Determining the functional relationship between epigenetic and physical chromatin domains in Drosophila

doi:10.21203/rs.3.rs-4901395/v1

Download PDF

Short Report

Determining the functional relationship between epigenetic and physical chromatin domains in Drosophila

https://doi.org/10.21203/rs.3.rs-4901395/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Eukaryotic genomes are hierarchically organized into topologically associating domains (TADs) that regulate genome function. In Drosophila, TADs strongly correlate with epigenetic domains marked by either active or repressive chromatin marks. This tight correlation may suggest causality between the deposition of the epigenomic marks and formation of TADs. However, it is still unknown whether the epigenome is a major driving force for TAD formation and structure and what is their functional importance. Here we directly address these questions by perturbing the position of Polycomb response elements (PREs), which act as nucleation sites for Polycomb group proteins (PcG) to deposit H3K27me3 and form epigenetic domains in the dac TAD. First, we remove the two PREs in the dac domain wiping out the H3K27me3 mark from the TAD and study the consequences on the physical property of the domain. Second, by shifting the position of one of the PREs, we create an ectopic PcG epigenetic domain and analyze how this affects TAD formation. Together, these two experimental perturbations indicate that, although TADs correlate very well with chromatin states in Drosophila, the epigenomic signature of Polycomb TADs does not play a causal role in the establishment of TADs. On the other hand, physical domains have an important impact on the formation of epigenetic domains, as they can restrict spreading of the repressive histone marks and of PRE loops.

Chromatin

genome organization

epigenome

TADs

Drosophila

PcG proteins

Although the organization of the genome into topologically associating domains (TADs) is well established, the processes that form TADs are still unclear. CTCF/cohesin loop extrusion is emerging as a key mechanism for the formation of TADs in mammals [1]. In contrast, genome organization in Drosophila appears to depend primarily on the partitioning of chromatin state domains, whereas insulator proteins are not required for the establishment of TADs in early Drosophila embryos [2]. Drosophila TAD borders do not correspond to highly localized loop anchor elements, but the interaction between active domains flanking inactive domains may lead to a loose topology based on chromatin state domains rather than on precise loop anchor elements [3, 4]. Indeed, in Drosophila, a strong correlation exists between 3D chromatin structure and the epigenome, and partitioning of chromatin states between domains of histone acetylation and methylation, in particular H3K27me3, has been proposed to contribute to TAD formation [3, 5, 6]. In the same line of evidence, high-resolution Hi-C experiments suggested that TAD organization in Drosophila reflects the switch between active and inactive chromatin states [7]. Finally, super resolution microscopy coupled to DNA labeling (3D-STORM, ORCA) confirmed the strong correlation of TADs with epigenetic domains at a single cell level [8–10].

Altogether, these lines of evidence strongly suggest that histone modifications are a major driving factor for TAD formation in Drosophila, but this hypothesis has never been directly tested due to an important experimental bottleneck: to directly address the importance of a chromatin state for TAD formation and structure it is necessary to completely erase this signature without globally affecting gene expression. Here we overcome this challenge by employing two complementary perturbations using CRISPR/Cas9 genome engineering, where we either erase the epigenetic signature defining a specific Polycomb-associated TADs or create a new epigenetic domain and analyze its consequences on TAD formation.

The key developmental regulator gene dachshund (dac) is located within a TAD spanning around 180 kb. The chromatin state within this TAD is defined by the presence of PcG proteins and the hallmark of PcG-dependent gene silencing, H3K27me3 (Fig. 1a). This epigenetic signature corresponds to the previously identified Polycomb (also defined as “blue”) chromatin state [11] and borders of the epigenetic domain correlates with borders of the physical domain.

