The focus of this thesis has been the phylogenomics and evolution of the Alphaproteobacteria. This is a very diverse group which encompasses bacteria from intraceullar parasites, such as the Rickettsiales, to freeliving bacteria such as the most abundant bacteria on earth, the SAR11. The genome sizes of the Alphaproteobacteria range between 1 Mb and 10 Mb. This group is also connected to the origin of the mitochondria. Several studies have placed the SAR11 clade together with the Rickettsiales and mitochon- dria. Here I have shown that this placement is an artifact of compositional heterogeneity. When choosing genes or sites less affected by heterogeneity we find that the SAR11-clade instead groups with free-living alphaproteobacteria. Gene-content analysis showed that SAR11 was missing several genes for recombination and DNA-repair. The relationships within the SAR11- clade has also been examined and questioned. Specifically, we found no support for placing the taxon referred to as HIMB59 within the SAR11. Ocean metagenomes have been investigated to determine whether the SAR11-clade is a potential relative of the mitochondria. No such relationship was found. Further I have shown how important it is to take the phylogenetic relationships into account when doing statistical analyzes of genomes. The evolution of LD12, the freshwater representative of SAR11, was investigated. Phyloge- nies and synonymous substitution frequencies showed the presence of three distinct subclades within LD12. The recombination to mutation rate was found to be extremely low. This is re- markable in light of the very high rate in the oceanic SAR11. This is may be due to adaptation to a more specialized niche. Finally we have compared structure-based and sequence-based methods for orthology pre- diction. A high fraction of the orfan proteins were predicted to code for intrinsically disordered proteins. Many phylogenetic methods are sensitive to heterogeneity and this needs to be taken into ac- count when doing phylogenies. There have been at least three independent genome reductions in the Alphaproteobacteria. The frequency of recombination differ greatly between freshwater and oceanic SAR11. Forces affecting the size of bacterial genomes and mechanisms of evolu- tionary change depend on the environmental context.

Bacteria are mediators of biogeochemical cycles and are in this way vital for maintaining life on earth. Their distribution, abundance and functioning are driven by environmental heterogeneity and dynamic change in abiotic and biotic factors. Both, freshwater lakes and oceans play central roles in the global carbon cycle and bacteria in these systems perform many services for the ecosystems, such as the transfer of organic carbon from primary producers to higher trophic levels. With estimated relative abundances up to 50% of the total bacterioplankton, the Alphaproteobacteria lineage SAR11 is the most abundant group of aquatic bacteria. It is globally distributed and can be partitioned into multiple sub-clades, one of which is exclusive to freshwaters. Until recently, the distribution, abundance and ecological role of this freshwater SAR11 named LD12 was unknown. The aim of the thesis was to study the drivers and mechanisms that influence the dynamics of aquatic bacterial communities in general and the SAR11 and LD12 groups in particular. The thesis consists of environmental surveys of a mesotrophic Lake Erken and the western Southern Ocean, an experiment and a data-mining exercise to reveal the phylogenetic structure of the SAR11 lineage on various temporal and spatial scales. The analysis of a long-term bacterioplankton community survey in lake Erken provided insights about the dynamics of the entire bacterial community and the LD12 population over an annual cycle. The results demonstrate that LD12 can be an equally abundant member of freshwater communities as marine SAR11 in oceans. LD12 featured strong seasonality and was positively coupled to environmental conditions indicative for an oligotrophic lifestyle. LD12 as well as other dominant lake bacterioplankton also maintained stable populations throughout spatial and temporal varying environments, but at high phylogenetic resolution, habitat preferences were revealed, particularly in response to oxygen concentrations. The later was not the case in LD12 as a single ribotype dominated. This is in stark contrast to the habitat partitioning with light availability, depth and water masses observed for marine SAR11 subclades in the Southern Ocean. The global data-mining corroborated that LD12 as a group was much less diverse than SAR11 furthermore, suggesting that the marine-freshwater barrier acted as a population bottleneck. My work shows that bacterial populations can respond in very different ways to environmental drivers, highlight the importance of highly resolved temporal and spatial scales as well as the need for high phylogenetic resolutions to target ecologically coherent populations.

