Su(Hw) interacts with Combgap to establish long-range chromatin contacts

doi:10.21203/rs.3.rs-3014225/v2

Download PDF

Article

Su(Hw) interacts with Combgap to establish long-range chromatin contacts

https://doi.org/10.21203/rs.3.rs-3014225/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this latest preprint version

Background

Insulator-binding proteins play a critical role in genome architecture by forming and maintaining contact domains. While the involvement of several IBPs in organising chromatin architecture in Drosophila has been described, the specific contribution of the Suppressor of Hairy wings (Su(Hw)) insulator-binding protein to genome topology remains unclear.

Results

In this study, we provide evidence for the existence of long-range interactions between chromatin bound Su(Hw) and Combgap, which was first characterised as Polycomb response elements binding protein. Loss of Su(Hw) binding to chromatin results in the disappearance of Su(Hw)-Combgap long-range interactions and in a decrease in spatial self-interactions among a subset of Su(Hw)-bound genome sites. Our findings suggest that Su(Hw)-Combgap long-range interactions are associated with active chromatin rather than Polycomb-directed repression. Furthermore, we observe that the majority of transcription start sites that are down-regulated upon loss of Su(Hw) binding to chromatin are located within 2 kb of Combgap peaks and exhibit Su(Hw)-dependent changes in Combgap and transcriptional regulators’ binding.

Conclusions

This proof-of-concept study demonstrates that Su(Hw) insulator binding protein can form long-range interactions with Combgap, Polycomb response elements binding protein, and that these interactions are associated with active chromatin factors rather than with Polycomb dependent repression.

Biological sciences/Molecular biology/Chromatin/Chromatin structure

Biological sciences/Molecular biology/Transcription

Insulator binding protein

Su(Hw)

Combgap

chromatin architecture

transcription

long-range interactions

Understanding the contribution of chromatin architecture to the correct spatio-temporal program of cell development is currently a crucial question in chromatin biology. Recent advances in chromosome conformation capture techniques have provided many insights into genome architecture organisation. However, it is still unclear how chromatin architecture is established and translated into a cell-specific developmental program.

Insulator-binding proteins (IBPs) are involved in the formation and maintenance of genome topology. In Drosophila, several IBPs enriched in topologically associated domain (TAD) borders have been described(1, 2, 3, 4, 5). Recent studies suggest that distinct IBP motifs and their combinations can define the TAD borders(6). Since many Drosophila IBPs are characterised by properties of homo- or hetero-oligomerisation(1, 7), it was suggested that these IBPs could be involved in maintaining TAD borders via the formation of long-range interactions (LRIs) between them. However, the role of IBPs in establishing TAD borders might be overestimated. Recent studies demonstrated that dCTCF and BEAF32 knockdowns disturb a relatively small number of TAD borders(5, 6). The contribution of these proteins to the regulation of gene expression may not only be due to the maintenance of TAD borders but also to the formation of specific regulatory long-range chromatin interactions between non-border IBP sites(5, 8).

IBPs can influence transcription by determining the nuclear or genomic context to which IBP-bound chromatin is exposed. For example, dCTCF promotes both repressed transgene localisation to Polycomb bodies and active transgene recruitment to transcription factories(9). BEAF32 can regulate the genome context of chromatin regions with which it interacts, enabling the formation of H3K27me3 micro-domains in euchromatin regions(8). IBPs frequently localise next to Polycomb-response elements (PRE)(10, 11). Recent studies suggest that IBPs can potentiate PRE activity through stimulation of pairing-sensitive silencing (PSS) and bring PREs on each homolog chromosome in close proximity(12). Further studies of IBP-regulated LRIs will provide a better understanding of how insulator proteins are involved in the implementation of cellular genetic programs.

Here we address the question of the Suppressor of hairy wings (Su(Hw)) IBP’s role in the formation of specific LRIs. We demonstrate the ability of Su(Hw) to form LRIs with Combgap, which was first introduced as a protein involved in the recruitment of PcG group proteins to chromatin(13, 14). Using wild-type and Su(Hw) loss of function Drosophila ovaries, we examine the genomic context and functional output of these LRIs.

Su(Hw) enables the indirect binding of Combgap to chromatin

Loss of Su(Hw) binding to chromatin is not lethal for flies; however, it causes female sterility with a mid-stage arrest of oogenesis and egg chamber degeneration at approximately stage 9(15, 16). Thus, Drosophila ovaries present the most biologically relevant organ to study Su(Hw) functioning. In our work we used su(Hw)^v/E8 heteroallelic mutant flies (onwards, Su(Hw) loss of function or Su(Hw)^LOFflies), which exhibit a complete loss of Su(Hw) binding to chromatin due to a point mutations in su(Hw) coding region for zinc finger 7(16). Female flies carrying su(Hw)^v/E8 alleles are sterile and suffer egg chamber degeneration at stage 9(15, 16). To compare wild type and Su(Hw)^LOF ovaries in further experiments, only ovaries of recently eclosed 15-hour-old flies, which contain egg chambers no older than stage 8, were dissected (the same way as in (17)).

In our previous study(17), we found that Su(Hw) ChIP-Seq peaks in the ovaries include a cluster of peaks that lack the Su(Hw) DNA-binding motif (hereon we refer to this cluster as indirect Su(Hw) peaks). Instead, indirect Su(Hw) peaks contain a GTGT-motif, which was previously associated with the binding of Combgap, a protein involved in the recruitment of PcG group proteins(13, 14). Here, we verified the binding of Combgap to these indirect Su(Hw) peaks both in wild-type and Su(Hw)^LOF ovaries using ChIP-Seq with antibodies against Combgap (Fig. 1a-c). The majority of indirect Su(Hw) peaks are included in the set of Combgap peaks: 78% of indirect Su(Hw) peaks (515 out of 659 peaks) intersect with Combgap peaks (Fig. 1d). However, in addition to binding to indirect Su(Hw) peaks, we noticed that Combgap also binds to a portion of direct Su(Hw) peaks (492 peaks out of 3166 direct Su(Hw) peaks), and this binding is Su(Hw)-dependent (Fig. 1a-b, Supplementary Fig. 1a). The binding of Combgap to direct and indirect Su(Hw) ChIP peaks was additionally verified in ChIP experiments coupled with qPCR (Fig. 1c). We selected 100 Su(Hw) peaks showing the strongest Combgap binding, and analysed the DNA motifs present in this set using the MEME-ChIP program(18). A strong enrichment was only found for the Su(Hw) DNA motif, not the Combgap DNA motif, indicating the favouring of Su(Hw)-mediated Combgap recruitment to these sites (Supplementary Fig. 1b).

