Ovo is a master regulator of the piRNA pathway in animal ovarian germ cells

The gene-regulatory mechanisms controlling the expression of the germline PIWI- interacting RNA (piRNA) pathway components within the gonads of metazoan species remain largely unexplored. In contrast to the male germline piRNA pathway, which in mice is known to be activated by the testis-specific transcription factor A-MYB, the nature of the ovary-specific gene-regulatory network driving the female germline piRNA pathway remains a mystery. Here, using Drosophila as a model, we combine multiple genomics approaches to reveal the transcription factor Ovo as the master regulator of the germline piRNA pathway in ovaries. The enforced expression of Ovo in somatic cells activates germline piRNA pathway components, including the ping-pong factors Aubergine, Argonaute-3, and Vasa, leading to assembly of peri-nuclear cellular structures resembling nuage bodies of germ cells. Cross-species ChIP-seq and motif analyses demonstrate Ovo binding to genomic CCGTTA motifs within the promoters of germline piRNA pathway genes, suggesting a regulation by Ovo in ovaries analogous to that of A-MYB in testes. Our results also show consistent engagement of the Ovo transcription factor family at ovarian piRNA clusters across metazoan species, reflecting a deep evolutionary conservation of this regulatory paradigm from flies to humans.

In the nuage bodies of Drosophila germ cells (Figure 1a), ping-pong amplification operates via the PIWI proteins Argonaute-3 (Ago3) and Aubergine (Aub) which cleave piRNA precursor and transposon transcripts in an alternating loop, hence shaping and amplifying the piRNA pool against active transposons 3,25 .Ago3 in complex with a sense piRNA recognizes and cleaves cluster transcripts through sequence complementarity generating antisense pre-piRNAs that are loaded into Aub 3,25 .Antisense piRNA-loaded Aub in turn detects and cleaves target transposon mRNAs, thereby forming a new Ago3 complex loaded with a sense piRNA, and completing the cycle 3,25 .This process is assisted by other nuage-localised protein components, namely, the DEAD-box RNA helicase Vasa (Vas), the putative nuclease Squash (Squ), and Tudor domain proteins Tejas (Tej), Tapas, Qin, Krimper (Krimp) and Spindle-E (Spn-E) [26][27][28][29] .This is in sharp contrast to piRNA biogenesis in somatic cells, which lack nuage bodies and ping-pong amplification, and where piRNA precursors are transcribed from unistrand clusters (e.g., flamenco), exported to the cytoplasm by a canonical machinery, and processed into piRNAs via Zucchini-mediated biogenesis 30 .
The gene-regulatory network underlying the control of the germline-specific piRNA pathway in the ovaries of metazoan species have been largely unexplored.The specific expression patterns of piRNA pathway genes in germ cells could be controlled by a positive regulation, such as activation by germline-specific transcription factors (TFs), or via a negative regulation through repression by soma-specific factors.To date, clues from previous studies have pointed towards the existence of both mechanisms.
A study in mouse testes have identified the TF A-MYB as the transcriptional activator of piRNA pathway components during spermatogenesis, including piRNA factors such as Piwil1 (Miwi) and the germline pachytene piRNA clusters 31 .This study revealed the existence of a positive transcriptional regulator of the male germline piRNA pathway in metazoan testes, yet such transcriptional regulatory pathways controlling the expression of the female germline piRNA pathway in the ovaries of metazoan species, including Drosophila, remain to be identified.On the other hand, earlier work has suggested the existence of negative regulation of germline piRNA factors in Drosophila somatic cells [32][33][34] .Here, germline-specific components of the piRNA pathway such as aub, ago3, and vas were upregulated upon deletion of the tumour suppressor gene lethal (3) malignant brain tumor [l(3)mbt].However, the mechanism underlying this regulation remained unclear.
In this study, using Drosophila as a model, we uncover key elements of the generegulatory network controlling the female germline piRNA pathway.In a systematic analysis integrating multiple approaches, including single-cell RNA-seq, ATAC-seq, ChIP-seq as well as depletion and overexpression screens, we identify the transcription factor Ovo as the key positive transcriptional regulator of the germline piRNA pathway in ovaries.Enforced expression of Ovo in somatic cells activates expression of the germline piRNA pathway genes including the ping-pong cycle components Aub, Ago3 and Vas, leading to formation of peri-nuclear cellular structures mimicking the nuage bodies of germ cells.This is orchestrated through binding of Ovo to conserved CCGTTA motifs within the promoters of these genes.We also reveal the mechanistic link between l(3)mbt, ovo and the germline piRNA pathway genes.In addition, we show that the ancient Ovo-binding motifs are highly enriched within the germline piRNA clusters in ovaries of metazoan species ranging from insects to humans.ChIP-seq experiments show that fly Ovo and its mouse and human orthologs, OVOL2, are recruited to the motifs within promoters of the piRNA pathway genes and ovarian piRNA clusters.Interestingly, the same CCGTTA motifs are also bound by the male-specific transcription factor A-MYB in testes of vertebrates to control the male germline piRNA pathway 31 .Our results suggest that Ovo is an ovary-specific counterpart of A-MYB, performing a role in ovaries analogous to that of A-MYB in vertebrate testes.Overall, our results reveal gene-regulatory interactions between the ovary-specific DNA-binding TFs and the ancient CCGTTA cis-regulatory elements underlying the control of the germline piRNA pathway in ovaries across the metazoan species.

