Deep learning-guided discovery of a narrow-spectrum antibiotic against Acinetobacter baumannii

Acinetobacter baumannii is a nosocomial Gram-negative pathogen that often displays multidrug-resistance due to its robust outer membrane and its ability to acquire and retain extracellular DNA. Moreover, it can survive for prolonged durations on surfaces and is resistant to desiccation. Discovering new antibiotics against A. baumannii has proven challenging through conventional screening approaches. Fortunately, machine learning methods allow for the rapid exploration of chemical space, increasing the probability of discovering new chemical matter with antibacterial activity against this burdensome pathogen. Here, we screened ~7,500 molecules for those that inhibited the growth of A. baumannii in vitro. We trained a deep neural network with this growth inhibition dataset and performed predictions on the Drug Repurposing Hub for structurally novel molecules with activity against A. baumannii. Through this approach, we discovered abaucin, an antibacterial compound with narrow-spectrum activity against A. baumannii, which could overcome intrinsic and acquired resistance mechanisms in clinical isolates. Further investigations revealed that abaucin perturbs lipoprotein trafficking through a mechanism involving LolE, a functionally conserved protein that contributes to shuttling lipoproteins from the inner membrane to the outer membrane. Moreover, abaucin was able to control an A. baumannii infection in a murine wound model. Together, this work highlights the utility of machine learning in discovering new antibiotics and describes a promising lead with narrow-spectrum activity against a challenging Gram-negative pathogen. Transcriptomic analysis of A. baumannii ATCC 17978 when treated with novel antibiotic compound