To validate the ChIP-Seq findings, we tested an interaction between Su(Hw) and Combgap in co-immunoprecipitation experiments (co-IPs) in nuclear protein extract from Drosophila S2 cells. Antibodies against Su(Hw) successfully co-precipitated Combgap from the nuclear extract and vice versa (Fig. 1e). Moreover, recent immunoaffinity purification of the Combgap interactome coupled with high throughput mass spectrometry (IP/LC-MS) revealed statistically significant enrichment of Su(Hw) and its partners Mod(mdg4) and CP190, further supporting the existence of protein-protein interaction between Su(Hw) and Combgap(19).

Su(Hw)-bound Combgap is associated with active chromatin rather that Polycomb-directed repression

Combgap was initially identified as a protein capable of recruiting the Ph subunit of the PRC1 transcription repression complex to chromatin(13). However, in addition to its association with Polycomb group proteins, Combgap ChIP-Seq peaks have also been found to colocalise with RNA polymerase II pausing factors and the transcription start sites (TSSs) of active genes(19). This prompted us to investigate the chromatin features associated with Su(Hw)-Combgap interactions.

To address whether H3K27me3 Polycomb-dependent chromatin mark is associated with Su(Hw)-Combgap interactions in Drosophila ovary we utilised the H3K27me3 ChIP-Seq from 512c germline cells nuclei, FACS-sorted from the ovaries of 3-6 days old females of Drosophila melanogaster(20). We examined the distribution of H3K27me3 on 492 direct Su(Hw) ChIP peaks, intersecting with Combgap, 515 indirect Su(Hw) ChIP peaks, intersecting with Combgap, and on H3K27me3 broad domains, annotated from this H3K27me3 ChIP-Seq (Supplementary Fig. 2a). We observed that there is no H3K27me3 enrichment on both direct and indirect Su(Hw) peaks with Combgap, compared to H3K27me3 broad domains. Therefore, we suggest that the Su(Hw)-Combgap interaction is not associated with Polycomb-mediated heterochromatin establishment.

To investigate the potential link between Su(Hw)-Combgap interactions and RNAP II pausing factors, we performed ChIP-Seq experiments on wild-type and Su(Hw)^LOF ovaries using antibodies against the NELF E subunit of NELF, the Paf1 positive elongation factor, and the Rpb3 RNAP II subunit. We found significant Su(Hw)-dependent enrichment of NELF E, Paf1, and Rpb3 at direct Su(Hw) peaks containing Combgap (Fig. 2a, b, Supplementary Fig. 2b). The binding of NELF E, Paf1 and Rpb3 to direct Su(Hw) ChIP peaks was additionally verified in ChIP experiments coupled with qPCR (Fig. 2c). Notably, direct Su(Hw) peaks bound to NELF E (326 peaks), Paf1 (398 peaks), and Rpb3 (277 peaks) in wild-type Drosophila ovaries account for only 10.3%, 12.6%, and 8.7% of all direct Su(Hw) peaks (3166 peaks), respectively. On the other hand, 85% of direct Su(Hw)-NELF E, 80% of direct Su(Hw)-Paf1, and 94% of direct Su(Hw)-Rpb3 peaks also intersect with Combgap peaks (Fig. 2d). These observations suggest that the binding of RNA polymerase II and pausing factors to direct Su(Hw) peaks is related to Combgap rather than being intrinsic to Su(Hw).

The majority of Su(Hw)^LOFdown-regulated promoters are located within 2 kb of Combgap peaks

It is well-known that disruption of Su(Hw) binding to chromatin results in abnormal transcription in Drosophila ovaries(21, 22), however, the mechanism mediating Su(Hw) impact on gene transcription remains unclear. To investigate whether the interactions between Su(Hw) and Combgap are associated with transcriptional regulation, we first identified all differentially expressed (DE) genes in Su(Hw)^LOF ovaries compared with wild-type ovaries. We purified mRNA from the corresponding ovaries and performed RNA sequencing (RNA-Seq) on the cDNA libraries obtained (RNA-Seq was performed in two biological replicates per each genotype). We identified 636 transcripts that were significantly DE in Su(Hw)^LOF mutants compared to WT (with adjusted p-value < 0.01 and fold-change > 2) (Supplementary Fig. 3a-b, Supplementary Table 1). Notably, the median expression of Su(Hw)^LOF up- and down-regulated genes differs significantly, with the expression of Su(Hw)^LOF up-regulated genes being much lower than the median expression of Su(Hw)^LOF down-regulated genes (Fig. 3a).

Previous studies reported that 35% of genes mis-regulated in su(Hw)^-/- mutant ovaries have Su(Hw) peaks located inside or within 2 kb upstream or downstream of these genes(21). Applying the same criteria to our data, we found that 34.4% (218) of DE transcripts are located within 2 kb of direct Su(Hw) ChIP-Seq peaks. We also compared the proportion of DE TSSs intersecting with direct Su(Hw) and Combgap ChIP-Seq peaks, analysing Su(Hw)^LOF up- and down-regulated TSSs separately (Fig. 3b). We observed that 17.1% (56 out of 328) of Su(Hw)^LOF up-regulated TSSs are located within 2 kb of direct Su(Hw) peaks, whereas only 3.9% (12 out of 308) of Su(Hw)^LOF down-regulated TSSs have direct Su(Hw) peaks within a distance of 2 kb. Surprisingly, 78.3% (247 out of 308) of Su(Hw)^LOF down-regulated TSSs have Combgap ChIP-Seq peaks within 2 kb, while only 41.2% (135 out of 328) of both Su(Hw)^LOF up-regulated TSSs and all D. melanogaster TSSs localise within 2 kb of Combgap ChIP-Seq peaks. This observation suggests that Combgap may be involved in the transcription regulation of Su(Hw)^LOF down-regulated TSSs. Interestingly, the median expression levels of Su(Hw)^LOF down-regulated genes with TSSs within 2 kb of direct Su(Hw) and Combgap ChIP-Seq peaks are similar, while those of Su(Hw)^LOF up-regulated genes vary significantly (Fig. 3c). Su(Hw)^LOF up-regulated genes containing direct Su(Hw) peaks within 2 kb show notably lower median expression compared with Su(Hw)^LOF up-regulated genes whose TSSs colocalise with Combgap.

