Eukaryotic domestication of a bacterial immune protein following horizontal transfer

Abstract Many components of eukaryotic innate immunity originated from bacterial immune systems. However, it has been unclear how eukaryotes acquire these genes, why eukaryotes have sampled only certain families of bacterial proteins, and how these components become domesticated into eukaryotic physiology. Here, we discovered a recent instance of bacteria-eukaryote horizontal transfer and used it to characterize the genetic and biochemical changes that accompanied HGT. We focus on TIR domains, which are widespread yet potentially costly immune modules that are commonly associated with inflammation and/or cell death. By generating an atlas of TIR diversity across the tree of life, we phylogenetically categorized the domains and uncovered highly diverged TIR families found in eukaryotes. This analysis also allowed us to identify the TirBCD protein family of amoeba, which has been horizontally acquired and is closely related to the bacterial immune protein TIR-STING. Across their short eukaryotic history, the amoeba genes have acquired introns, evolved distinct patterns of gene expression, and engaged in evolutionary patterns of duplication and divergence typical of eukaryotic immune genes. While the TIR domain was transferred into amoebae, the genomic locus did not contain other components of a bacterial operon nor were regulatory domains transferred into the TIR protein. Nevertheless, TirC retains biochemical and physiological similarities to TIR-STING. TirC is a highly potent NADase, capable of spontaneously oligomerizing into large complexes and depleting cellular NAD even at very low protein concentrations. When expressed in yeast or E. coli , TirC is spontaneously active and highly toxic, illustrating the dangers of autoimmunity following TIR protein movement into novel hosts. In contrast, amoebae tolerated high TirC expression with no disruption in cell size, growth, or behavior. Single, double, and triple knock out mutants of amoeba tirBCD are viable and display modest defects in their ability to phagocytose bacteria, implying that the co-opted bacterial TIR domain may regulate eukaryotic host-microbe interactions. Overall, this study uncovers an informative example of recent eukaryotic TIR evolution that captures features of both bacterial and eukaryotic immunity. In addition, we expect that the TIR domain atlas will be useful to researchers in many model systems as they explore the vast diversity of TIR molecular and cellular functions.