The compounds described in this paper were

The compounds described in this paper were prepared using a modular approach that allowed diversification of R or R at the final step (). Route A involved amide coupling of -butyl 4-(methylamino)piperidine-1-carboxylate with phenyl acetic acids (step a) using 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride (DMTMM) followed by Boc deprotection to give the piperidine (step b). Subsequent alkylation (step c) provided compounds & . Route B involved alkylation of isomeric --butyl -methyl--(4-piperidyl)carbamate followed by Boc deprotection to give the piperidine intermediate that was acylated to provide compounds & –. Compound was synthesised using route B with a reductive amination of the corresponding aldehyde (step d) as the final step. Other examples (–) were synthesised by elaboration of functional groups (). Amide was synthesised from an ester intermediate through hydrolysis and coupling with methylamine (steps d, e) whereas the reversed amide was synthesised through reduction of a nitro group followed by coupling with an alkaline phosphatase inhibitor chloride (steps g, h). Fluoro substituted aromatic examples & were synthesised via formation of the bromo intermediates followed by conversion to sulphone (step i) using copper catalysis. For the fluorinated piperidine examples (), the known (,)-3-fluoro-piperazine intermediate was methylated (step a) and the carboxybenzyl (Cbz) protecting group removed under hydrogenating conditions (step b). Coupling with the required phenyl acetic acid using HATU (2-(7-Aza--benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate) (step c) and deprotection of the Boc group (step d) afforded an intermediate that could be alkylated (step e) to afford the final compounds & –. Compound was synthesised from the enantiomeric (,) intermediate using the same synthetic sequence. For the methoxy analogues, the known (,)-3-methoxy-piperazine intermediate was carboxybenzyl protected (step f) and then methylated (step g). Removal of the carboxybenzyl protecting group (step h) was followed by coupling with the required phenyl acetic acid (step i) using DMTMM. Deprotection of the ethyl carbamate was carried out using trimethylsilyl iodide (step j) and the product was then alkylated (step k) to afford the final compound . Compound was synthesised from the enantiomeric (,) intermediate using the same synthetic sequence. Truncation of the -ethyl to an -methyl amide improved the in vitro metabolic profile (Cl=3μL/min/mg) whilst maintaining potency (EC=456nM), LLE (3.8) and hERG (IC=3μM; now eightfold relative to primary target). We were keen to improve potency but not at the cost of additional lipophilicity and therefore had LLE as a key optimization parameter. Exploration of the R substituent is shown in . Fluorination of the aryl ring ( & ) resulted in an increase in LLE score (LLE=4.2 & 3.9, respectively) and efficacies (88% & 73%, respectively) relative to the unsubstituted analogue . Although the hERG affinities had remained unaltered, the increase in primary potency resulted in an improved selectivity ratio (>20-fold in both cases). C and N linked amides were tolerated ( & ) although neither offered significant advantages over the sulphone . The -linked tetrazole was the most potent compound identified (EC=74nM) and had the highest efficacy (132%) with an LLE of 4.1. Although this compound showed increased hERG affinity (1.2μM), the increased potency resulted in an increased hERG selectivity (16×). Substitution of the aryl ring was crucial to activity as exemplified by compound which was inactive in the in vitro cAMP assay, despite being considerably more lipophilic than the other compounds investigated. Exploration of the R substituent is shown in with the R substituent fixed as the -tetrazole as in compound (EC=74nM; LLE=4.1). Incorporation of a pyridyl N ( & ) reduced potency in line with the reduction in lipophilicity (LLE scores of 4.1, 4.0 & 3.9 for compounds , & , respectively) but with efficacy remaining high. Consequently, although reductions in hERG were observed this did not result in increased margins. Fluorination of the aryl ring () resulted in a modest potency increase and decrease in hERG affinity leading to increased selectivity (53-fold) however, increased susceptibility to metabolism was observed (Cl=14μL/min/mg). The CF unit could be replaced by polar heterocycles such as the oxadiazole with potency loss broadly in line with lipophilicity reduction (LLE=4.3) but with efficacy maintained and selectivity against hERG remaining at similar levels (15-fold) to compound . Benzothiazole was identified as an interesting group that had a similar profile to the trifluoroaryl with high LLE (4.3) and similar selectivity against hERG (17-fold). Attempts to saturate the aryl ring () had the desired effect of reducing the absolute affinity for hERG (IC=7.6μM), but led to a reduction in LLE and an increase in turnover in human microsomes (Cl=169μL/min/mg).