To further clarify the chromatin features of genes mis-regulated in Su(Hw)^LOF ovaries, we analysed MNase-Seq data available for follicular cells from egg chamber stages 1–8(23) and FAIRE-Seq data for wild-type and Su(Hw)^LOF ovaries(17). Su(Hw)^LOF down-regulated TSSs are located in open-chromatin regions, exhibiting high levels of nucleosome phasing (Fig. 3d). On the other hand, Su(Hw)^LOF up-regulated TSSs do not have well-positioned nucleosomes and show depletion of FAIRE signal. The high median level of transcription, nucleosome phasing, and open state of promoter chromatin (Fig. 3c, d) indicate that Su(Hw)^LOF down-regulated genes are actively transcribed genes(24, 25). In contrast, Su(Hw)^LOF up-regulated genes (specifically those containing direct Su(Hw) peaks within 2 kb of their TSSs) possess features of genes that are repressed or inactive in wild-type ovaries, in general.

Interestingly, we observed that Combgap ChIP-Seq peaks located within 2 kb of Su(Hw)^LOF down-regulated TSSs are enriched with BEAF32 and M1BP DNA motifs, in addition to Combgap's own DNA-binding motifs (Supplementary Fig. 3c). For Combgap ChIP-Seq peaks located within 2 kb of Su(Hw)^LOF up-regulated TSSs, we found enrichment of Combgap and GAF DNA motifs. These observations imply that in addition to Combgap and Su(Hw), other DNA binding proteins may be involved in the regulation of Su(Hw)^LOF DE genes.

Su(Hw) ChIP-Seq peaks form long-range chromatin interactions with Combgap ChIP-Seq peaks

In a previous study, it was demonstrated that BEAF32 IBP ChIP-Seq peaks lacking BEAF32’s DNA-binding motif reflect the LRIs between chromatin-bound BEAF32 and its partner proteins(26). To investigate whether there are LRIs between Su(Hw) and Combgap ChIP-Seq peaks, we conducted in situ Hi-C on the ovaries dissected from wild-type and Su(Hw)^LOF flies using the four-cutter restriction enzyme MboI. In situ Hi-C was performed in two biological replicates per each genotype. The Hi-C data obtained totalled 266-276 million raw pair-end reads per genotype with 66-70 million of valid Hi-C pairs per genotype (the statistics for the read alignments, mapped reads and valid Hi-C pairs can be found in Supplementary Fig. 4 and Supplementary Table 2).

We estimated the average LRIs between different pairs of ChIP-Seq peak sets genome-wide using aggregation plots with the coolpup.py program(27). As a positive control, we tested the average spatial interactions between Rpb3 (a subunit of RNAPII) peaks and observed a strong enrichment of Rpb3-Rpb3 LRIs (Supplementary Fig. 5a). We also observed significant enrichment of Combgap-Combgap LRIs (Fig. 4a). This is consistent with the finding that Combgap is capable of recruiting the Ph subunit of the PRC1 complex, which is often found at the anchors of chromatin loops in Drosophila(13, 28). Importantly, both Rpb3-Rpb3 and Combgap-Combgap LRIs remained intact in Su(Hw)^LOF flies, indicating that these LRIs are independent of Su(Hw). Subsequently, we estimated the average spatial interactions within the set of direct Su(Hw) ChIP-Seq peaks. We observed clear enrichment of LRIs between direct Su(Hw) ChIP-Seq peaks but were unable to detect any differences in these LRIs in the Su(Hw)^LOF background (Supplementary Fig. 5a).

In our previous study, we demonstrated that direct Su(Hw) ChIP-Seq peaks in Drosophila ovaries can be classified into two clusters based on enrichment with the FAIRE signal and active chromatin features, such as chromatin remodelers and Gcn5 histone acetyltransferase (hereon referred to as Su(Hw) direct FAIRE+ peaks), or lack such enrichment (Su(Hw) direct FAIRE- peaks)(17). We analysed average LRIs for these two clusters separately. We observed LRIs between Su(Hw) direct FAIRE+ peaks that disappear in the absence of Su(Hw) (Fig. 4a). Interestingly, for Su(Hw) direct FAIRE- peaks, LRIs appeared to be independent of Su(Hw) binding to chromatin. When we analysed the average insulation profiles for these groups of peaks, we observed that Su(Hw) direct FAIRE+ peaks present strong insulation scores (Fig. 4b). In Su(Hw)^LOF, the insulation score for Su(Hw) direct FAIRE+ peaks decreased, suggesting that Su(Hw) can define local insulation at the domain borders where it binds. Unlike Su(Hw) direct FAIRE+ peaks, Su(Hw) direct FAIRE- peaks do not show any insulation score enrichment, which implies that they do not colocalise with the domain borders.

To investigate the spatial interactions between Combgap and Su(Hw), we excluded all peaks from the Combgap peak set that intersect with direct Su(Hw) ChIP-Seq peaks. We observed the existence of LRIs between Combgap and Su(Hw) direct FAIRE+ peaks, but not Su(Hw) direct FAIRE- peaks (Fig. 4a). In Su(Hw)^LOF flies, these LRIs disappeared, indicating the importance of Su(Hw) for their formation. These Su(Hw)-dependent LRIs between Su(Hw) direct FAIRE+ peaks and Combgap are also evident when examining Hi-C maps (Fig. 4c, Supplementary Fig. 5b-e).

