Plant miRNA evolution update

So here’s another figure I’ve prepared for the Plant Cell review article I am writing. This is an update on the patterns of miRNA conservation across land plants according solely to the information present in miRBase release 21.

For this analysis, I first placed each land plant miRNA family in miRBase 21 into one of eight broad plant groups (Eudicots-Rosids, Eudicots-Asterids, Basal Eudicots, Monocots, Basal Angiosperms, Gymnosperms, Lycophytes, and Bryophytes). I then defined a conserved miRNA family as one which had at least one high-confidence annotation, and which was annotated in two or more of these broad groups. By this definition, there are just 36 conserved families out of the more than 2,000 that are currently annotated.

grid_fig

This figure also highlights a major issue in the large-scale analysis of land plant miRNA conservation .. highly unequal sampling density within the different groups. Rosids and monocots have received the most attention, with large numbers of species represented, and high numbers of overall annotations (barcharts at the top). In contrast, basal eudicots, basal angiosperms, lycophytes, and bryophytes are much less well-sampled, with each group represented at present by just one species in miRBase 21. Inferring secondary losses in some of these lineages (basal eudicots, basal angiosperms, lycophytes) is NOT believable.

This chart also illustrates how important the high-quality miRNA annotation set for bryophytes is (the sole species represented is Physcomitrella patens). Based on the presence of high-confidence annotations in both Physcomitrella and one or more angiosperm group, we can confidently say that nine families (miR156, miR160, miR166, miR171, miR319, miR390, miR477, miR529, and miR535) were most likely present in the last common ancestor of all land plants. Another three families (miR167, miR395, and miR408) might also belong in the ultra-conserved set, but they are less certain because their Physcomitrella annotations are not high-confidence. Another two families (miR396 and miR482) clearly predate divergence of all seed plants. Other patterns are less certain because of the frequent presence of annotations that are not (yet) known to be high confidence. I think this again highlights the need for a systematic review of miRNA annotations, based on re-analysis of all available small RNA-seq data with a single, high-confidence MIRNA identification methodology. We are working toward this goal in my lab.

Leave a Reply

Your email address will not be published. Required fields are marked *