In defense of Apocynaceae: inference on evolution of pyrrolizidine alkaloids from evolution of an enzyme in their biosynthetic pathway, homospermidine synthase

Abstract

AbstractIf plant defenses experience positive selection when they first evolve, evidence of selection on a gene can place the origin of novel defenses on a phylogeny. Prior research on pyrrolizidine alkaloids (PAs), rare specialized metabolites of Apocynaceae, reconstructed evolution of homospermidine synthase (HSS), an enzyme of PA biosynthesis, to infer a single origin of PAs by showing a single duplication in the MRCA of all known PA-producing species that gave rise to an enzyme with the predicted motif (VXXXD) of a functionally optimized HSS. We follow up by testing the effect of amino acid motif on HSS function, revisiting motif evolution inHSS-like genes, and testing for selection to infer evolution of HSS function. Some evidence supports a single origin of PAs: an IXXXD HSS, similar in function to VXXXD HSS, evolved in the MRCA of all PA-producing species; loss of optimized HSS function occurred multiple times via pseudogenization and perhaps via evolution of an IXXXN motif. Other evidence indicates multiple origins: the VXXXD motif, present in all PA-producing species, evolved two or three times independently; branch-site models suggest that the ancestral IXXXD branch was not under positive selection while some VXXXD branches were; there is no evidence of relaxed selection on putatively impairedHSS-like genes (IXXXN motif); substitutions at sites experiencing positive selection occurred on multiple branches in theHSS-like gene tree, suggesting that HSS function may have been optimized in multiple independent lineages. We discuss how complexity of genotype-phenotype maps confounds inference of phenotypic from genic evolution.