To validate the findings of our Hi-C experiments, we examined the available Hi-C datasets(29) for the OSC cell line, which originates from somatic cells of the Drosophila ovary. Our analysis revealed that the long-range interactions observed in our Hi-C for wild-type ovaries are also present in the Hi-C data from OSC cells (Supplementary Fig. 6).

Changes in chromatin architecture in Su(Hw)^LOF background correlate with transcription mis-regulation

To investigate whether the changes in chromatin architecture correlate with transcriptional mis-regulation in Su(Hw)^LOF ovaries, we analysed the average spatial interaction between Su(Hw) direct FAIRE+ peaks and Combgap localised within 2 kb of Su(Hw)^LOF DE genes (Supplementary Fig. 7). In wild-type ovaries, we observed only weak interactions between these regions at distances ranging from 50-200 kb, which may be due to the shorter range of transcriptional regulatory interactions (such as enhancer-promoter interactions) on average. Indeed, it was shown that the majority of enhancers characterised are within 10 kb of their target gene and only a few can act at distances > 50 kb(30). Unfortunately, the analysis of interactions at distances of less than 50 kb is not possible with the Hi-C matrices obtained due to the high basal level of contacts at such distances.

Analysing Hi-C maps for wild-type and Su(Hw)^LOF ovaries, we observed that the disruption of Su(Hw)-Combgap LRIs correlated with transcriptional mis-regulation of nearby genes in a Su(Hw)^LOF background (Fig. 5, Supplementary Fig. 8). It is known that changes in LRIs can affect enhancer-promoter communication within interacting domains(4, 5). In the OSC cell line, which is derived from somatic ovary cells, all DNA elements that can act as enhancers are annotated(31). We observed that Su(Hw)-Combgap LRIs often localise near OSC enhancers (Fig. 5, Supplementary Fig. 8), suggesting that changes in Su(Hw)-Combgap LRIs may impact the enhancer-promoter interaction networks of DE genes in the somatic cells of the ovary.

To identify changes in the binding of transcriptional regulators to mis-regulated genes in Su(Hw)^LOF, we analysed the ChIP-Seq distribution of Combgap, chromatin remodelers(17), the Rpb3 subunit of RNA polymerase II, the NELF E subunit of NELF complex, and the Paf1 positive elongation factor binding along mis-regulated genes (Fig. 6a, Supplementary Fig. 9). We found that there are no significant changes in the distribution of these proteins along Su(Hw)^LOF up-regulated genes (Supplementary Fig. 9). However, for Su(Hw)^LOF down-regulated genes that have Combgap located within 2 kb of their TSSs, we observed an increase in the binding of Combgap, Rpb3, and ISWI ATPase of the ISWI chromatin remodeler family, along with a decrease in binding of Brm ATPase of the SWI/SNF family and CHD1 ATPase of the CHD family. For Su(Hw)^LOF down-regulated genes that have direct Su(Hw) peaks within 2 kb of their TSSs, we observed a substantial decrease in binding of Brm ATPase and Mi2 ATPase of the CHD chromatin remodeler family, along with a reduction in Paf1 binding. It is known that ISWI and Mi-2, in addition to their participation in transcription activation(32, 33), can contribute to the establishment of inactive chromatin state(34, 35, 36, 37), while Brm and CHD1 are mainly involved in chromatin opening resulting in transcription activation(38, 39). The increased ISWI binding at Su(Hw)^LOF down-regulated genes on Su(Hw)^LOF background suggests that ISWI can act as a co-repressor of this group of genes, while decreased Brm, Mi2 and CHD1 levels correlate with these remodelers’ role as transcriptional co-activators.

Our study provides the first direct evidence of the participation of the Su(Hw) IBP in maintaining chromatin architecture in Drosophila ovaries. Consistent with previous research on BEAF32 IBP indirect chromatin binding(7), we show that indirect binding of Su(Hw) to chromatin can reflect the long-range chromatin interactions in which Su(Hw) participates. Previously, we identified a cluster of indirect Su(Hw) ChIP-Seq peaks in Drosophila ovaries that contain a GTGT-motif instead of the Su(Hw) DNA-binding motif(17). Here, we found that these indirect Su(Hw) peaks are occupied by Combgap, a protein known to recruit PcG group proteins to chromatin(13, 14). We showed that Combgap also binds to a subset of direct Su(Hw) peaks in a Su(Hw)-dependent manner. Our in situ Hi-C experiments revealed that Su(Hw)-dependent long-range chromatin interactions occur between Combgap and a part of direct Su(Hw) ChIP-Seq peaks (Fig. 6b).

Intriguingly, Combgap-bound Su(Hw) peaks exhibit features of active rather than inactive chromatin regions, while, on average, the majority of Su(Hw) peaks colocalise within repressed chromatin types(21, 40) (Fig. 2a-c, Supplementary Fig. 2a). We believe that the long-range interactions between Combgap and Su(Hw) are the reason why active chromatin factors such as RNA polymerase II, pausing factors, and chromatin remodelers(17, 40, 41) can be found in direct Su(Hw) ChIP-Seq peaks. However, more research is needed to fully address this suggestion. In particular, Combgap depletion from the ovary will clarify its role in the recruitment of active chromatin factors to Su(Hw) sites.

