Gibberellins (GAs) are key phytohormones that regulate growth, development and environmental responses in angiosperms. From an evolutionary perspective, all major steps of GA biosynthesis are conserved among vascular plants, while GA biosynthesis intermediates such as ent-kaurenoic acid (KA) are also produced by bryophytes. Here we show that in the liverwort Marchantia polymorpha, KA and GA₁₂ are synthesized by evolutionarily conserved enzymes, which are required for developmental responses to far-red light (FR). Under FR-enriched conditions, mutants of various biosynthesis enzymes consistently exhibited altered thallus growth allometry, delayed initiation of gametogenesis, and abnormal morphology of gamete-bearing structures (gametangiophores). By chemical treatments and liquid chromatography–tandem mass spectrometry (LC-MS/MS) analyses, we confirmed that these phenotypes were caused by the deficiency of some GA-related compounds derived from KA, but not bioactive GAs from vascular plants. Transcriptome analysis showed that FR enrichment induced the up-regulation of genes related to stress responses and secondary metabolism in M. polymorpha, which was largely dependent on the biosynthesis of GA-related compounds. Due to the lack of canonical GA receptors in bryophytes, we hypothesize that GA-related compounds are commonly synthesized in land plants but were co-opted independently to regulate responses to light quality change in different plant lineages during the past 450 million years of evolution.

Plants synthesize gibberellin (GA), a diterpenoid hormone, via ent-kaurenoic acid (KA) oxidation. GA has not been detected in the moss Physcomitrium patens despite its ability to synthesize KA. It was recently shown that a KA metabolite, 3OH-KA, was identified as an active regulator of protonema differentiation in P. patens. An inactive KA metabolite, 2OH-KA, was also identified in the moss, as was KA2ox, which is responsible for converting KA to 2OH-KA. In this review, we mainly discuss the GA biosynthetic gene homologs identified and characterized in bryophytes. We show the similarities and differences between the OH-KA control of moss and GA control of flowering plants. We also discuss using recent genomic studies; mosses do not contain KAO, even though other bryophytes do. This absence of KAO in mosses corresponds to the presence of KA2ox, which is absent in other vascular plants. Thus, given that 2OH-KA and 3OH-KA were isolated from ferns and flowering plants, respectively, vascular plants may have evolved from ancestral bryophytes that originally produced 3OH-KA and GA.