We used CRISPR/Cas9 genome engineering to delete the two Polycomb response elements (PREs) that act as nucleation sites for PcG protein binding, which is expected to create the epigenetic chromatin domain. The resulting line was defined as “Double PRE mutant”. CUT&RUN experiments in both, embryos and 3rd instar larval leg discs showed that the deletion of both PREs results in the absence of the repressive histone mark, H3K27me3, specifically at the dac gene locus, whereas other H3K27me3 domains are unaffected (Fig. 1b, Fig. S1a). To analyze the effect of the loss of the PcG-associated epigenetic signature on gene expression, we performed RNA-seq analysis in Double PRE mutant imaginal leg discs. Surprisingly, the erasure of the repressive chromatin mark did not significantly affect the expression of genes (including dac) within or flanking the dac TAD, except the CG5888 gene, which is moderately upregulated (Fig. S1b). RNA-FISH experiments revealed that activation of CG5888 only occurs in a smaller subset of cells of the larval leg disc (Fig. S1c). These results show a large uncoupling between the gene expression changes and the PcG epigenetic chromatin signature at the dac TAD, an important feature which allows the uncoupling between structural consequences of PRE deletion and transcription. We next performed Hi-C experiments of wild type (WT) or Double PRE mutant imaginal leg discs (Fig. 1c) in order to analyze whether the loss of the epigenetic signature affects the formation and structure of the underlying TAD. Comparison of Hi-C maps and quantification of insulation scores of the TAD borders revealed that global TAD structure and its boundaries are not significantly changed upon the erasure of the underlying chromatin state, indicating that the presence of H3K27me3 and PcG proteins is not essential to define TAD borders (Fig. 1d,e). However, the loss of H3K27me3 correlates with changes in the intra-TAD organization. As previously shown [12], the PRE loop is disrupted in absence of PRE sequences in the locus (Fig. 1f,g). Remarkably, a global decrease of intra-TAD interactions was observed in the Double PRE mutant compare to the control, which might reflect reduced chromatin compaction within the TAD upon loss of H3K27me3 (Fig. 1h). The reduction in global intra-TAD interactions is not solely due to the loss of PRE interactions, as no difference in global intra-TAD interactions has been observed in a mutant that causes Polycomb loop disruption without erasing of H3K27me3 mark (gypsy 2 mutant, Fig. 1h) [13]. In summary, these results showed that the epigenetic signature of a Polycomb TAD is not essential for the formation of a physical domains and its boundary, but might regulate intra-TAD interactions that impact chromatin compaction rather that TAD boundary delimitation.

As a complementary strategy to dissect the functional link between epigenetic domains and physical domains, we used a previously described fly line where the endogenous PRE1 is deleted (ΔPRE1) [12] and created from this line another one where we inserted the same PRE1 sequence upstream to the right dac TAD border, in flanking black [11] chromatin that has no Polycomb binding despite being transcriptionally inactive (PRE1_UP line). Importantly, the distance of the new PRE insertion to the endogenous PRE2 at the dac TSS is the same as the distance (65kb) between the two endogenous PREs (Fig. 2a). This experimental setup allowed us to examine, whether the two PRE sequences (endogenous PRE2 and ectopic PRE1) can create a repressive chromatin domain, as it is the case for the endogenous dac locus, and whether this affects TAD structure or borders of the endogenous domain. At the same time, this allowed us to test the importance of TADs for the formation of epigenetic domains.

CUT&RUN experiments showed that insertion of the PRE sequence upstream to the dac locus results in the recruitment of the PRC1 component PH, indicating that the PRE is functional (Fig. 2b). Notably, PcG levels at the ectopic PRE1 insertion site are comparable to PcG levels bound to the endogenous PRE1. However, the spreading of H3K27me3 from the ectopically inserted PRE1 is strongly limited, resulting in a small H3K27me3 domain that does not exceed 10 kb (Fig. 2b). Of note, a similar extent of H3K27me3 levels is observed when we insert the PRE2 sequence in the upstream domain (Fig. S2). Since the PRE2 sequence has stronger PRE activity in terms of PcG binding levels and silencing activity in transgenic reporter assays compared to PRE1 [12], these data suggest that the reduced H3K27me3 spreading from the ectopically inserted PRE1 is not due to the lower affinity of PcG proteins to the PRE1 sequence. This indicates that the endogenous PRE at the dac TSS and an ectopically inserted PRE in its upstream TAD do not cooperate to form a large repressive H3K27me3 domain, as it is the case when they are located within the same TAD.

