Identification of a Novel Series of Potent TrkA Receptor Tyrosine Kinase Inhibitors

A novel series of N-(3-(6-substituted-aminopyridin-3-yloxy)phenyl)-2-oxo-3-phenylimidazolidine-1-carboxamides targeting TrkA receptor tyrosine kinase was identified. SAR study of the series allowed us to design and synthesize compounds possessing inhibitory activity of TrkA kinase enzyme in the low nanomolar range with low residual activity against c-Met and with no significant activity against VEGFR2.

Tropomyosin-related kinases (Trks) are receptor tyrosine kinases normally expressed in neuronal tissue where they play important role in both development and function of the nervous system [1]. The Trk receptor family is composed of three members (A, B, and C) activated by specific ligands called neurotrophins (NT). Each Trk receptor contains an extracellular domain (ligand binding), a transmembrane region, and an intracellular domain (including kinase domain) which upon binding of their respective ligand (nerve growth factor (NGF) for TrkA, brain-derived growth factor (BDNF) and NT-4/5 for TrkB, and NT3 for TrkC) triggers oligomerization of the receptors, phosphorylation of specific tyrosine residues in the kinase domain, and downstream signal transduction pathways, including survival, proliferation, and differentiation in normal and neoplastic neuronal cells [2][3][4]. While Trks are expressed at low levels outside the nervous system in the adult, deregulation of TrkA and TrkB and their cognate ligands has been described in numerous types of cancers including prostate, breast, colorectal, ovarian, lung, pancreas, melanoma, thyroid, and neuroblastoma and occurs mainly through wild type receptor overexpression, activation, amplification, and/or mutation [5,6]. Importantly, increased Trks activation in tumor tissues correlates with an aggressive phenotype and poor clinical outcome [6].
We have previously reported on four novel series of kinase inhibitors with a 2-oxo-3-phenylimidazolidine-1carboxamide head group based on thieno [  [19], quinoline [20], pyrimidine, and pyridine scaffolds [21], exemplified by structures 1-4, respectively (Table 1). Bi-cyclic RTK inhibitors 1 and 2 strongly inhibited both c-Met and VEGFR2 enzymes while compounds 3 and 4 showed some preference for the c-Met kinase enzyme. In our efforts to identify new kinase inhibitors with a different kinase inhibitory profile from the ones of compounds 1-4 (and their analogues), we viewed compound 4 as a valuable starting point for further exploration. As the first exercise, we changed the substitution pattern between the central phenyl ring and the pyridine core of compound 4 as illustrated in Figure 2 (transposition from 1,4-to 1,3-substitution pattern like in isothiazole 5n from Pfizer, thiazole 20h from BMS, and pyrimidine AZ-23 from AstraZeneca), thus synthesizing compound 5. This compound was roughly ten times weaker than the parent compound 4 against both c-Met and VEGFR2 enzymes, but surprisingly showed potent inhibition of TrkA when tested at 0.1 μM. As a consequence, we initiated the SAR study around compound 5 as a prototype of selective TrkA inhibitors. The IC 50 of compound 5 was determined to be 4 nM against the TrkA human enzyme.
A straightforward synthesis of the target molecule 5 was developed and is illustrated in Scheme 1. (The details for the synthesis and the characterization of all the new compounds are well described in [21].) 5-Bromo-2-nitropyridine was reacted with 3-aminophenol in the presence of cesium carbonate in acetonitrile at room temperature. The choice of the base, solvent, and temperature is critical in order to obtain the intermediate A in fairly good yield; otherwise multiple products are formed. Intermediate A was subsequently reacted with 2-oxo-3-phenylimidazolidine-1carbonyl chloride generated in situ from the reaction between 1-phenylimidazolidin-2-one and triphosgene to afford intermediate B. After reduction of the nitro group and subsequent derivatization of 6-aminopyridine intermediate C, compound 5 was obtained in good yield.
To continue with the SAR study around compound 5, we explored the replacement of the oxygen linker by keeping unchanged the 2-oxo-3-phenylimidazolidine-1-carboxamide head group, which is presumed to occupy the hydrophobic back pocket of the kinase catalytic domain, and 6-acetamidopyridine scaffold, which is supposed to  interact with the ATP catalytic site in the hinge region of the target kinase enzyme ( Table 2). Not surprisingly, the replacement of the oxygen linker in compound 5 by an ureido, ethynyl, sulfamoyl, or carbamoyl spacer (compounds 6-9) completely abolished the activity against TrkA kinase enzyme. We next investigated the presence of a small substituent located on the central phenyl ring (positions 2 and 4) of compound 5 (Table 3). In contrast with either thiazole 20h from Bristol-Myers Squib [16] or AZ-23 from AstraZeneca [17,18] (Figure 1), in which the presence of a small substituent on the central phenyl or pyridinyl ring improves the activity and the selectivity against the TrkA kinase enzyme, in our series of compounds that trend was not observed. Thus, compounds 10-12 turned out to be inactive against TrkA.
We finally examined the influence of each extremity of compound 5 where R 1 is presumed to extend into the solvent exposed area and R 2 is thought to occupy the hydrophobic back pocket of the kinase catalytic domain-both sites potentially capable of modulating physicochemical and pharmacokinetic properties of the inhibitors (compounds 13-17, Table 4). Removal of the acetyl group from compound 5 afforded compound 13 with a lower potency against TrkA, probably due to the absence of a strong hydrogen bond donor (NH amide) and/or a hydrogen bond acceptor (carbonyl amide) which is presumed to interact with the hinge region of the kinase binding domain. Introduction of relatively small electronegative fluorine in the para position of the phenyl ring is well tolerated (compound 14). Finally, replacement of the acetamido group by a functionalized ureido moiety is also accepted, giving rise to potent TrkA inhibitors (compounds 15-17) and allowing us to incorporate water solubilizing groups.