Evolutionary timing of the origins of mesophilic sulphate reduction and oxygenic photosynthesis: a phylogenomic dating approach

ABSTRACT Until recently, the deep‐branching relationships in the bacterial domain have been unresolved. A new phylogenetic approach (termed compartmentalization) was able to resolve these deep‐branching relationships successfully by using a large number of genes from whole genome sequences and by reducing long branch attraction artefacts. This new, well‐resolved phylogenetic tree reveals the evolutionary relationships between diverse bacterial groups that leave important traces in the geological record. It shows that mesophilic sulphate reducers originated before the Cyanobacteria, followed by the origination of sulphur‐ and pyrite‐oxidizing bacteria after oxygen became available in the biosphere. This evolutionary pattern mirrors a similar pattern in the Palaeoproterozoic geological record. Sulphur isotopic fractionation records indicate that large‐scale bacterial sulphate reduction began in marine environments around 2.45 billion years ago (Ga), followed by rapid oxygenation of the atmosphere about 2.3 or 2.2 Ga. Oxygenation was then followed by increasing oceanic sulphate concentrations (probably owing to pyrite oxidation and continental weathering), which then resulted in the disappearance of banded iron formations by 1.8 Ga. The similarity between the phylogenetic and geological records suggests that the geochemical changes observed on the Palaeoproterozoic Earth were caused by major origination events in the mesophilic bacteria, and that these geochemical changes then caused additional origination events, such as aerobic respiration. If so, then constraints on divergence dates can be established for many microbial groups, including the Cyanobacteria, mesophilic bacteria, mesophilic sulphate reducers, methanotrophs, several anoxygenic phototrophs, as well as for mitochondrial endosymbiosis. These dates may also help to explain a large number of other changes in the geological record of the Neoarchean and Palaeoproterozoic Earth. This hypothesis, however, does not agree with the finding of cyanobacterial and eukaryote lipids at 2.7 Ga, and suggests that further work needs to be done to elucidate the discrepancies in both these areas.