Figures and data
NPC-DamID, a new method to probe NPC-genome interactions at the nuclear periphery
a) Schematic representation of in vitro DamID. The cells are semi-permeabilized with a low concentration of digitonin (0.001-0.0025% w/v) so the plasma membrane is preferentially permeabilized, leaving the nuclear membrane intact. Dam-fused nuclear import factor Dam-Impβ is supplied to the cells along with the other transport factors and SAM (s-adenosyl methionine), efficiently targeting Dam-Impβ to the NPCs and subsequently methylating NPC-associated chromatin. In the next steps, cells were lysed to extract genomic DNA to prepare the DamID-seq libraries. b) Coomassie gel of recombinantly expressed and purified Dam-Impβ (molecular weight 132 kDa) through size exclusion chromatography (lane 1-4), lane M is protein marker. c) EtBr-stained agarose gel showing the enzymatic activity of Dam-Impβ. d) Targeting of fluorescently labeled Dam-Impβ and wt Impβ labeled with Alexa 647 to NPCs in HeLa cells after semi-permeabilization and NPC-DamID assay. WGA was added to the assay to block the NPCs. The lack of nuclear rim signal for labeled Dam-Impβ in the presence of WGA shows specific localization of Dam-Impβ at the periphery through NPCs. The scale bar is 10 µm. The arrows indicate the planes used for line plots. The dashed lines in the line plot indicate the limits of the nuclear peripheral. e) Schematic representation of Dam-Impβ methylated DNA imaging in the vicinity of NPCs using fluorescently labeled Dpn8. f) dSTORM images of IMR90 cells treated with Dpn8 after NPC-DamID assay. The cells show co-staining of Dpn8 (staining for NPC-interacting genomic regions) and Nup153 (marking NPCs). The scale bar is 500 nm.
NPC-DamID is a robust and reproducible method to study NPC interaction with chromatin.
a) Normalized NPC-DamID DNA sequencing tracks for Dam alone, Dam-Impβ, Dam-Impβ with WGA, wt Impβ, and Lamin B1 for the genomic region around the PVT1 gene in HeLa cells. b) Normalized NPC-DamID DNA sequencing tracks for Dam alone, Dam-ΔN44Impβ, Dam-Impβ, Dam-ΔN44Impβ with RanGTP, Dam-Impβ with RanGTP, and Dam-Nup153 for the genomic region around the ITGB1 gene in IMR90 cells. c) Sequencing reads distribution centered around Dam-ΔN44Impβ peaks for Dam-ΔN44Impβ, Dam-Impβ, Dam-ΔN44Impβ+ RanGTP, Dam-Impβ+RanGTP, Dam-Nup153 for IMR90 and NPC-bound Dam-Nup98 in HeLa cells.
Genomic DNA interacting with NPCs contains mainly regulatory elements.
a) Distribution of H3K27Ac, H3K4Me3, H3K9Me3 ChIP-seq reads centered around NPC-DamID peaks (Dam-ΔN44Impβ) in IMR90 cells. b) Percentage overlap of NPC-DamID peaks with the genome-wide annotation of IMR90 from the 18-state ChromHMM model. c) Transcriptional output comparison for genes associated with NPCs versus the same number of randomly selected other genes (NPC-bound genes are shown in Table S1, ** p< 0.01).
NPCs are consistently interacting with SEs across multiple cell types.
a) Pie charts show the overlap of NPC-DamID peaks with SEs and non-SE regions for HeLa, IMR90, and C2C12 myotubes. b) Bar plots for cell type consensus score of all SEs, NPC-associated SEs (NPC-SE), and SEs not associated with NPCs (non-NPC-SE) in IMR90 cells. c) Pie chart showing that 70% of NPC-SE have hierarchical organization while 30% are non-hierarchical. d) The occupancy of NPC in four enhancer groups: hub (n=1,050) and non-hub (n=7,874) enhancers within hierarchical NPC-SEs, regular enhancers in IMR90, and randomly selected genome regions as control. P values were calculated using Fisher’s exact test.