To examine how TAD structure, their borders and PRE chromatin contacts are affected upon shifting the position of the PREs, we performed Hi-C experiments in the line where the PRE positions is shifted (PRE1_UP). Comparison of Hi-C maps revealed that endogenous TAD structures are largely maintained upon deletion of endogenous PRE1 (ΔPRE1) and insertion of PRE1 in the upstream TAD (PRE1_UP) (Fig. 2c), although a slight but significant decrease in the insulation score between the endogenous dac TAD and the upstream TAD can be observed upon insertion of the ectopic PRE (Fig. 2d,e).

To analyze the effect of a TAD boundary placed between two PRE anchors on their ability to form chromatin loops, we quantified chromatin contacts between the endogenous PRE2 (dac TSS) and the PRE1 in its endogenous and ectopic integration site in the upstream TAD in each fly line. As expected, the endogenous PRE loop is disrupted when the PRE 1 is deleted within the dac TAD in both ΔPRE1 and PRE1_UP (Fig. 2f, left). On the other hand, a slight increase in chromatin contacts is observed between the endogenous PRE2 and the PRE1 inserted in the upstream TAD in PRE1_UP, although chromatin interactions between these genomic loci stay much lower than those of the endogenous PRE loop (Fig. 2f, right). This indicates that the two PREs, when positioned in different TADs, have a significantly reduced potential to interact with each other. In agreement, endogenous PRE loops are largely limited to PREs located within the same TAD [12, 14]. This suggests that endogenous TAD structures or borders may limit PRE communication and function. Additionally, the presence of multiple insulator binding sites between the two PREs (Fig. S2) might interfere with the ability of the two PREs to efficiently establish a larger epigenetic chromatin domain, when located in different physical domains. However, we note that insulator binding sites do not precisely colocalize with borders of the ectopic H3K27me3 domain (Fig. S2), making it unlikely that they directly account for the reduced spreading of H3K27me3 from the ectopic PRE.

Interestingly, by analyzing the differential score enrichment of interactions in ΔPRE1 line versus PRE1_UP line (Fig. 2g), we observed increased interactions of the whole dac TAD with a smaller region around the upstream PRE insertion side. These increased contacts might reflect the high affinity for homotypic interactions between H3K27me3 marked chromatin domains, which has been previously reported in flies and mammals [15–17], resulting in the compartmentalization of repressive chromatin domains. This in turn may explain the slightly reduced insulation activity of the boundary region between the two TADs (Fig. 2d).

In conclusion, the two experimental approaches show that although TAD structures are tightly linked to chromatin states in Drosophila, the epigenetic mark H3K27me3 is not essential for TAD formation and delimitation. However, H3K27me3 influences intra-TAD interactions and contribute to TAD compartmentalization. Conversely, physical domains impact the formation of epigenetic domains and the interaction of PREs.

Fly work and generation of mutant flies by CRISPR/Cas9 genome engineering

All flies were raised on standard corn meal yeast extract medium at 25°C. CRISPR/Cas9 mutant fly lines Double and ΔPRE1 are described in [12].

Sequences of gRNAs used to create fly lines PRE1_UP and PRE2_UP are described in Table S1. Sense and antisense oligonucleotides were annealed and phosphorylated by the T4 polynucleotide kinase (NEB#M0201S) before being inserted inside a pCFD3 plasmid (Addgene #49410) previously digested by BbsI (NEB#R0539S). To create the pHD-dsRED donor plasmid (Addgene) containing a removable (floxed) 3XP3-dsRED construct flanked by loxP sites and DNA fragments having homology to the target regions (homology arms) serving as template for homology-directed repair, 1.5 kb genomic DNA fragments were amplified by PCR (Table S1) and inserted into the pHD-dsRED plasmid using the GIBSON assemble (kit NEBuilder NEB#E2621S).

