Abstract
PIWI proteins and PIWI-interacting RNAs (piRNAs) play a central role in repressing transposable elements in animal germ cells. It is thought that piRNAs are mainly produced from discrete genomic loci named piRNA clusters, which often contain many “dead” transposon remnants from past invasions and have heterochromatic features. In the genome of silkworm ovary-derived cultured cells called BmN4, a well-established model for piRNA research, torimochi was previously annotated as a unique and specialized genomic region that can capture transgenes and produce new piRNAs bearing a trans-silencing activity. However, the sequence identity of torimochi has remained elusive. Here, we carefully characterized torimochi by utilizing the updated silkworm genome sequence and the long-read sequencer MinION. We found that torimochi is in fact a full-length gypsy-like LTR retrotransposon, which is exceptionally active and has massively expanded its copy number in BmN4 cells. Many copies of torimochi in BmN4 cells have features of open chromatin and the ability to produce piRNAs. Therefore, torimochi may represent a young, growing piRNA cluster, which is still “alive” and active in transposition yet capable of trapping other transposable elements to produce de novo piRNAs.
Author summary
Transposons are DNA sequences that can jump from one region to another in the genome. Since transposon insertions can disrupt the genetic information, the hosts need to silence their activity especially in germ cells. PIWI-interacting RNAs (piRNAs) act as the sequence-specific guide to suppress transposons. Many piRNAs are generated from discrete genomic regions, called piRNA clusters, which often contain a variety of fragmented, “dead” transposons. piRNA clusters are thought to act as genomic storage devices that memorize non-self sequences to be suppressed. However, resent findings suggest that piRNA clusters are not stationary entities but can flexibly appear and disappear. BmN4 cells, derived from silkworm ovaries, have been widely used in the piRNA research. A previous study using BmN4 cells identified torimochi (the Japanese word for birdlime) as a representative piRNA cluster, which can trap foreign sequences and produce piRNAs. However, the detailed properties of torimochi have remained unclear. Here, we found that torimochi is in fact a retrotransposon, which is exceptionally active and has massively expanded its copy number in BmN4 cells. Therefore, torimochi may represent a young, growing piRNA cluster, which is still “alive” and active in jumping yet capable of trapping other transposons to produce new piRNAs.
Introduction
Transposons are DNA sequences that can migrate from one region to another in the genome. Since transposon insertions can disrupt the genetic structure, it is necessary for the hosts to suppress transposon activity especially in germ cells, where the genome is inherited to the next generation [1]. PIWI-interacting RNAs (piRNAs) and PIWI proteins play a central role in suppressing transposons in the germline [2, 3]. PIWI proteins use piRNAs as the sequence-specific guide to repress target transposons in two ways: transcriptional gene silencing (TGS) and post-transcriptional gene silencing (PTGS). piRNA-mediated TGS is achieved by heterochromatic histone modifications and DNA methylation [4, 5]. On the other hand, piRNA-mediated PTGS relies on the endoribonucleolytic activity of PIWI proteins [6–9]. The 3′ fragments of the cleavage products by piRNA-guided PIWI protein are incorporated into another PIWI protein and processed into new piRNAs [10, 11]. These new piRNAs further guide the cleavage of complementary RNAs (i.e., the same strand as the original piRNA), leading to the production of piRNAs with the same sequence as the original piRNAs. This cleavage-dependent piRNA biogenesis, called the “ping-pong” amplification cycle, is thought to be broadly conserved among many animals including flies, mice, zebrafish, sponges, and silkworms [5, 10, 12–14].
Most piRNA sequences are densely mapped to discrete genomic regions, called piRNA clusters [10]. The piRNA clusters often contain a variety of fragmented, “dead” transposon remnants deriving from past invasions, producing piRNAs that can recognize and repress currently active transposons [7, 9, 11, 15, 16]. In this way, piRNA clusters are thought to act as genomic storage devices that keep the information of non-self sequences to be suppressed.
The flamenco and 42AB loci in Drosophila melanogaster represent the most well-studied piRNA clusters [10]. piRNAs derived from these piRNA clusters are produced in different ways: flamenco is a uni-strand piRNA cluster that produces Piwi-bound piRNAs through the phased piRNA biogenesis pathway, whereas 42AB is a dual-strand piRNA cluster that produces Aubergine (Aub)/Ago3-bound piRNAs through the ping-pong piRNA biogenesis pathway [10, 11, 17]. Despite this difference, both piRNA clusters possess trimethylation of Lys9 on histone H3 (H3K9me3), a typical mark of heterochromatin [18]. These large heterochromatic piRNA clusters were originally proposed to provide the immunity to homologous transposon copies across the genome in trans [19, 20]. Indeed, flamenco is essential for transposon repression and oogenesis [21]. However, it was recently shown that 42AB and some other dual-strand piRNA clusters are evolutionarily labile and mostly dispensable for transposon suppression, raising the possibility that dispersed elements in individual transposons can mediate silencing of themselves in cis [22]. Moreover, it has been reported that new euchromatic transposition of transposons can act as a source of piRNAs [23, 24]. Such euchromatic piRNA clusters are found not only in flies but also in mice and mosquitos [5, 23, 25]. These findings suggest that piRNA clusters are not stationary entities but can flexibly appear and disappear.
BmN4 cells, derived from silkworm ovaries, retain the active piRNA pathway including the ping-pong amplification cycle, and thus have been widely used in piRNA research [14]. Silkworms have no homolog of Drosophila Piwi, the PIWI protein that is specialized for the phased piRNA biogenesis pathway. Instead, silkworm Siwi participates in both the ping-pong and phased piRNA biogenesis pathways [26]. A previous study using BmN4 cells identified torimochi (the Japanese word for birdlime) as a piRNA cluster in silkworms [27]. Torimochi was regarded as a unique and specialized piRNA cluster in the BmN4 genome, which traps foreign transgenes and produces piRNAs that can repress homologous transgenes in other genomic regions in trans [27]. However, the detailed sequence properties of torimochi have remained unclear.
The previous torimochi study was conducted using the old silkworm genome sequence originally published in 2008 [27, 28]. At that time, the silkworm genome contained many unassembled regions, even within torimochi. Accordingly, there was only one genomic region to which torimochi-derived piRNAs were mappable, and the apparent difference in the chromosomal position of torimochi between BmN4 cells and silkworm ovaries was thought to be due to a large rearrangement of the corresponding genomic region [27]. However, the silkworm genome sequence has recently been updated with completing most of the previously unassembled regions [29]. In this study, we carefully reexamined the torimochi region to understand why it can trap foreign transgenes for silencing. Using the updated silkworm genome sequence and the long-read sequencer MinION, we found that torimochi is in fact a full-length gypsy-type transposon, which has massively expanded its copy number in BmN4 cells compared to silkworms. Moreover, our single-nucleotide polymorphism (SNP) and chromatin immunoprecipitation (ChIP) analyses indicated that many copies of torimochi in BmN4 have features of open chromatin and the ability to produce piRNAs. Therefore, torimochi may serve as a model for young piRNA clusters, which are still “alive” and active in transposition, can trap other transposons, and produce de novo piRNAs.