Ovo is co-expressed with germline piRNA pathway genes in Drosophila ovary
In a previous study on male piRNA factors in mouse testes, the testis-specific transcription factor A-MYB was identified as a regulator of male germline piRNA pathway genes 31 .To identify a potential master regulator of the female germline piRNA pathway in the Drosophila ovary, we performed a co-expression analysis using three separate single-cell RNA-seq datasets from Drosophila ovaries to determine the ovarian TFs co-expressed with the germline-specific piRNA pathway genes, aub, vas, and ago3 (Figure 1b and Supplementary Dataset 1).Two of the scRNA-seq datasets were from adult ovaries 35,36 and one was from larval ovaries 37 .Clustering of each of the datasets revealed distinct germline clusters showing specific expression of germline markers (e.g., vas, nos) and germline-specific piRNA pathway genes (i.e., aub, vas, ago3, tej, qin, krimp, boyb, nxf3, moon, boot, rhi, del, and cuff) (Figures 1b  and 1c, Supplementary Figures 1a-d).Correlation analysis (Pearson's r) identified ovo as the most significantly co-expressed TF with germline-specific piRNA pathway genes in all three datasets (Pearson's r>0.7 in adult flies and Pearson's r>0.9 in larva, adjusted p<1.0x10 -300 ) (Figure 1d and Supplementary Dataset 1).Interestingly, the second most significant co-expressed TF in adult ovaries was kipferl (Figure 1d), a recently identified zinc-finger protein that serves as a co-factor for the germline piRNA factor, Rhino, which is required for piRNA production from most dual-strand piRNA clusters 38 .Of note, the second top-ranking TF in larval ovaries was trf2 (Figure 1d), which has been shown to assist Moonshiner with transcription of dual-strand piRNA clusters 18 .
During Drosophila oogenesis, germline stem cells (GSCs) at the tip of germarium continuously generate egg chambers 39,40 , thus the germ cell cluster identified by scRNA-seq represents an aggregate of stages throughout germline development.To delineate the TFs co-expressed with germline piRNA pathway genes across oogenesis stages, we isolated and re-clustered the germ cell cluster and performed a germline-specific correlation analysis (Figure 2 and Supplementary Figures 2).Using early (bgcn and bam) and late markers (osk) of oogenesis [41][42][43] , we computed a pseudotime of germline development and tracked the expression pattern of the germline piRNA pathways genes from GSCs to nurse cells (Figures 2a and 2b, and Supplementary Figures 2a and 2b).Correlation analysis revealed ovo as the topranking TF co-expressed with the germline piRNA pathway genes over the course of germline development (adjusted p<3.4x10 -5 , Figure 2c, Supplementary Figure 2c, and Supplementary Dataset 1).

Ovo is the principal germline transcription factor in the Drosophila ovary
Differential expression analyses between germline and somatic cell clusters in the three separate ovarian single-cell RNA-seq datasets showed ovo as the most enriched germline TF (Supplementary Figures 3a and Supplementary Dataset 2).To validate this finding, we next aimed to experimentally identify all germline-enriched TFs within Drosophila ovaries that could potentially be responsible for promoting germline fate along with the germline piRNA programme.
To delineate germline-specific TFs, we first performed RNA-seq on FACS-sorted germline (vas-GFP+) and somatic cells (vas-GFP-) from vas-GFP Drosophila ovaries (Figure 3a).A differential RNA-seq analysis between the sorted germline and somatic cells (DESeq2) revealed ovo as the top germline-enriched TF (~87-fold, p<1.0x10 -300 , ovo-B isoform matching NM_080338 transcript, Figure 3a and Supplementary Dataset 2).Cross-tissue RNA-seq analysis showed ovaries are the most expressed site for ovo (Supplementary Figure 3b).This result corroborates the hypothesis that ovo is the principal ovarian germline TF, potentially controlling a germline-specific expression program encompassing the piRNA pathway in ovarian germ cells.To confirm our findings, we additionally performed RNA-seq on the ovarian germline/somatic co-culture line (fGS/OSS) and compared it to purely somatic ovarian cell line (OSCs), which similarly revealed ovo as the most enriched germline-specific TF (~155-fold, p<1.0x10 -300 , Figure 3b).

Enforced ovo expression in ovarian somatic cells activates the germline piRNA pathway components
To determine the gene-regulatory effects of Ovo genome-wide, we performed RNAseq in ovarian somatic cells (OSCs) following ectopic ovo expression (overexpression vector carrying FLAG-tagged ovo-B isoform driven by the act5C promoter) and compared it to OSCs transfected with an empty vector.Western blotting confirmed the presence of Ovo-B-FLAG protein at the expected size (~114 kDa) (Figure 4a).Immunofluorescence showed that overexpressed Ovo-B-FLAG localized to the nucleus (Figure 4b and Supplementary Figure 4a).Differential RNA-seq analysis (DESeq2) showed that ectopic Ovo expression significantly upregulated ~60-70% of the germline-specific piRNA pathway genes in OSCs, namely, ago3, aub, tej, qin, vas, moon, boot and nxf3 (>1.5-fold, adjusted p<0.05, Figure 4c and Supplementary Figures 4b and 4c and Supplementary Dataset 2).The strongest effects were seen on aub and ago3 (16-fold and 6-fold, adjusted p<1.0x10 -41 , Figure 4c and Supplementary Figures 4b and 4c).In contrast to components of the germline piRNA pathway, the soma-specific fs(1)Yb was downregulated upon Ovo overexpression, with all other somatic factors involved in biogenesis and transcriptional gene silencing (TGS) not changing in expression (Figure 4c and Supplementary Figure 4c).
Using immunofluorescence, we observed the appearance of Aub, Ago3, and Vas proteins following ectopic Ovo expression in OSCs which assembled into peri-nuclear "nuage"-like structures (Figures 4d-f and Supplementary Figure 4d).Co-localization analysis showed that Aub, Ago3, and Vas proteins assembled as foci around nuclei in a ring-shaped manner (Figures 4e and 4f), resembling the nuage structures of the Drosophila germ cells where piRNAs are processed by the ping-pong cycle.Of note, we confirmed that these "nuage"-like structures were distinct from the somatic Ybbodies of OSCs (Supplementary Figure 4e).
A total of 693 genes were upregulated by ectopic Ovo expression in OSCs (log2FC>0.6,p.adj<0.1;DESeq2), 182 of which (~26%) showed strong germlineenriched expression in ovaries (vas-GFP log2FC>1, p.adj<0.01) (Supplementary Dataset 2).Ovo could account for the regulation of at least ~18% of all germlineenriched genes in ovaries (182 out of 1,019 genes).Notably, among the top germline targets of Ovo was nanos (nos) (Supplementary Figure 4f), a gene essential for germline formation and maintenance 44,45 .Overall, our results show that Ovo upregulates the majority of the germline-specific piRNA pathway genes when ectopically overexpressed in OSCs.Next, we sought to decipher mechanisms of Ovo regulation to understand how it activates piRNA factor expression exclusively in germ cells.

