The majority of human infections originate from animals, however in many cases we have limited knowledge of when or how these infections first emerged. Knowledge of this timeline could help us understand the ecological and evolutionary processes that lead to spillover. Bartonella spp. bacteria are a great model system for studying the evolution of parasites and host-specificity. These bacteria have a prolific ability to adapt to specific hosts and evolve into separate clades that infect different orders of mammals. Additionally, early Bartonella species and ancestors may have been symbionts of arthropods, including ants, bees, and mites, and later evolved to parasitize mammals. This study seeks to determine when this transition from symbiont to parasite occurred and whether the timing of Bartonella diversification corresponds to the diversification of mammals during the Cretaceous Period (80-92 million years ago, MYA). We compiled molecular data from nine genetic loci from known Bartonella species and additional cultures from bats to build a phylogenetic tree. We parameterized a molecular clock for the 16S rRNA gene using mutation rates in bacterial symbionts of arthropods and assigned diffuse prior distributions for the mutation rates of the eight other loci. Molecular clocks were allowed to vary across branches of the tree by assigning a lognormal prior for each locus. We then estimated a Bayesian phylogenetic tree and projected the age of the most recent common ancestor of all Bartonella species known to infect mammals.
Results/Conclusions
The results show that mutation rates vary by at least an order of magnitude over the nine sequenced loci, with 16S rRNA having the lowest rate at 4.5x10-10 substitutions/site/year. The combined molecular clocks from all loci show that Bartonella emerged as a parasite of mammals 88 MYA (95% credible interval 52-136 MYA), during the diversification of the crown clades of mammals 86 MYA (95% confidence interval 80-92 MYA). Additionally, the timing of diversification and biogeography of particular clades of bat-, feline-, and ruminant-associated Bartonella species correspond strongly to the timing of diversification and biogeography in the mammalian hosts. Overall, this study shows that Bartonella evolves slowly but can become highly host-specific over time and with sufficient phylogenetic and geographic isolation of hosts. This approach could be useful for understanding the diversification and evolution of host ranges in other bacterial and viral parasites.