PR000933 (Project)

Description:Antibiotics have enormous impact on microbial communities, yet many studies focused on microbiome and health are confounded by limited information about antibiotic exposure. Given that antibiotics have diverse pharmacokinetic and antibacterial properties, investigating the type, concentration and location would help us understand what microbes encounter at an infection site. Here, we developed Liquid Chromatography Mass Spectrometry (LC-MS) methods to detect 18 antibiotics in clinical samples using cystic fibrosis sputum as an application. Three liquid extraction methods were compared with spike-in controls of antibiotics on the Waters Acquity Quattro Premier XE. Extraction with dithiothreitol captured the most antibiotics and was then used to detect antibiotics in sputum samples from 11 people with cystic fibrosis in comparison to self-reported antibiotic use. For the sputum samples, the Quattro Premier XE method detected nanomolar or micromolar concentrations of 16 antibiotics. In 71 of the 158 sputum samples, at least one antibiotic that was not reported by the subject was detected by the LC-MS method, a discordance largely explained by alternating usage and the long half-life of azithromycin. For ~37% of samples, antibiotics were marked by the subject but were not detected, which has implications for in vivo antibiotic penetrance and efficacy. Our study takes the first step in detecting antibiotics at the infection site, with future goals of incorporating accurate antibiotic data into microbiome studies.