Results and discussion
The piRNA cluster torimochi is a full-length gypsy-type transposon
In the original silkworm genome sequence published in 2008, torimochi was annotated as a specialized piRNA cluster located on chromosome 11, because no highly homologous region was found elsewhere in the genome [27, 28]. Moreover, the chromosome 11 torimochi itself contained a large unassembled region. However, recent improvements in sequencing technology have determined the sequences of many previously unassembled regions in the silkworm genome [29]. Using the updated silkworm genome sequence published in 2019, we performed a nucleotide BLAST search using the known torimochi sequence as a query. We found that there are at least three copies of torimochi-like sequences, not only on chromosome 11 but also on chromosomes 12 and 24 (Fig 1A). Alignment of these three copies of torimochi-like sequences suggested that they are full-length gypsy-type retrotransposons, which even retain the conserved LTR sequences at both ends (Fig 1B). Therefore, torimochi is not a unique sequence in the genome as was thought in the past, but should be now interpreted as a gypsy-type transposon with multiple copies in the genome.
Fig 1. torimochi is a full-length, gypsy-type transposon.
(A) Schematic representation of the three copies of torimochi in the recently published silkworm genome. The copy on chromosome 24 has a deletion in the region encoding F protein. (B) Structure of torimochi on chromosome 12. There are LTR sequences at both ends, with retrotransposon gag protein (GAG), retroviral aspartyl protease (RVP), reverse transcriptase (RTase), integrase, and F protein (Baculovirus F protein) encoded inside. RVP, RTase, and integrase are encoded as polyproteins without stop codons. Black boxes show the reading frame and the numbers below indicate the amino acid numbers. (C) Distribution of torimochi-derived piRNAs in BmN4 cells and silkworm ovaries. Although torimochi on chromosome 11 is used as a representative, piRNAs produced from torimochi on other chromosomes can be mapped in the same way due to sequence similarity. The positive and negative values on the y-axis indicate the coverage of piRNAs in the sense and antisense directions, respectively (the same hereafter). (D) Distribution of piRNAs derived from the regions inside and outside of torimochi in the recently published silkworm genome. (E–G) Genomic PCR using primers outside of torimochi on chromosome 11 (E), chromosome 12 (F), and chromosome 24 (G). The genome DNAs of BmN4 cells and silkworm strains p50T, N4, and C108T and B. mandarina (Bma) were used. Black, white arrows indicate the band lengths with and without torimochi, respectively. The PCR amplification efficiency of the full-length torimochi on chromosome 12 was extremely low due to its sequence context (see S1E Fig). (H) Estimation of the copy number of torimochi by qPCR of genomic DNA. The copy number of rp49 (a gene on an autosomal chromosome) was normalized to 1.
We found that torimochi-derived piRNAs are highly expressed in BmN4 cells but not in silkworm ovaries, suggesting that torimochi has gained the piRNA-producing ability in BmN4 cells (Fig 1C). Most (~73%) of the torimochi-derived small RNAs were ~26–32 nucleotides long (S1A Fig), which showed both the ping-pong signature (S1B Fig) and the head-to-tail phasing signature (S1C Fig). Therefore, they can be regarded as bona fide piRNAs. To further characterize the piRNA-generating activity of torimochi, we compared the patterns of piRNA production inside and outside of torimochi for the three insertion sites on the BmN4 genome. Every insertion site showed clear boundaries for piRNA production; piRNAs were densely mapped within the torimochi region, whereas the outside region was almost devoid of piRNAs (Fig 1D). This observation suggests that torimochi is a full-length transposon that acts as a piRNA-producing cassette.
The 2019 silkworm genome was based on the p50T strain of silkworm [29]. Given that torimochi is a full-length transposon that could in theory retain transposition activity, the actual locations of torimochi may be different between the genomes of P50T strain and BmN4 cells. Therefore, we performed quantitative genomic PCR of three silkworm strains (p50T, N4, and C108T), Bombyx mandarina (Bma, a wild progenitor of Bombyx mori), and BmN4 cells. We found that the torimochi insertion status at the above-identified three genomic locations greatly varies among strains. In fact, the BmN4 genome lacks torimochi insertion at any of the three above-mentioned locations (Fig 1E–1G). Nevertheless, when we measured the copy number of torimochi by qPCR, we found that torimochi in BmN4 cells has ~25-fold more copies than the ribosomal protein 49 (rp49) gene on the autosomes, which is remarkably higher than those in the B. mori strains and B. mandarina (Fig 1H). These results suggest that torimochi is an active transposon unit that keeps transpositioning in the silkworm genome and that BmN4 cells have gained massive copy numbers of torimochi presumably during the establishment of the cell line and/or cell passage.
Detailed annotation of torimochi in the BmN4 genome by utilizing MinION sequencer
To investigate where torimochi copies are inserted in the BmN4 genome, we opted to use the long-read sequencer MinION. We processed the MinION reads to search for the junctions between the known torimochi sequences and other genomic regions in BmN4 cells (Fig 2A). Among them, we identified at least nine sites that have reliable junctions at both ends and with a reasonable genomic distance. Genomic PCR showed two distinct lengths of bands—with and without torimochi—for eight of those nine sites, indicating that torimochi is inserted in a heterozygous manner at each site (S2 Fig). For the insertion in chromosome 22, bands of incomplete lengths were observed, so we decided to exclude it from further analyses for simplicity (S2 Fig). We then obtained the precise sequences of these eight torimochi insertions by Sanger sequencing of PCR-amplified fragments and constructed a phylogenetic tree together with the above-identified three torimochi insertion sequences in the p50T silkworm genome. We found that all the eight torimochi sequences in the BmN4 genome are similar to each other, belonging to one clade that is distant from the three torimochi copies in p50T strain (Fig 2B). This suggests that expansion of torimochi in BmN4 cells has occurred relatively recently. In the eight homologous torimochi sequences in the BmN4 genome, we could identify 61 SNP positions that can help distinguish them (Fig 2C). At each position of these SNPs, we analyzed the nucleotide composition of BmN4 piRNAs using previously published data [26]. We found piRNA and mRNA sequence polymorphisms at all the SNP positions (Fig 2D–2G), suggesting that torimochi piRNAs and mRNAs are produced from many different torimochi loci, rather than a single locus. Comparison of the piRNA production patterns inside and outside of torimochi for these eight insertion sites showed clear boundaries for piRNA production (S3A–S3H Fig). These observations support the model that torimochi is a full-length, multi-copy transposon, each copy of which acts as a piRNA-producing cassette.
Fig 2. Detection of novel torimochi inserts in the BmN4 genome.