L(3)mbt suppresses the germline piRNA pathway by repressing ovo in ovarian somatic cells
Early clues regarding the regulation of germline piRNA pathway components in Drosophila came from studies that deleted the l(3)mbt gene in somatic tissue, including the brain, ovary and ovary-derived OSCs, which uniformly resulted in the upregulation of several germline-specific genes such as the ping-pong components Aub, Vas, and Ago3 [32][33][34] .The molecular mechanism underlying this activation has remained unknown.L(3)mbt has been shown to interact with chromatin and cause histone compaction leading to suppression and insulation of gene expression [46][47][48] .While a direct regulation by L(3)mbt could explain prior observations, we hypothesized that the deletion of L(3)mbt could also upregulate one or several specific TFs that in turn control the expression of germline piRNA pathway genes.
To determine the precise mechanistic link between L(3)mbt and germline piRNA pathway components, we performed RNA-seq (n=3 replicates) and ATAC-seq (n=2 replicates) on wild-type (WT) OSCs and OSCs carrying a deletion of l( 3)mbt [Δl(3)mbt] (Figure 5a).We aimed to uncover regulatory elements, transcription factors, and genes that are differentially regulated and could be responsible for the activation of the germline piRNA pathway components upon L(3)mbt deletion.
Differential expression analysis of DNA-binding TFs (DESeq2 on RNA-seq data) between Δl(3)mbt and wild-type OSCs revealed ovo as the top upregulated TF upon l(3)mbt deletion (~97-fold, adjusted p value <1.0x10 -300 ) (Figure 5a).The degree of ovo upregulation was markedly higher compared to other TFs (Figure 5a).To assess if this expression difference was due to increased accessibility at the ovo promoter upon l(3)mbt loss, we analysed differential chromatin accessibility using ATAC-seq data (DiffBind) from Δl(3)mbt and wild-type OSCs (Figures 5b-d).This analysis uncovered 207 genomic regions with significantly increased accessibility upon l(3)mbt deletion (>1.5-fold,FDR q <0.05) of which 111 were at gene promoters (±1kb of TSS) (Figures 5b and 5c).Out of these, 30 were germline-specific and included only one germline TF, ovo (~2-fold, FDR q <0.01) (Figures 5d).The only other TF that showed increased promoter accessibility in Δl(3)mbt-OSCs was bigmax, which was not germline-enriched in ovaries.Among the piRNA pathway genes, only vas showed increased promoter accessibility in Δl(3)mbt-OSCs and thus could potentially be a piRNA factor that is directly regulated by L(3)mbt, with others likely controlled via Ovo.
We speculated that direct binding of L(3)mbt to the ovo promoter could be responsible for the decreased chromatin accessibility in somatic cells, leading to its soma-specific transcriptional repression.Indeed, ChIP-seq data from OSCs showed strong L(3)mbt binding specifically at the promoter region of ovo (Figure 5e), which increased in accessibility upon l(3)mbt deletion.This suggests that direct L(3)mbt binding to the ovo promoter in somatic cells represses its expression.Elimination of l(3)mbt consequently leads to increased accessibility at the ovo promoter and results in drastically increased expression of ovo, mimicking its germline expression in the Drosophila ovary (Figure 5e).We further confirmed our results by reanalysing recently published RNA-seq data from l(3)mbt knockdown in OSCs 49 which also showed ovo as the most significantly upregulated gene (Supplementary Figure 5a).3)mbt binding to the promoter of the ovo gene (n=2 replicates from distinct samples; merged; data from 49 ).RNAseq tracks below showing the expression of the germline ovo isoform in Δl(3)mbt OSCs and the Drosophila ovary germline cells (n=4 replicates from distinct samples; merged).(f) Differential gene expression (Deseq2) between the control and Δl(3)mbt fly ovaries, both lacking germline due to tud maternal mutations (tud M ) (n=3 replicates from distinct samples; RNA-seq data from 33 ).(g) L(3)mbt ChIP-seq from OSCs showing absence of L(3)mbt binding at the germline aub promoter (n=2 replicates from distinct samples; merged; data from 49 ).Ovo ChIP-seq showing Ovo binding to the aub promoter (n=2 replicates from distinct samples; merged; data from ENCODE, whole-fly) (h) L(3)mbt and Ovo ChIP-seq peaks within the promoters of the ping-pong genes aub, ago3, and vas (±1kb of TSS).(i) Model for the L(3)mbt and Ovo regulation of the germline piRNA pathway genes.(j) Ovo siRNA knockdown experiments in Δl(3)mbt-OSCs (RT-qPCR; n=3 replicates from distinct samples; p-value: *<0.01, one-tailed two-sample t-test).(k) Volcano plot showing downregulation of the germline-specific piRNA pathway genes on day 2 of ovo siRNA knockdowns in Δl(3)mbt-OSCs using differential RNA-seq analysis (Deseq2) between ovo and renilla siRNA knockdowns (n=3 replicates from distinct samples).
To extend our findings to an in vivo setting, we analysed previously published RNAseq data from Lehmann and colleagues from l(3)mbt mutant and control ovaries of transgenic flies that lack germ cells due to maternal tud mutations (tud M ) 33 .This allowed us to perform differential expression analysis between Δl(3)mbt and control ovaries that were devoid of germline cells (Figure 5f).This analysis identified ovo as the second most significantly upregulated TF upon l(3)mbt deletion in vivo within somatic cells of fly ovaries (Figure 5f).It was previously shown that L(3)mbt forms a complex with Lint-O at chromatin to silence gene expression 49 .Differential expression analysis of the RNA-seq data from lint-O and control knockdowns in OSCs, revealed ovo as the top upregulated TF upon Lint-O depletion (Supplementary Figure 5b), suggesting that Lint-O together with L(3)mbt forms a repressive complex at the ovo promoter in somatic cells, which we confirmed using Lint-O ChIP-seq data in OSCs (Supplementary Figure 5c).
We then analysed if L(3)mbt could be exerting its effects on germline piRNA pathway genes, such as aub, indirectly via regulation of Ovo.Most importantly, using Ovo ChIPseq, we observed strong Ovo binding at the promoters of germline-specific piRNA pathway genes where L(3)mbt binding was lacking (Figures 5g and 5h), suggesting that L(3)mbt was not directly controlling their expression.We therefore hypothesized that L(3)mbt indirectly represses the expression of germline-specific piRNA pathway genes in somatic cells via its regulation of ovo expression (Figure 5i).

