Tuberculosis is an infectious disease that presents an enormous health and economic burden for the global community. One of the biggest challenges in combatting the TB epidemic is increasing rates of resistance to antimicrobial drugs. Resistance often occurs from the acquisition of mutations in the Mycobacterium tuberculosis genome and can affect protein structure and function, and in turn drug efficacy. Mutations in the GidB and EmbB genes lead to resistance against the key first and second line drugs Streptomycin and Ethambutol, respectively. There is considerable natural variation already present in both genes, and we currently lack a molecular understanding of how changes in the protein sequence lead to resistance, hindering our ability to use clinical genomic sequencing to diagnose resistance. We have applied a structural approach in order to better identify resistance variants in these genes, and their molecular consequences.
Manual data curation and genomic analysis of variants in circulating Thailand strains has identified approximately 60 distinct resistance mutations and 33 susceptible variants in GidB, and over 150 different variants in EmbB, with strong genomic and experimental evidence. The structures of both proteins was homology modelled (sequence identities >50%), and the two drugs docked and minimised into the structures. The identified mutations were then mapped onto the structures. This revealed that mutations in GidB were located throughout the entire protein, while those in EmbB were more highly localised to a specific region where the drug bound.
The structural and biophysical consequences of the mutations were then analysed using the mCSM predictive platform. By characterising how these mutations affected protein folding and stability, drug binding and other protein interactions, we can identify how each mutation affected protein function and led to resistance. This will aid pathogen surveillance and the development of new drugs less prone to resistance.