Batch cultures of Candidatus Pelagibacter ubique were grown under iron-, organosulfur-, and nitrogen-limiting conditions to understand how this ubiquitous marine bacterium responds to and interacts with environments where growth is limited by the availability of these nutrients. Global gene expression was monitored using microarrays and quantitative mass spectrometry to observe both transcriptional and post-transcriptional responses to nutrient limitation. Iron- and nitrogen-limited cultures were characterized by increased transcription and translation of transporters involved in acquisition of the limiting nutrient, whereas organosulfur-limited cultures were not. Methionine synthesis genes downstream of S-adenosyl methionine riboswitches were up-regulated in mRNA and protein during organosulfur-replete stationary phase. Comparative genomics also revealed Ca. Pelagibacter to be the only genus among the free-living Alphaproteobacteria to lack a P[subscript II]-mediated nitrogen regulatory pathway – a pathway which may be complemented in Ca. P. ubique by putative riboswitches and a citric acid cycle able to bypass the glutamate precursor 2-oxoglutarate. Overall, the results of this study provide insight into the regulatory and metabolic processes of this ecologically significant organism, and enable better interpretation of metatranscriptomic and meta-proteomic surveys by identifying sfuC and amtB as likely biomarkers for iron and nitrogen limitation, respectively, in natural Ca. P. ubique populations. / Graduation date: 2012 / Access restricted to the OSU Community at author's request from Dec. 13, 2011 - Dec. 13, 2013

Life in oceans originated about 3.5 billion years ago where microbes were the only life form for two thirds of the planet’s existence. Apart from being abundant and diverse, marine microbes are involved in nearly all biogeochemical processes and are vital to sustain all life forms. With the overgrowing number of data arising from culture-independent studies, it became necessary to improve culturing techniques in order to obtain pure cultures of the environmentally significant bacteria to back up the findings and test hypotheses. Particularly in the ultra-oligotrophic Red Sea, the ubiquitous SAR11 bacteria has been reported to account for more than half of the surface bacterioplankton community. It is therefore highly likely that SAR11, and other microbial life that exists have developed special adaptations that enabled them to thrive successfully. Advances in conventional culturing have made it possible for abundant, unculturable marine bacteria to be grown in the lab. In this study, we analyzed the effectiveness of the media LNHM and AMS1 in isolating marine bacteria from the Red Sea, particularly members of the SAR11 clade. SAR11 strains obtained from this study AMS1, and belonged to subgroup 1a and phylotype 1a.3. We also obtained other interesting strains which should be followed up with in the future. In the long run, results from this study will enhance our knowledge of the pelagic ecosystem and allow the impacts of rising temperatures on marine life to be understood.

Given the high salinity, prevailing annual high temperatures, and ultra-oligotrophic conditions in the Red Sea isolation and characterization of important microbial groups thriving in this environment is important in understanding the ecological significance and metabolic capabilities of these communities. By using a high-­throughput cultivation technique in natural seawater amended with minute amounts of nutrients, members of the rare biosphere (PS1), methylotrophic Betaproteobacteria (OM43), and the ubiquitous and abundant SAR11 group (Pelagibacterales), were isolated in pure culture. Phylogenetic analyses of Red Sea isolates along with comparative genomics with close representatives from disparate provinces revealed ecotypes and genomic differentiation among the groups. Firstly, the PS1 alphaproteobacterial clade was found to be present in very low abundance in several metagenomic datasets form divergent environments. While strain RS24 (Red Sea) harbored genomic islands involved in polymer degradation, IMCC14465 (East (Japan) Sea) contained unique genes for degradation of aromatic compounds. Secondly, methylotrophic OM43 bacteria from the Red Sea (F5, G12 and H7) showed higher similarities with KB13 isolate from Hawaii, forming a ‘H-­RS’ (Hawaii-­Red Sea) cluster separate from HTCC2181 (Oregon isolate). HTCC2181 members were shown to prevail in cold, productive coastal environments and had an nqrA-­F system for energy generation by sodium motive force. On the contrary, H-­RS cluster members may be better adapted to warm and oligotrophic environments, and seem to generate energy by using a proton-­translocating NADH:Quinone oxidoreductase (complex I; nuoA-­N subunits). Moreover, F5, G12, and H7 had unique proteins related to resistance to UV, temperature and salinity, in addition to a heavy metal ‘resistance island’ as adaptive traits to cope with the environmental conditions in the Red Sea. Finally, description of the Red Sea Pelagibacterales isolates from the Ia (RS39) and Ib (RS40) subgroups, principally revealed unique putative systems for iron uptake and myo-inositol utilization in RS39, and a potential phosphonates biosynthetic pathway present in RS40. The findings presented here reflect how environments influence the genomic repertoire of microbial communities and shows novel metabolisms and putative pathways as unique adaptive qualities in diverse microbes encompassing from rare to predominant bacterioplankton groups from the Red Sea.