Ovo regulates the germline piRNA pathway genes via binding to conserved CNGTTA motifs
To capture Ovo binding events that lead to activation of the germline piRNA pathway genes, we performed Ovo ChIP-seq (n=2) following enforced ovo expression in OSCs (n=4769 peaks, FDR<0.05,>5-fold enrichment).Using Ovo ChIP-seq data from adult female flies (ENCODE 50 ) we were also able to map in vivo Ovo binding events (n=4,477 peaks) across the Drosophila genome (Figure 6a).We performed ATAC-seq in Drosophila ovaries to locate Ovo binding regions corresponding to open chromatin (n=4,007) and closed chromatin (n=470) within ovaries (Figure 6a).Ectopic Ovo expression in OSCs recapitulated ~65% of the in vivo binding events observed in flies of which the majority (~93%) were within open chromatin in ovaries (Figure 6a).
The top scoring de novo motif within in vivo Ovo ChIP-seq peaks in flies was CCGTTA (MEME-ChIP, e-value=1.3 -10 ), while the secondary motif was CNGTTA (MEME-ChIP, e-value=8.3 -6 ) (Figure 6b).Interestingly, the CCGTT core of the motif was the top hit within the ectopic Ovo ChIP-seq peaks following enforced ovo expression in OSCs (MEME-ChIP, e-value=7.8 -06 ) (Figure 6b) and CNGTT was the fourth strongest hit in vivo (Supplementary Figures 6c).These results suggest that CCGTTA is the most preferred Ovo binding site in vivo, with changes to the second and last nucleotides often tolerated.The majority of Ovo binding events (~79%) were proximal to promoters (<1kb to the TSS), while only 1.12% were distal or intergenic (Figure 6c).~95% of Ovo binding events at promoter proximal regions corresponded to open chromatin in ovaries (based on ATAC-seq peaks) (Supplementary Figures 6a and 6b).The genes upregulated in response to enforced ovo expression in OSCs showed >2-fold stronger Ovo ChIP-seq binding signals and peak enrichments in their promoters (±1kb of TSS, p<0.01) compared to the genes that did not change in expression (Figure 6d and Supplementary Figure 6d).Correspondingly, higher numbers of Ovo motifs were observed at the upregulated Ovo target gene promoters when compared to genes that did not change in expression (~1.6-fold, p<0.01;Supplementary Figures 6e-f).
We then looked at Ovo binding events within promoters of the piRNA pathway genes (Figures 6e and 6f).The average Ovo binding signal (input-normalized ChIP-seq) was >2-fold stronger (p<0.05) at promoters of the germline-specific piRNA pathway genes when compared to somatic or shared piRNA factors (Figures 6e).This was generally true for all germline genes, which showed on average ~2-fold stronger Ovo ChIP-seq signal (p<4.1x1.0 -8 ) at their promoters (±0.5kb of TSS) when compared to the somatic or shared genes.Moreover, ~90% of the germline-specific piRNA pathway genes had an Ovo motif ±0.5kb of their TSS overlapping an Ovo ChIP-seq peak summit while only ~30% of somatic factors did (Supplementary Figure 4c).Multiple sequence alignments of the Ovo binding sites within promoters of the germline piRNA pathway genes showed high conservation of the CNGTTA motifs across six Drosophila species (D. melanogaster, D. simulans, D. sechellia, D. yakuba, D. erecta, D. suzuki), sharing a common ancestor about ~13 million years ago 51 , suggesting an evolutionary conserved gene-regulatory mechanism behind Ovo interaction with CNGTTA motifs (Figure 6f).These results indicate that direct binding to CNGTTA motifs is a major conserved mechanism through which Ovo controls expression of the germline piRNA pathway genes in the Drosophila species.

Vertebrate homologs of Ovo bind to ovarian piRNA pathway components
Most vertebrates have three homologs of the fly ovo gene (e.g., mouse Ovol1, Ovol2, and Ovol3) with the Ovol2 paralog involved in the development of primordial germ cells (PGCs) 52,53 .Clustering of all vertebrate TF motifs in the JASPAR database (2,022 vertebrates CORE) revealed high similarities between the motifs of Ovo family TFs and, remarkably, A-MYB (Figure 7a), a testis-specific TF previously reported to control transcription of both the germline piRNA pathway genes and piRNA clusters in male mice.We hypothesized that Ovo homologs could provide the ovary-specific equivalent of this regulation in females, controlling germline piRNA clusters in ovaries in addition to piRNA pathway genes via the same CCGTTA motifs.
To test this hypothesis, we used mouse OVOL2 ChIP-seq data from female mouse primordial germ cell-like cells (PGCLCs; day 2 of induction) under transgenic expression of mouse Ovol2 52 .We observed OVOL2 binding to CCGTTA motifs within or near promoters of the piRNA pathway genes (e.g., Tdrkh, Tdrd7, Pld6) (Figure 7b and Supplementary Figure 7b) which also showed upregulation in response to Ovol2 expression in mouse ESCs or PGCLCs (Supplementary Figures 7b and 7c).To find out whether mouse OVOL2 also binds to ovary piRNA clusters, we defined ovaryspecific, testis-specific, and shared piRNA clusters using small RNA-seq data from mouse ovaries and testes (>1 RPKM normalized counts mapping to the piRNA cluster regions; data from 54 ; coordinates obtained via proTRAC 55 ) and checked OVOL2 binding at these regions (Figure 7c and Supplementary Figure 7a).Our analysis revealed strong OVOL2 binding events to CCGTTA motifs within regions corresponding to ovary-specific piRNA clusters (Figure 7c and Supplementary Figure 7a).These OVOL2 binding events within motifs were found near ends or the centres of ovary piRNA clusters (Figure 7a and Supplementary Figure 7a).OVOL2 ChIP-seq peaks showed a significant enrichment at ovarian piRNA clusters but were absent from testis-specific piRNA clusters (Figure 7a and Supplementary Figures 7a and 7d).
Next, we asked whether there was evidence of human OVOL2 engagement at human ovarian piRNA clusters.We exploited small RNA-seq data from human fetal ovaries (data from 56 ) and analysed OVOL2 motif occurrences within human ovarian piRNA clusters (Figures 7d).Ranking of the human ovary piRNA clusters with small RNAseq expression using data and coordinates defined in 56 (Figure 7d and Supplementary Figure 7e) revealed that the most highly expressed piRNA cluster in human fetal ovaries was cluster 71, which accounted for 54% of all piRNAs reads in ovaries (Figures 7d and Supplementary Figure 7e).The piRNA cluster 71 showed 2.5-fold enrichment for OVOL2 motifs over the genomic background (p<0.01, Figure 7d).The enriched OVOL2 motifs were highly concentrated near the 5' end of cluster 71, suggesting a potential impact on promoter activity (Figure 7d).To determine if OVOL2 was physically associated with these motifs, we used OVOL2 ChIP-seq data from human induced pluripotent stem cells (iPSCs; WTC11, data from ENCODE 57 ).Indeed, OVOL2 ChIP-seq showed strong binding to the promoter region of the ovarian piRNA cluster 71 (Supplementary Figure 7f).Interestingly, piRNA cluster 14, the second highest ranking ovarian piRNA cluster, showed several strong OVOL2 binding regions across the cluster locus (Supplementary Figure 7f).Similar to the binding pattern observed for mouse OVOL2 ChIP-seq, human OVOL2 ChIP-seq showed binding near promoters of the piRNA pathway genes (e.g., TDRKH, TDRD7, TDRD3, PIWIL4) when overexpressed in the human iPSCs (Figure 7b and Supplementary Figure 7b).Moreover, human OVOL2 binding near promoters of piRNA pathway genes (e.g., TDRKH, PIWIL4, and TDRD7) occurred at orthologous regions in the mouse genome where mouse OVOL2 showed corresponding binding events (Figure 7b).These orthologous OVOL2 binding events occurred at Ovo motifs that were highly conserved across the mammalian species based on multiple sequence alignments (Figure 7b).Overall, our results indicate that the Ovo TF family interactions with CCGTTA motifs in regulation of ovarian piRNA pathway components is a gene-regulatory feature that is conserved from flies to vertebrates.

