Compounds were screened for their activity against
Compounds were screened for their activity against the hERG channel, with the 2,3-Cl compounds more active here than the 2,4-F compounds. The homopiperazine examples (, , , ) showed increased hERG activity relative to the corresponding piperazines. The cell data for the 6,7-dimethoxyquinoline compounds showed that other anilines were also tolerated. In addition to determining potency, the choice of aniline also influenced the overall physical properties and PK. For example, compounds with a 2,4-substituted aniline typically had the best PK profiles. To explore further the relative importance of in vitro potency, physical properties (especially protein binding) and PK on in vivo activity, the aniline substituent was re-evaluated with the 6-methylpiperazine, 7-ethoxy scaffold (–, ). Enhanced in vitro potency had been observed upon moving from 7-MeO to 7-EtO, and a set of 7-PrO examples (–) was also prepared, to see if this trend would continue. While tolerated in terms of potency, there was no further improvement to justify the increase in lipophilicity. Introducing a 7-methoxyethoxy group (–), however, gave compounds that retained the potency of the 7-EtO analogues, but with higher aqueous solubility and reduced Celecoxib protein binding. Several examples from , including the 7-methoxyethoxy compounds, showed excellent activity in the PD model at both 2 and 6h. Compounds in , were prepared as shown in . For the 7-MeO (–) and 7-EtO compounds (– and –) condensation of the appropriate 4-bromoaniline with diethylethoxymethylene malonate followed by a one-pot cyclization/chlorination gave the chloroquinoline esters. For the 7-F quinolines (–), it was necessary to perform the cyclization step at 250°C in diphenyl ether, followed by a separate chlorination step. Aniline additions were typically straightforward, but required higher temperatures for the less nucleophilic anilines such as 2,3-Cl. The methylpiperazine was introduced under standard Buchwald–Hartwig conditions, and final compounds were prepared by conversion of the quinoline ester to the amide by treatment with formamide and then methoxide. 6-Methoxy, 7-piperazine compounds (–) were prepared in a similar fashion from 3-bromo-4-methoxyaniline. An alternative route was used to prepare compounds with either no substituent (–), isopropoxy (–) or methoxyethoxy (–) at the 7-position, in which the amine was installed prior to the cyclization step. Quinolines unsubstituted at the 7-position were prepared from commercially available 4-(4-methylpiperazino)aniline. The other alkoxy substituents were derived from 2-chloro-5-nitrophenol, via alkylation, nucleophilic substitution of the chloronitrobenzene with -methylpiperazine, and reduction of the nitro group to give the required aniline. Where the amine substituent was commercially available, compounds in were prepared from the appropriate 6-bromoquinoline ester by standard Buchwald–Hartwig coupling conditions, followed by amidation of the ester. Where the amine substituents were either unavailable, or unsuitable for the Pd-coupling step, compounds were prepared by derivatization of the unsubstituted piperazine or homopiperazine amides, accessed in turn from coupling of the Boc protected amines, as shown in . A reductive amination reaction with [(1-ethoxycyclopropyl)oxy]trimethylsilane gave the cyclopropyl derivatives and . As with example , compound was found to have an extremely selective profile when evaluated in a panel of ∼85 kinases (). Apart from moderate activity versus ARK5, the compound was essentially inactive against all the other kinases screened, including the other class III RTKs: c-Kit, Flt3 and PDGFR α. In conclusion, 3-amido-4-anilinoquinolines are potent inhibitors of CSF-1R, with excellent kinase selectivity. The introduction of cyclic amines such as -methylpiperazine at the 6-position gives compounds with attractive physical properties and PK profiles. Several of these compounds show good activity in our mouse PD model. The amidoquinoline series may therefore be able to test the hypothesis that a CSF-1R inhibitor will impact tumor progression through an effect on tumor-associated macrophages. Examples have been selected for further biological profiling, the results of which will be published in due course.