NPCs are hubs for chromatin structural proteins.
a) Representative images of proximity ligation assay (PLA) between the nucleoporin TPR and BRD4 upon treatment with siRNA against Luciferase control (top panel) or TPR (bottom panel) in U2OS cells. Scale bar is 20 μm. Images are maximum intensity projections of confocal z-slices. b) Bar plot of mean fluorescence intensity of PLA signal per nucleus for TPR interaction with BRD4, Med1, P300, and CTCF in control and TPR knockdown cells. Data normalized to the average value of the control (siLuc) sample. c) Bar plot of the mean number of PLA puncta per nucleus for TPR interaction with BRD4, Med1, P300, and CTCF in control and TPR knockdown cells. For b and c, n>250 nuclei per condition, error bars represent a 95% Confidence Interval, and a t-test was used to calculate statistical significance (**** p< 0.0001). d) Distribution of reads of CTCF, BRD4, PolII, and P300 ChIP-seq centered around NPC-DamID peaks in IMR90 cells shown as a metagene profile and heatmap.
Phase separation properties of Nups are important for their interaction with other SE hub proteins.
a) PONDR prediction for intrinsically disordered regions (IDR) in the structure of Med1 and Nup153. A score over 0.5 signifies a lack of structure. b) Confocal images of HeLa cells transfected with GFP fused with Med1-IDR or Nup153-IDR before and after 30s of treatment with 1.5% hexanediol (HD). c) Bar plots of mean fluorescence intensity of PLA signal per nucleus in DMSO and HD treated cells. The protein interaction pairs are mentioned on top of each plot. Data normalized to average PLA intensity of DMSO sample. d) Bar plots of the mean number of PLA puncta per nucleus in DMSO and HD treated cells for different protein interacting pairs mentioned on top of the plot. For c and d, n>300 nuclei per condition, error bars represent 95% Confidence internal, t-test was used to calculate statistical significance (ns not significant, * p<0.05, *** p<0.001, **** p< 0.0001).
Synthetic LacO array mediated Nup IDR hubs can recruit other chromatin structural proteins.
a) Scheme of a cell containing an array of LacO repeats (∼50,000) which can be decorated with the binding of LacI fused to a fluorescent protein (mCherry), IDRs, and nuclear localization signal (NLS). b) Confocal representative images of U2OS cells containing LacO arrays transfected with LacI fused to mCherry alone (mCherry-NLS-LacI), mCherry-Med1-IDR-NLS-LacI, or mCherry-Nup153-IDR-NLS-LacI. The scale bar is 1 µm. c) Confocal images of U2OS cells harboring LacO arrays and co-transfected with EGFP-NLS and mCherry-Nup153-IDR-LacI or EGFP-PolII CTD52-NLS and mCherry-Med1 IDR-NLS-LacI or EGFP-Med1-NLS and mCherry-Nup153-IDR-NLS-LacI or EGFP-PolII CTD52-NLS and mCherry-Nup153-IDR-NLS-LacI. d) Bar plot of the percentage of mCherry-LacI positive nuclei showing EGFP colocalization for images in c). The number of total nuclei counted is shown alongside the plot.
a) Coomassie gel for recombinantly expressed and purified Dam protein. b) Coomassie gel for recombinantly expressed and purified Dam-Impβ protein. c) Coomassie gel for recombinantly expressed and purified nuclear transport factors-Ran (Lane 1), Impβ (Lane 2), Impα (Lane 3), NTF2 (Lane 4), NLS-GFP-MBP cargo (Lane 5). d) Fluorescent gel scan of elutions for labeled Impβ (Lanes 1-5) and Dam-Impβ (Lanes 6-10) with Alexa 488 succinimide dyes after size exclusion chromatography e) Activity assay for Dam and Dam-Impβ with varying concentrations of SAM-5 μM (+) and 7.5 μM (++). f) Confocal images of the nuclear transport assay of NLS-MBP-GFP using Dam-Impβ in the presence and absence of RanGDP/GTP. The scale bar is 15 µm. g) Coomassie gel for recombinantly expressed and purified Dpn8. h) Fluorescent gel scan for Dpn8 labeled with Alexa 488. i) Dot blot assay for binding specificity of GFP-Dpn8 with methylated (Me) and unmethylated (UnMe) lambda DNA. j) Gel shift assay for Dpn8 binding to methylated lambda dsDNA (Me DNA) with increasing concentration of Dpn8. k) Confocal images of Hela cells labeled with Flag-Dpn8 (2μM) after NPC-DamID assay. Dam alone shows a uniform nucleoplasmic signal while Dam-LBR and Dam-Impβ show enrichment of signal at the nuclear periphery for LADs and NPC-interacting DNA, respectively. Impβ alone and Dam-Impβ with WGA (blocking agent for NPCs) show no Dpn8 signal. The scale bar is 5 µm.