Ancient Ovo motifs are hallmarks of ovarian piRNA clusters in metazoans
To test whether engagement of Ovo to the ovary piRNA clusters is a general conserved feature of animal ovarian germ cells, we analysed Ovo motif enrichment within ovary piRNA clusters across multiple metazoan species.We characterized ovary piRNA clusters in species by their RPKM normalized counts (>1) mapping to the piRNA cluster regions (proTRAC piRNA cluster coordinates 55,58 ) using ovary small RNA-seq datasets from human (Homo sapiens), crab-eating macaque (Macaca fascicularis), mouse (Mus musculus), golden hamster (Mesocricetus auratus), cow (Bos taurus), zebrafish (Danio rerio), buff-tailed bumblebee (Bombus terrestris), African malaria mosquito (Anopheles gambiae), Arizona bark scorpion (Centruroides sculpturatus), and Pacific oyster (Crassostrea gigas).We then calculated the enrichments of OVOL2 motifs within the ovary piRNA clusters of each species.Our results revealed significant enrichment of OVOL2 motifs at ovary piRNA clusters in all ten species when compared to genomic backgrounds (p<0.05,Wilcoxon Signed-Rank Test, Figure 7d).Overall, our results suggest a conserved regulatory mechanism utilizing CCGTTA motifs underpinning the expression patterns of germline piRNA pathway in metazoan species, where Ovo/OVOL2 interaction with the motifs in ovaries serves as a female counterpart of the male-specific regulation by A-MYB in testes (Figure 7e).(c) OVOL2 ChIP-seq from female mouse PGCLCs (day 2 of PGCLC induction; overexpressing transgenic mouse Ovol2a; data from 52 ) showing OVOL2 binding to the CCGTTA motifs (OVOL2 motifs) within the mouse ovary-specific piRNA clusters (small RNA-seq data from 54 ).(d) Ranking of the human piRNA clusters by their expression levels in human fetal ovaries (rpkm; data from 56 ).The genomic browser below depicting OVOL2 motifs (red; UCSC squish track display) and the number of OVOL2 motifs per 100 kb at the top-expressed human ovary piRNA cluster (#71).Human fetal ovary and fetal testis small RNA-seq signals show the relative production of the piRNAs from the cluster #71 (rpm; data from 56 ).Sodium periodate treatment enriches for piRNAs.(e) Numbers of OVOL2 motifs per 10 kb compared to the genomic backgrounds across ten metazoan species.Ovarian piRNA clusters were identified using ovary small RNA-seq data from each species (Supplementary Dataset 3).The classification tree is based on the NCBI taxonomy database.Error bars indicate standard error of the mean.p-value: *<0.05,Wilcoxon Signed-Rank Test.(f) Model depicting a conserved regulation of the piRNA pathway via interactions of the CCGTTA motifs with Ovo family TFs in animal ovaries and A-MYB in animal testes.

