The chromatin remodeler Ino80 mediates RNAPII pausing site determination

Abstract

Background: Promoter-proximal pausing of RNA polymerase II (RNAPII) is a critical step for the precise regulation of gene expression. Despite the apparent close relationship between promoter-proximal pausing and nucleosome, the role of chromatin remodeler governing this step has mainly remained elusive.

Results: Here, we report highly confined RNAPII enrichments downstream of the transcriptional start site in Saccharomyces cerevisiae using PRO-seq experiments. This non-uniform distribution of RNAPII exhibits both similar and different characteristics with promoter-proximal pausing in Schizosaccharomyces pombe and metazoans. Interestingly, we find that Ino80p knockdown causes a significant upstream transition of promoter-proximal RNAPII for a subset of genes, relocating RNAPII from the main pausing site to the alternative pausing site. The proper positioning of RNAPII is largely dependent on nucleosome context. We reveal that the alternative pausing site is closely associated with the + 1 nucleosome, and nucleosome architecture around the main pausing site of these genes is highly phased. In addition, Ino80p knockdown results in an increase in fuzziness and a decrease in stability of the + 1 nucleosome. Furthermore, the loss of INO80 also leads to the shift of promoter-proximal RNAPII toward the alternative pausing site in mouse embryonic stem cells.

Conclusions: Based on our collective results, we hypothesize that the highly conserved chromatin remodeler Ino80p is essential in establishing intact RNAPII pausing during early transcription elongation in various organisms, from budding yeast to mouse.

Keywords: + 1 nucleosome; AID system; Alternative pausing site; PRO-seq; Promoter-proximal RNAPII pausing; The chromatin remodeler Ino80p.

Fig. 1.

PRO-seq reveals a non-uniform distribution at TSS genome-wide in S. cerevisiae. The gene set used in the previous study [28] was used (N = 3403) for analysis. a Scaled composite profile shows the median intensity of our Ctrl PRO-seq data generated in Ino80p-AID cells cultured without auxin treatment. Regions between 300 bp downstream from TSS and 300 bp upstream of TES were scaled to 60 bins. Outside of these regions were reflected as a 10-bp bin. b Heatmaps display previously published wt (w303a) PRO-seq data (GSM1974983) and our Ctrl PRO-seq data or wt (BY4741) PRO-seq data around TSS. Genes were sorted by the PR intensity of our Ctrl data. Signals reflect the 10-bp bin. c Scatterplots represent the correlation of PR and GB intensity and PI between the indicated data. d, e Average profiles show the median intensity of GSM1974983. Genes were grouped into quartiles based on PI of GSM1974983 (d) or Ctrl (e) data, where Q1 represents the highest PI. f Venn diagram indicates the overlap between paused and not paused genes defined based on GSM1974983 or Ctrl data. P value was calculated using the hypergeometric distribution. g, h Average profiles display the paused and not paused genes defined based on GSM1974983 (g; 177 paused genes and 2770 not paused genes) or Ctrl (h; 1815 paused genes and 980 not paused genes) data. The green line in the GSM1974983 plot (g) indicates 1655 genes defined as paused genes in only Ctrl data. The green line in the Ctrl plot (h) represents 160 genes defined as paused genes in both data. i Genome browser view of PRO-seq signal for representative genes among overlapped paused genes in both data. PRO-seq data were generated using combined biological replicates

Fig. 2

Features of the non-uniform distribution of transcription elongation in S. cerevisiae. a Average profile shows median PRO-seq intensity around TSS (left). In total, 2599 paused and 1990 not paused genes were classified among 5697 total filtered protein-coding genes (right). b Genome browser view of PRO-seq signals for representative paused (top) or not paused genes (bottom). c Average profile indicates median MNase-seq (GSM3304635) intensity centered on TSS of paused and not paused genes. d Boxplots exhibit the relationship of PRO-seq PR density (left) and PI (right) to the bottom (0–20%), middle (40–60%), and top (80–100%) pentiles of gene activity. Only paused genes were considered. Asterisks represent statistically significant differences based on Mann-Whitney U test. e Cumulative curve analyzes the distance from TSS to P1 at paused genes. The red dotted lines represent the 25th (76 bp) and 75th (156 bp) percentiles. f Average profiles show median MNase-seq intensity (GSM3304635) at TSS (left) and median PRO-seq intensity at the + 1 dyad (right). The + 1 dyad was defined by the improved nucleosome-positioning algorithm (iNPS) using existing MNase-seq data (GSM3304635) [57]. The two dotted red lines represent the 25th (Q1; 76 bp) and 75th (Q3; 156 bp) percentiles of P1 (left), and the two dotted black lines represent the expected position of the + 1 nucleosome (right; 75 bp upstream and downstream of + 1 dyad). All PRO-seq data represent the Ctrl data that were generated in auxin-untreated Ino80p-AID cells and combined biological replicates were used. For average profiles, medians reflect the 10-bp bin

