Origin, Phylogeny and Structure of Light-driven Protochlorophyllide Reduction

The reduction of protochlorophyllide represents one key steps in the complex biosynthesis pathway yielding (bacterio)chlorophyll. In oxygenic phototrophs such as plants, algae and cyanobacteria this step is catalysed by light-driven protochlorophyllide (Pchlide) oxidoreductases (LPORs). In an evolutionary context, it is currently widely accepted that LPORs were first invented in cyanobacteria and were then transferred via endosymbiosis to plants and algae. In contrast to this long held paragdigm, we have recently identified a functional light-driven LPOR in the anoxygenic alpha-proteobacterium Dinoroseobacter shibae DFL12T (unpublished data, see preliminary work). This raises far reaching questions regarding the evolution of chlorophyll biosynthesis and photosynthesis and suggests that additional functionally diverse LPORs might be found in e.g. unculturable strains and/or specialized microbes. The proposal aims to exploit the yet unexplored sequence diversity of metagenomic/environmental sources to elucidate the evolutionary origin, phylogeny, as well as the structural and functional basis of light-driven protochlorophyllide reduction. To achieve these goals, it is necessary to combine the interdisciplinary expertise of microbiologists, molecular biologists, biochemists, bioinformaticians and structural biologists. While in vivo and in vitro functional and structural data (working groups Krauss and Drepper) will provide broad insight into functional adaptation and mechanistic aspects of light-driven Pchlide reduction by the LPOR family of enzymes, the evolutionary perspective which will be established by the working group of Arndt von Haeseler is expected to provide a new dimension and enriched scope.

Light-dependent protochlorophyllide oxidoreductases (LPORs) are key enzymes for the synthesis of (bacterio)chlorophyll, the major pigment needed for both anoxygenic and oxygenic photosynthesis. Their evolutionary history is therefore intricately linked to the evolution of photosynthesis. At present LPORs are assumed to have evolved from short- chain dehydrogenases, emerging first among cyanobacteria as a consequence of rising oxygen levels in the primordial atmosphere. While we have recently shown the presence of a functional LPOR in the aerobic anoxygenic phototrophic bacterium (AAPB) Dinoroseobacter shibae, and attributed its presence to an isolated horizontal gene transfer (HGT) event from cyanobacteria, we now provide evidence for the widespread presence of genuine LPORs in AAPB a-, ß- proteobacteria and Gemmatimonadetes. When we inferred a phylogenetic tree for the identified LPOR sequences, interestingly, we observed that all proteobacteria and Gemmatimonadetes species cluster in a monophyletic within cyanobacteria. We elaborated several evolutionary scenarios that could account for the observed tree topology. A monophyletic AAPB-LPORs clade within the cyanobacteria might point towards an ancient HGT event as a possible cause for the presence of LPORs among AAPBs. However, the expected placement of the clade closer to the root is not observed and a massive loss is necessary to account for the absence of LPOR in other AAPBs. More importantly, as validated in the lab experiments, a group of newly identified LPORs possess quite distinct biochemical properties compared to cyanobacteria and plant LPORs. Similar concerns are valid for the hypothesis that LPOR originated in AAPBs and was then transferred to cyanobacteria. Distinct biochemical properties and uncertainty in the position of the clade on the phylogenetic tree rose yet another hypothesis. Convergent evolution could be responsible for the occurrence of two biochemically distinct groups of LPORs. The enzyme could be developed by cyanobacteria and AAPBs separately as a result of adaptation to changing environmental conditions. Despite this hypothesis is appealing, there are no methods that could test it. We also performed an extensive analysis of genomic islands (GIs), the clusters of consecutive genes likely obtained via HGT. The analysis of GIs also did not provide an immediate support for HGT hypothesis. Discovering new proteobacteria LPORs has the potential to improve the present trees and would likely provide more information about their origin, thus contributing to our understanding of the evolution of photosynthesis in general. Moreover, a development of strategies for distinguishing of HGT and convergent evolution is necessary to fully resolve the question.