Polynucleobacter necessarius,a model for genome reduction in both free-living and symbiotic bacteria

We present the complete genomic sequence of the essential symbiont Polynucleobacter necessarius (Betaproteobacteria), which is a valuable case study for several reasons. First, it is hosted by a ciliated protist, Euplotes; bacterial symbionts of ciliates are still poorly known because of a lack of extensive molecular data. Second, the single species P. necessarius contains both symbiotic and free-living strains, allowing for a comparison between closely related organisms with different ecologies. Third, free-living P. necessarius strains are exceptional by themselves because of their small genome size, reduced metabolic flexibility, and high worldwide abundance in freshwater systems. We provide a comparative analysis of P. necessarius metabolism and explore the peculiar features of a genome reduction that occurred on an already streamlined genome. We compare this unusual system with current hypotheses for genome erosion in symbionts and free-living bacteria, propose modifications to the presently accepted model, and discuss the potential consequences of translesion DNA polymerase loss.Symbiosis, defined as a close relationship between organisms belonging to different species (1), is a ubiquitous, diverse, and important mechanism in ecology and evolution (e.g., refs. 2–4). In extreme cases, through the establishment of symbiotic relationships, quite unrelated lineages can functionally combine their genomes and generate advantageous emergent features or initiate parasite/host arms races. Ciliates, common unicellular protists of the phylum Ciliophora, are extraordinary receptacles for prokaryotic ecto- and endosymbionts (5, 6) that provide varied examples of biodiversity and ecological roles (6). Nevertheless, most of these symbionts are understudied, partially owing to the scarcity of available molecular data and the absence of sequenced genomes. Yet, thanks to their various biologies and the ease of sampling and cultivating their protist hosts, they are excellent potential models for symbioses between bacteria and heterotrophic eukaryotes. Until recently this field was dominated by studies on endosymbionts of invertebrates, especially insects (e.g., ref. 7), although unicellular systems like amoebas (e.g., refs. 8 and 9) have been shown to be suitable models.Polynucleobacter necessarius was first described as a cytoplasmic endosymbiont of the ciliate Euplotes aediculatus (10, 11). Further surveys detected its presence in a monophyletic group of fresh and brackish water Euplotes species (12, 13). All of the investigated strains of these species die soon after being cured of the endosymbiont (10, 12, 13). In the few cases in which P. necessarius is not present, a different and rarer bacterium apparently supplies the same function (12, 14). No attempt to grow symbiotic P. necessarius outside their hosts has yet been successful (15), strongly suggesting that the relationship is obligate for both partners, in contrast to most other known prokaryote/ciliate symbioses (6).Thus, the findings of many environmental 16S rRNA gene sequences similar to that of the symbiotic P. necessarius (16) but belonging to free-living freshwater bacteria came as a surprise. These free-living strains, which have been isolated and cultivated (17), are ubiquitous and abundant in the plankton of lentic environments (17, 18). They are smaller and do not show the most prominent morphological feature of the symbiotic form: the presence of multiple nucleoids, each containing one copy of the genome (10, 11). It is clear that free-living and endosymbiotic P. necessarius are not different life stages of the same organism (15). Nevertheless, these strikingly different bacteria, occupying separate ecological niches, exhibit >99% 16S rRNA gene sequence identity, and phylogenetic analyses fail to separate them into two distinct groups (15). Rather, several lines of evidence point to multiple, recent origins of symbiotic strains from the free-living bacterial pool (14, 15).Thus, the Euplotes–Polynucleobacter symbiosis provides a promising system for the study of changes promoting or caused by the shift to an intracellular lifestyle. The remarkably small (2.16 Mbp) genome of the free-living strain QLW-P1DMWA-1 has been sequenced and studied, especially for features that would explain the success of this lineage in freshwater systems worldwide (19, 20). Phylogenies based on the 16S rRNA gene (13, 14) and multiple-gene analyses (19, 21, 22) consistently cluster Polynucleobacter with bacteria of the family Burkholderiaceae (Betaproteobacteria), either in a basal position or as the sister group of Ralstonia and Cupriavidus.Here we provide the complete genomic sequence of a symbiotic P. necessarius harbored in the cytoplasm of E. aediculatus and present a comparative analysis of the two sequenced Polynucleobacter genomes, addressing the possible biological basis of the Euplotes–Polynucleobacter symbiosis. We also provide insights into the evolution of the unique two-step genome reduction in this bacterial species: the first step involving streamlining in a free-living ancestor and the second a more recent period of genome erosion confined to the symbiotic lineage.