Parallel Recruitment of Multiple Genes into C(4) Photosynthesis

During the diversification of living organisms, novel adaptive traits usually evolve through the co-option of preexisting genes. However, most enzymes are encoded by gene families, whose members vary in their expression and catalytic properties. Each may therefore differ in its suitability for recruitment into a novel function. In this work, we test for the presence of such a gene recruitment bias using the example of C(4) photosynthesis, a complex trait that evolved recurrently in flowering plants as a response to atmospheric CO(2) depletion. We combined the analysis of complete nuclear genomes and high-throughput transcriptome data for three grass species that evolved the C(4) trait independently. For five of the seven enzymes analyzed, the same gene lineage was recruited across the independent C(4) origins, despite the existence of multiple copies. The analysis of a closely related C(3) grass confirmed that C(4) expression patterns were not present in the C(3) ancestors but were acquired during the evolutionary transition to C(4) photosynthesis. The significant bias in gene recruitment indicates that some genes are more suitable for a novel function, probably because the mutations they accumulated brought them closer to the characteristics required for the new function.