(A) Definition of inserts in this paper. When the junctions between the two genomic locations are less than 100 bp apart at the genomic location 1 and are consistent in orientation, the region between the junctions at the genomic location 2 is considered as an insert. (B) Phylogenetic tree of the new torimochi insert sequences in BmN4 cells and the torimochi copies present in p50T strain. Numbers indicate bootstrap probabilities. (C) List of SNPs in each torimochi in BmN4 cell. The numbers below indicate the position of the SNPs in torimochi on chromosome 3. (D–G) Distribution of SNPs at each position in sense (D, F) and antisense (E, G) piRNAs (D, E) and mRNAs (F, G) deriving from the new torimochi inserts in BmN4 cells. SNPs position indicates the order of the SNPs shown in (C). The read coverages of the SNPs site are shown above the graphs.
In the previous torimochi paper, a GFP transgene, introduced via the piggyBac transposon, was trapped by torimochi and silenced by de novo piRNAs in BmN4 cells [27]. To identify which torimochi copy newly annotated in the BmN4 genome the GFP transgene was actually inserted into, we analyzed the genome of #8 cells, which produce abundant GFP-derived piRNAs [27], using MinION. We found two reads that cover the GFP cassette, torimochi, and the neighboring silkworm genome at the 11,929 kb position on chromosome 13 (S3I Fig). Mapping of piRNAs showed that piRNA are produced mainly within the torimochi cassette rather than the surrounding regions (S3J Fig), just like other torimochi copies on different chromosomes (S3A–S3H Fig). These data suggest that, unlike the original assumption that torimochi is a unique, specialized piRNA cluster, the torimochi unit that captured the GFP transgene in #8 cells is merely one of many torimochi copies in the BmN4 genome.
Comprehensive identification of transposons activated in the BmN4 genome
Given that torimochi turned out to be an active, multi-copy transposon, we decided to extend our analysis to other transposons in silkworms. We searched for novel transpositions in the BmN4 genome in an unbiased manner, using MinION sequencing and the same criteria as for torimochi (Fig 2A). In this analysis, we defined “novel transpositions” when those sequences that existed in the p50T genome (published in 2019) are found elsewhere in the BmN4 genome. We identified ~700 new inserts that have reliable junctions at both ends. The estimated size of these inserts was mostly in the range of 102 to 104 base pairs, so we decided to focus on the 597 inserts that fall within this range (Fig 3A). These predicted inserts can be either full-length transposons, fragments of transposons, or transposon-unrelated sequences. Therefore, we categorized them into groups by sequence homology; when locally homologous sequences were bridged by a longer insert(s) (e.g., sequence 4), we considered them all as one group (Fig 3B). Then, we further focused on 20 groups that include at least five inserts (Fig 3C). Of these, remarkably, Group 2 that had the second most insert sites represented torimochi (and its fragments), while Group 1 with the highest number of inserts turned out to represent LINE transposon (and fragments). For each group, the longest sequence was selected as the representative and annotated by nucleotide BLAST with the B. mori transposon database (S2 Table). As a result, they were identified to be LINE or SINE (Groups 1, 7, 9, 11, 15), known transposons (Groups 5, 6, 10, 12, 13, 16, 18, 19), torimochi (Group 2), or novel sequences (Groups 3, 4, 8, 14, 17, 20) (S2 Table). We found that LTR transposons like torimochi and transposons with Inverted Terminal Repeat (ITR) tend to retain their full-length copies upon new transposition (Figs 3D, S4E, S4G, S4H, S4I, S4K and S4L), whereas non-LTR transposons have produced incomplete copies fragmented from the 5’ side (S4A–S4C Fig). These results suggest that the new insert sequences identified in this study are “alive” and active transposons that have transposed relatively recently. For these novel transposons, we named group 3 as mejiro (Japanese white-eye, which is often trapped by birdlime [torimochi]; see S5 Fig), group 4 as TRAS-lbm (TRAS-like in BmN4; homologous to TRAS transposon), non-autonomous (transposase-lacking) transposons of group 8 as kotaro, 14 as kojiro, 17 as kotetsu (popular Japanese cat names; cats jump up and down, attracted by catnip [transposase from another autonomous transposon]), and group 20 wao (wandering pao; homologous to pao transposon), respectively (S4A, S4B, S4E, S4G, S4I and S4L Fig).
Fig 3. Comprehensive identification and grouping of novel insertion sequences.
(A) Estimated lengths of novel inserts in the BmN4 genome. Blue dotted lines indicate 100 bp and 10,000 bp. (B) Schematic diagram of grouping. Segments with the same color are homologous to each other. For example, sequences 1 and 6 do not have direct homology but are bridged by sequence 4, thereby categorized into the same group. (C) The numbers of inserts in each group. Those groups that had 5 or more insert sites were sorted. Group 2 represents torimochi. (D) Domain structure of the representative sequence of group2 (torimochi) (top) and the region where it is originally annotated in the p50T genome (left). The lengths of the inserts found in BmN4 cells are shown at the bottom.
torimochi is the most specifically activated transposon in BmN4 cells
Next, we compared the published genome sequence of the p50T strain and our MinION-based genome sequence of BmN4 cells. We found that p50T already has >70,000 copies of Group 1 (LINE-associated) insertions but only 10 copies of Group 2 insertions (including three full-length torimochi and seven partial torimochi; S3 Table). To quantitatively evaluate the expansion of insertions in BmN4 cells, we calculated the ratio of the total genomic areas between the “old” copies already existing in the p50T genome and the “new” copies specifically found in the BmN4 genome for each insertion group. Notably, the ratio was the highest for torimochi, indicating that torimochi is the most massively expanded insertion in BmN4 cells (Fig 4A). At the same time, we found that not only torimochi but also many other insertions (Groups 3, 8, 17, 4, 19 and 20) have increased their copies in BmN4 cells, most of which correspond to the novel transposons identified in this study (Fig 4A).
Fig 4. torimochi is the most active transposon in BmN4 cells.
(A) Ratios between “new” areas in BmN4 cells and “old” areas in p50T strain for the inserts in each group. torimochi (group 2) is the most highly activated insert group in BmN4 cells. T (torimochi), N (novel transposons), K (known transposon), R (SINE/LINE) are shown (top). (B) The amounts of piRNAs derived from the inserts in each group in ovaries and BmN4 cells. The order is the same as in (A). Active inserts tend to produce more piRNAs in BmN4 cells than in ovaries. (C) The amounts of mRNAs derived from the inserts in each group in ovaries and BmN4 cells. The order is the same as in (A). Active inserts tend to produce more mRNAs in BmN4 cells than in ovaries. (D) Enrichment of the H3K27ac euchromatin mark and the H3K9me3 heterochromatin mark compared to IgG (control) of the inserts in each group BmN4 cells. The order is the same as in (A). (E) Enrichment of the H3K27ac euchromatin mark and the H3K9me3 heterochromatin mark for 1811 known transposon sequences.
