Novel prokaryotic system employing previously unknown nucleic acids-based receptors

Inferring strain-level mutational drivers of phage-bacteria interaction phenotypes

The enormous diversity of bacteriophages and their bacterial hosts presents a significant challenge to predict which phages infect a focal set of bacteria. Infection is largely determined by complementary -and largely uncharacterized- genetics of adsorption, injection, and cell take-over. Here we present a machine learning (ML) approach to predict phage-bacteria interactions trained on genome sequences of and phenotypic interactions amongst 51 Escherichia coli strains and 45 phage λ strains that coevolved in laboratory conditions for 37 days. Leveraging multiple inference strategies and without a priori knowledge of driver mutations, this framework predicts both who infects whom and the quantitative levels of infections across a suite of 2,295 potential interactions. The most effective ML approach inferred interaction phenotypes from independent contributions from phage and bacteria mutations, predicting phage host range with 86% mean classification accuracy while reducing the relative error in the estimated strength of the infection phenotype by 40%. Further, transparent feature selection in the predictive model revealed 18 of 176 phage λ and 6 of 18 E. coli mutations that have a significant influence on the outcome of phage-bacteria interactions, corroborating sites previously known to affect phage λ infections, as well as identifying mutations in genes of unknown function not previously shown to influence bacterial resistance. While the genetic variation studied was limited to a focal, coevolved phage-bacteria system, the method's success at recapitulating strain-level infection outcomes provides a path forward towards developing strategies for inferring interactions in non-model systems, including those of therapeutic significance.

AtbFinder Diagnostic Test System Improves Optimal Selection of Antibiotic Therapy in Persons with Cystic Fibrosis

Cystic fibrosis (CF) is characterized by mutations of CFTR that lead to increased viscous secretions, bacterial colonization, and recurrent infections. Chronic Pseudomonas aeruginosa infection in persons with CF is associated with progressive and accelerated lung function decline despite aggressive antibiotic treatment. We report the management of respiratory infections in persons with CF with antibiotic therapy that was based on the recommendations of AtbFinder, a novel, rapid, culture-based diagnostic test system that employs a novel paradigm of antibiotic selection. AtbFinder mimics bacterial interactions with antibiotics at concentrations that can be achieved in affected tissues or organs and models conditions of interbacterial interactions within polymicrobial biofilms. This open-label, single-arm, investigator-initiated clinical study was designed to identify the efficacy of antibiotics selected using AtbFinder in persons with CF. Microbiological and clinical parameters were assessed following the change of antibiotic therapy to antibiotics selected with AtbFinder between January 2016 and December 2018 and retrospectively compared with clinical data collected between January 2013 and December 2015. We enrolled 35 persons with CF (33 with chronic P. aeruginosa colonization). Antibiotics selected using AtbFinder resulted in clearance of P. aeruginosa in 81.8% of subsequent cultures, decreased pulmonary exacerbations from 1.21 per patient per annum to 0, and an increase in predicted percent predicted forced expiratory volume in 1 s up to 28.4% from baseline. The number of systemic antibiotic courses used in patients after switching to the AtbFinder-selected therapy was reduced from 355 to 178. These findings describe the superiority of antibiotic regimens selected with AtbFinder compared with routine antimicrobial susceptibility testing.

Overcoming Antibiotic Resistance with Novel Paradigms of Antibiotic Selection

Conventional antimicrobial susceptibility tests, including phenotypic and genotypic methods, are insufficiently accurate and frequently fail to identify effective antibiotics. These methods predominantly select therapies based on the antibiotic response of only the lead bacterial pathogen within pure bacterial culture. However, this neglects the fact that, in the majority of human infections, the lead bacterial pathogens are present as a part of multispecies communities that modulate the response of these lead pathogens to antibiotics and that multiple pathogens can contribute to the infection simultaneously. This discrepancy is a major cause of the failure of antimicrobial susceptibility tests to detect antibiotics that are effective in vivo. This review article provides a comprehensive overview of the factors that are missed by conventional antimicrobial susceptibility tests and it explains how accounting for these methods can aid the development of novel diagnostic approaches.