Our in situ Hi-C experiments revealed that direct Su(Hw) sites exhibit heterogeneity in their ability to support long-range interactions (Fig. 4a, b; Supplementary Fig. 5a). Previous DNA-motif analysis of Su(Hw) binding sites together with the analysis of available ChIP-Seq data has shown that Su(Hw) binding sites can be categorised into distinct groups, based on their intersection with active chromatin features(40). Our ChIP-Seq data from wild-type and Su(Hw)^LOF Drosophila ovaries confirmed this observation. We showed that direct Su(Hw) ChIP-Seq peaks can be classified into two clusters: those enriched with FAIRE signal and active chromatin proteins (Su(Hw) direct FAIRE+ peaks), and those lacking these features (Su(Hw) direct FAIRE- peaks)(17). In situ Hi-C experiments demonstrated that both Su(Hw)-dependent LRIs and domain border insulation are only characteristic of Su(Hw) direct FAIRE+ peaks (Fig. 4a, b). Su(Hw) direct FAIRE- peaks do not colocalise with domain boundaries and retain the ability to self-interact even in the absence of Su(Hw). We hypothesise that Su(Hw) direct FAIRE- peaks retain the ability to self-interact in the absence of Su(Hw) due to other proteins maintaining the interactions between these chromatin regions. Supporting this, previous studies have shown that, while Su(Hw) is enriched at the borders and inside lamina-associated domains (LADs), knockdown of Su(Hw) only causes minimal changes in the interactions of these chromatin regions with the nuclear lamina(42), meaning that LADs can retain their self-interaction in a Su(Hw)^LOF background.

Our data suggest that the involvement of Su(Hw) in transcriptional regulation(21, 22), particularly its role in activating the transcription of certain genes, may be related to its ability to establish long-range interactions. Su(Hw)^LOF up-regulated genes, containing direct Su(Hw) peaks within 2 kb of their TSSs, have closed chromatin at their promoters, do not bind transcriptional regulators, and have low median transcription level in wild-type Drosophila ovaries (Fig. 3d, Supplementary Fig. 9). We believe that these genes reflect the direct property of Su(Hw) to repress transcription(21, 22). On the other hand, Su(Hw)^LOF down-regulated genes, containing Combgap or direct Su(Hw) peaks within 2 kp of their promoters, have open chromatin on TSSs and Su(Hw)-dependent binding of some transcriptional regulators (Fig. 3d, Fig. 6a). We suggest that the long-range interactions between Combgap and direct Su(Hw) peaks are involved in the formation of the local environment regulating the transcription of Su(Hw)^LOF down-regulated genes (Fig. 5, Supplementary Fig. 8, Fig. 6a).

Collection of the Drosophila ovaries

The flies of Oregon-R-modENCODE (the wild-type control, corresponds to Bloomington stock 25211) and su(Hw)^v/E8 (a kind gift of A. Golovnin laboratory) stocks were used. All flies were raised at 25°C on standard agar medium. Ovaries were dissected from recently eclosed 15 hour-old flies (contain only egg chamber stages 1-8) the same way as in (17). During dissection, we thoroughly controlled the correctness of the stages of egg chambers in the ovaries, collected for the analysis.

Nuclear protein extract and immunoprecipitation

Drosophila Schneider cell line 2 (S2) cells were used for nuclear protein extract preparation. S2 cells were maintained at 25 °C in Schneider’s insect medium (Sigma) containing 10% FBS (HyClone). The S2 cells nuclear protein extract was obtained as described previously(43). Immunoprecipitations were performed as previously described(43).

ChIP, ChIP-Seq, ChIP-Seq analysis

For the ChIP experiments ovaries were collected in PBS buffer (50 pairs per ChIP), cross-linked with 1% of formaldehyde for 5 min and then incubated for 5 min with 125 mM Glycine. After that, ovaries were washed with PBS buffer for three times. The remaining ChIP protocol was performed as described previously(41, 44). The primers used for qPCR analysis of ChIP experiments are listed in Supplementary Table 3.

ChIP-Seq libraries were obtained using the NEBNext Ultra^TM II DNA library preparation kit (New England Biolabs). Only the library fragments of 250-500 bp were subjected to next generation sequencing (NGS). NGS was performed by Evrogen (evrogen.ru) and Genetico (genetico.ru) on the Illumina NovaSeq6000 sequencer. For each of the ChIP‐Seq libraries approximately 4-12 millions of unique single-end reads were obtained. The single‐end reads in FastQ format were mapped to the Drosophila genome assembly dm6 using Bowtie2(45) (Galaxy Version 2.3.4.3) and filtered (with minimum MAPQ quality score = 5).

BigWig files were generated using bamCoverage(46) (Galaxy Version 3.5.1.0.0) with scores representing number of reads per kilobase per million (RPKM). The final BigWig files (representing the protein-binding profiles) were obtained using bigwigcompare(46) (Galaxy Version 3.5.1.0.0) as a ratio of ChIP signal to Input. The peaks of proteins’ binding were defined by MACS2 callpeak(47) (Galaxy Version 2.1.1.20160309.6) with the following parameters: -gsize '120,000,000' -keep-dup '1'-qvalue '0.01' -mfold '5' '50' --bw ‘350’ 2 > &1 > macs2_stderr. Corresponding input DNA was used as a control for peak calling.

Visualization of ChIP-Seq data in the heatmaps and pile-up profiles was performed on the Galaxy-P platform(48). All ChIP-Seq data obtained in the current study were deposited into the Gene Expression Omnibus, GSE231576 (token for access is wdenqwectdsvncd).

The definition of DNA motifs for different ChIP-Seq peaks was performed with MEME suite 5.4.1(18, 49). The motifs were searched in the regions flanking the ChIP-Seq summit by 250 bp.

The intersection of the MACS2-called ChIP-Seq peaks was performed using the regions flanking the ChIP-Seq summit by 150 bp, with the bedtools Intersect intervals(50) (Galaxy Version 2.30.0+galaxy1).

For the ovaries of egg chamber stages 1-8 we used the following the ChIP-Seq and MNase-Seq data previously described: ChIP-Seq for Su(Hw), Brm, ISWI, Mi2, CHD1(17) (GSE168894), H3K27me3 ChIP-Seq from 512c germline cells nuclei, FACS-sorted from the ovaries of 3-6 days old Drosophila females(20) and MNase-Seq data for follicular cells from egg chambers stages 1–8(23) (NCBI-SRA BioProject SRP057811). We have not used any figures or text from the manuscripts previously published, only analysed the data deposited at the free public access databases.

As Polycomb response elements (PREs) for S2 cell line we used the same set of regions as in (19).