The PRE1 and PRE2 sequences were PCR amplified and introduced into the donor plasmid cut by SpeI and BglII using GIBSON cloning (Table S1). To generate mutant fly lines, gRNA-containing pCFD3 and pHD-dsRED donor plasmids were injected into flies expressing Cas9 in the germline (vas-Cas9(X) RFP-; Bloomington stock #55821). Injections and dsRED screening was performed by BestGene (https://www.thebestgene.com/). To remove the dsRED reporter construct, mutant flies were crossed with a fly line expressing CRE recombinase (Bloomington stock #34516).

To generate the PRE1_UP and PRE2_UP mutant line gRNAs targeting the PRE upstream region and corresponding donor plasmid were injected into ΔPRE1 mutant lines previously generated and expressing Cas9 (vas-Cas9(III)). Genotypes of mutant fly lines were confirmed by PCR genotyping and sequencing analysis of the mutated region.

CUT&RUN

CUT&RUN experiments were performed as described by Kami Ahmad in protocilas.io (https://dx.doi.org/10.17504/protocols.io.umfeu3n) with minor modifications. 50 leg discs were dissected in Schneider medium, centrifuged for 3 min at 700g and washed twice with wash+ buffer before addition of Concanavalin A-coated beads. Embryos were collected and homogenized using a glass douncer in Nuclear extraction buffer (20mM HEPES, 10mM KCl, 0.1% TritonX, 20% Glycerol), then washed once with wash+ buffer before addition of Concanavalin A-coated beads.

MNase digestion (pAG-MNase Enzyme from Cell Signaling) was performed for 30 min on ice. After ProteinaseK digestion, DNA was recovered using SPRIselect beads and eluted in 50ul TE. DNA libraries for sequencing were prepared using the NEBNext® Ultra™ II DNA Library Prep Kit for Illumina. Sequencing (paired-end sequencing 150bp, approx. 2Gb/sample) was performed by Novogene (https://en.novogene.com/).

CUT&RUN Analysis

Fastq files were aligned using Bowtie 2 (v 2.4.2) [19] with the following parameters: --local ---very-sensitive-local --no-unal --no-mixed --no-discordant --phred33 -I 10 -X 700. The samples for WT and ΔPRE1 were aligned to the D. melanogaster reference genome dm6 in https://s3.amazonaws.com/igenomes.illumina.com/ and samples for the PRE1_UP condition were aligned to a modified genome. First, the 1.091 bp region chr2L:16,422,435-16,423,525 including the PRE1 was removed from the dm6 reference genome and replaced with a string of 1091 n’s (undetermined nucleotides). Next, the removed 1,091bp-long sequence was inserted at position chr2L:16,560,363 bp to create the modified genome. SAM files were compressed into BAM files using samtools (v 1.16.1) [20] and reads with low mapping quality (Phred score <30) were discarded. Duplicate reads were removed using sambamba markdup (v 1.0.0) [21] with the following parameters: -r --hash-table-size 500000 --overflow-list-size 500000. For visualization, replicates were merged using samtools merge with default parameters and reads per kilo base per million mapped reads (RPKM)-normalized bigWig binary files were generated using the bamCoverage (v 3.5.5) function from deepTools2 [22] with the following parameters: --normalizeUsing RPKM --ignoreDuplicates -e 0 -bs 10. Two biological replicates were used per condition and the IgG condition was used as a control for the differential enrichment analysis in the 131 Drosophila Polycomb domains [15] performing the DESeq2 method from the “DiffBind” R package (v 3.12.0) (Table S2).

RNA-seq experiments

Third instar larval leg discs were dissected in Schneider medium on ice. Total RNA was extracted using TRIzol reagent. RNA purification was performed using the RNA Clean & Concentrator kit (Zymo Research, #R1015). Finally, poly-A RNA selection, library preparation and Illumina sequencing (20M paired-end reads, 150nt) were performed by Novogene (https://en.novogene.com/). All experiments were performed in triplicate.

