The Piwi-piRNA (PIWI-interacting RNA) complex (PiwipiRISC) in Drosophila ovarian somatic cells represses transposons transcriptionally to maintain genome integrity; however, the underlying mechanisms remain obscure. Here, we reveal that DmGTSF1, a Drosophila homolog of gametocyte-specific factor 1 (GTSF1) (which is required for transposon silencing in mouse testes), is necessary for Piwi-piRISC to repress target transposons and neighboring genes. DmGTSF1 depletion affected neither piRNA biogenesis nor nuclear import of PiwipiRISC. DmGTSF1 mutations caused derepression of transposons and loss of ovary follicle layers, resulting in female infertility. We suggest that DmGTSF1, a nuclear Piwi interactor, is an integral factor in Piwi-piRISCmediated transcriptional silencing.

The establishment of body axes in multicellular organisms requires accurate control of microtubule polarization. Mutations in Drosophila PIWI-interacting RNA (piRNA) pathway genes often disrupt the axes of the oocyte. This results from the activation of the DNA damage checkpoint factor Checkpoint kinase 2 (Chk2) due to transposon derepression. A piRNA pathway gene, maelstrom (mael), is critical for the establishment of oocyte polarity in the developing egg chamber during Drosophila oogenesis. We show that Mael forms complexes with microtubuleorganizing center (MTOC) components, including Centrosomin, Mini spindles, and gTubulin. We also show that Mael colocalizes with aTubulin and gTubulin to centrosomes in dividing cyst cells and follicle cells. MTOC components mislocalize in mael mutant germarium and egg chambers, leading to centrosome migration defects. During oogenesis, the loss of mael affects oocyte determination and induces egg chamber fusion. Finally, we show that the axis specification defects in mael mutants are not suppressed by a mutation in mnk, which encodes a Chk2 homolog. These findings suggest a model in which Mael serves as a platform that nucleates other MTOC components to form a functional MTOC in early oocyte development, which is independent of Chk2 activation and DNA damage signaling.

PIWI-interacting RNAs (piRNAs) mediate transcriptional and post-transcriptional silencing of transposable element (TE) in animal gonads. In Drosophila ovaries, Piwi-piRNA complexes (Piwi-piRISCs) repress TE transcription by modifying the chromatin state, such as by H3K9 trimethylation. Here, we demonstrate that Piwi physically interacts with linker histone H1. Depletion of Piwi decreases H1 density at a subset of TEs, leading to their derepression. Silencing at these loci separately requires H1 and H3K9me3 and heterochromatin protein 1a (HP1a). Loss of H1 increases target loci chromatin accessibility without affecting H3K9me3 density at these loci, while loss of HP1a does not impact H1 density. Thus, Piwi-piRISCs require both H1 and HP1a to repress TEs, and the silencing is correlated with the chromatin state rather than H3K9me3 marks. These findings suggest that Piwi-piRISCs regulate the interaction of chromatin components with target loci to maintain silencing of TEs through the modulation of chromatin accessibility.

Piwi-interacting RNAs (piRNAs) suppress transposon activity in animal germ cells. In the Drosophila ovary, primary Aubergine (Aub)-bound antisense piRNAs initiate the ping-pong cycle to produce secondary AGO3-bound sense piRNAs. This increases the number of secondary Aub-bound antisense piRNAs that can act to destroy transposon mRNAs. Here we show that Krimper (Krimp), a Tudor-domain protein, directly interacts with piRNA-free AGO3 to promote symmetrical dimethylarginine (sDMA) modification, ensuring sense piRNA-loading onto sDMA-modified AGO3. In aub mutant ovaries, AGO3 associates with ping-pong signature piRNAs, suggesting AGO3's compatibility with primary piRNA loading. Krimp sequesters ectopically expressed AGO3 within Krimp bodies in cultured ovarian somatic cells (OSCs), in which only the primary piRNA pathway operates. Upon krimp-RNAi in OSCs, AGO3 loads with piRNAs, further showing the capacity of AGO3 for primary piRNA loading. We propose that Krimp enforces an antisense bias on piRNA pools by binding AGO3 and blocking its access to primary piRNAs.

The PIWI-interacting RNA (piRNA) pathway preserves genomic integrity by repressing transposable elements (TEs) in animal germ cells. Among PIWI-clade proteins in Drosophila, Piwi transcriptionally silences its targets through interactions with cofactors, including Panoramix (Panx) and forms heterochromatin characterized by H3K9me3 and H1. Here, we identified Nxf2, a nuclear RNA export factor (NXF) variant, as a protein that forms complexes with Piwi, Panx, and p15. Panx-Nxf2-P15 complex formation is necessary in the silencing by stabilizing protein levels of Nxf2 and Panx. Notably, ectopic targeting of Nxf2 initiates co-transcriptional repression of the target reporter in a manner independent of H3K9me3 marks or H1. However, continuous silencing requires HP1a and H1. In addition, Nxf2 directly interacts with target TE transcripts in a Piwidependent manner. These findings suggest a model in which the Panx-Nxf2-P15 complex enforces the association of Piwi with target transcripts to trigger co-transcriptional repression, prior to heterochromatin formation in the nuclear piRNA pathway. Our results provide an unexpected connection between an NXF variant and small RNA-mediated co-transcriptional silencing. The EMBO JournalKensaku Murano et al F Density plots for normalized H3K9me3 ChIP-seq signals over the consensus sequence from mdg1, gypsy (targeted by Piwi-piRNA, orange), and roo (not targeted by Piwi-piRNA, blue) TEs in EGFP-, Piwi-, Panx-, and Nxf2-KD OSCs. G Boxplots as in (D) showing fold changes in the H3K9me3 levels of Piwi-piRNA-targeted and un-targeted TEs upon the indicated KD. H Northern blotting for Idefix-piRNA, traffic jam (tj)-piRNA, and esiRNA sl-1 (control) on total RNA isolated from OSCs with the indicated KD.