You are here

NEWS

News / 01.19.17

Detecting warning signs of antibiotic resistance in tuberculosis

Susanna M. Hamilton, Broad Communications
Credit : Susanna M. Hamilton, Broad Communications
By Karen Zusi

A genomic analysis of more than 5,000 clinical Mycobacterium tuberculosis strains from around the world sheds light on early signs of antibiotic resistance and suggests potential strategies to slow its progression.

Antibiotic-resistant tuberculosis poses a significant and growing threat to global health. Approximately five percent of tuberculosis strains worldwide are considered multidrug-resistant, indicating that the bacteria have evolved resistance to two front-line antibiotics: isoniazid and rifampicin. Treating patients with multidrug-resistant tuberculosis can take longer than 18 months with a highly toxic combination of antibiotics, and is often unsuccessful — and the occurrence of these infections is rising.

The most common rapid diagnostic for drug-resistant tuberculosis only tests for rifampicin resistance, which is then used as a proxy for multidrug resistance. But now, a global analysis led by researchers from the Broad Institute of MIT and Harvard and funded by the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health, reveals that in strains with multidrug resistance, mutations conferring resistance to isoniazid overwhelmingly arise before rifampicin resistance. The study, published in Nature Genetics, suggests that looking specifically for isoniazid resistance could prevent the further evolution of multidrug resistance by providing a roadmap to treat the initial infection more effectively.

Treating typical, or “drug-susceptible,” tuberculosis requires a cocktail of four drugs (isoniazid, rifampicin, ethambutol, and pyrazinamide) taken orally for at least six months. However, if a patient harboring an isoniazid-resistant tuberculosis strain is treated with the standard isoniazid-containing regimen, the cocktail is less effective. The course of drugs simply feeds rifampicin to the bacterium without killing it, which can spur the evolution of rifampicin resistance and the development of a multidrug-resistant strain.

A previous Broad-led analysis detected the emergence of isoniazid resistance before other mutations in a smaller study of 337 Mycobacterium tuberculosis genomes from South Africa. In this latest work, the team — including Broad researchers Abigail Manson, Keira Cohen, Bruce Birren, and Ashlee Earl — studied more than 5,000 M. tuberculosis genomes gathered from new sequence data and previously published studies (supported by NIAID and other funders) to see if this pattern was also observed worldwide, and which specific mutations were the early harbingers of resistance emergence.

Using whole genome sequence information, the researchers reconstructed the strains’ evolutionary history to determine how drug resistance evolved step by step. In over 90 percent of cases, across all geographic regions and timeframes, isoniazid resistance (primarily through the S315T katG mutation) arose prior to the emergence of mutations indicating rifampicin resistance in strains where the two occurred.

It’s not clear why isoniazid resistance through katG S315T more commonly evolves before rifampicin resistance, but the team believes the phenomenon may be related to this mutation’s minor impact on the bacterium’s fitness. Other mutations that confer isoniazid resistance often deal a fatal blow to the katG catalase, but S315T still preserves the function of the protein.

The team hopes their results will encourage the development of new rapid tests for drug resistance incorporating the katG S315T mutation—which will allow the medical community to test early for isoniazid resistance in tuberculosis cases. The goal is to enable clinicians to prescribe an appropriate course of antibiotics without isoniazid if the infection is resistant, and, by identifying the earliest signs, help to prevent further evolution and spread of drug resistance.

This project was funded in part through NIAID grant U19AI110818, contract HHSN272200900018C, and contract HHSN2722000900050C.