15-37: IC50=340nM). An oxymethylene linker resulted in a slight loss of the binding affinity (15-42: IC50=38nM), and a substantial loss in the functional potency (15-42: IC50=180nM). The oxyethylene and oxypropylene linkers maintained the potent binding affinity (15-43: IC50=12nM, 15-44: IC50=11nM), while the functional activity of these compounds was weak (15-43: IC50=130nM, 15-44: IC50=85nM).