RNA-seq processing and differential analysis

RNA-seq data quality was assessed using FastQC (v0.12.1). Stranded RNA-seq data were mapped to the Drosophila melanogaster dm6 genome (FlyBase dmel_r6.34) using Subread-align (Subread v2.0.6) with default parameters. Samtools merge was used to merge bam files from the same sample sequenced in different lanes. (Samtools v1.16.1). Aligned sequencing reads mapped to gene transcripts were counted using featureCounts (Subread v 2.0.6) with -s 2 (reverse stranded) and default parameters. The gene transcript annotation file was obtained from FlyBase (release 6.34). Prior to statistical analysis, genes with fewer than 10 reads (cumulating all samples analyzed) were removed. Differentially expressed genes (DEGs) were identified using the DESeq2 R package (Table S3). Genes with adjusted p-value<0.01 (using the Benjamini-Hochberg FDR method) and |log2FC|>1 were considered differentially expressed. Volcano plots were generated using the “EnhancedVolcano” R package (DOI: 10.18129/B9.bioc.EnhancedVolcano).

Hi-C

Hi-C experiments were performed using the EpiTect Hi-C Kit (Quiagene#59971). All Hi-C experiment were performed in two or three independent experiments using 50 3^rd instar imaginal leg discs. Briefly, discs were homogenized and fixed in activated Buffer T and 2% Formaldehyde using Tissue Masher tubes (Biomasher II (EOG-sterilized) 320103 Funakoshi). Tissue was digested by adding 25ul Collagenase I and II (40 mg/ml) for 1h at 37°C. Samples were centrifuged and supernatant was carefully aspirated, leaving ~250 μl of solution in the tube. Then 250ul QIAseq Beads equilibrated to room temperature were added to bind nuclei to the beads and all subsequent reactions were performed on the beads according to the manufactures protocol. Libraries were sequenced at BGI (https://www.bgi.com/) PE 150 (approx. 400 million reads per replicate).

Hi-C analysis

Raw data from Hi-C sequencing were processed by using the "shHiC2" pipeline. Sequencing statistics are summarized in Table S4. Valid interactions were stored in a database using the “misha” R package (https://github.com/msauria/misha-package). Extracting the valid interactions from the misha database, the "shaman" R package [https://bitbucket.org/tanaylab/shaman] has been used for computing the Hi-C expected models, Hi-C scores with parameters k=250 and k_exp=500 (Fig. 1c and 2c), and differential Hi-C interaction scores with parameters k=250 and k_exp=250 and per each comparison down-sampling the compared datasets to have the same number of valid-pairs in chr2L (Fig. 1f and 2g). Specifically, Hi-C scores quantify the contact enrichment (positive values) or depletion (negative values) of each bin of the map with respect to a statistical model used to evaluate the expected number of counts. To generate this expected model, we (randomized) shuffled the observed Hi-C contacts using a Markov Chain Monte Carlo-like approach per chromosome [23]. Shuffling is done such that the marginal coverage and decay of the number of observed contacts with the genomic distance are preserved, but any features of genome organization (e.g, TADs or loops) are not. These expected maps were generated for each biological replicate separately and contain twice the number of observed cis-contacts. Next, the score for each contact in the observed contact matrix was calculated using k nearest neighbors (kNN) strategy [23]. In brief, the distributions of two-dimensional Euclidean distances between the observed contact and its nearest k_exp neighbors in the pooled observed and pooled expected (per cell type) data are compared, using Kolmogorov–Smirnov D statistics to visualize positive (higher density in observed data) and negative (lower density in observed data) enrichments. These D-scores are then used for visualization (-100 to +100 scale) and are referred to as Hi-C scores in the text. Accordingly, the color-scale of the Hi-C scores comprises both positive and negative values. When computing the Differential Hi-C scores maps the reference dataset (WT in Fig. 1f and ΔPRE1 in Fig. 2g) were used as the expected model.

