Fig 2: Characterization of the retained intron events following SDE2 depletion. (A) Comparison of the number and percentage of increased significant RI events (P<0.05) in knockdown between ENCODE data and SDE2 data. The percentages represent the fraction of introns in the whole genome affected by the knockdown of the corresponding splicing factor. (B) Histogram of the Location of the RIs between significant (Sig) and non-significant (Non-sig) events. On the X-axis, ‘0’ corresponds to the 5′ end of the transcript and the ‘1’ corresponds to the 3′ end of the transcript. Position of the RI is determined using the center of the RI. Dashed lines mark the median of the fraction in each group. (C) Boxplot of the Length of the RIs among Sig, Non-sig and Non-AS group. The number of events in each group is annotated. Length is log10 transformed on Y-axis. (ns: not significant, ***P < 0.001, ****P < 0.0001, Mann–Whitney U test). (D) Boxplot of the GC content of the RIs among Sig, Non-sig, Non-AS-short and Non-AS-long groups. The number of events in each group is annotated (**** P < 0.0001, Mann–Whitney U test). (E) Boxplot of the splice site scores on both the 5′ and 3′ end of the RIs among the four groups. The plot is only showing the distribution of positive scores, but the statistical test included all of the data (ns: not significant, ***P < 0.001, ****P < 0.0001, Mann–Whitney U test).

Fig 3: SDE2 is an RNA binding protein that interacts with ncRNAs. (A) Western blot of SDE2, GAPDH, and H4 following cellular fractionation of HeLa, U2OS, 293 FT and RPE-1 cell lines. GAPDH is used as a marker for the cytoplasmic fraction and H4 is used as a marker for the nuclear fraction. (B) Radiographic image of 32P labelled RNA from CLIP of SDE2 and IgG in HeLa cells, in both uncrosslinked (-UV) and crosslinked (+UV) conditions. (C) Western blot of PTex samples in HeLa cells from both uncrosslinked (-UV) and crosslinked (+UV) conditions. U2AF1 is used as a positive control and DNA-binding protein H4 is used as a negative control. (D) eCLIP usable reads were identified as either ‘unique,’ or ‘repetitive,’ and were plotted as percentages of total usable reads for both IPs and inputs for SDE2 and U2AF1. (E) All (unique and repetitive) elements were filtered for information content >100. Unique non-repetitive elements correspond to mRNA components including unique proximal intron, unique 3′ UTR, unique 5′ UTR, unique intergenic and unique noncoding proximal intron. These unique counts were averaged between IP replicates and plotted at the family level for both SDE2 and U2AF1. (F) Using a custom bioinformatic pipeline, repetitive elements from SDE2 and U2AF1 eCLIP were identified at the gene level by filtering for raw counts >100 and Log2 fold change >0 in duplicate IP conditions relative to respective inputs. Normalized counts were calculated and averaged between replicates and plotted against the average Log2 fold change between replicates. Two genes from SDE2 and four genes from U2AF1 were excluded in (F) because their data points were outside the axis limits. The omission did not alter the overall analysis or interpretation of the data.

Fig 4: SDE2 depletion increases intron retention. (A) Reciprocal immunoprecipitation of SDE2, SF3B1, U2AF1 and Cactin, with a 1% input from whole cell extract in HeLa cells. Tubulin is used as a negative IP control. (B) Western blot of SF3B1, SDE2 and GAPDH on a membrane containing 16 human tissue samples. (C) Overview of the AS events identified using our custom pipeline after 3 days of SDE2 depletion via siRNA. Pie chart represents the percentage of each AS type. The total number of significant AS events is annotated at the bottom. (D) Bar plot demonstrating the absolute number of each AS type. (E) Schematic representation of one gene containing an RI event identified in our analysis pipeline; exons are depicted by peach boxes, the retained intron is depicted with a thick black line, and the exons flanking the RI event are depicted by larger orange colored boxes. Location of primers used to amplify RI events are shown as gray arrows. (F and G) DNA gels depicting intron retention events following RT-PCR. RNA was extracted from siCTRL and siSDE2 cells and target genes were amplified by RT-PCR in HeLa cells (F and H) and U2OS cells (G and I) with (+RT) and without (-RT) reverse transcriptase. RI events and spliced events are labeled. Graphs show mean percent intron retention in each condition relative to total transcript (intron retained + spliced) ± SD. RFNG does not demonstrate a dAS RI event and serves as a negative control. n=3 for all quantifications and experiments, P-values denoted by comparing siCTRL versus siSDE2-1 or siSDE2-2 by two-way ANOVA followed by Dunnet's multiple comparison test, P ≤ 0.01 denoted by **, P ≤ 0.001 denoted by ***, and P ≤ 0.0001 denoted by ****, ns = not significant. For RI events with different levels of P-values between siSDE2-1 and siSDE2-2, the least significant value is shown.

Fig 5: SDE2 directly interacts with C/D snoRNAs and the methyltransferase FBL to promote ribosome biogenesis. SDE2 also interacts with the U2 snRNP complex and is required for efficient pre-mRNA splicing.