Extended Data Fig. 5: irCLIPv2 supports a multi-RBP model on the same RNA molecule.
From: irCLIP-RNP and Re-CLIP reveal patterns of dynamic protein assemblies on RNA
(a) Workflow of irCLIPv2 experiments of three RNP subzones for 3 RDAPs and 4 RBPs belonging to HNRNPs protein family in HEK293T. (b) Nitrocellulose images indicate the infrared signal of irCLIP-RNP ligations for the 3 RDAPs (green) and 4 RBPs (dark fuchsia) in HEK293T. Blue, light-green, and green boxes highlight the gel sectioning of the three RNP subzones (n = 2 biologically independent experiments). L: ladder. (c) Pie chart indicates the percentages of total significant binding regions of 3 RDAPs (green) and 4 RBPs (dark fuchsia). (d) Heatmap indicates the number of significant regions per bin along each chromosome for RDAPs (green) and RBPs (dark fuchsia). Genes distribution across chromosomes are displayed in shades of blue. (e) Venn diagram indicates the overlap between irCLIPv2-identified regions for HNRNPC and previously published binding regions identified with easyCLIP22. (f) Lines indicate the normalized coverage at significant binding regions in the three RNP subzones and no-UV samples (blue: #1; light-green: #2; green: #3; grey: no-UV) for RBPs (dark fuchsia) and RDAPs (green). Coverage on the negative strand was reversed to be represented in a 5’ to 3’ orientation. An extension of +/− 50 bp was applied from the highest RBP peaks in the RNP subzone #1. (g) Schematics of cDNA synthesis early truncation resulting in signal 3’-shift due to RDAPs co-binding. (h) Lines indicate the slope density distribution calculated from average normalized RT stop counts of significant regions for RDAPs (green) and RBPs (dark fuchsia). One standard deviation (SD, dotted lines) was used to classify the significant regions according to their occurrence in the three RNP subzones. Blue/light-green/green boxes: RNP subzones #1-3. (i) Stacked bars indicate the percentage of significant regions in transcriptomic landscape features for RBPs (green) and RDAPs (dark fuchsia). Blue/light-green/green boxes: RNP subzones #1-3. (j) Schematics of RDAP motif analysis across RNP subzones in the regions directly downstream of highest RBP peaks. 250 bp, 150 bp, and 75 bp fragments were used as input in the AME tool for RNP subzones #1 to #3, respectively. (k) Heatmap indicates the enriched motif -log10(P-values) as determined by AME analysis46 of high-MM proteins in the three RNP subzones of RDAPs (green) and RBPs (dark fuchsia). Blue/light-green/green boxes: RNP subzones #1-3. (l) Raincloud plots indicate the distribution of average log2FC against no-UV samples for HNRNPC regions categorized as low (blue), medium (light green), or high (green) distances in their corresponding RNP subzone. one-way ANOVA with one-sided Tukey’s multiple comparison tests. Only the significance of high-categorized regions against low and medium categories related to the same protein or against the ALYREF long-distance category is shown (n = number of regions specified in the plot). Boxes indicate the interquartile range (IQR) with 25th and 75th percentile box limits; whiskers denote the 1.5 × IQR from the 25th and 75th percentiles; center lines: medians; black dots: outliers; red dots: mean values. (m) Representative immunoblot image of n = 3 biologically independent experiments indicates the protein signal of HNRNPC and KHSRP after native RNA pull-down of biotinylated HNRNPC-KHSRP flanking regions in HEK293T cells. (n) Bars indicate the fold change against scramble control of HNRNPC and KHSRP protein signal after native RNA pull-down of biotinylated HNRNPC-KHSRP flanking regions in HEK293T cells. Data are mean ± SD; KHSRP-HNRNPC P-value = 5.3e-3**; KHSRP-PTBP2 P-value = 4.0e-4***; HNRNPC-HNRNPC P-value = 1e-3***; HNRNPC-PTBP2 P-value = 2.0e-4*** – unpaired two-tailed Student’s t-test against scramble control 1 and 2 (n = 3 biologically independent experiments).