For each condition, the Hi-C interaction quantifications at the dac PRE loop (Fig. 1g, 2f) were performed by considering the Hi-C scores between two regions of 6 kb: in Fig. 1g. and Fig 2f (left) we considered the locations of the two endogenous PREs (PRE1 in chr2L:16,419,514-16,425,515bp and PRE2 in chr2L:16,482,929-16,488,930bp) in Fig 2f (right) we considered the locations of the endogenous PRE2 and the insertion site of the relocated PRE1 (PRE2 in chr2L:16,482,929-16,488,930bp and PRE1_UP in chr2L:16,557,363-16,563,364bp). Each of the comparisons of the Hi-C interaction quantifications at the dac PRE loop was performed between the WT (control line) and the indicated mutant line. An unpaired two-sided Wilcoxon statistical test (H0: true median shift is equal to 0. The two variables are not normally distributed) was used to estimate the reported p-values. The annotation of the Polycomb-associated TADs in chr2L in [15] was used to compute the number of Hi-C interactions intra-PcG-TAD, which were then divided by the number of Hi-C interactions intra-PcG-TAD in the WT (control line). The distributions of Log2 of these ratios are shown in the violin- and boxplots of Fig. 1h. The boxplots show: central line, median; box limits, 75th and 25th percentiles; whiskers, 1.5×interquartile range. An unpaired two-sided Wilcoxon statistical test (H0: true median shift is equal to 0. The two variables are not normally distributed) was used to estimate the reported p-values.

The insulation scores [24] were computed on the observed Hi-C datasets binned at 2kb resolution with windows of 100, 150, 200, 250, and 300kb resulting in five values per bin and were stored in the misha database using an in-house R script. The mean and standard deviation per each of the 2kb-bins were computed were used for the plots in Fig. 1d and 2d. The quantification of the insulation scores (IS) at the TAD borders was performed applying the pair-wise statistical comparison of the five IS quantifications per 2kb-bins. The p-values in Fig. 1d and 2d resulted from a Welch t-test (H0: true difference in means is equal to 0. The variances of the samples are thought not to be equal) between the WT condition and each of the mutant condition at the corresponding locus. All plots of Hi-C maps, Hi-C interaction scores comparisons, insulation score (IS) profiles with window 100kb, p-values of IS comparisons were obtained with in-house R scripts provided in Github (see Availability of data and materials).

RNA-FISH

RNA FISH was performed as described in Denaud et al, in press 10.1038/s41594-024-01375-7)

Acknowledgments

We thank all members of the Cavalli lab for their feedback on the manuscript and help with dissections.

Author’s contributions

B.S. and G.C conceived and designed the study. B.S. generated mutant fly lines and performed RNA-seq and CUT&RUN experiments. S.D. performed Hi-C and RNA-FISH experiments. G.S. established CUT&RUN protocol and performed computational analysis of CUT&RUN experiments and RNA-seq experiments. M.D.S. and G.P performed computational analysis of Hi-C experiments. B.S. and G.C wrote the manuscript with input from all the authors.

Funding

This work was supported by grants from the European Research Council (Advanced Grant 3DEpi), the European CHROMDESIGN ITN project (Marie Skłodowska-Curie grant agreement No 813327), the European E-RARE NEURO DISEASES grant “IMPACT”, by the Agence Nationale de la Recherche (PLASMADIFF3D, grant N. ANR-18-CE15-0010, LIVCHROM, grant N. ANR-21-CE45-0011), by the Fondation ARC (EpiMM3D), the Fondation pour la Recherche Médicale (EQU202303016280), by the MSD Avenir Foundation (Project GENE-IGH), and by the French National Cancer Institute (INCa, PIT-MM grant N. INCA-PLBIO18-362)!

Competing interest declaration

The authors declare no competing interests.

Availability of data and materials

All genomic data have been deposited on GEO: GSE274157, GSE274158 and GSE274159 Hi-C data for WT and gypsy 2 are available on GEO: GSE247377. All original code was deposited on GitHub (https://github.com/cavallifly/Denaud_et_al_2024/tree/main). and is publicly available as of the date of publication. Any additional information required to reanalyze the data reported in this paper is available from the corresponding author upon request.

