Search

Extrachromosomal genetic elements such as bacterial endosymbionts and plasmids generally exhibit AT-contents that are increased relative to their hosts’ DNA. The AT-bias of endosymbiotic genomes is commonly explained by neutral evolutionary processes such as a mutational bias towards increased A+T. Here we show experimentally that an increased AT-content of host-dependent elements can be selectively favoured. Manipulating the nucleotide composition of bacterial cells by introducing A+T-rich or G+C-rich plasmids, we demonstrate that cells containing GC-rich plasmids are less fit than cells containing AT-rich plasmids. Moreover, the cost of GC-rich elements could be compensated by providing precursors of G+C, but not of A+T, thus linking the observed fitness effects to the cytoplasmic availability of nucleotides. Accordingly, introducing AT-rich and GC-rich plasmids into other bacterial species with different genomic GC-contents revealed that the costs of G+C-rich plasmids decreased with an increasing GC-content of their host’s genomic DNA. Taken together, our work identifies selection as a strong evolutionary force that drives the genomes of intracellular genetic elements toward higher A+T contents. [ABSTRACT FROM AUTHOR]

Endosymbionts are organisms that live inside the cells of other species. This lifestyle is ubiquitous across the tree of life and is featured by unicellular eukaryotes, prokaryotes, and by extrachromosomal genetic elements such as plasmids. Given that all of these elements dwell in the cytoplasm of their host cell, they should be subject to similar selection pressures. Here we show that strikingly similar features have evolved in both bacterial endosymbionts and plasmids. Since host and endosymbiont are often metabolically tightly intertwined, they are difficult to disentangle experimentally. We propose that using plasmids as tractable model systems can help to solve this problem, thus allowing fundamental questions to be experimentally addressed about the ecology and evolution of endosymbiotic interactions. [ABSTRACT FROM AUTHOR]