SSUP-72/PINN-1 coordinates RNA-polymerase II 3' pausing and developmental gene expression in C. elegans

Abstract

During exit from Caenorhabditis elegans (C. elegans) L1 developmental arrest, a network of growth- and developmental genes is activated, many of which are organized into operons where transcriptional termination is uncoupled from mRNA 3'-end processing. CDK-12-mediated Pol II CTD S2 phosphorylation enhances SL2 trans-splicing at downstream operonic genes, preventing premature termination and ensuring proper gene expression for developmental progression. Using a genetic screen, we identified the SSUP-72/PINN-1 module as a suppressor of defects induced by CDK-12 inhibition. Loss of SSUP-72/PINN-1 bypasses the requirement for CDK-12 in post-embryonic development. Genome-wide analyses reveal that SSUP-72, a CTD S5P phosphatase, affects Pol II 3' pausing and regulates intra-operon termination. Our findings establish SSUP-72/PINN-1 as a key regulator of Pol II dynamics, coordinating operonic gene expression and growth during C. elegans post-embryonic development.

Fig. 1. Mutations of the SSUP-72/PINN-1 module suppress the developmental arrest resulting from inhibition of CDK-12as.

A Schematic overview of the EMS mutagenesis and screen for suppressors of the L1 developmental arrest resulting from the inhibition of the CDK-12as (analogue-sensitive) mutant. B, C Transmitted light images of the indicated strains and conditions. Images were taken 96 h after depositing individual P0 worms at L4 stage. Blue arrows point to the P0, yellow arrows to L1 arrested worms. D Left, Western blot analysis of Pol II CTD-S2P levels at L1 stage in the indicated strains and conditions. Uncropped scan in Supplementary Fig. 2A. Right, quantification of CTD S2P intensities normalized to total protein signal. The intensities for the wild type were set to 1. Significance was tested using a two-tailed unpaired t-test test (ns not significant, ** = <0.01). Measurements were taken from 3 independent replicates. Error bars represent standard deviation. E Worm length as a proxy for growth, measured after 72 h of CDK-12as inhibition, starting with staged L1 worms of the indicated strains and conditions. Measurements were taken from two independent replicates. F Transmitted light images of the indicated strains and conditions. Images were taken 96 h after depositing individual P0 worms at L4 stage. G Quantification of Western blot analysis of CTD S5P levels normalized to total protein in L1-staged worms from the indicated strains. The intensities for the wild type were set to 1. Significance was tested using a two-tailed unpaired t-test (** = <0.01). Measurements were taken from three independent replicates. Error bars represent standard deviation. Error bars represent standard deviation.

Fig. 2. The phosphatase SSUP-72 is required for efficient pausing of Pol II at the 3′ end.

A Left panel: Metagene analysis of GRO-seq signal at non-operonic protein-coding genes (n = 16411). The GRO-seq signal was averaged for all genes and aligned at their annotated transcription start site (TSS) and transcription end site (TES) for the indicated strains and conditions. +I denotes growth in the presence of 2 µM 3MB-PP1. The plot was generated using data from three replicates. The red box highlights the 3′ end region where pausing defects are observed. Right panel: Ratio of GRO-seq signal from ssup-72[E22K] over wild type, aligned at the TES, highlighting the lower Pol II density at the 3′ end in the ssup-72[E22K] single mutant compared to the wild type. B Empirical cumulative distribution plots of the processivity score on individual genes in the indicated strains and conditions, as in (A). The plot was generated from three replicates of GRO-seq data. C Boxplot showing processivity scores for all isolated protein-coding genes, in the indicated strains and conditions. Worms were grown in the presence of 2 µM 3MB-PP1. Significance was tested using a two-tailed Mann–Whitney–Wilcoxon test (ns not significant, * = <0.05, *** = <0.001). The plot was generated from three replicates of GRO-seq data. Boxplots display median (line), first, and third quartiles (box), and 90th/10th percentile values (whiskers).

Fig. 3. The downregulation of SSUP-72 restores efficient expression of SL2 trans-spliced mRNAs that are downregulated upon CDK-12 inhibition.

