The Expansion of Animal MicroRNA Families Revisited

Abstract

MicroRNAs are important regulatory small RNAs in many eukaryotes. Due to their small size and simple structure, they are readily innovated de novo. Throughout the evolution of animals, the emergence of novel microRNA families traces key morphological innovations. Here, we use a computational approach based on homology search and parsimony-based presence/absence analysis to draw a comprehensive picture of microRNA evolution in 159 animal species. We confirm previous observations regarding bursts of innovations accompanying the three rounds of genome duplications in vertebrate evolution and in the early evolution of placental mammals. With a much better resolution for the invertebrate lineage compared to large-scale studies, we observe additional bursts of innovation, e.g., in Rhabditoidea. More importantly, we see clear evidence that loss of microRNA families is not an uncommon phenomenon. The Enoplea may serve as a second dramatic example beyond the tunicates. The large-scale analysis presented here also highlights several generic technical issues in the analysis of very large gene families that will require further research.

A Uniform System for the Annotation of Vertebrate microRNA Genes and the Evolution of the Human microRNAome

“…For example, Hertel & Stadler (68; see also 57, 90) used miRBase in its entirety with no quality control, despite the large body of work, consistent with our own analyses (Figure 3b), suggesting that it contains numerous false positive miRNAs (22, 25, 77, 85, 101, 143, 146, 153), including almost all of the supposed aberrant miRNAs found by Kenny et al (78). Further, in both Hertel & Stadler (68) and Kenny et al (78) no attempt was made to analyze the secondary structure of putative orthologs to bona fide miRNAs (which, when examined, do not support a miRNA assignment), and no small RNA data sets support their processing. Finally, no attempt was made to distinguish between false negatives versus genuine absences (146), an issue that also plagues the study of Thomson et al (147).…”

“…Thus, approximately 10% of the human miRNA repertoire at the family level and at the gene level have been lost over time. This difference between gene gain and loss is the same order of magnitude recorded by Tarver et al (144) across the animal kingdom, as opposed to other studies that have suggested much higher rates of loss (57, 68, 78, 90, 147). Each of these latter studies has flaws, greatly biasing the results.…”

Rapid Evolutionary Rates and Unique Genomic Signatures Discovered in the First Reference Genome for the Southern Ocean Salp,Salpa thompsoni(Urochordata, Thaliacea)

“…For example, bursts in the number of novel miRNA families occurred in the base of vertebrates (48 miRNA families acquired) and in the lineages leading to therian mammals (63 miRNA families acquired) and primates (414 miRNA families acquired) (reviewed in Candiani 2012). However, subsequent studies of two of the three Tunicate classes, appendicularians ( O. dioica ) and ascidians ( C. intestinalis and C. savignyi ), revealed evidence of both loss and acquisition of new miRNA families, albeit the latter at rates far lower that observed in primates (Fu et al 2008; Dai et al 2009; Shi et al 2009; Hendrix et al 2010; Candiani 2012; Hertel and Stadler 2015). These data support the hypothesis that while conserved miRNAs likely define the “robustness” of developmental programs (e.g., temporally) (Fu et al 2008), lineage specific morphological and developmental novelty is coupled with novel miRNA acquisition across urochordates (Candiani 2012; Hertel and Stadler 2015).…”

Automated detection of ncRNAs in the draft genome sequence of a colonial tunicate: the carpet sea squirt Didemnum vexillum

“…Different miRNAs reported by [8] and hits homologous to Ensembl miRNA queries sequences, were not classified in a correct way by covariance models as noted at Additional file 8. For further validation manual curation was performed, against miRNAs families.…”