Discussion
Transcription factors and co-activators controlling the male germline piRNA pathway have been previously described in vertebrate testes 31,59,60 ; however, a female-specific counterpart of such a regulatory network controlling the expression of the female germline piRNA pathway in insect and vertebrate ovaries has remained an enigma.The identification of ovary-specific TFs that control the germline-specific piRNA pathway in female germ cells provides crucial insights into the regulation of transposon repression during oogenesis in animals.
In our study we revealed Ovo as the principal regulator of ~70% of the germlinespecific piRNA pathway genes in Drosophila.The transcription factor Ovo locus encodes both somatic and germline isoforms that are driven by distinct promoters and were once thought to be two distinct genes: the somatic shavenbaby (svb) and the germline ovo (isoforms ovo-A and ovo-B) (Figure 5e).The somatic svb is expressed only in embryonic, larval, and pupal epidermis cells, while the germline ovo-B is the major and the only essential isoform required for viability of female Drosophila germ cells 53,61 .Our results indicate that enforced ovo-B expression activates the germline piRNA pathway components in the ovarian somatic cells (Figure 4) and l(3)mbt specifically represses the expression of ovo-B in the somatic cells by occupying the promoter region responsible for germline expression (Figure 5e).
Our results point towards a gene-regulatory model of Ovo in the female germline piRNA pathway in ovaries where transcription of germline piRNA pathway genes (e.g., aub, ago3 and vas) could be directed by Ovo family TFs, analogous to the model previously suggested for A-MYB in the regulation of the male germline piRNA pathway in mice testes (Figure 7f) 31 .Similar to A-MYB, Ovo and its homologs auto-regulate their own expression [62][63][64] , and as our motif and ChIP-seq results reveal, show strong binding to germline piRNA pathway components.Our experiments show that Ovo is indeed able to activate expression of germline piRNA pathway components when ectopically expressed in fly somatic cells.This includes the nuage components Aub, Ago3 and Vas which are activated in expression in somatic cells in the presence of ectopic Ovo and assemble to form cellular structures resembling the nuage bodies of germ cells (Figures 4d-f).Our model is further supported by the multi-species analyses of OVOL2 ChIP-seq in mouse PGCLCs and human iPSCs revealing strong OVOL2 binding to the ovary piRNA clusters (Figure 7), thus indicating a high degree of conservation of this gene-regulatory mechanism across the animal kingdom.Previous work analysing ovarian piRNA clusters 65 reported an observation of strong A-MYB motif enrichments at ovarian piRNA clusters in macaques leading them to suggest a paradoxical involvement of the testis-specific A-MYB in driving transcription of ovarian clusters in ovaries; however, this enrichment is now compatible with our findings given that OVOL2 and A-MYB utilize the same motifs and these motifs would recruit OVOL2 in ovaries.
Ovo is well known for its conserved role in germline development in animals ranging from flies to mice 53 .Germline development in flies and mammals follow distinct pathways 66 .In flies, the preformation model states that the germline is established through maternally deposited germ cell determinants within oocytes while mammalian PGCs develop according to epigenetic/inductive mechanisms (epigenesis) where external cues dictate their developmental trajectory 66,67 .Importantly, Ovo in flies is maternally deposited as a component of the germplasm and later pole cells, which establish the fly PGCs 68 , while mammalian OVOL2 acts downstream of BMP signalling to control cell fate decisions during mammalian PGC specification in the epiblast 53,69 , thus suggesting that Ovo family TFs control animal PGC development and germline piRNA pathway expression via both intrinsic and inductive mechanisms.
Interestingly, ~36% of the mouse OVOL2 binding sites were also occupied by A-MYB in testes as both OVOL2 and A-MYB recognize the same core CCGTTA motif sequence (Figure 7a).This could be a mechanism driving the expression of the shared germline piRNA clusters in both ovaries and testes (Supplementary Figure 7a); however, other mechanisms such as motif affinity, chromatin accessibility or methylation must account for expression of sex-specific piRNA clusters.In our motif analysis, we noticed a pseudo-palindromic extension to the GTT core in the human A-MYB motif (Figure 6a) which could prefer A-MYB over OVOL2 at testis-specific clusters.Additionally, we observed that ~52% of the OVOL2 binding events within the mouse ovary piRNA clusters occurred at the CpG islands (Figure 7c and Supplementary Figure 7a) which could be linked to sex-specific DNA methylation patterns in developing PGCs.TF binding sites are enriched at hypomethylated regions that evade the first wave of default de novo DNA-methylation 70 , which starts from day 13 of embryonic development (E13.5) in male mouse PGCs 71 and coincides with the appearance of the pre-pachytene piRNAs 54 ; however, female PGCs undergo de novo methylation only later after birth 72 .Temporal differences in DNA methylation and binding patterns by sex-specific TFs that direct female PGC development such as OVOL2 could therefore protect against methylation to control sex-specific piRNA cluster transcription.Of note, the mouse Ovol2 gene encodes both the repressor isoform Ovol2a and the activator isoform Ovol2b.The binding patterns between OVOL2A and OVOL2B ChIP-seq were indistinguishable from each other at piRNA pathway components in mice, thus the mechanism driving the ovary-specific piRNA pathway could also depend on the interplay of these isoforms.
In fly ovaries, Ovo is continuously present in the nucleus of the germline cells and maternal Ovo persists in the embryo until zygotic Ovo is expressed; thus, Ovo binding sites could be potentially marking genomic locations important during the transition from one generation to the next 61 .Ovo persists in ovarian germ cells and plays a role in female germline sex determination by controlling expression of Otu and Sxl 73 .Our findings reveal its concurrent role in regulation of the female germline piRNA pathway.However, the testis-specific regulator controlling the male germline piRNA pathway in fly testes remains to be identified.The fly Myb TF is unlikely to act analogously to the vertebrate A-MYB, as it does not recognize the CCGTTA motifs (based on de novo motifs in ENCODE ChIP-seq data) and was reported to function as a weak repressor of piRNA factors in OSCs 49 ; therefore, a different testis-specific TF candidate concurrently controlling spermatogenesis and male piRNA pathway likely exists in fly testes.
Intriguingly, OVO-family TFs and MYB-family TFs belong to different TF lineages and harbour distinct DNA-binding domains yet are capable of binding and competing for the same genomic CCGTTA motifs 64 , thus illustrating an example of convergent evolution where unrelated classes of DNA-binding domains evolve to bind to the same DNA elements.Under this model, the CCGTTA elements represent ancient evolutionary conserved cis-regulatory elements that interact with female germlinespecific Ovo-family TFs in ovaries and male germline-specific A-MYB in testes of animals that in combination with co-factors such as TCFL5 60 and other epigenetic mechanisms (e.g., chromatin accessibility, DNA-methylation, histone marks, L(3)mbt) control the germline piRNA pathway.Moreover, Ovo and Ovo-like TFs are comprised of both repressor and activator isoforms which could interplay to control the piRNA pathway in a sex-and tissue-specific manner.Overall, our results reveal a conserved gene-regulatory mechanism involving interactions of the ovary-and testis-specific TFs with the CCGTTA cis-regulatory elements behind the regulation of the germline piRNA pathway in animal ovaries and testes.

Fly stocks and handling
All flies were kept at 25°C on standard cornmeal food or propionic food.Control w 1118 flies were a gift from the University of Cambridge Department of Genetics Fly Facility.Transgenic vas-GFP fly strain (w[*]; TI{TI}vas[AID:EGFP]) was reported in 76 and obtained from Bloomington Drosophila Stock Center (#76126).

Ovary dissections and cell dissociation
Flies were fed with yeast extract 2-3 days before the dissections.Ovaries were dissected into 1 mL ice-cold PBS and centrifuged at 500 g (4°C) for 5 min.Cells were dissociated using 5 mg/mL trypsin (Sigma-Aldrich, T1426-50MG) and 2.5 mg/mL collagenase-A (Sigma-Aldrich, 10103586001) in Ringer's solution (600 µL for 148 ovaries) for 1 hr.at 800 rpm (30°C) in a thermomixer and passed through a 100μm cell strainer followed by addition of 400 µL of Schneider's Medium (Thermo Fisher Scientific, 21720001) containing 10% FBS and centrifuged at 500 g (4°C) for 5 min.The supernatant was discarded and 600 µL of PBS was added to the pellet to wash by centrifugation at 500 g (4°C) for 5 min.The pellet was resuspended in 500 µL of PBS.Dissociation of ~30 ovaries resulted in ~500,000 cells.

FACS sorting
The w 1118 strain flies and transgenic vas-GFP flies were fed with yeast extract 3 days prior to dissections.Ovaries from control w 1118 strain flies and transgenic vas-GFP flies were then dissected in ice-cold PBS and cells dissociated as described above.Cells were resuspended in PBS and sorted into GFP-positive and GFP-negative populations with FACS Aria cell sorter (BD Biosciences).Approximately ~148 ovaries from dissections of ~74 vas-GFP flies gave ~36,000 Vas-GFP-positive cells after FACS sorting and 19 ovaries from the w 1118 strain flies were dissociated and used as controls.
RNA isolation and RT-qPCR RNA was isolated using RNeasy Mini Kit (QIAGEN, 74106) with RNase-Free DNase Set (QIAGEN, 79254) for DNase digestion during RNA purification as per manufacturer's instruction.Reverse transcription was performed with 100 ng to 1 µg RNA using SuperScript™ IV Reverse Transcriptase (Thermo Fisher Scientific, 18090010).RT-qPCR was performed on 1:10 diluted cDNA using Fast SYBR™ Green Master Mix (Thermo Fisher Scientific, 4385610) with Bio-Rad C1000 Thermal Cycler.Primers were designed against exon-exon junctions for genes with introns and rpL32 was used as an internal standard (Supplementary Dataset 3).Relative expression was analysed using the delta-delta Ct method 77 .ethanol precipitation.The purified ChIP DNA and 220 ng of whole cell extract DNA were used to prepare ChIP-seq libraries using NEBNext® Ultra™ II DNA Library Prep Kit for Illumina (NEB, E7645S) with PCR amplifications carried out for 16 cycles.The ChIP-seq libraries were assessed for fragment size distribution using Agilent 2100 bioanalyzer on a High Sensitivity DNA Kit, quantified with KAPA Library Quantification Kit for Illumina (Kapa Biosystems, KK4873), and sequenced on a NovaSeq 6000 System with 2 × 150 bp reads on SP flow cell.

