The timing of eukaryotic evolution: Does a relaxed molecular clock reconcile proteins and fossils?

TLDR

Nucleotide and amino acid sequences improve the timing of evolutionary events, yet molecular estimates are often older than fossils, likely due to limited genes/species and pervasive rate variation. The study compared concatenated amino acid sequences of 129 proteins from 36 eukaryotes to estimate divergence times of major eukaryotic clades. A Bayesian relaxed molecular clock calibrated with six paleontological constraints was applied to account for rate variation and fossil uncertainty. The relaxed clock estimates place eukaryotic kingdom diversification at 950–1,259 Mya, animal–choanoflagellate split at 761–957 Mya, and protostome–deuterostome split at 642–761 Mya, with results robust to priors and calibrations and indicating bilaterian diversification ~100 Mya older than the Cambrian.

Abstract

The use of nucleotide and amino acid sequences allows improved understanding of the timing of evolutionary events of life on earth. Molecular estimates of divergence times are, however, controversial and are generally much more ancient than suggested by the fossil record. The limited number of genes and species explored and pervasive variations in evolutionary rates are the most likely sources of such discrepancies. Here we compared concatenated amino acid sequences of 129 proteins from 36 eukaryotes to determine the divergence times of several major clades, including animals, fungi, plants, and various protists. Due to significant variations in their evolutionary rates, and to handle the uncertainty of the fossil record, we used a Bayesian relaxed molecular clock simultaneously calibrated by six paleontological constraints. We show that, according to 95% credibility intervals, the eukaryotic kingdoms diversified 950-1,259 million years ago (Mya), animals diverged from choanoflagellates 761-957 Mya, and the debated age of the split between protostomes and deuterostomes occurred 642-761 Mya. The divergence times appeared to be robust with respect to prior assumptions and paleontological calibrations. Interestingly, these relaxed clock time estimates are much more recent than those obtained under the assumption of a global molecular clock, yet bilaterian diversification appears to be approximately 100 million years more ancient than the Cambrian boundary.