Most mRNAs up-regulated in the scd6Δedc3Δ mutant are not iESR transcripts and exhibit Scd6/Edc3 functional redundancy in repression of transcript abundance.

(A) Venn diagram of overlap between the 741 mRNAs up-regulated in the scd6Δedc3Δ mutant vs. WT and the 283 induced ESR mRNAs, indicating fold-enrichment and P value of overlap determined by the hypergeometric distribution. (B) Venn diagram of overlaps involving all 591 non-iESR mRNAs derepressed in abundance by either the scd6Δ, edc3Δ, or scd6Δedc3Δ mutations. (C) Notched box-plot analyses of log2 changes in mRNA abundance (log2ΔmRNA) determined by DESeq2 analysis between the indicated mutants vs. WT for mRNAs belonging to the specified sectors of the Venn diagram in (B). Un-logged median values are indicated at the bottom. (D) Gene browser image for MDH2 presented as in Fig. S3C.

Most mRNAs up-regulated in the scd6Δedc3Δ mutant are also derepressed by dhh1Δ and pat1Δ.

(A) Venn diagram of overlap between all 591 non-iESR mRNAs derepressed in abundance by the scd6Δ, edc3Δ, or scd6Δedc3Δ mutations (from Fig. 1B) and 1018 non-iESR mRNAs up-regulated by the dhh1Δ, pat1Δ, or pat1Δdhh1Δ mutations identified previously (Vijjamarri et al. 2023a). (B) Notched box-plots of log2ΔmRNA between the indicated mutants vs. WT for mRNAs in the three sectors specified in (A). (C) Hierarchical clustering analysis of log2ΔmRNA values conferred by the indicated mutations vs. WT for 784 of the 794 mRNAs up-or down-regulated in dhh1Δ vs. WT cells for which RNA-Seq data was obtained in all four strains and with log2ΔmRNA values >-5 and <5, conducted with R heatmap.2 function from R ‘gplots’ library, using default hclust hierarchical clustering algorithm. Spearman coefficients (ρ) and associated P values are given for the indicated correlation analyses. (D) Venn diagram of overlaps between the 591 non-iESR mRNAs up-regulated by the scd6Δ/edc3Δ mutations vs. WT (from Fig. 1B) and the indicated 431 and 590 non-iESR mRNAs up-regulated by dhh1Δ or pat1Δ vs. WT, respectively, identified previously (Vijjamarri et al. 2023a). (E) Notched box-plots of log2ΔmRNA between the indicated mutants vs. WT for the 431 non-iESR mRNAs up-regulated by dhh1Δ vs. WT shown in panel (D).

Evidence that impaired decapping rather than increased transcription is a key driver of changes in mRNA abundance in scd6Δedc3Δ cells.

(A) Hierarchical clustering analysis of log2ΔmRNA values conferred by the indicated mutations vs. WT for 741 of the 1052 mRNAs up-or down-regulated in scd6Δedc3Δ vs. WT cells for which RNA-Seq data was obtained in all five strains and with log2ΔmRNA values >-5 and <5, conducted as in Fig. 2C, showing Spearman coefficients and P values for indicated correlations. (B) Notched box-plots of log2ΔmRNA between the indicated mutants vs. WT for the 554 non-iESR mRNAs up-regulated by scd6Δedc3Δ vs. WT (shown in Fig. 1B). (C) Ratios of capped to total mRNA abundance in TPMs (Relative C/T) in WT or dhh1Δ cells plotted for all mRNAs or the 554 non-iESR mRNA_ up_s6,e3 or 526 non-rESR mRNA_down_s6,e3 transcripts dysregulated by scd6Δedc3Δ vs. WT. (D) Notched box-plots of the codon-protection index (CPI) for all mRNAs or for the sets of mRNAs up-or down-regulated by scd6Δedc3Δ vs. WT, including or excluding ESR transcripts, as indicated. (E) Notched box-plots showing log2 changes in absolute mRNA abundance from ERCC spike-in normalized RNA-Seq (left) or absolute Rpb1 occupancies averaged over the CDSs from S. pombe chromatin spike-in normalized Rpb1 ChIP-Seq (right) in scd6Δedc3Δ vs. WT cells for all mRNAs or the 554 or 526 non-ESR mRNAs up-or down-regulated, respectively, by scd6Δedc3Δ vs. WT.

Most mRNAs translationally up-regulated in scd6Δedc3Δ cells exhibit Scd6/Edc3 functional redundancy for repressing TE and show correlated changes in TE and protein abundance.