To investigate if these BmN4-specific new insertions have the ability to produce piRNAs like torimochi, we compared piRNA production from these insertions in silkworm ovaries and BmN4 cells (Fig 4B). We found that most of the “new” insertions (left) show significantly increased piRNA production specifically in BmN4 cells, whereas the piRNA levels from “old” insertions such as LINE transposon (right) were comparable between silkworm ovaries and BmN4 cells (Fig 4B). This increase of piRNA levels correlated well with the increase of mRNA levels in BmN4 cells (Fig 4C). Previous ChIP-seq analysis revealed that torimochi is a transcriptionally active region with high levels of euchromatic modifications [30]. We therefore investigated the degree of histone modifications for these “new” inserts. Remarkably, these “new” insertions with piRNA-producing ability showed enrichment of the euchromatin mark H3K27ac and the depletion of the heterochromatin mark H3K9me3 (Fig 4D), compared to their average levels in all sequences in the silkworm transposon database (Fig 4E). These results suggest that these “new” insertions have an open chromatin structure, as previously reported for torimochi [27, 30]. Nevertheless, it is noteworthy that, among those “new” insertions, torimochi has the highest activity in transposition and the most open chromatin structure (Fig 4A and 4D), which may explain why the GFP transgene was trapped by torimochi for piRNA production in the previous study [27]. This is not limited to the GFP transgene; we identified as many as 36 insertions of transposons with at least one end forming a well-defined junction with torimochi (S5A Fig). Notably, we found that the Group 3 transposon mejiro, the second most active transposon in BmN4 cells, is inserted into seven different sites within torimochi (S5A Fig), one of which was confirmed by PCR to be BmN4-specific (S5B and S5C Fig.). Thus, torimochi is active in trapping not only exogenously introduced transgenes but also endogenous transposons.
To determine when torimochi became active, we compared torimochi insertion sites using our standard line of BmN4 cells (originally from Kyushu University; referred to as BmQ here) and another line of BmN4 cells that had been maintained in Yamaguchi University for decades and then have maintained in our laboratory for ~10 years [31] (referred to as BmY here). Initial screening of the torimochi insertion sites in the genome of each BmN4 line identified 26 sites in BmQ and 17 sites in BmY, only two of which appeared to be common in both lines. However, it was possible that our standard criteria for defining the torimochi insertion sites were too stringent. Therefore, we directly mapped the reads of the MinION sequence at each identified insertion site and calculated the ratio of reads with or without torimochi insertion (S6A Fig; note that the BmN4 genome is highly polyploidized, so each chromosomal copy may or may not contain the torimochi sequence at each defined insertion site). As a result, all the 26 insertion sites defined in the BmQ genome turned out to contain the torimochi insertion sequence even in the BmY genome at similar presence/absence ratios, and vice versa for 16 out of the 17 insertion sites defined in the BmY genome (S6B Fig). There was one insertion site that was present in the BmY genome but not in the BmQ genome. The results suggest that the majority of torimochi transpositions had occurred more than a few decades ago, most likely upon establishment of the original BmN4 cell line, and yet new insertion events are still ongoing.
torimochi is transcribed by the BmN4 cell-specific transcription system
It is known that well-established piRNA clusters in Drosophila (e.g., the 42AB cluster) have a specialized system for transcriptional activation. However, those specialized transcriptional activators such as the Rhino-Deadlock-Cutoff (RDC) are conserved only within the Drosophila genus, and thus the transcriptional activation systems of piRNA clusters are likely to be different in different organisms [32–34]. Keeping this in mind, we asked if the transcription mechanism of torimochi is any different from other piRNA-producing transposons in BmN4 cells. Although specific transcriptional activators of piRNAs clusters remain unknown in silkworms (as in many other animals except for Drosophila), there is a method to differentiate BmN4 cells into adipocyte-like cells as performed in mammalian 3T3-L1 cells [35]. Therefore, we decided to differentiate BmN4 cells into adipocytes so that they lose their “germline-ness” [35]. As expected, the expression of the adipocyte marker BmFABP1 (Fatty Acid-Binding Protein 1) was markedly increased (Fig 5A), while the expression levels of piRNA-related factors such as Vasa were decreased (Fig 5B). Importantly, transcription of torimochi was drastically reduced by adipocyte differentiation (Fig 5C), whereas most other transposons, including those piRNA-producing transposons in BmN4 cells, remained unrepressed or rather increased by differentiation (Fig 5C and 5D). These findings suggest that, even among those piRNA-producing transposons in BmN4 cells, torimochi has started to gain a specialized, germline-specific transcriptional activation system.
Fig 5. Downregulation of torimochi by differentiation of BmN4 cells into adipocytes.
(A) Relative expression levels of BmFABP1 (Fatty Acid-Binding Protein 1) during differentiation of BmN4 cells into adipocytes, calculated from the RNA-seq data. The expression levels at day 0 (untreated) were used as the baseline. BmFABP1 is upregulated by differentiation. (B) Relative expression levels of piRNA-related genes during differentiation of BmN4 cells into adipocytes. The crowded parts are labeled in the order of the relative expression levels at day 20. The expression levels at day 0 (untreated) were used as the baseline. Many piRNA-related factors are downregulated by differentiation. (C) Relative expression levels of the insert groups identified in Fig 4 during differentiation of BmN4 cells into adipocytes. The crowded parts are labeled in the order of the relative expression levels at day 20. The expression levels at day 0 (untreated) were used as the baseline. torimochi is downregulated by differentiation, whereas other transposons are upregulated. (D) Box plots of the relative expression levels of transposons during adipocyte differentiation. 579 transposons with the RPKM values greater than 10 in any of the libraries on day 0, 10, or 20 were plotted. Center line, median; box limits, upper and lower quartiles; whiskers, 1.5 × interquartile range; points, outliers.
In this study, we found that torimochi is not a specialized piRNA cluster but in fact a full-length gypsy-type transposon that is exceptionally active in BmN4 cells. It was recently reported that large piRNA clusters are evolutionarily labile and can be deleted without compromising transposon regulation in Drosophila [22]. Thus, piRNA clusters in the genome are not as static as previously thought but can flexibly appear and disappear. Indeed, torimochi does not serve as a source of piRNAs in silkworm ovaries but has massively expanded its copy number and gained the activity to produce piRNAs in BmN4 cells (Fig 1). Torimochi has the open chromatin structure and can trap foreign transgenes as well as endogenous transposons (Figs 4, S3I and S5). Moreover, our SNP analysis showed that piRNAs are produced from many different copies of torimochi in the BmN4 genome. Notably, unlike other transposons, torimochi seems to be integrated into a specialized, BmN4-specific transcription system (Fig 5). Therefore, torimochi in BmN4 cells may represent a young, growing piRNA cluster, which is still “alive” and active in transposition but capable of trapping other transposable elements to produce de novo piRNAs. Moreover, we successfully identified six novel transposons that have expanded their copies and gained the piRNA-producing activity specifically in BmN4 cells, just like torimochi (S4 Fig). Future studies should focus on how transposons gain the ability of piRNA production and what hallmark features active piRNA clusters have.
