Angiosperms Are Unique among Land Plant Lineages in the Occurrence of Key Genes in the RNA-Directed DNA Methylation (RdDM) Pathway

The RNA-directed DNA methylation (RdDM) pathway can be divided into three phases: 1) small interfering RNA biogenesis, 2) de novo methylation, and 3) chromatin modification. To determine the degree of conservation of this pathway we searched for key genes among land plants. We used OrthoMCL and the OrthoMCL Viridiplantae database to analyze proteomes of species in bryophytes, lycophytes, monilophytes, gymnosperms, and angiosperms. We also analyzed small RNA size categories and, in two gymnosperms, cytosine methylation in ribosomal DNA. Six proteins were restricted to angiosperms, these being NRPD4/NRPE4, RDM1, DMS3 (defective in meristem silencing 3), SHH1 (SAWADEE homeodomain homolog 1), KTF1, and SUVR2, although we failed to find the latter three proteins in Fritillaria persica, a species with a giant genome. Small RNAs of 24 nt in length were abundant only in angiosperms. Phylogenetic analyses of Dicer-like (DCL) proteins showed that DCL2 was restricted to seed plants, although it was absent in Gnetum gnemon and Welwitschia mirabilis. The data suggest that phases (1) and (2) of the RdDM pathway, described for model angiosperms, evolved with angiosperms. The absence of some features of RdDM in F. persica may be associated with its large genome. Phase (3) is probably the most conserved part of the pathway across land plants. DCL2, involved in virus defense and interaction with the canonical RdDM pathway to facilitate methylation of CHH, is absent outside seed plants. Its absence in G. gnemon, and W. mirabilis coupled with distinctive patterns of CHH methylation, suggest a secondary loss of DCL2 following the divergence of Gnetales.