With the narrow SAR of
With the narrow SAR of the tail favouring the ortho sulfone group, we examined the remaining position open to substitution at the pyrazole 3-position, in an attempt to at least moderate the excessive human PPB (). Lipophilic groups in this position maintained huge shifts in potency in the absence or presence of plasma proteins. Only the polar pyridyl substitution () was capable of maintaining high potency in the presence of albumin. In this case we used a substitute assay of running the GTPγS binding in the presence of 1% human serum albumin (HSA). This assay allowed a much higher throughput than the ESC assay and we always observed a good correlation between the two (data not shown). At this point, through the tail SAR we had found an optimum substitution pattern in the ortho-sulfonylbenzyl tail. This in turn had led to an optimisation of the substituents of the pyrazole core, resulting in potent inhibitor . However, it was evident that these substitution patterns of both tail and core were quite specific and that we were left with little further room for manoeuvre. We could however see parallels with a series of thiazole-4-acetic autophagy inhibitor compounds reported by 7TM, in which the equivalent R group of the core was also 4-pyridyl. The tail groups of this series were dibenzyl derivatives, so we decided to pursue this line of investigation. For some of these compounds, alternative synthetic routes were useful (). If the correct diketone was available, this was cyclised with hydrazine () and then alkylated and brominated (in either order) to give the bromopyrazole . Alternatively, to allow introduction of the R group at a later stage of the synthesis the iodobromopyrazoles were synthesised. Alkylation (→) proceeded with very high regioselectivity in this case. The R group was then introduced via a Suzuki reaction. With the purified regioisomer of the bromopyrazole in hand, the acetic acid head group was introduced via a palladium-catalysed Negishi reaction under microwave conditions. Finally, acid deprotection was carried out with HCl in dioxane, as when the R tail group was benzhydryl, standard TFA treatments also cleaved the tails. In terms of this new SAR, the 4-pyridyl group at the pyrazole 3-position was preferable (). Again, we found that small changes to the first compound of the series lost activity. For example, extending the hydrogen-bond acceptor capacity of the pyridyl group to a pyridone () lost all potency. In the end, the benzhydryl tail motif was unable to reach the same levels of potency as seen with the sulfone tails. A selection of the compounds throughout the series, namely , , , , , and were also tested for their selectivity against both the DP1 receptor and the thromboxane A2 receptor. All compounds were selective, showing <40% inhibition at 10μM in all cases. We also characterised the dissociation kinetics of some of these compounds, following some reports of the potential for slow off-rate compounds from the CRTh2 receptor. Briefly, membranes were incubated for 1h with the test compound at 10×IC. We then initiated compound dissociation by adding a huge excess of PGD (100μM and 0.1nM S-GTPγS) so that once the test compound dissociated, re-binding was prevented by mass-action law. We could then perform the standard GTPγS assay at intervals event to follow the loss of binding activity over time. Using this assay, benzyl-substituted compounds and showed rapid dissociation half-lives of 2 and 12min respectively. Benzhydryl analogue was somewhat slower with a half-life of around 30min.