Materials and methods
Cell lines
BmN4 cells (BmQ provided by Chisa Yasunaga-Aoki, Kyushu University, and BmY provided from Dr. Jun Kobayashi, Yamaguchi University) were cultured at 27°C in IPL-41 medium (Applichem) and TC-100 medium (Applichem) supplemented with 10% fetal bovine serum, respectively.
Search for torimochi sequences in the genome and their annotation
torimochi sequences were BLAST-searched in SilkBase (http://silkbase.ab.a.u-tokyo.ac.jp/) using the sequence information described in a previous paper [27] as a query in the updated silkworm genome [29]. Since there were homologous regions on multiple chromosomes, the sequences were extended to the 5′ and 3′ ends, and then the common regions and Long Terminal Repeat (LTR) regions were identified by BLAST. The obtained nucleotide sequences were analyzed by pfam to estimate the domain structure of the encoded proteins (http://pfam.xfam.org/).
Sequence analysis of deposited small RNAs for known torimochi
Small RNA sequence of BmN4 cells (DRR181866) and silkworm ovaries (DRR086591) were reported previously [26, 36]. Informatics analysis of small RNAs was performed as reported previously [37]. In brief, the 3′-adaptor sequences were identified and removed by cutadapt 2.6 with Python 3.7.0 with “—minimum-length 20” parameter [38] Reads that could be mapped by bowtie 1.2.3 [39] to the Bombyx mori genome [29] up to two mismatches were used to calculate the mapping rate and to normalize each library. Sam files were converted to bam files by SAMtools [40], then to bed files, and the coverage of each nucleotide was calculated by BEDTools [41]. The torimochi sequence on chromosome 11 was used as a representative of torimochi sequences. The 10 kbs regions before and after the sequences homologues to the chromosome 11 torimochi sequence were extracted using BEDTools fastaFromBed [41]. Two nucleotide mismatches and multi-maps were allowed when mapping small RNAs to those sequences using bowtie [39]. Of the mapped results, only small RNAs of ≥23 nucleotides were extracted and treated as piRNAs, and the coverage of each nucleotide was calculated using the “-d, -s, -S” options of coverageBed in bedtools. For the calculation of the ping-pong signature, torimochi-mapped small RNAs (23–32 nucleotides) were extracted and the frequency of overlap at the 5′ end was calculated. The z-score was calculated by dividing the difference from the mean scores of the background by the standard deviation of the scores the background, defined as distances of 1–9 and 11–20 nucleotides. To calculate the phased signature, torimochi-mapped small RNAs of ≥23 nucleotides were extracted and the frequency of the distance between the 3′ end and the downstream 5′ end was calculated. The z-score was calculated by dividing the difference from the mean scores of the background by the standard deviation of the scores of the background, defined as distances of 0 and 2–19 nucleotides. Results were imported into R and graphed.
Genome extraction and genomic PCR
Genomic DNA was extracted from BmN4 cells using the Blood and Tissue kit (QIAGEN) according to the protocol of the manufacturer. Genomic DNA from Bombyx mori and Bombyx mandarina was provided by Takashi Kiuchi. PCR was performed by step-down PCR using KOD One (TOYOBO), and the PCR products were analyzed in 1% agarose with 1 kb DNA Ladder (New England Biolabs). Primers used in this study are shown in S1 Table.
Quantitative PCR for measuring the copy number of torimochi
Genomic DNA was used as a template for quantitative PCR analysis on a Thermal Cycler Dice Real Time System (TaKaRa) using KAPA SYBR Fast qPCR Kit (KAPA Biosystems) and the primers shown in S1 Table Absolute quantification of the copy number was performed using a known amount of plasmid containing the primer region. The primer set corresponding to the rp49 gene on the autosome was used as a control.
MinION-seq and mutation detection
Genomic DNA was extracted from BmN4 cells using Blood & Cell Culture DNA Kit (QIAGEN), sequencing libraries were prepared using Rapid Sequencing Kit (Oxford Nanopore), and the data were acquired by MinION using flow cell (version 9). The obtained sequences were mapped to the silkworm genome by ngmlr, and the variants of genomic structure were detected by svim with “–nanopore” option [42, 43]. Then, the structural variants of BND (breakends; connection of distal genomic positions) were extracted, and the 597 inserts that have reliable junctions at both ends were obtained by R script. Blastn was performed using the silkworm genome, known transposon sequences, and the 597 new insert sequences. Groups were created by combining sequences with e.values of less than 1e-100 among the new inserts (Fig 3B). Then, for each group, the regions in the silkworm genome that showed e.values of less than 1e-100 were identified as the regions of homology. The longest insert sequence in each group was used as the representative sequence and its sequence annotation was performed by pfam (http://pfam.xfam.org/). For novel transposon sequences that did not show homology to known transposon sequences, Sanger sequencing was performed, and the determined sequences were registered in the repository (the accession numbers: LC742510–5).
Cloning of torimochi and phylogenetic tree analysis
The sequences outside each torimochi copy were used to design the primers to distinguish it from other copies. PCR was performed by using inside-outside and outside-inside primer sets for each torimochi copy, and the PCR products were purified using NucleoSpin Gel and PCR Clean-up (MACHEREY-NAGEL) and then sequenced. torimochi sequences were clustered by Clustal W in MEGA X [44] and the phylogenetic trees were generated by the maximum likelihood method with bootstrapping.
SNPs detection
The sequence of each torimochi copy was combined and aligned with MEGA X [44], and SNP-sites [45] was used to generate a list of polymorphisms. The piRNA sequence library and the RNA sequence library divided into 33 nucleotides each [26] were mapped to the torimochi sequence on chromosome 3 using bowtie, allowing two mismatches and multi-mapping [39]. Sam files were converted to bam files by SAMtools [40], and the SNP information of piRNAs was extracted using a custom R script.
Detection of EGFP transgene inserted torimochi
The genomic sequence of the GFP repressed BmN4 line [27] was obtained by minION and analyzed by svim [43], and the IDs of the reads mapped to both GFP and torimochi were extracted by using the information in “candidates_breaksends.bed” of svim output. The region on the genome corresponding to the extracted reads were identified using BLAST on SilkBase (http://silkbase.ab.a.u-tokyo.ac.jp/), and the sequence was reconstructed for the following analysis. The piRNA sequences of the GFP repressed line were mapped to the reconstructed sequence using bowtie, allowing one mismatch and multi-mapping. Of the mapped reads, only those with ≥23 nucleotides were used to calculate the coverage of each nuelcotide using coverageBed and the “-d, -s, -S” options [41].
Mapping of ChIP-seq, mRNA-seq, and piRNA-seq to novel inserts and transposons
Previously published libraries of ChIP-seq, mRNA-seq and piRNA-seq [26, 27, 30, 36, 46–48] were mapped to the newly identified inserts using hisat2 (ChIP-seq, mRNA-seq) and bowtie (piRNA-seq), and the numbers of length-corrected mapped reads were calculated [39, 49]. Sam files were converted to bam files by SAMtools [40], then to bed files, and the coverage of each nucleotide was calculated by BEDTools [41].