Fly lines are available on request.

da Costa-Nunes JA, Noordermeer D. TADs: Dynamic structures to create stable regulatory functions. Curr Opin Struct Biol. 2023;81:102622.
Cavalheiro GR, Girardot C, Viales RR, Pollex T, Cao TBN, Lacour P, Feng S, Rabinowitz A, Furlong EEM. CTCF, BEAF-32, and CP190 are not required for the establishment of TADs in early Drosophila embryos but have locus-specific roles. Sci Adv. 2023;9:eade1085.
Matthews NE, White R. Chromatin Architecture in the Fly: Living without CTCF/Cohesin Loop Extrusion? Alternating Chromatin States Provide a Basis for Domain Architecture in Drosophila. BioEssays. 2019;41:e1900048.
Szabo Q, Jost D, Chang JM, Cattoni DI, Papadopoulos GL, Bonev B, Sexton T, Gurgo J, Jacquier C, Nollmann M, et al. TADs are 3D structural units of higher-order chromosome organization in Drosophila. Sci Adv. 2018;4:eaar8082.
El-Sharnouby S, Fischer B, Magbanua JP, Umans B, Flower R, Choo SW, Russell S, White R. Regions of very low H3K27me3 partition the Drosophila genome into topological domains. PLoS ONE. 2017;12:e0172725.
Ulianov SV, Khrameeva EE, Gavrilov AA, Flyamer IM, Kos P, Mikhaleva EA, Penin AA, Logacheva MD, Imakaev MV, Chertovich A, et al. Active chromatin and transcription play a key role in chromosome partitioning into topologically associating domains. Genome Res. 2016;26:70–84.
Rowley MJ, Nichols MH, Lyu X, Ando-Kuri M, Rivera ISM, Hermetz K, Wang P, Ruan Y, Corces VG. Evolutionarily Conserved Principles Predict 3D Chromatin Organization. Mol Cell. 2017;67:837–e852837.
Boettiger AN, Bintu B, Moffitt JR, Wang S, Beliveau BJ, Fudenberg G, Imakaev M, Mirny LA, Wu CT, Zhuang X. Super-resolution imaging reveals distinct chromatin folding for different epigenetic states. Nature. 2016;529:418–22.
Mateo LJ, Murphy SE, Hafner A, Cinquini IS, Walker CA, Boettiger AN. Visualizing DNA folding and RNA in embryos at single-cell resolution. Nature. 2019;568:49–54.
Szabo Q, Donjon A, Jerkovic I, Papadopoulos GL, Cheutin T, Bonev B, Nora E, Bruneau BG, Bantignies F, Cavalli G. Regulation of single-cell genome organization into TADs and chromatin nanodomains. Nat Genet. 2020;52:1151–7.
Filion GJ, van Bemmel JG, Braunschweig U, Talhout W, Kind J, Ward LD, Brugman W, de Castro IJ, Kerkhoven RM, Bussemaker HJ, van Steensel B. Systematic protein location mapping reveals five principal chromatin types in Drosophila cells. Cell. 2010;143:212–24.
Ogiyama Y, Schuettengruber B, Papadopoulos GL, Chang JM, Cavalli G. Polycomb-Dependent Chromatin Looping Contributes to Gene Silencing during Drosophila Development. Mol Cell. 2018;71:73–e8875.
Denaud S, Bardou M, Papadopoulos GL, Grob S, Di Stefano M, Sabaris G, Nollmann M, Schuettengruber B, Cavalli G. A PRE loop at the dac locus acts as a topological chromatin structure that restricts and specifies enhancer promoter communication. Nature Structural & Molecular Biology 2024, in press.
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;114:8764–9.
Sexton T, Yaffe E, Kenigsberg E, Bantignies F, Leblanc B, Hoichman M, Parrinello H, Tanay A, Cavalli G. Three-dimensional folding and functional organization principles of the Drosophila genome. Cell. 2012;148:458–72.
Bonev B, Mendelson Cohen N, Szabo Q, Fritsch L, Papadopoulos GL, Lubling Y, Xu X, Lv X, Hugnot JP, Tanay A, Cavalli G. Multiscale 3D Genome Rewiring during Mouse Neural Development. Cell. 2017;171:557–e572524.
Kraft K, Yost KE, Murphy SE, Magg A, Long Y, Corces MR, Granja JM, Wittler L, Mundlos S, Cech TR, et al. Polycomb-mediated genome architecture enables long-range spreading of H3K27 methylation. Proc Natl Acad Sci U S A. 2022;119:e2201883119.
Loubiere V, Papadopoulos GL, Szabo Q, Martinez AM, Cavalli G. Widespread activation of developmental gene expression characterized by PRC1-dependent chromatin looping. Sci Adv. 2020;6:eaax4001.
Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9:357–9.
Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R. Genome Project Data Processing S: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25:2078–9.
Tarasov A, Vilella AJ, Cuppen E, Nijman IJ, Prins P. Sambamba: fast processing of NGS alignment formats. Bioinformatics. 2015;31:2032–4.
Ramirez F, Ryan DP, Gruning B, Bhardwaj V, Kilpert F, Richter AS, Heyne S, Dundar F, Manke T. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 2016;44:W160–165.
Olivares-Chauvet P, Mukamel Z, Lifshitz A, Schwartzman O, Elkayam NO, Lubling Y, Deikus G, Sebra RP, Tanay A. Capturing pairwise and multi-way chromosomal conformations using chromosomal walks. Nature. 2016;540:296–300.
Crane E, Bian Q, McCord RP, Lajoie BR, Wheeler BS, Ralston EJ, Uzawa S, Dekker J, Meyer BJ. Condensin-driven remodelling of X chromosome topology during dosage compensation. Nature. 2015;523:240–4.

