Genome-wide analysis ofEscherichia colifitness determinants in a cross-feeding mutualism withRhodopseudomonas palustris

Microbial interactions abound in natural ecosystems and shape community structure and function. Substantial attention has been given to cataloging mechanisms by which microbes interact, but there is a limited understanding of the genetic landscapes that promote or hinder microbial interactions. We previously developed a mutualistic coculture pairingEscherichia coliandRhodopseudomonas palustris, whereinE. coliprovides carbon toR. palustrisin the form of glucose fermentation products andR. palustrisfixes N2gas and provides nitrogen toE. coliin the form of NH4+. The stable coexistence and reproducible trends exhibited by this coculture make it ideal for interrogating the genetic underpinnings of a cross-feeding mutualism. Here, we used random barcode transposon sequencing (RB-TnSeq) to conduct a genome-wide search forE. coligenes that influence fitness during cooperative growth withR. palustris. RB-TnSeq revealed hundreds of genes that increased or decreasedE. colifitness in a mutualism-dependent manner. Some identified genes were involved in nitrogen sensing and assimilation, as expected given the coculture design. The other identified genes were involved in diverse cellular processes, including energy production and cell wall and membrane biogenesis. Additionally, we discovered unexpected purine cross-feeding fromR. palustristoE. coli, with coculture rescuing growth of anE. colipurine auxotroph. Our data provide insight into the genes and gene networks that can influence a cross-feeding mutualism and underscore that microbial interactions are not necessarily predictablea priori.IMPORTANCEMicrobial communities impact life on earth in profound ways, including driving global nutrient cycles and influencing human health and disease. These community functions depend on the interactions that resident microbes have with the environment and each other. Thus, identifying genes that influence these interactions will aid the management of natural communities and the use of microbial consortia as biotechnology. Here, we identified genes that influencedEscherichia colifitness during cooperative growth with a mutualistic partner,Rhodospeudomonas palustris. Although this mutualism centers on the bidirectional exchange of essential carbon and nitrogen,E. colifitness was positively and negatively affected by genes involved in diverse cellular processes. Furthermore, we discovered an unexpected purine cross-feeding interaction. These results contribute knowledge on the genetic foundation of a microbial cross-feeding interaction and highlight that unanticipated interactions can occur even within engineered microbial communities.