Optimized protocols for assessing libraries of poorly soluble sortase A inhibitors for antibacterial activity against medically-relevant bacteria, toxicity and enzyme inhibition.

Increasing antimicrobial resistance is a major global health concern. Conventional antibiotics apply selection pressures, which promote the accumulation of resistant microbes. Anti-virulence strategies, in contrast, are less potent antimicrobials, but are less likely to select for resistance, can be combined with existing antibiotics to improve their activity, and in some cases can overcome antimicrobial resistance towards other antimicrobials. Sortase A inhibitors (SrtAIs) represent an exciting example of this class; however, many reported examples demonstrate poor water solubility, which complicates their biological assessment and activity. This includes reports that use antimicrobial concentrations of organic solvents or conditions that fail to solubilise these compounds for minimal inhibitory concentration (MIC) assessments. Herein, we report the first study to optimise screening processes for a library of prospective SrtAIs (trans-chalcone (TC), berberine (BR), curcumin (CUR), and quercetin (QC)), including comparative assessment of the effects of various co-solvent concentrations, along with comparative assessment of their antimicrobial activities against multiple disease relevant bacterial strains (methicillin-sensitive and resistant S. aureus, E. coli, and P. aeruginosa), inhibition of the sortase A enzyme, and toxicity towards mammalian cells (HEK-293), using these optimised conditions. Optimal solubility with minimal effect on bacterial viability was observed in the presence of 5% (v/v) dimethyl sulfoxide (DMSO)-Mueller-Hinton Broth. Three antimicrobial susceptibility tests (broth microdilution, agar dilution, and disk diffusion) were assessed for their ability to accurately determine minimal inhibitory concentration (MIC) data for each SrtAI. Broth microdilution and agar dilution were both effective; however, the broth microdilution assay required the addition of a colorimetric metabolic indicator (resazurin) to enable simple and reliable MIC determination due to the development of precipitants over time. In contrast, disk diffusion did not provide reliable zone of inhibition data. Identical MIC data was observed with methicillin-sensitive and -resistant S. aureus (MRSA; ATCC43300), with lower potency activity against E. coli and P. aeruginosa. Under these conditions, TC and CUR demonstrated significant toxicity towards human embryonic kidney (HEK-293) cells, with QC showing less toxicity and BR limited-to-no toxicity at its MIC. Overall, the findings of this work provide optimised processes, which will prove useful for the study of other poorly soluble antimicrobial agents and SrtAIs. The obtained data suggests that BR should be considered in preference to the other SrtAIs for the development of new antimicrobial formulations, based on its superior antimicrobial and SrtA inhibition potency, and greatly reduced toxicity.