RNA-Seq, RNA-Seq analysis

For extraction of RNA the ovaries of 15 hour old wild-type (oregon) and su(Hw)^v/E8 females were collected in PBS buffer (50 pairs per sample), in two biological repeats. Total RNA was extracted with the TRI reagent (Ambion). PolyA comprising RNA fraction was isolated and prepared for sequencing with the NEBNext Ultra^TM II Directional RNA Library Prep Kit. NGS was performed by Evrogen (evrogen.ru) with the Illumina NovaSeq6000 sequencer. For each RNA‐Seq library approximately 15-32 millions of unique single‐end reads were obtained. Obtained reads were mapped to the Drosophila genome assembly dm6 using HISAT2(51) (Galaxy Version 2.2.1+galaxy0). BigWig files were generated using bamCoverage(46) (Galaxy Version 3.3.2.0.0) with scores representing number of reads per kilobase per million (RPKM). Differential analysis was employed by limma(52) (Galaxy Version 3.48.0+galaxy1). Obtained RNA-Seq data were deposited into the Gene Expression Omnibus, GSE231576 (token for access is wdenqwectdsvncd).

In situ Hi-C, Hi-C analysis

For the in situ Hi-C experiments, ovaries were collected in PBS buffer (200 pairs per Hi-C), cross-linked with 1% of formaldehyde for 5 min and then incubated for 5 min with 125 mM Glycine. Then ovaries were washed with 1xPBS buffer for three times. The remaining in situ Hi-C protocol was performed as described previously(53). In situ Hi-C was performed in two biological replicates per each genotype. Hi-C libraries were obtained using the NEBNext Ultra^TM II DNA library preparation kit (New England Biolabs). Only the library fragments of 200-600 bp were subjected to NGS sequencing. NGS was performed by Genetico (genetico.ru) on the Illumina NovaSeq6000 sequencer. For each genotype in total 266-276 million of raw paired-end reads were obtained (the statistics for the read alignments, mapped reads and valid Hi-C pairs generated in hicBuildMatrix 3.4.2 for raw contact matrices can be found in Supplementary Fig. 4 and Supplementary Table 2).

Hi-C data was processed using the HiCExplorer suite version 3.4.2(54). The paired‐end reads in FastQ format were separately mapped to the Drosophila genome assembly dm6 using Bowtie2(45) (Galaxy Version 2.4.2+galaxy0), with the following options: --sensitive-local --reorder. The raw contact Hi-C matrices in .h5 format with the bins size of 4 kb were generated in hicBuildMatrix(54) (Galaxy Version 3.4.2.0). The percentage of mapped reads, self-ligation, same-fragment, self-circle and duplicates for the generated in hicBuildMatrix raw contact matrices is presented on Supplementary Fig. 4. The biological replicates of the raw contact Hi-C matrices were pooled for each genotype. The resulting raw contact matrices were normalized using hicCorrectMatrix (Imakaev’s iterative correction), after running hicCorrectMatrix in diagnostic mode to extract optimal minimal and maximal values for the filterThreshold option (we removed the lowest 1% and the highest 0.05% of interchromosomal contacts to avoid vastly underrepresented or overrepresents regions). Insulator scores were obtained on normalized matrices with 4 kb bin sized using hicFindTADs with q-value 0.01 cut-off. Following the recommendations of the authors of the program, we set --minDepth 12000 –maxDepth 40000 and --step 4000 for the 4 kb binned matrices. To generate genomic regions views we used pyGenomeTracks 3.5.1.

Averaged spatial interactions between different sets of ChIP-Seq peaks were calculated using coolpup.py program v0.9.5(27). The minimal and maximal distances of interactions were set at 200 kb and 1000 kb, correspondingly (except for the average spatial interactions on Supplementary Fig. 7, for this we used 50-200 kb distance range). The pad size was set at ±40 kb or ±24 kb around the central pixel and 10 randomly shifted control regions were used to normalise the signal for local background.

The following previously described Hi-C datasets for the OSC Drosophila cells, treated with EGFP dsRNA(29), were used: GSM4790413, GSM4790414, GSM4790421, GSM4790422, GSM4790427, GSM4790428 from GSE158082. The following FASTQ datasets were combined together and processed the same way as described above, except for the 6kb bin size used to generate the matrix and the following setting of hicFindTADs: --minDepth 18000 --maxDepth 60000 and --step 6000.

Antibodies

For the co-immunoprecipitations and ChIP-Seq experiments we used previously described antibodies against Combgap(19), Su(Hw)(17), Paf1, NELF E and Rpb3(55).

Data availability

All sequencing data have been deposited at GEO repository GSE231576.

Funding

This work was supported by the grant 075-15-2019-1661 from the Ministry of Science and Higher Education of the Russian Federation and by the grant № 20-14-00269 from Russian Science Foundation.

Competing interests

Authors declare no conflict of interest.

Authors’ contributions

M.M. designed and supervised the project, N.V. performed the immunoprecepitation experiments and ChIP, M.M. performed ChIP-Seq, in situ Hi-C and RNA-Seq experiments, M.M., A.K, N.V., M.E., and D.C. performed bioinformatics analysis of ChIP-Seq data, M.M. performed bioinformatics analysis of in situ Hi-C and RNA-Seq, M.M. wrote the manuscript text and prepared the figures. All authors read and approved the final manuscript.

Acknowledgements

We thank the Center for Precision Genome Editing and Genetic Technologies for Biomedicine, IGB RAS for the experimental equipment. We are very grateful to Nikolay Zolotarev for his valuable advices.

1. Zolotarev N, Fedotova A, Kyrchanova O, Bonchuk A, Penin AA, Lando AS, et al. Architectural proteins Pita, Zw5,and ZIPIC contain homodimerization domain and support specific long-range interactions in Drosophila. Nucleic Acids Res. 2016;44(15):7228-41.

2. Chathoth KT, Mikheeva LA, Crevel G, Wolfe JC, Hunter I, Beckett-Doyle S, et al. The role of insulators and transcription in 3D chromatin organization of flies. Genome Res. 2022;32(4):682-98.

