Impacts of gene duplication in the evolution of symbiotic root nodule symbiosis in legumes.

Root nodule symbiosis (RNS) is found in approximately 16-18 widely-separated lineages within the "nitrogen-fixing nodulation clade (NFNC)". Although modeling of trait gain and loss across approximately 13,000 species within the rosid group indicates multiple gains and losses, there is no consensus about whether RNS had a single or multiple origins; and our understanding is fragmentary regarding the molecular mechanisms underlying those changes. Evolution of a new organ and functions involves many thousands of genes; but the evolutionary histories for many of these genes may be uninformative regarding RNS evolution. A portion of the genes, however, are likely to be derived from prior gene duplications and to have acquired new functions or to have come under new regulatory patterns. Whole genome duplications (WGDs) could conceivably enable the necessary neo- or sub-functionalization for new roles in the nodule. All species that exhibit RNS share a history of several ancient WGDs; but the last such common WGD for these species was the "gamma" paleohexaploidy that occurred early in the core eudicot lineage, ~120 Mya. This presents a puzzle: If legume RNS within the NFNC only arose in the Late Cretaceous, several tens of millions of years after the gamma event, what explains the long, seemingly quiescent interval and the many eudicot lineages without RNS? This study focuses on a collection of gene superfamilies with additional independent WGDs that appear to have occurred in the interim period, after the gamma triplication and prior to the evolution of RNS, identifying several that are both essential for RNS and that show evidence of critical roles of both ancient WGDs and more recent local duplications.