Detection of transposition of transposon inside torimochi
The BmN4 genome sequences were mapped to the known transposons and newly identified transposons in this paper by ngmlr, and the variants of genomic structure were detected by svim with “-nanopore” option [42, 43]. Then, the structural variants of BND of torimochi were extracted and the 36 inserts that have reliable junctions at at least one end of transposon and inside of torimochi were identified. The results were visualized using the circlize package in R. A representative insertion was confirmed by PCR (S1 Table).
Adipocyte differentiation of BmN4 cells
For the induction of adipocyte differentiation, BmN4 cells were cultured in IPL-41 medium supplemented with 10% fetal bovine serum, 10 μg/ml insulin (SIGMA), 0.5 μM 3-isobutyl-1-methylxanthine (SIGMA) and 250 μM dexamethasone (SIGMA). Medium was changed twice a week and collected 10 and 20 days after the drug addition to prepare RNA sequence libraries. These libraries were mapped to the silkworm genome and genemodel, transposon, and the novel inserts identified in this paper using Hisat2 with the default parameters, and the mapped reads were calculated by coverageBed of bedtools [41, 49]. Silkworm homologs of piRNA-related genes were identified using blast of silkbase. Since the genemodel for BmAgo3 was splited into three (KWMTBOMO01223-5) and the genemodel for Tudor was splited into two (KWMTBOMO06482-3), the RPKM was calculated by combining them.
Comparison of torimochi insertion sites in different lines of BmN4 cells
The genome sequence of BmY was also sequenced by MinION in the same manner as described above, and the insert position of torimochi was identified by svim. Sequences with and without torimochi inserted at these insert positions were prepared using R scripts. MinION sequencing data for BmQ and BmY were mapped to these sequences using bwa-mem with “ont2d” option. The presence or absence of the reads that cover at least 100 bps at both sides of a given junction site was used as a criterion.
Supporting information
(A) The length distribution of torimochi-derived small RNAs in BmN4 cells. (B) Schematics for the ping-pong signature (C) and the phasing signature (D). (C) The fraction of 5’-5’ overlapped piRNAs (y-axis) at the indicated length (x-axis) for torimochi-derived piRNAs in the small RNA libraries from BmN4 cells. The Z-scores at the 10 nt overlap (ping-pong signature) are shown in the top-right of the graphs. (D) The fraction of 3’-5’ distance for torimochi-derived piRNAs in naive BmN4 cells or BmN4 cells knocked out for Trimmer, the exonuclease that trims the 3’-end of precursor piRNAs for maturation (TriKO). Z1 denotes the z score at position 1. A strong tail-to-head phasing signature (at position = 1) was detected in torimochi-derived piRNAs from TriKO cells. (E) In the chr12 region of the p50T genome, torimochi is nested within another transposon, LINE/R1. This LINE/R1 element is extremely GC-rich and repetitive, which likely hinders the PCR amplification of the full-length sequence in Fig 1F. On the other hand, a strong band was detected at ~2.0 kb for BmN4 (Fig 1F), which indicates the absence of both torimochi and LINE/R1 in this region.
(TIFF)
(A–I) Genomic PCR of the torimochi inserts newly identified in BmN4 cells. The genome DNAs of BmN4 cells and silkworm strains p50T, N4, and C108T and B. mandarina (Bma) were used. Black and white arrows indicate the band lengths with and without torimochi, respectively. (G) The genome of the N4 strain has a torimochi insertion at the same site.
(TIFF)
(A–H) Distribution of torimochi-derived piRNAs in BmN4 cells and ovaries. The boundaries of torimochi are shown by gray lines. (I) Schematic diagram of torimochi, GFP transgene, and the outside genomic regions on chromosome 13. The MinION reads that span these regions are shown by black lines. (J) piRNA production from the torimochi copy on chromosome 13 in BmN4 cells and ovaries
(TIFF)
(A–L) Domain structure of the representative sequence of each group (top), the region where it is originally annotated in the p50T genome (left), and the lengths of the inserts found in BmN4 cells (bottom). ORF: ORFs that could not be annotated by Pfam. Names of newly identified transposons are indicated in red.
(TIFF)
(A) Transposons inserted in torimochi and their locations. Transposons with at least one end (within 100 bp) forming a junction with a sequence inside torimochi were considered to be inserted into torimochi. Seven out of the 36 inserts accounted for mejiro (S4A Fig). The red line shows the insertion whose presence was confirmed by genomic PCR in S5B and S5C Fig. (B) The insertion of mejiro within torimochi shown as red in S5A Fig. The primer sets used in C are shown by black bars. (C) Genomic PCR shows the transposon insertion specifically in BmN4 cells.
(TIFF)
(A) Definition of torimochi insertion. The presence or absence of the reads that cover at least 100 bps at both sides of a given junction site was used as a criterion in (B). (B) The torimochi insertion sites identified in the BmQ (left) or BmY (right) genome. The ratio of MinION reads from BmQ (top) or BmY (bottom) cells with or without torimochi sequence was plotted for each identified insertion site.
(TIFF)
(PDF)
The longest sequence in each group as representative sequence was annotated by BLASTn. Green or gray rows are those that could not be illustrated in S3 Fig due to many variations in the sequence. Green group’s representative sequences are LINE/SINE and gray are others.
(PDF)
The "old_area" represents the length of homologous regions in the p50T genome, and the "new_area" represents the total length of the inserts identified in this study.
(PDF)
Acknowledgments
We thank Yutaka Suzuki for sequencing for differentiated BmN4 cells, Takashi Kiuchi for providing silkworm genomic DNA and Kaori Kiyokawa for technical assistance. We thank all the members of the Tomari laboratory for discussion and critical comments on the manuscript.
Data Availability
The sequencing data obtained in this study are available under the accession number DRA014297 (https://ddbj.nig.ac.jp/resource/sra-submission/DRA014297), DRA015290 (https://ddbj.nig.ac.jp/resource/sra-submission/DRA015290) (long DNA-seq). RNA sequences of differentiated BmN4 cells are available under the accession number DRA005335 (https://ddbj.nig.ac.jp/resource/sra-submission/DRA005335). Small RNA sequences of BmN4 cells (DRR181866), GFP repressed BmN4 cell line (DRR000974) and silkworm ovary (DRR086591), RNA sequences of BmN4 cells (DRR178519) and silkworm ovary (DRR186503), and ChIP sequences of IgG-R (DRR001873), H3K9me3 (DRR001878) and H3K27ac (DRR018643) antibodies were reported previously (Yokoi et al. 2021 (doi:10.3390/insects12060519); Izumi et al. 2020 (doi:10.1038/s41586-020-1966-9); Katsuma et al. 2021 (doi:10.1002/arch.21761), 2019 (doi:10.1002/2211-5463.12698); Kawaoka et al. 2012 (doi:10.1261/rna.029777.111), 2013 (doi:10.1093/nar/gks1275); Shoji et al. 2015 (doi:10.1093/nar/gku862)). The sequences of the transposons newly identified in this paper have been deposited in DDBJ under the following accession numbers: LC742510–5 (http://getentry.ddbj.nig.ac.jp/getentry/na/LC742510/ http://getentry.ddbj.nig.ac.jp/getentry/na/LC742511/ http://getentry.ddbj.nig.ac.jp/getentry/na/LC742512/ http://getentry.ddbj.nig.ac.jp/getentry/na/LC742513/ http://getentry.ddbj.nig.ac.jp/getentry/na/LC742514/ http://getentry.ddbj.nig.ac.jp/getentry/na/LC742515/). All code required for bioinformatics analysis in this paper is available at https://github.com/keishoji/torimochi.