3. Bag I, Chen S, Rosin LF, Chen Y, Liu CY, Yu GY, et al. M1BP cooperates with CP190 to activate transcription at TAD borders and promote chromatin insulator activity. Nat Commun. 2021;12(1):4170.

4. Kaushal A, Dorier J, Wang B, Mohana G, Taschner M, Cousin P, et al. Essential role of Cp190 in physical and regulatory boundary formation. Sci Adv. 2022;8(19):eabl8834.

5. Kaushal A, Mohana G, Dorier J, Ozdemir I, Omer A, Cousin P, et al. CTCF loss has limited effects on global genome architecture in Drosophila despite critical regulatory functions. Nat Commun. 2021;12(1):1011.

6. Ramirez F, Bhardwaj V, Arrigoni L, Lam KC, Gruning BA, Villaveces J, et al. High-resolution TADs reveal DNA sequences underlying genome organization in flies. Nat Commun. 2018;9(1):189.

7. Vogelmann J, Le Gall A, Dejardin S, Allemand F, Gamot A, Labesse G, et al. Chromatin insulator factors involved in long-range DNA interactions and their role in the folding of the Drosophila genome. PLoS Genet. 2014;10(8):e1004544.

8. Heurteau A, Perrois C, Depierre D, Fosseprez O, Humbert J, Schaak S, et al. Insulator-based loops mediate the spreading of H3K27me3 over distant micro-domains repressing euchromatin genes. Genome Biol. 2020;21(1):193.

9. Li HB, Ohno K, Gui H, Pirrotta V. Insulators target active genes to transcription factories and polycomb-repressed genes to polycomb bodies. PLoS Genet. 2013;9(4):e1003436.

10. Kyrchanova O, Kurbidaeva A, Sabirov M, Postika N, Wolle D, Aoki T, et al. The bithorax complex iab-7 Polycomb response element has a novel role in the functioning of the Fab-7 chromatin boundary. PLoS Genet. 2018;14(8):e1007442.

11. Gruzdeva N, Kyrchanova O, Parshikov A, Kullyev A, Georgiev P. The Mcp element from the bithorax complex contains an insulator that is capable of pairwise interactions and can facilitate enhancer-promoter communication. Mol Cell Biol. 2005;25(9):3682-9.

12. Erokhin M, Gorbenko F, Lomaev D, Mazina MY, Mikhailova A, Garaev AK, et al. Boundaries potentiate polycomb response element-mediated silencing. BMC Biol. 2021;19(1):113.

13. Ray P, De S, Mitra A, Bezstarosti K, Demmers JA, Pfeifer K, et al. Combgap contributes to recruitment of Polycomb group proteins in Drosophila. Proc Natl Acad Sci U S A. 2016;113(14):3826-31.

14. Bredesen BA, Rehmsmeier M. DNA sequence models of genome-wide Drosophila melanogaster Polycomb binding sites improve generalization to independent Polycomb Response Elements. Nucleic Acids Res. 2019;47(15):7781-97.

15. Klug WS, Bodenstein D, King RC. Oogenesis in the suppressor of hairy-wing mutant of Drosophila melanogaster. I. Phenotypic characterization and transplantation experiments. J Exp Zool. 1968;167(2):151-6.

16. Baxley RM, Soshnev AA, Koryakov DE, Zhimulev IF, Geyer PK. The role of the Suppressor of Hairy-wing insulator protein in Drosophila oogenesis. Dev Biol. 2011;356(2):398-410.

17. Vorobyeva NE, Erokhin M, Chetverina D, Krasnov AN, Mazina MY. Su(Hw) primes 66D and 7F Drosophila chorion genes loci for amplification through chromatin decondensation. Sci Rep. 2021;11(1):16963.

18. Machanick P, Bailey TL. MEME-ChIP: motif analysis of large DNA datasets. Bioinformatics. 2011;27(12):1696-7.

19. Chetverina D, Vorobyeva NE, Mazina MY, Fab LV, Lomaev D, Golovnina A, et al. Comparative interactome analysis of the PRE DNA-binding factors: purification of the Combgap-, Zeste-, Psq-, and Adf1-associated proteins. Cell Mol Life Sci. 2022;79(7):353.

20. DeLuca SZ, Ghildiyal M, Pang LY, Spradling AC. Differentiating Drosophila female germ cells initiate Polycomb silencing by regulating PRC2-interacting proteins. Elife. 2020;9.

21. Soshnev AA, Baxley RM, Manak JR, Tan K, Geyer PK. The insulator protein Suppressor of Hairy-wing is an essential transcriptional repressor in the Drosophila ovary. Development. 2013;140(17):3613-23.

22. Melnikova L, Elizar'ev P, Erokhin M, Molodina V, Chetverina D, Kostyuchenko M, et al. The same domain of Su(Hw) is required for enhancer blocking and direct promoter repression. Sci Rep. 2019;9(1):5314.

23. Liu J, Zimmer K, Rusch DB, Paranjape N, Podicheti R, Tang H, et al. DNA sequence templates adjacent nucleosome and ORC sites at gene amplification origins in Drosophila. Nucleic Acids Res. 2015;43(18):8746-61.

24. Gaertner B, Johnston J, Chen K, Wallaschek N, Paulson A, Garruss AS, et al. Poised RNA polymerase II changes over developmental time and prepares genes for future expression. Cell Rep. 2012;2(6):1670-83.

25. Weber CM, Henikoff JG, Henikoff S. H2A.Z nucleosomes enriched over active genes are homotypic. Nat Struct Mol Biol. 2010;17(12):1500-7.

26. Liang J, Lacroix L, Gamot A, Cuddapah S, Queille S, Lhoumaud P, et al. Chromatin immunoprecipitation indirect peaks highlight long-range interactions of insulator proteins and Pol II pausing. Mol Cell. 2014;53(4):672-81.

27. Flyamer IM, Illingworth RS, Bickmore WA. Coolpup.py: versatile pile-up analysis of Hi-C data. Bioinformatics. 2020;36(10):2980-5.

28. Eagen KP, Aiden EL, Kornberg RD. Polycomb-mediated chromatin loops revealed by a subkilobase-resolution chromatin interaction map. Proc Natl Acad Sci U S A. 2017;114(33):8764-9.

