A platform for detecting cross-resistance in antibacterial drug discovery.

BACKGROUND: To address the growing antibiotic resistance problem, new antibacterial drugs must exert activity against pathogens resistant to agents already in use. With a view to providing a rapid means for deselecting antibacterial drug candidates that fail to meet this requirement, we report here the generation and application of a platform for detecting cross-resistance between established and novel antibacterial agents. METHODS: This first iteration of the cross-resistance platform (CRP) consists of 28 strains of defined resistance genotype, established in a uniform genetic background (the SH1000 strain of the clinically significant pathogen Staphylococcus aureus). Most CRP members were engineered through introduction of constitutively expressed resistance determinants on a low copy-number plasmid, with a smaller number selected as spontaneous resistant mutants. RESULTS: Members of the CRP collectively exhibit resistance to many of the major classes of antibacterial agent in use. We employed the CRP to test two antibiotics that have been proposed in the literature as potential drug candidates: γ-actinorhodin and batumin. No cross-resistance was detected for γ-actinorhodin, whilst a CRP member resistant to triclosan exhibited a 32-fold reduction in susceptibility to batumin. Thus, a resistance phenotype that already exists in clinical strains mediates profound resistance to batumin, implying that this compound is not a promising antibacterial drug candidate. CONCLUSIONS: By detecting cross-resistance between established and novel antibacterial agents, the CRP offers the ability to deselect compounds whose activity is substantially impaired by existing resistance mechanisms. The CRP therefore represents a useful addition to the antibacterial drug discovery toolbox.