Fig. 14

A hypothetical evolutionary scenario for the metazoan eTudor domain proteins essential to the piRNA pathway. The predicted evolution of Tudor protein function and domain composition is depicted. In the first stage, the core Tudor domain is involved in recognition of methylarginine in histone tails, with an unknown prokaryotic role. Next, this domain is combined with an N-terminal α-helix in an ancestral SMN-like splicing factor in the eukaryotic stem lineage, binding SDMA in spliceosomal Sm proteins. Subsequently, but also during the eukaryotic stem phase, this type of Tudor domain was inserted into an SNase domain in the ancestor of the Tudor-SN-like family, forming the eTudor domain. The Tudor-SN protein class became involved in RNAi and splicing, primarily binding SDMA in spliceosomal Sm proteins and potentially in other proteins as well. During the diversification of eukaryotes, in ancestral choanoflagellates, the eTudor domain was fused to the DCD domain, forming a putative mRNA maturation factor binding SDMA in spliceosomal Sm proteins with eTudor and m⁵C RNA with DCD. This type of protein may play a role in mRNA nuclear export, self- vs. non-self-discrimination, and protein degradation, with N-terminal Znf domains that might bind ubiquitin or RNA. At the next step, after the divergence of the loricate and non-loricate choanoflagellates, duplication of the eTudor domain in the non-loricate choanoflagellates allowed elaboration of the function of the eTudor-DCD protein, freeing one eTudor domain from the conserved function of preferential Sm protein binding and allowing it to interact preferentially with other proteins containing SDMA, eventually including Piwi-related Argonautes, but perhaps not yet at this stage. The involvement of the CAPAM-like methyltransferase domain is not necessarily part of the piRNA pathway origins, but the possibility is suggested by the data. In the final link to the extant piRNA systems, we propose a hypothetical factor, potentially still binding methylated Sm proteins, that evolved to target mRNA via interaction with piRNA/Argonaute complexes rather than binding m⁵C. These complexes now discriminate spliced and modified self from non-self/aberrant mRNA, regulating nuclear export, as well as transcription via Argonaute-directed chromatin modification, a conserved feature of eukaryotic cells also extensively performed by plants. N-terminal Znfs contribute to eTudor localization to perinuclear germ granules and possibly link them to the ubiquitination and protein degradation system to provide even more control over expression