Four hypotheses for the inter-relationships of four taxa, the sipunculans and the three annelids: the clitellate Lumbricus and the two polychaetes Nereis and Capitella. (a) Morphological cladistic analysis (Rouse & Fauchald 1997) suggests that Annelida is monophyletic, as is Polychaeta, and that the last common ancestor of Nereis and Capitella is the last common ancestor of all living polychaetes. (b,c) Molecular studies suggest that annelids are paraphyletic with respect to sipunculans and that polychaetes are paraphyletic with respect to clitellates, although the exact position of key taxa varies: (b) some studies suggest a basal position for Capitella (Rousset et al. 2007); while others (c) suggest a basal position for Nereis (Colgan et al. 2006; Struck et al. 2007). (d) The fossil record suggests that annelids are monophyletic with respect to sipunculans, but that polychaetes are paraphyletic with respect to clitellates. Note that no indication is given whether Lumbricus is more closely related to Nereis or to Capitella, just that the presence of biramous parapodia is primitive for Annelida. Black circle indicates the annelid crown group; grey circles the sipunculan and annelid total groups. Crosses indicate the age of stem-group members of each of the two phyla (Conway Morris 1979; Huang et al. 2004; Conway Morris & Peel 2008); note that the Burgess Shale (Middle Cambrian) polychaetes are shown as a simple polytomy for illustrative purposes only. Abbreviations are as follows: EC, Early Cambrian; MC, Middle Cambrian; LC, Late Cambrian; EO, Early Ordovician.

Exploration of the annelid crown group for annelid-specific miRNAs. (a) Phylogenetic perspective of the annelid taxa explored by Northern analysis and (b,c) genome walking in relation to Nereis and Capitella, according to Rouse & Fauchald (1997). (b) Northern analysis using probes derived from the sequences of miR-2688 and miR-10c-antisense against the total RNA derived from the indicated taxon (see electronic supplementary material, file 2). Note that all annelids queried express miR-2688 as an approximately 22 nt RNA, but only Abarenicola appears to express miR-10c-antisense. (c) Predicted secondary structure of miR-2688 from the polychaete annelid Chaetopterus sp. derived from genome walking (§2).

miRNAs suggest that annelids are monophyletic with respect to the sipunculan P. agassizii, but that polychaetes are paraphyletic with respect to the clitellate Lumbricus sp. (a) miRNA family 2686, transcripts of which were found only in Capitella and in Lumbricus. In both taxa, multiple mature reads were found (see alignment), and in Capitella all three genes are embedded in the same genomic trace (shown in bold capital letters) and are likely to be processed from a single polycistronic transcript. The star sequence of miR-2686 was also found in the Capitella small RNA library (shown in bold small letters both on the polycistron and in the structure of miR-2686a). (b) Predicted secondary structure (top, miR-1995) and Northern analyses (bottom) for miRNAs restricted to the sipunculan + annelid clade. As expected from the 454 library reads (electronic supplementary material, file 1), transcripts of miR-1997 were detected in the sipunculan Phascolosoma (P) and the two annelids Nereis (N) and Lumbricus (L), but not in the mollusc Mytilus (M). In addition, transcripts of miR-2688 were only detected in the two annelids, and transcripts of miR-10c-antisense were detected only in Lumbricus; transcripts of miR-1990 were only detected in the bivalve, but not in the sipunculan nor in any of the annelids, consistent with the 454 read data (electronic supplementary material, file 1). (c) Maximum parsimony analysis of 71 miRNA families in 13 metazoan taxa using the demosponge Amphimedon as the outgroup (electronic supplementary material, file 3). The maximum parsimony analysis with all characters unordered and unweighted finds 11 equally shortest trees at 107 steps (CI = 0.66; RI = 0.75). Bremer support values are shown at each node. Black circles indicate that the miRNA family is known to occur in that taxon; empty circles indicate that the miRNA family was not cloned in the library (e.g. Chaetoderma) or not found in the genomic traces (e.g. Lottia) in that specific taxon. miRNAs not detected in our Capitella small RNA library, but found in the genome, are shown in grey. Note that lowly expressed miRNAs are often absent from miRNA libraries (e.g. miR-124), but present in genomic sequences when available (electronic supplementary material, file 1).