Dating the impossible: the origin and divergence of the largest infection on Earth

Wolbachia are common intracellular bacteria of many arthropod and nematode species; they are responsible for likely the most widespread and evolutionary significant infection on Earth. Wolbachia are generally acquired by maternal transmission, but they often move across species boundaries, impairing the use of host divergences as independent calibration to infer their evolutionary history. Here we present a first systematic effort to estimate the origin and the divergence of the Wolbachia infection by coupling the molecular clock analysis of a genomic and an MLST data sets with three types of disputed calibration priors: a root prior for splits within Proteobacteria, the recently proposed Nomada host-symbiont co-divergence, and an adjusted mutation rate gathered from Drosophila melanogaster and Nasonia species. We first perform a hierarchical based model comparison on the MLST data set to select the best clock, replacement, and tree priors; we then used the best prior combination to calibrate both the genomic and the MLST data sets using different combinations of calibrations. Although divergence estimates are characterized by a high degree of uncertainty and strongly depend on the use of certain calibrations, most results point towards a pre-Cambrian origin of Wolbachia (its split from Ehrlichia). Diversification of Wolbachia supergroups is more recent, from the late Paleozoic to the Mesozoic: this is compatible with a long-fuse model of evolution, and suggests a co-radiation with holometabolan insects. Our results provide a comprehensive first effort to date Wolbachia evolution: results are model, prior, and data set dependent and reveal the difficulties embedded in the molecular dating of such elusive bacteria characterized by the absence of reliable calibration priors and a clear evolutionary history