A Volcano plot showing cdk-12as-dependent changes in transcript levels as determined by RNA-seq (CDK-12as inhibited compared to the wild-type). +I denotes growth in the presence of 2 µM 3MB-PP1. Purple dots highlight 2+ encoded mRNAs within operons (n = 2155). The dotted lines indicate a Fold Change of ±1.5 (vertical) and a p-value < 0.05 (horizontal) (Wald test). The plot was generated from 3 independent replicates. B Bar graphs showing the relative percentage of cdk-12as-dependent changes (RNA-seq) by the operonic position (non-operonic: isolated genes, n = 16411; +1 Genes: First position genes within operons, n = 1430; 2+ Genes: Genes at position 2 and over within operons, n = 2155). Chi-square statistical test (*** = p-value < 0.001). The plot was generated from 3 independent replicates. C Schematic of the suppression of the 2+ gene downregulation in cdk-12as;ssup-72[E22K] compared to cdk-12as. The proportion of 2+ genes that is downregulated in cdk-12as worms compared to wild type is shown in green. The proportion of 2+ genes affected by the inhibition of CDK-12as that is upregulated in cdk-12as;ssup-72[E22K] compared to cdk-12as is shown in orange. D Comparison between mRNA cdk-12as-dependent changes in mRNA levels (x-axis: cdk-12as/wild-type) and the cdk-12as;ssup-72[E22K] suppressive changes (y-axis: cdk-12as;ssup-7[E22K]/cdk-12as) (RNA-seq). +I denotes growth in the presence of 2 µM 3MB-PP1. 2+ genes (n = 2155) are highlighted with purple dots. The dotted lines indicate a Fold Change of ±1.5. The plot was generated from three independent replicates. E Comparison between cdk-12as-dependent changes in mRNA levels (x-axis: cdk-12as/wild-type) and the cdk-12as;ssup-72[E22K]-dependent changes (y-axis: cdk-12as;ssup-7[E22K]/wild-type) (RNA-seq). +I denotes growth in the presence of 2 µM 3MB-PP1. +1 genes (n = 1430) are highlighted with orange dots. The dotted lines indicate a Fold Change of ±1.5. The plot was generated from three independent replicates. F, G Volcano plots showing ssup-72-dependent changes (F: ssup-72[C13S], G: ssup-72[E22K]) in transcript levels as determined by RNA-seq. +1 genes (n = 1430) are highlighted with orange dots. The dotted lines indicate a Fold Change of ±1.5 (vertical) and a p-value < 0.05 (horizontal) (Wald test). The plot was generated from three independent replicates. H Correlation between the SL2 score (SL2/SL2 + SL1) in ssup-72[E22K] mutant and the wild type. The red dotted line is the linear regression. Only genes with at least 10 matching splice-leader reads were kept for the analysis (n = 1739).

Fig. 4. The downregulation of SSUP-72 inhibits premature Pol II transcription termination at position 2 and over within operons.

A Left panel: Metagene analysis of GRO-seq signal for the first genes (+1 genes) within operons (n = 1430) in the indicated strains and conditions. +I denotes growth in the presence of 2 µM 3MB-PP1. Right panel: Metagene analysis of GRO-seq signal for the 2+ genes within operons (n = 2155). The red box highlights the gene bodies showing the restoration of a wild-type level of Pol II occupancy during elongation in the cdk-12as;ssup−72[E22K] strain. The plots were generated using data from three replicates. The GRO-seq signal was averaged and aligned at the annotated transcript start sites (TSS) and transcript end sites (TES). B. Top panel: schematic of a three genes operon structure showing the genes and outrons. Middle panel: Metagene fold change analysis of cdk-12as over wild-type GRO-seq signal. Bottom panel: Metagene fold change analysis of cdk-12as;ssup-72[E22K] over cdk-12as GRO-seq signal. Only operons harboring three genes (n = 296) were considered. The metagene plots were generated using data from three replicates. The GRO-seq signal was averaged and aligned at the annotated transcript start sites (TSS) and transcript end sites (TES). C Metagene Analysis of ssup-72[E22K] over wild-type (N2) GRO-seq signal aligned around annotated transcript end sites (TES). Non-operonic genes (n = 16411), first genes within operons (n = 1430), 2+ genes within operons (n = 2155). The plots were generated from three independent replicates. D, E Volcano plots showing the differential expression of SL2 outrons (n = 1621) in ssup-72 mutants (D: ssup-72[S39E], E: ssup-72[S39A]) compared to the wild type. Orange dots highlight significantly increased outron detection, and blue dots highlight significantly decreased outron detection. The dotted lines indicate a fold change of ±1.5 (vertical) and a p-value < 0.05 (horizontal) (Wald test). The plots were generated from three independent replicates.