Funding Statement
This work is supported by MEXT | Japan Society for the Promotion of Science (JSPS) 18H05271 to YT; 16KT0064 to SK; 15J00469 to KS; 19K06484 to KS; 22K15082 to KS. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.Ernst C, Odom DT, Kutter C. The emergence of piRNAs against transposon invasion to preserve mammalian genome integrity. Nat Commun. 2017. Dec 10;8(1):1411. doi: 10.1038/s41467-017-01049-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Klattenhoff C, Theurkauf W. Biogenesis and germline functions of piRNAs. Development. 2008. Jan;135(1):3–9. doi: 10.1242/dev.006486 [DOI] [PubMed] [Google Scholar]
- 3.Ghildiyal M, Zamore PD. Small silencing RNAs: an expanding universe. Nat Rev Genet. 2009. Feb;10(2):94–108. doi: 10.1038/nrg2504 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sienski G, Dönertas D, Brennecke J. Transcriptional Silencing of Transposons by Piwi and Maelstrom and Its Impact on Chromatin State and Gene Expression. Cell. 2012. Nov;151(5):964–80. doi: 10.1016/j.cell.2012.10.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Aravin AA, Sachidanandam R, Bourc’his D, Schaefer C, Pezic D, Toth KF, et al. A piRNA pathway primed by individual transposons is linked to de novo DNA methylation in mice. Mol Cell. 2008. Sep 26;31(6):785–99. doi: 10.1016/j.molcel.2008.09.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Li C, Vagin V v, Lee S, Xu J, Ma S, Xi H, et al. Collapse of germline piRNAs in the absence of Argonaute3 reveals somatic piRNAs in flies. Cell. 2009. May 1;137(3):509–21. doi: 10.1016/j.cell.2009.04.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Reuter M, Berninger P, Chuma S, Shah H, Hosokawa M, Funaya C, et al. Miwi catalysis is required for piRNA amplification-independent LINE1 transposon silencing. Nature. 2011. Dec 8;480(7376):264–7. doi: 10.1038/nature10672 [DOI] [PubMed] [Google Scholar]
- 8.Wang W, Yoshikawa M, Han BW, Izumi N, Tomari Y, Weng Z, et al. The initial uridine of primary piRNAs does not create the tenth adenine that is the hallmark of secondary piRNAs. Mol Cell. 2014. Dec;56(5):708–16. doi: 10.1016/j.molcel.2014.10.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.de Fazio S, Bartonicek N, di Giacomo M, Abreu-Goodger C, Sankar A, Funaya C, et al. The endonuclease activity of Mili fuels piRNA amplification that silences LINE1 elements. Nature. 2011. Dec 8;480(7376):259–63. doi: 10.1038/nature10547 [DOI] [PubMed] [Google Scholar]
- 10.Brennecke J, Aravin A, Stark A, Dus M, Kellis M, Sachidanandam R, et al. Discrete small RNA-generating loci as master regulators of transposon activity in Drosophila. Cell. 2007. Mar;128(6):1089–103. doi: 10.1016/j.cell.2007.01.043 [DOI] [PubMed] [Google Scholar]
- 11.Gunawardane LS, Saito K, Nishida KM, Miyoshi K, Kawamura Y, Nagami T, et al. A slicer-mediated mechanism for repeat-associated siRNA 5’ end formation in Drosophila. Science. 2007. Mar 16;315(5818):1587–90. doi: 10.1126/science.1140494 [DOI] [PubMed] [Google Scholar]
- 12.Houwing S, Kamminga LM, Berezikov E, Cronembold D, Girard A, van den Elst H, et al. A role for Piwi and piRNAs in germ cell maintenance and transposon silencing in Zebrafish. Cell. 2007. Apr 6;129(1):69–82. doi: 10.1016/j.cell.2007.03.026 [DOI] [PubMed] [Google Scholar]
- 13.Grimson A, Srivastava M, Fahey B, Woodcroft BJ, Chiang HR, King N, et al. Early origins and evolution of microRNAs and Piwi-interacting RNAs in animals. Nature. 2008. Oct 1;455(7217):1193–7. doi: 10.1038/nature07415 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kawaoka S, Hayashi N, Suzuki Y, Abe H, Sugano S, Tomari Y, et al. The Bombyx ovary-derived cell line endogenously expresses PIWI/PIWI-interacting RNA complexes. RNA. 2009. Jul;15(7):1258–564. doi: 10.1261/rna.1452209 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nishida KM, Saito K, Mori T, Kawamura Y, Nagami-Okada T, Inagaki S, et al. Gene silencing mechanisms mediated by Aubergine piRNA complexes in Drosophila male gonad. RNA. 2007. Sep 5;13(11):1911–22. doi: 10.1261/rna.744307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Saito K, Nishida KM, Mori T, Kawamura Y, Miyoshi K, Nagami T, et al. Specific association of Piwi with rasiRNAs derived from retrotransposon and heterochromatic regions in the Drosophila genome. Genes Dev. 2006. Aug 15;20(16):2214–22. doi: 10.1101/gad.1454806 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Malone CD, Brennecke J, Dus M, Stark A, McCombie WR, Sachidanandam R, et al. Specialized piRNA pathways act in germline and somatic tissues of the Drosophila ovary. Cell. 2009. May 1;137(3):522–35. doi: 10.1016/j.cell.2009.03.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rangan P, Malone CD, Navarro C, Newbold SP, Hayes PS, Sachidanandam R, et al. piRNA production requires heterochromatin formation in Drosophila. Curr Biol. 2011. Aug 23;21(16):1373–9. doi: 10.1016/j.cub.2011.06.057 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Levine MT, Malik HS. Learning to protect your genome on the fly. Cell. 2011. Dec 23;147(7):1440–1. doi: 10.1016/j.cell.2011.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Malone CD, Hannon GJ. Small RNAs as guardians of the genome. Cell. 2009. Feb 20;136(4):656–68. doi: 10.1016/j.cell.2009.01.045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mével-Ninio M, Pelisson A, Kinder J, Campos AR, Bucheton A. The flamenco Locus Controls the gypsy and ZAM Retroviruses and Is Required for Drosophila Oogenesis. Genetics. 2007. Apr 1;175(4):1615–24. doi: 10.1534/genetics.106.068106 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gebert D, Neubert LK, Lloyd C, Gui J, Lehmann R, Teixeira FK. Large Drosophila germline piRNA clusters are evolutionarily labile and dispensable for transposon regulation. Mol Cell. 2021. Oct;81(19):3965–3978.e5. doi: 10.1016/j.molcel.2021.07.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Shpiz S, Ryazansky S, Olovnikov I, Abramov Y, Kalmykova A. Euchromatic Transposon Insertions Trigger Production of Novel Pi- and Endo-siRNAs at the Target Sites in the Drosophila Germline. PLoS Genet. 