Evolution of coumaroyl conjugate 3‐hydroxylases in land plants: lignin biosynthesis and defense

Multiple adaptations were necessary when plants conquered the land. Among them were soluble phenylpropanoids related to plant protection and lignin necessary for upright growth and long‐distance water transport. Cytochrome P450 monooxygenase 98 (CYP98) catalyzes a rate‐limiting step in phenylpropanoid biosynthesis. Phylogenetic reconstructions suggest that a single copy of CYP98 founded each major land plant lineage (bryophytes, lycophytes, monilophytes, gymnosperms and angiosperms), and was maintained as a single copy in all lineages but the angiosperms. In angiosperms, a series of independent gene duplications and losses occurred. Biochemical assays in four angiosperm species tested showed that 4‐coumaroyl‐shikimate, a known intermediate in lignin biosynthesis, was the preferred substrate of one member in each species, while independent duplicates in Populus trichocarpa and Amborella trichopoda each showed broad substrate ranges, accepting numerous 4‐coumaroyl‐esters and ‐amines, and were thus capable of producing a wide range of hydroxycinnamoyl conjugates. The gymnosperm CYP98 from Pinus taeda showed a broad substrate range, but preferred 4‐coumaroyl‐shikimate as its best substrate. In contrast, CYP98s from the lycophyte Selaginella moellendorffii and the fern Pteris vittata converted 4‐coumaroyl‐shikimate poorly in vitro, but were able to use alternative substrates, in particular 4‐coumaroyl‐anthranilate. Thus, caffeoyl‐shikimate appears unlikely to be an intermediate in monolignol biosynthesis in non‐seed vascular plants, including ferns. The best substrate for CYP98A34 from the moss Physcomitrella patens was also 4‐coumaroyl‐anthranilate, while 4‐coumaroyl‐shikimate was converted to lower extents. Despite having in vitro activity with 4‐coumaroyl‐shikimate, CYP98A34 was unable to complement the Arabidopsis thaliana cyp98a3 loss‐of‐function phenotype, suggesting distinct properties also in vivo.