Overexpression screen
Candidate genes were amplified from Drosophila melanogaster ovary cDNA library using KOD Hot Start DNA Polymerase (Merck Millipore, 71086-4) and cloned into overexpression vectors under act5 promoter with either N-terminal or C-terminal FLAG tags using Gibson Assembly® Master Mix (NEB, E2611L) at 50 °C for 1 hr.9 μl of Mix & Go Competent Cells (Strain DH5 Alpha) were thawed on ice and transformed with 1 μl of diluted (2.5x in H2O) Gibson Assembly reaction by incubation for 5 min on ice before plating on LB plates containing the appropriate antibiotic.Colony PCR was performed to identify the colonies harbouring the ligated constructs followed by their inoculation into 3 mL broth containing 100 μg/mL ampicillin that were shaken overnight at 37 °C.The constructs were purified using QIAGEN Plasmid Plus Kits and Sanger sequenced to check for mutations and confirm the correct ligation.The constructs were then transfected into WT OSCs and Δl(3)mbt-OSCs with Nucleofector Kit V (Lonza, VVCA-1003) using T-029 program and a NucleofectorTM 2b Device (Lonza).Cells were passaged 1 day prior to transfections and were 70-80% confluent at the time of transfections.Ectopic ovo overexpression vector carrying the FLAG-tagged ovo-B isoform matching NM_080338 transcript was used.

siRNA knockdown experiments
Sense and antisense 21 nt siRNA sequences for target genes were designed using the DSIR tool (http://biodev.cea.fr/DSIR/DSIR.html)(Supplementary Dataset 3).The designed sequences were ordered from IDT and resuspended in 400 μl RNase-free water.Equal volumes of the resuspended sense and antisense siRNA were mixed then added to an equal amount of 2x annealing buffer (60 mM potassium acetate; 200 mM HEPES, pH 7.5).The mix was boiled for 5 min at 75 °C, then ramped down to 25 °C (-0.1 °C/second) to anneal the siRNA sequences into the siRNA duplex (100 μM final concentration =200 pmoles). 2 μL of the final 100 μM siRNA duplex was mixed with 100 μL of Nucleofection solution V (Lonza, VVCA-1003) and transfected into 10 million cells using a NucleofectorTM 2b Device and program T-029 (Lonza).Cells were plated into 12-well plates and the media was changed after 24 hr.RNA was either harvested at 48 hr or nucleofection was repeated after 48 hr and RNA was harvested at 96 hr.

Western blots
Proteins were extracted from 1-3 x 10 6 cells, washed in PBS (300 g, 5 min), and lysed in ice-cold RIPA buffer (Thermo Scientific, 89900) containing cOmplete™ Protease Inhibitor Cocktail EDTA-free tablet (Roche) (1 tablet per 20 mL of RIPA) under rocking condition at 4 °C for 30 min.The lysate was centrifuged 20,000g at 4 °C for 20 min and supernatant was transferred to a new tube.The protein concentrations were measured using Direct Detect® Spectrometer.15 mg/mL of protein was used with NuPAGE 4X LDS Sample Buffer and NuPAGE™ 10X Sample Reducing Agent (Thermo Fisher Scientific) in 20 μL total volume and denatured at 70°C for 10 min followed by protein gel electrophoresis using NuPAGE™ 4-12%, Bis-Tris, 1.5mm 10 well, Mini Protein Gels and XCell SureLock Mini-Cell Electrophoresis System (Thermo Fisher Scientific) at 180V for 1hr at 4°C cold room.Precision Plus Protein™ All Blue Prestained Protein Standards (Bio-Rad, 1610373) was run in parallel as a ladder.Blotting was performed for 7 min with iBlot™ 2 Transfer Stacks, nitrocellulose, mini using iBlot™ 2 Gel Transfer Device.The nitrocellulose membrane was blocked for 1 hr at room temperature under gentle agitation using Intercept® blocking buffer (TBS) (LI-COR Biosciences).The membrane was rinsed with TBS buffer and incubated with the primary mouse monoclonal ANTI-FLAG® M2 antibody (Merck, F1804-200UG) (1:2,500) and the primary rabbit polyclonal anti-Tubulin (DM1A+DM1B) (Abcam, ab18251) (1:5,000) in TBS with 0.1% Tween 20 (TBST) overnight at 4 °C under gentle agitation.The membrane was washed 3 x 5-10 min with TBST and incubated with the secondary antibodies IRDye® 800CW Donkey anti-Mouse IgG Secondary Antibody (1:5,000) (LI-COR Biosciences, 926-32212) and IRDye® 680RD Donkey anti-Rabbit IgG Secondary Antibody (1:20,000) (LI-COR Biosciences, 926-68073) in TBST for 1 hr room temperature under gentle agitation.The membrane was washed 3 x 10 min with TBST and additional washed with TBS before visualizing with the Odyssey CLx Infrared Imaging System (LI-COR Biosciences).
ovo-FLAG construct) relative to the nucleofection with an empty vector in OSCs, and the foldenrichments of the same genes in the Vas-GFP+ germline cells compared to the Vas-GFP-somatic cells from the FACS-sorted transgenic Vas-GFP ovaries (Deseq2, RNA-seq n=3 replicates from distinct samples).The piRNA pathway-specific genes are highlighted in yellow; †=ubiquitous genes that are not specific to the piRNA pathway; # =somatic-specific piRNA pathway genes; *<0.01, adjusted p-values from Deseq2; normalized gene counts (Deseq2) are shown for Vas-GFP-(soma) and Vas-GFP+ (germ) cells (hsp83 in black is excluded due to high expression levels).The presence of Ovo motifs and Ovo ChIP-seq peaks (ENCODE) within 500 bp +\-of gene TSS is shown with dots on the right side of the table.Transparent dots represent non-specific motifs and peaks within 500 bp +\-of gene TSS that are closer to a different/neighbouring gene.

Figure 1 .
Figure 1.Ovo is the top transcription factor co-expressed with the germline piRNA pathway genes in the Drosophila ovary.(a) Model depicting germline piRNA pathway in Drosophila germ cells.(b) UMAP clustering of Drosophila ovary single-cell RNA-seq datasets (adult ovary 1 from 35 ; adult ovary 2 from 36 ; LL3 larva ovary from 37 ).Expression of the germline marker vasa is shown as the marker of the germline clusters.(c) Dot plot showing expressions of the germline-specific, shared, and soma-specific piRNA pathway genes across the clusters identified for the adult ovary 2 dataset (40 µm cell strainer; clusters from the panel a; cluster 4 is the germline cluster).(d) Ranking of DNA-binding proteins and transcription factors by their average expression correlation values (Pearson's r) with the expression of the germline piRNA pathway factors aub, vas, and ago3.The colour scale indicates correlation p-values adjusted with Bonferroni correction for multiple testing.