2014. Feb 6;10(2):e1004138. doi: 10.1371/journal.pgen.1004138 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Olovnikov I, Ryazansky S, Shpiz S, Lavrov S, Abramov Y, Vaury C, et al. De novo piRNA cluster formation in the Drosophila germ line triggered by transgenes containing a transcribed transposon fragment. Nucleic Acids Res. 2013. Jun;41(11):5757–68. doi: 10.1093/nar/gkt310 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.George P, Jensen S, Pogorelcnik R, Lee J, Xing Y, Brasset E, et al. Increased production of piRNAs from euchromatic clusters and genes in Anopheles gambiae compared with Drosophila melanogaster. Epigenetics Chromatin. 2015. Dec 27;8(1):50. doi: 10.1186/s13072-015-0041-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Izumi N, Shoji K, Suzuki Y, Katsuma S, Tomari Y. Zucchini consensus motifs determine the mechanism of pre-piRNA production. Nature. 2020. Feb 13;578(7794):311–6. doi: 10.1038/s41586-020-1966-9 [DOI] [PubMed] [Google Scholar]
- 27.Kawaoka S, Mitsutake H, Kiuchi T, Kobayashi M, Yoshikawa M, Suzuki Y, et al. A role for transcription from a piRNA cluster in de novo piRNA production. RNA. 2012. Feb;18(2):265–73. doi: 10.1261/rna.029777.111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.The International Silkworm Genome Consortium. The genome of a lepidopteran model insect, the silkworm Bombyx mori. Insect Biochem Mol Biol. 2008. Dec;38(12):1036–45. doi: 10.1016/j.ibmb.2008.11.004 [DOI] [PubMed] [Google Scholar]
- 29.Kawamoto M, Jouraku A, Toyoda A, Yokoi K, Minakuchi Y, Katsuma S, et al. High-quality genome assembly of the silkworm, Bombyx mori. Insect Biochem Mol Biol. 2019. Apr;107:53–62. doi: 10.1016/j.ibmb.2019.02.002 [DOI] [PubMed] [Google Scholar]
- 30.Kawaoka S, Hara K, Shoji K, Kobayashi M, Shimada T, Sugano S, et al. The comprehensive epigenome map of piRNA clusters. Nucleic Acids Res. 2013. Feb 1;41(3):1581–90. doi: 10.1093/nar/gks1275 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Katsuma S, Shimada T. Comparative studies of Bombyx mori nucleopolyhedrovirus infection in BmN4 cell lines maintained in different laboratories. J Insect Biotechnol Sericology. 2012;81:7–12. [Google Scholar]
- 32.Andersen PR, Tirian L, Vunjak M, Brennecke J. A heterochromatin-dependent transcription machinery drives piRNA expression. Nature. 2017. Sep 23;549(7670). doi: 10.1038/nature23482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Mohn F, Sienski G, Handler D, Brennecke J. The rhino-deadlock-cutoff complex licenses noncanonical transcription of dual-strand piRNA clusters in Drosophila. Cell. 2014. Jun 5;157(6):1364–79. doi: 10.1016/j.cell.2014.04.031 [DOI] [PubMed] [Google Scholar]
- 34.Zhang Z, Wang J, Schultz N, Zhang F, Parhad SS, Tu S, et al. The HP1 homolog rhino anchors a nuclear complex that suppresses piRNA precursor splicing. Cell. 2014. Jun 5;157(6):1353–63. doi: 10.1016/j.cell.2014.04.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Akiduki G, Imanishi S. Establishment of a lipid accumulation model in an insect cell line. Arch Insect Biochem Physiol. 2007. Nov;66(3):109–21. doi: 10.1002/arch.20203 [DOI] [PubMed] [Google Scholar]
- 36.Katsuma S, Shoji K, Suzuki Y, Iwanaga M. Potential for small RNA production against Bombyx mori latent virus in Bombyx mori ovaries. Arch Insect Biochem Physiol. 2021. Feb 1;106(2):e21761. doi: 10.1002/arch.21761 [DOI] [PubMed] [Google Scholar]
- 37.Shoji K, Suzuki Y, Sugano S, Shimada T, Katsuma S. Artificial “ping-pong” cascade of PIWI-interacting RNA in silkworm cells. RNA. 2017;23(1):86–97. doi: 10.1261/rna.058875.116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011. May 2;17(1):10. [Google Scholar]
- 39.Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009;10(3):R25. doi: 10.1186/gb-2009-10-3-r25 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009. Aug 15;25(16):2078–9. doi: 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26(6):841–2. doi: 10.1093/bioinformatics/btq033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Sedlazeck FJ, Rescheneder P, Smolka M, Fang H, Nattestad M, von Haeseler A, et al. Accurate detection of complex structural variations using single-molecule sequencing. Nat Methods. 2018. Jun 30;15(6):461–8. doi: 10.1038/s41592-018-0001-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Heller D, Vingron M. SVIM: structural variant identification using mapped long reads. Bioinformatics. 2019. Sep 1;35(17):2907–15. doi: 10.1093/bioinformatics/btz041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: Molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol. 2018. Jun 1;35(6):1547–9. doi: 10.1093/molbev/msy096 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Page AJ, Taylor B, Delaney AJ, Soares J, Seemann T, Keane JA, et al. SNP-sites: rapid efficient extraction of SNPs from multi-FASTA alignments. Microb Genom. 2016. Apr 1;2(4):e000056. doi: 10.1099/mgen.0.000056 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Yokoi K, Tsubota T, Jouraku A, Sezutsu H, Bono H. Reference Transcriptome Data in Silkworm Bombyx mori. Insects. 2021. Jun 3;12(6):519. doi: 10.3390/insects12060519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Katsuma S, Shoji K, Sugano Y, Suzuki Y, Kiuchi T. Masc-induced dosage compensation in silkworm cultured cells. FEBS Open Bio. 2019. Sep 1;9(9):1573–9. doi: 10.1002/2211-5463.12698 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Shoji K, Hara K, Kawamoto M, Kiuchi T, Kawaoka S, Sugano S, et al. Silkworm HP1a transcriptionally enhances highly expressed euchromatic genes via association with their transcription start sites. Nucleic Acids Res. 2015. Jan;42(18):11462–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Kim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat Biotechnol. 2019. Aug 2;37(8):907–15. doi: 10.1038/s41587-019-0201-4 [DOI] [PMC free article] [PubMed] [Google Scholar]