A trans-piRNA network and transcriptional antagonism shape piRNA cluster function

Abstract

PIWI-interacting (pi)RNAs protect animal germlines from transposable elements (TEs) by guiding their sequence-specific repression. In Drosophila germline, piRNAs are encoded in distinct genomic regions, piRNA clusters (piCs), that are transcribed by a non-canonical machinery that is anchored on chromatin by the HP1 paralogue Rhino. Studies of transgenic piCs revealed that piRNA biogenesis depends on cytoplasmic inheritance of piRNAs, however, whether native piCs require trans-generational piRNA transmission remained unknown. Here, we used two approaches to show that cytoplasmic inheritance of cognate piRNAs is critical for piRNA biogenesis. Our analyses reveal that individual piCs form a tightly interconnected network linked by trans-acting piRNAs that reinforce biogenesis. According to the transposon trap model, the content of piCs is updated by integration of novel TEs leading to production of piRNA guides against integrated transposons. However, we found that transcription driven by promoters integrated into piCs disrupt local piRNA biogenesis by removing Rhino and antagonizing non-canonical transcription of piRNA precursors. Thus, newly inserted transposons might suppress piRNA production before they become domesticated by the piRNA pathway calling for a revision of the trap model. Together, our results reveal that piC activity is shaped by transcriptional competition and a dynamic interplay between individual piCs connected into a common network. HighlightsO_LICytoplasmic inheritance of maternal piRNAs is required for piRNA biogenesis in the next generation C_LIO_LITrans-acting piRNA ensure robustness of piRNA biogenesis and repression by connecting individual clusters into a functional network C_LIO_LIPing-pong amplification provides non-uniform processing of different regions inside piCs C_LIO_LIGenes can remain active inside piRNA clusters antagonizing Rhino-dependent non-canonical transcription and piRNA biogenesis, challenging the transposon trap model C_LI