Genetic Enablers Underlying the Clustered Evolutionary Origins of C-4 Photosynthesis in Angiosperms

The evolutionary accessibility of novel adaptations varies among lineages, depending in part on the genetic elements present in each group. However, the factors determining the evolutionary potential of closely related genes remain largely unknown. In plants, CO2-concentrating mechanisms such as C4 and crassulacean acid metabolism (CAM) photosynthesis have evolved numerous times in distantly related groups of species, and constitute excellent systems to study constraints and enablers of evolution. It has been previously shown for multiple proteins that grasses preferentially co-opted the same gene lineage for C4 photosynthesis, when multiple copies were present. In this work, we use comparative transcriptomics to show that this bias also exists within Caryophyllales, a distantly related group with multiple C4 origins. However, the bias is not the same as in grasses and, when all angiosperms are considered jointly, the number of distinct gene lineages co-opted is not smaller than that expected by chance. These results show that most gene lineages present in the common ancestor of monocots and eudicots produced gene descendants that were recruited into C4 photosynthesis, but that C4-suitability changed during the diversification of angiosperms. When selective pressures drove C4 evolution, some copies were preferentially co-opted, probably because they already possessed C4-like expression patterns. However, the identity of these C4-suitable genes varies among clades of angiosperms, and C4 phenotypes in distant angiosperm groups thus represent genuinely independent realizations, based on different genetic precursors.