No competing interests reported.

SupplFiguresGB120824.pdf
Figure S1: (a) MA plot showing the log2 fold change of H3K27me3 in Double PRE mutant versus WT as a function of the log2 mean read concentration in the 131 Drosophila Polycomb domains in embryo (left) and third instar larval leg imaginal disc (right). Significance was calculated using the DESeq2 method with two replicates per line (cut-offs: FDR<0.05 and |log2FC|>0.58). The dac Polycomb domain is the only significant domain in both the embryo and leg imaginal disc and is shown in blue. (b) Volcano plot showing the number of differentially expressed genes in Double PRE mutant vs WT in third instar larval leg discs. Differential expression analysis was performed using DESeq2 and three replicates per condition (cut-off: adjusted p-value<0.01 and |log2FC|>1). The genes located in the dac Polycomb domain are indicated. (c) RNA FISH images of WT and Double PRE mutant third instar imaginal leg discs (dacgene: violet, CG5888 gene: green). White bars indicates 20 micrometers. Figure S2: Snapshots of the H3K27me3 CUT&RUN profiles in WT, PRE1_UP and PRE2_UP flies together with Insulator binding sites (modENCODE, class I and II).
TableS1.xlsx
Table S1: List of Primers used in the study.
TableS2.xlsx
Table S2: Differential enrichment analysis of H3K27me3 in embryo and third instar larval leg imaginal disc of Double PRE mutant versus WT flies in the 131 Drosophila Polycomb domains.
TableS3.xlsx
Table S3: Differential expression analysis results and DESeq2 normalized counts of Double PRE mutant versus WT flies.
TableS4.xlsx
Table S4: Hi-C alignment statistics reporting on the number of sequenced reads and the number of valid-pairs genome-wide, on chr2L, and on the portion of the chr2L containing the dac TAD (chr2L:10,000,000-20,000,000)

Download PDF

Reviewers agreed at journal
05 Nov, 2024
Reviews received at journal
24 Sep, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers invited by journal
29 Aug, 2024
Editor assigned by journal
16 Aug, 2024
Submission checks completed at journal
13 Aug, 2024
First submitted to journal
12 Aug, 2024

You are reading this latest preprint version

Determining the functional relationship between epigenetic and physical chromatin domains in Drosophila

Status:

Version 1

Abstract

Figures

Background

Results and Discussion

Conclusions

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1