Fig. 3

Ino80p knockdown causes the transition of RNAPII pausing at genes with the alternative pausing site. a Average profile indicates median PRO-seq intensity in Ino80p-AID cells under Ctrl and KD conditions for paused genes. b Boxplots represent the relative read ratio (log₁₀) at the significantly increasing peaks (red indicates the ratio of KD versus Ctrl and green indicates the ratio of Rescue versus Ctrl) at each replicate. To analyze the relative distance to P1 from each peak (x-axis), we first calculated the distance either from TSS to P1 or P1 to TSS + 250 bp, and divided each into 10 bins, and mapped the peak to the corresponding bins. Former and later bins were referred to as − 10 to − 1 and 1 to 10 depending on the distance from P1, respectively. To display the significantly enriched bins, bins with less than 5% of the total increasing peaks were excluded; in the end, 386 peaks for replicate 1 and 383 peaks for replicate 2 out of 618 peaks at 467 genes were used. c Average profiles exhibit median PRO-seq intensity at TSS of shift-to-5′ genes (left). The arrows and dotted lines represent the median of P1 (blue, 142 bp) and P2 (red, 56 bp). Boxplot shows the distance from TSS to the indicated pausing sites (right). d Average profiles represent median PRO-seq intensity for no-shift and shift-to-5′ genes in Ctrl data. The two dotted lines represent the 10th (− 22 bp) and 90th (− 144.8 bp) percentiles of P2 relative to P1 for shift-to-5′ genes. e Average profiles display median PRO-seq intensity (left), and boxplots depict the smoothed PRO-seq intensity (right) at the indicated pausing sites of shift-to-5′ genes. f Genome browser view of PRO-seq signal for representative shift-to-5′ genes. All data except for (a) were generated using combined biological replicates. For average profiles, medians reflect the 10-bp bin. Asterisks represent statistically significant differences based on Wilcoxon signed rank test

Fig. 4

Proper localization of pausing is closely associated with the regulation of + 1 nucleosome by Ino80p. a Average profile shows median MNase-seq intensity in the auxin-untreated Ino80p-AID cells (GSM3304635) at P1 of no-shift and shift-to-5′ genes (left). The green line exhibits the median of 1000 bootstrap samples (Each sample containing the same number of genes to the original population and was obtained by resampling allowing replacement). Boxplot represents the distance from the + 1 dyad to the indicated pausing sites for shift-to-5′ genes (right). b Average profiles depict median MNase-seq intensity at P1 of shift-to-5′ genes in the auxin-untreated (Ctrl) or -treated (KD) data. GSM3304635 and GSM3304637 were used for the left panel, and GSM3177778 and GSM3177779 were used for the right panel (high MNase). For (a) and (b), the two dotted lines indicate the 10th (− 22 bp) and 90th (− 145 bp) percentiles of P2 relative to P1. c, d Average profiles display median PRO-seq intensity for the indicated samples around the + 1 dyad. The two dotted lines in profiles represent the expected position of the + 1 nucleosome (75 bp upstream and downstream of the + 1 dyad). e Genome browser view of PRO-seq signals for representative genes whose pausing sites were shifted in the 5′ direction in both Ino80p-KD and arp5Δ. All PRO-seq data were generated using combined biological replicates. Only genes with nucleosomes overlapped with H3K4me3 ChIP-seq enrichment (GSM2507874) were used in an effort to exclude false-positive nucleosomes (260 no-shift genes and 251 shift-to-5′ genes upon Ino80p-KD; 398 shifted to 5′ genes in arp5Δ). For average profiles, medians reflect the 10-bp bin. Asterisks represent statistically significant differences, as calculated using either Wilcoxon signed rank test or Mann-Whitney U test

Fig. 5

INO80 mediates RNAPII pausing site determination in mESCs. a Boxplots represent the relative read ratio (log₁₀) at the significantly increasing peaks (the ratio of siINO80 versus siEGFP) in a manner similar to Fig. 3b (2406 peaks for replicate 1 and 2408 peaks for replicate 2 out of 2962 peaks at 1890 genes). b Average profile displays median PRO-seq intensity in mESCs treated with siEGFP or siINO80 for shift-to-3′ genes (left). Boxplot shows the distance from 5′-peak to the indicated pausing sites (right). c Average profiles exhibit median PRO-seq intensity in siEGFP-treated mESCs at no-shift and shift-to-3′ genes. d Average profiles depict median PRO-seq intensity (left), and boxplots represent the smoothed PRO-seq intensity (right) at the indicated pausing sites for shift-to-3′ genes. e Average profile shows median MNase-seq intensity obtained from untreated mESCs (GSM2906312 and GSM2906313) around P1 of no-shift and shift-to-3′ genes (top). The green line indicates the median of 1000 bootstrap samples (Each sample containing the same number of genes to the original population and was obtained by resampling allowing replacement). Boxplot exhibits the distance from the + 1 dyad to the indicated pausing sites for shift-to-3′ genes (bottom). For (c) and (e), the two dotted lines represent the 10th (12.9 bp) and 90th (76.1 bp) percentiles of P2 relative to P1. f Average profile displays median PRO-seq intensity around the + 1 dyad for shift-to-3′ genes. For (e) and (f), only genes with nucleosomes overlapped with H3K4me3 ChIP-seq enrichment (GSM590111) were used in an effort to exclude false-positive nucleosomes (431 no-shift genes and 500 shift-to-5′ genes upon INO80-KD). g Genome browser view of PRO-seq signal for representative shift-to-3′ genes. All PRO-seq data were generated using combined biological replicates. For average profiles, medians reflect the 5-bp bin (PRO-seq) or the 10-bp bin (MNase-seq). Asterisks represent statistically significant differences based on Mann-Whitney U test

Fig. 6

Model of the function of Ino80p in mediating pausing site determination. Model depicts a regulatory role of Ino80p in proper localization of RNAPII pausing in a nucleosome context-dependent manner in S. cerevisiae. See “Discussion” for details