a) EtBr-stained agarose gel for testing enzymatic activity of Dam. The unmethylated lambda DNA (1 µg) was incubated with an increasing concentration of Dam in the presence of SAM for 30 min at RT. After 30 min, DpnII was added to digest the remaining unmethylated DNA after the enzyme reaction. We observe the efficient protection of lambda DNA from digestion by DpnII through methylation by Dam. b) Same as a) for Dam-Impβ. c) Time-dependent activity of Dam (reaction set up was same as in a). Dam (5 units) requires at least 20 min at RT for complete methylation of 2 µg of lambda DNA. d) Same as c) for Dam-Impβ. e) Left, Venn diagram of the overlap between NPC-DamID peaks and LADs in IMR90. Right, heatmap and metagene profile showing the distribution of Dam-ΔN44Impβ around LAD regions. f) the distance-dependent distribution of the read density of Dam-ΔN44Impβ, Dam-Impβ+WGA, Dam-ΔN44Impβ+WGA centered around Dam-Impβ peaks. g) Proportion of NPC-DamID peaks overlapping with Nup153-DamID peaks or Nup93 cut&run peaks in IMR90 cells. h) Snapshot of the 3D computational model of IMR90 using Chrom3D with LADs as a peripheral nuclear constraint. Dam-Impβ peaks and Dam-Nup153 peaks were then overlapped to observe their spatial distribution. i) Frequency histogram for the distance of the beads corresponding to genomic loci of the LADs, Dam-Impβ, or Dam-Nup153 peaks from the center of the nuclei computed through Chrom3D models. j) DNA-FISH-based determination of the nuclear localization of loci shown to interact with LADs, the NPC, or neither.
NPC-DamID is also applicable to tissue samples. a) Top, schematic representation of the NPC-DamID assay for live human pancreatic islet tissues. Bottom, the normalized profile for Dam alone and Dam-Impβ along with the tracks for histone marks-H3K27Ac and H3K4Me3 for human pancreatic islets are shown. b) Pie chart for H3K27Ac overlap with NPC-DamID peaks for human pancreatic islets.
a) Bar plots for cell type consensus score of all SEs, NPC-associated SEs (NPC-SE), and SEs not associated with NPCs (non-NPC-SE) in HeLa cells. b) GO biological process enrichment analysis for genes associated with NPC SEs and non-NPC SEs in IMR90 (fibroblasts) and C2C12 (myotube) cells. c-f) Bar plots showing fold-enrichment for CTCF (c), BRD4 (d), P300 (e), PolII (f) co-occupied NPC-associated hierarchical SEs (NPC-SE) with hub, non-hub, and regional enhancer (RE) over the rest of the genome. g) Scheme for the model of NPCs acting as structural hubs for the organization of SEs through interactions mediated by disordered regions of Nups, transcription factors, and coactivators.
a) Confocal images for PLA experiment between TPR and chromatin structural proteins P300, BRD4, SMC3. The scale bar is 10 μm. b) Scatter plots comparing co-occupancy between CTCF, BRD4, POL II, P300, and NPC-DamID across the genome. The Spearman correlation coefficient is also shown along with each comparison.
a) PONDR prediction for intrinsically disordered regions (IDR) in the structure of Nup98. A score over 0.5 signifies a lack of structure. b) Confocal images of HeLa cells transfected with GFP fused with Nup98-IDR before and after 30 s treatment with 1.5% hexanediol (HD). c) FRAP recovery plots for 1.5 µm2 bleach spot in HeLa nuclei transfected with GFP fused to Med1-IDR, Nup153-IDR, or Nup98-IDR. The solid black line represents the curve fit, and tau is the diffusion time calculated from the curve fits.
FRAP recovery plots of fluorescent spots for mCherry-NLS-LacI, mCherry-Med1(IDR)-NLS-LacI, and mCherry-Nup153(IDR)-NLS-LacI. The data points are gray, and the solid black line represents the fit. k1 and k2 are dissociationconstants.
