Retrotransposon Silencing by piRNAs: Ping-Pong Players Mark Their Sub-Cellular Boundaries

Germ cells of many animals exhibit characteristic cytoplasmic structures— termed germ granules or nuage—which are ribonucleoprotein (RNP) amorphous aggregates without limiting membranes and are often closely associated with nuclei or mitochondria [1]. In several model animals, such as Drosophila, Caenorhabditis elegans, and Xenopus, studies on germ granules have mainly focused on their asymmetric partitioning to prospective germ cells in early embryogenesis, leading to a supposition that the RNP structures contribute to the establishment of the germline. In mammals, in contrast, germ granules become discernible at later stages of germ cell differentiation, i.e., in spermatogenesis and oogenesis, and are not asymmetrically segregated. Thus, their possible function seems different from those in early embryos of other species. Despite the difference, germ granules of diverse animals, including mammals, not only share morphological similarities, but their molecular compositions are also conserved [2], suggesting a common and essential function in the germline, which however remained unanswered for many years. Recently, almost a century after the first description of germ granules [1], this longstanding enigma, or at least a part of it, is finally being unraveled. Accumulating evidence now points to a close association of germ granules with retrotransposon control and, especially, the piwi-small RNA pathway. Germ cells of many animals, from worms to mammals, are equipped with specific members of the argonaute subfamily, the piwi proteins, which associate with piwi-interacting small RNAs (piRNAs), and this small RNA pathway is critical for retrotransposon silencing in the germline (and gonadal soma in Drosophila) [3]. In mice, Mili/ Piwil2 and Miwi2/Piwil4, two mouse piwi members, are central to the feed-forward or ping-pong production of piRNAs from retrotransposon and other cellular transcripts in the male germline, and their lossof-function mutations lead to deregulation of Line-1 and IAP retrotransposons, resulting in male sterility with meiotic spermatocyte defects [4–6]. In this issue of PLoS Genetics, Aravin and colleagues in the Bortvin and Hannon groups report the remarkable finding that both piwi proteins exhibit distinct subcellular compartmentalization in fetal prospermatogonia/gonocytes, wherein retrotransposon silencing is established during male germline development [7]. MILI localizes to inter-mitochondrial cement (also called inter-mitochondrial material/ bar/cloud etc.), a form of germ granules commonly observed in divergent animals, and MIWI2 accumulates at processing bodies (P-bodies), an mRNP assembly widely conserved in eukaryotes from yeast to humans and whose presumed function is general RNA degradation/translational control, including those mediated by miRNAs and siRNAs [8]. These distinct MILI and MIWI2 granules contain tudor domain containing proteins TDRD1 and TDRD9, respectively, which also operate in the piwi pathway to suppress retrotransposons [9–12]. The discrete localization of MILI-TDRD1 to germ granules and MIWI2-TDRD9 to processing bodies indicates that the two RNP complexes, which are often found in close proximity, represent functionally separate assemblies of the small RNA machineries that likely co-operate and interdependently function in piRNA biogenesis and retrotransposon silencing. Aravin et al. add another key player, Maelstrom (MAEL), in the piwi-small RNA pathway. MAEL has a HMG box and a domain homologous to DnaQ-H 39-59 exonuclease, and is conserved from protists to mammals [13–16]. The authors show that the MAEL protein in mice is specifically colocalized with MIWI2-TDRD9 granules in the cytoplasm of prospermatogonia, in addition to their accumulation in the nucleus. The MIWI2-TDRD9-MAEL complex was shown to correspond to a subpopulation of processing bodies as identified by P-body markers DCP1a, DDX6, XRN1, and GW182. They named this subpopulation of processing bodies containing MIWI2-TDRD9-MAEL ‘‘piPbodies’’ and inter-mitochondrial cement localized with MILI-TDRD1 ‘‘pi-bodies’’. The latter term was recently also proposed for Drosophila nuage enriched with piwi pathway components, retroelement transcripts, and processing body components [17]. In mice, the MAEL localization is dependent on the Mili function, but not vice versa, similarly to the requirement of Mili for MIWI2 and TDRD9 localizations [10,18], and then Mael regulates the assembly of MIWI2 and TDRD9 onto piP-bodies. Thus, MAEL acts downstream of MILI and upstream of MIWI2-TDRD9 with respect to the subcellular compartmentalization in fetal prospermatogonia in mice. Previously, the authors reported that Mael gene–targeted mice are male-sterile and show a strong activation of Line-1 retrotransposon in postnatal testes [16]. Now, they extend their analysis of Mael mutants to fetal prospermatogonia and uncover a striking finding that piRNA production is severely impaired at embryonic day 16.5 (E16.5) with transposonderived piRNAs being virtually absent in Mael mutants, but the defect largely recovers at postnatal day 2 (P2), while secondary piRNAs, which preferentially load onto MIWI2, are under-represented by several-fold. Together with the precise colocalization of MAEL with MIWI2 at piP-bodies and their epistatic relationship,