29. Iwasaki YW, Sriswasdi S, Kinugasa Y, Adachi J, Horikoshi Y, Shibuya A, et al. Piwi-piRNA complexes induce stepwise changes in nuclear architecture at target loci. EMBO J. 2021;40(18):e108345.

30. Ghavi-Helm Y, Klein FA, Pakozdi T, Ciglar L, Noordermeer D, Huber W, et al. Enhancer loops appear stable during development and are associated with paused polymerase. Nature. 2014;512(7512):96-100.

31. Yanez-Cuna JO, Arnold CD, Stampfel G, Boryn LM, Gerlach D, Rath M, et al. Dissection of thousands of cell type-specific enhancers identifies dinucleotide repeat motifs as general enhancer features. Genome Res. 2014;24(7):1147-56.

32. Deuring R, Fanti L, Armstrong JA, Sarte M, Papoulas O, Prestel M, et al. The ISWI chromatin-remodeling protein is required for gene expression and the maintenance of higher order chromatin structure in vivo. Mol Cell. 2000;5(2):355-65.

33. Mathieu EL, Finkernagel F, Murawska M, Scharfe M, Jarek M, Brehm A. Recruitment of the ATP-dependent chromatin remodeler dMi-2 to the transcribed region of active heat shock genes. Nucleic Acids Res. 2012;40(11):4879-91.

34. Liu YI, Chang MV, Li HE, Barolo S, Chang JL, Blauwkamp TA, et al. The chromatin remodelers ISWI and ACF1 directly repress Wingless transcriptional targets. Dev Biol. 2008;323(1):41-52.

35. Scacchetti A, Brueckner L, Jain D, Schauer T, Zhang X, Schnorrer F, et al. CHRAC/ACF contribute to the repressive ground state of chromatin. Life Sci Alliance. 2018;1(1):e201800024.

36. Mugat B, Nicot S, Varela-Chavez C, Jourdan C, Sato K, Basyuk E, et al. The Mi-2 nucleosome remodeler and the Rpd3 histone deacetylase are involved in piRNA-guided heterochromatin formation. Nat Commun. 2020;11(1):2818.

37. Kunert N, Wagner E, Murawska M, Klinker H, Kremmer E, Brehm A. dMec: a novel Mi-2 chromatin remodelling complex involved in transcriptional repression. EMBO J. 2009;28(5):533-44.

38. Armstrong JA, Papoulas O, Daubresse G, Sperling AS, Lis JT, Scott MP, et al. The Drosophila BRM complex facilitates global transcription by RNA polymerase II. EMBO J. 2002;21(19):5245-54.

39. Bugga L, McDaniel IE, Engie L, Armstrong JA. The Drosophila melanogaster CHD1 chromatin remodeling factor modulates global chromosome structure and counteracts HP1a and H3K9me2. PLoS One. 2013;8(3):e59496.

40. Baxley RM, Bullard JD, Klein MW, Fell AG, Morales-Rosado JA, Duan T, et al. Deciphering the DNA code for the function of the Drosophila polydactyl zinc finger protein Suppressor of Hairy-wing. Nucleic Acids Res. 2017;45(8):4463-78.

41. Vorobyeva NE, Mazina MU, Golovnin AK, Kopytova DV, Gurskiy DY, Nabirochkina EN, et al. Insulator protein Su(Hw) recruits SAGA and Brahma complexes and constitutes part of Origin Recognition Complex-binding sites in the Drosophila genome. Nucleic Acids Res. 2013;41(11):5717-30.

42. van Bemmel JG, Pagie L, Braunschweig U, Brugman W, Meuleman W, Kerkhoven RM, et al. The insulator protein SU(HW) fine-tunes nuclear lamina interactions of the Drosophila genome. PLoS One. 2010;5(11):e15013.

43. Kovalenko EV, Mazina MY, Krasnov AN, Vorobyeva NE. The Drosophila nuclear receptors EcR and ERR jointly regulate the expression of genes involved in carbohydrate metabolism. Insect Biochem Mol Biol. 2019;112:103184.

44. Mazina M, Vorob'eva NE, Krasnov AN. [Ability of Su(Hw) to create a platform for ORC binding does not depend on the type of surrounding chromatin]. Tsitologiia. 2013;55(4):218-24.

45. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357-9.

46. Ramirez F, Ryan DP, Gruning B, Bhardwaj V, Kilpert F, Richter AS, et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 2016;44(W1):W160-5.

47. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008;9(9):R137.

48. Afgan E, Baker D, Batut B, van den Beek M, Bouvier D, Cech M, et al. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Res. 2018;46(W1):W537-W44.

49. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 2009;37(Web Server issue):W202-8.

50. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26(6):841-2.

51. Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015;12(4):357-60.

52. Liu R, Holik AZ, Su S, Jansz N, Chen K, Leong HS, et al. Why weight? Modelling sample and observational level variability improves power in RNA-seq analyses. Nucleic Acids Res. 2015;43(15):e97.

53. Rao SS, Huntley MH, Durand NC, Stamenova EK, Bochkov ID, Robinson JT, et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 2014;159(7):1665-80.

54. Wolff J, Rabbani L, Gilsbach R, Richard G, Manke T, Backofen R, et al. Galaxy HiCExplorer 3: a web server for reproducible Hi-C, capture Hi-C and single-cell Hi-C data analysis, quality control and visualization. Nucleic Acids Res. 2020;48(W1):W177-W84.

55. Mazina MY, Kovalenko EV, Derevyanko PK, Nikolenko JV, Krasnov AN, Vorobyeva NE. One signal stimulates different transcriptional activation mechanisms. Biochim Biophys Acta Gene Regul Mech. 2018;1861(2):178-89.

The authors declare no competing interests.

Vorobyevaetal2024Supplementaryfile.docx

Download PDF

Version 2

posted

You are reading this latest preprint version

Su(Hw) interacts with Combgap to establish long-range chromatin contacts

Status:

Version 2

Abstract

Figures

Background

Results

Discussion

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 2