Figure 2 .
Figure 2. Overexpression screen of the germline co-expressed candidates reveals Ovo as a positive regulator of piRNA pathway genes.(a) Diagram showing isolation and re-clustering of the germ cells (cluster 4) from adult ovary single-cell RNA-seq (dataset 2) and computation of the pseudotime trajectory by rooting the bgcn-expressing germline stem cells (GSCs) as the starting point.(b) Expression pattern of the early (bgcn) and late (osk) stage markers of germline differentiation along the pseudotime trajectory of the germline development shown together with the germline piRNA pathway genes (ago3 and aub) and the top co-expressed transcription factor ovo. (c) Ranking of the DNA-binding proteins and transcription factors by the average expression correlation (Pearson's r) with the germline piRNA pathway genes aub, vas, qin, and ago3 within the re-clustered germ cell cluster (cluster 4 in adult ovary scRNA-seq dataset 2).The colour scale shows correlation p-values adjusted with Bonferroni correction for multiple testing.(d) Overexpression screen in ovarian somatic cells (OSCs) using the top co-expressed candidates (TFs and chromatin-binding proteins) from the co-expression analysis (RT-qPCR, 48-72hr postnucleofection in OSCs, n=3 replicates from distinct samples, error bars indicate standard error of the mean).Overexpression of Ovo-B isoform (matching NM_080338) is indicated for ovo.

Figure 3 .
Figure 3. Ovo is the principal transcription factor of the Drosophila ovarian germ cells.(a) Differential gene expression of the DNA-binding TFs and chromatin proteins between the FACSsorted Vas-GFP+ (germline) and Vas-GFP-(somatic) cells from the transgenic Vas-GFP fly ovaries (Deseq2; RNA-seq n=3 replicates from distinct samples).(b) Differential gene expression of the DNA-binding TFs and chromatin proteins between the ovarian somatic cells (OSCs) and the ovarian germ/soma co-culture line (fGS/OSS) (Deseq2; RNA-seq n=4 replicates from distinct samples).

Figure 4 .
Figure 4. Enforced expression of ovo in ovarian somatic cells (OSCs) activates the germline piRNA pathway components leading to formation of cellular structures resembling nuage bodies of germ cells.(a) Western blot showing the presence of the Ovo-FLAG protein following ectopic expression in OSCs.(b) Immunofluorescence images showing the nuclear localization of the Ovo-FLAG protein in OSCs after nucleofection with the ovo-FLAG construct (48 hr; Ovo-B isoform, NM_080338 transcript).DAPI indicates DNA and Lamin indicates the nuclear envelope.(c) Differential gene expression between ovo-FLAG nucleofected OSCs relative to empty vector (Deseq2; RNA-seq n=3 replicates from distinct samples; the piRNA pathway genes labelled according to the colour key, TGS =transcriptional gene silencing).(d-e) Immunofluorescence images showing the appearance of the nuage components, Aub, Ago3, and Vas, as peri-nuclear nuage-like structures and foci (arrowheads in d) in the ovo-FLAG nucleofected OSCs (Piwi indicates the nucleoplasm, DAPI indicates DNA).(f) Colocalization of Aub, Ago3 and Vas proteins within the nuage-like bodies formed around the nuclei (DAPI) of the ovo-FLAG nucleofected OSCs.The fluorescence intensity along the white arrow inset is normalized to the highest value.

Figure 6 .
Figure 6.Ovo regulates germline piRNA pathway genes via binding to conserved CNGTTA motifs.(a) Heatmap of fly Ovo ChIP-seq peaks (n=2 replicates from distinct samples merged; ENCODE) corresponding to open and closed chromatin states in fly ovaries based on ovary ATACseq peaks (n=2 replicates from distinct samples merged).The binding pattern of the ectopic Ovo ChIP-seq following enforced ovo expression in OSCs (n=2 replicates from distinct samples) and ATAC-seq accessibility of OSCs (n=2 replicates from distinct samples) centred on the fly Ovo ChIPseq peaks are shown by the heatmaps on the right.The engagement of RNA polymerase II (Pol2) in ovaries is also depicted using the ovary Pol2 ChIP-seq signals (n=1; data from 18 ).(b) The topscoring de novo motifs discovered within Ovo ChIP-seq peaks using MEME-ChIP.(c) Genomic annotations (UCSC; Ensembl genes; dm6) of the fly Ovo ChIP-seq peaks using ChIPseeker.(d) Genomic profile of fly Ovo ChIP-seq signals (rpm) within 1kb ± of TSS of the upregulated,

Figure 7 .
Figure 7. Ovo binding sites are hallmarks of ovarian piRNA pathway components across the genomes of metazoan species.(a) Radial tree clustering of all human TF motifs in the JASPAR database (JASPAR 2022 vertebrates CORE; RSAT -matrix-clustering) showing a zoomed-in view of the motif group 48 containing A-MYB and OVOL2 motifs.Information content of each branch is taken from the dynamic logo forest in JASPAR matrix clustering.(b) Human and mouse OVOL2 ChIP-seq showing OVOL2 binding events at orthologous genomic regions corresponding to promoters of the piRNA pathway genes TDRD7 and TDRKH (rpm; merged n=2 replicates from distinct samples; mouse data is from day 2 of female mouse primordial germ celllike cell (PGCLC) induction overexpressing mouse Ovol2a 52 ; human data is from iPSC line overexpressing human OVOL2 57 ).Multiple sequence alignments of OVOL2 motifs at peak summits indicate high degree of conservation of Ovo binding sites across mammalian species (d) Immunofluorescence images showing the presence of Ago3 proteins as peri-nuclear nuage-like foci (arrowheads) in the OSCs nucleofected with the ovo-FLAG construct (DAPI indicates the nuclear DNA and Piwi indicates the nucleoplasm).(e) Immunofluorescence images showing distinct localization of the nuage-like foci (red, Aub) appearing within the ovo-FLAG nucleofected OSCs from the somatic Yb bodies (green) normally present in OSCs.(f) Plot showing the Ovo target genes (x-axis; log2 fold-change response to ectopic Ovo expression in OSCs) that are germline-enriched (y-axis; vas-GFP+) and have gene-regulatory functions (TF=transcription factor, CBP=chromatin binding protein).