(A) Venn diagram of overlap between the 184 and 42 mRNAs in the TE_up groups identified in the scd6Δedc3Δ vs. WT or scd6Δ vs. WT comparisons, respectively. (B) Notched box-plots of log2ΔTE values between the indicated mutants vs. WT for the mRNAs belonging to sectors (i) or (iii) of the diagram in (A). (C) Gene browser image for SPI1 presented as in Fig. S3C, except also giving the TE changes for each mutant vs. WT on the lower right. (D) Density scatterplot of log2ΔRPF values measured by ribosome profiling vs. log2Δprotein values measured by TMT-MS for 4339 mRNAs for which data were obtained in both analyses, indicating the Pearson correlation coefficient (r) and P-value of the correlation. (E-F) Notched box-plots of log2Δprotein values from TMT-MS analysis between the scd6Δedc3Δ mutant vs. WT for the 843 and 839 mRNAs belonging to the RPF_up or RPF_down groups, respectively (D), or the 184 and 152 TE_up or TE_down mRNA groups (E) determined for the scd6Δedc3Δ mutant vs. WT, or for all mRNAs, for which TMT-MS data was obtained.

Most mRNAs translationally up-regulated in the scd6Δedc3Δ mutant are also translationally derepressed by dhh1Δ or pat1Δ.

(A) Venn diagram of overlap between all 212 mRNAs translationally derepressed by scd6Δ, edc3Δ, or scd6Δedc3Δ (defined in Fig. 4A) or all 274 mRNAs translationally derepressed by dhh1Δ, pat1Δ, or pat1Δdhh1Δ vs. WT identified previously (Vijjamarri et al. 2023a). (B) Notched box-plots of log2ΔTE values between the indicated mutants vs. WT for the mRNAs belonging to the specified sectors of the diagram in (A). (C) Hierarchical clustering analysis of log2ΔTE values conferred by the indicated mutations vs. WT for 222 of the 336 mRNAs translationally up- or down-regulated in scd6Δedc3Δ vs. WT cells for which RNA-Seq and Ribo-Seq data were obtained in all five strains and with log2ΔTE values >-5 and <5 conducted as in Fig. 2C, including the Spearman coefficients (ρ) and P values for the indicated correlations. (D) Notched box-plots of log2ΔTE values between the indicated mutants vs. WT for the 54 mRNAs showing >2-fold TE increases conferred by both scd6Δedc3Δ and pat1Δdhh1Δ mutations vs. WT. (E) Relative Dhh1 occupancies from the Dhh1 RIP-seq experiments of Miller et al (2018) for the 212 and 274 mRNAs identified as TE_up in the scd6Δ, edc3Δ, or scd6Δedc3Δ mutants, or the dhh1Δ, pat1Δ, or pat1Δdhh1Δ mutants, vs. WT, respectively (cols 1-2), or for the 591 and 1018 mRNAs identified as mRNA_up in either the scd6Δ, edc3Δ, or scd6Δedc3Δ mutants, or the dhh1Δ, pat1Δ, or pat1Δdhh1Δ mutants, vs. WT, respectively (cols. 3-4).

Scd6/Edc3 post-transcriptionally repress proteins involved in respiration and suppress mitochondrial membrane potential in rich medium.

(A) Log2 changes in mRNA, RPFs, or TE conferred by the indicated double mutations vs. WT for 53 nuclear genes encoding mitochondrial proteins involved directly in oxidative phosphorylation. (B-C) Western blot analysis of 9 mitochondrial proteins and Gcd6 (examined as loading control) in WT and scd6Δedc3Δ strains, cultured in duplicate in YPD medium to OD600 of ∼0.6-0.8. WCEs were extracted under denaturing conditions and aliquots corresponding to 1X or 2X amounts of WCE were loaded in successive lanes for the two replicate cultures. Immune complexes were visualized with enhanced chemiluminescence (B). Signals for each protein were quantified, normalized to the corresponding signals for Gcd6 in the same extract and expressed relative to the resulting values for WT cells. Mean values and standard errors are plotted (C). (D) Western blot analysis of Cox2 in strains of the indicated genotypes in cells cultured as in (B). Cox2 signal intensity was normalized to total Coomassie-stained protein and the resulting relative Cox2 protein levels from three biological replicates were averaged and plotted. (E) Log2 changes in mRNA, RPFs, or TE conferred by the indicated double mutations vs. WT for 22 genes encoding enzymes of the glyoxylate cycle or fatty acid metabolism. (F) Expression of the CYC1-lacZ reporter on plasmid pLG265, lacking UAS1 and containing the optimized version of UAS2, UAS2UP1, in the WT strain grown on SC-Ura medium containing either 2% glucose or 3% glycerol/2% ethanol as carbon sources (i), or in WT and the indicated mutant strains on SC-Ura with 2% glucose (ii). β-galactosidase activity (nmoles of o-nitrophenyl-β-D-galactopyranoside (ONPG) cleaved per min per mg of total protein) was measured in whole cell extracts for 3 biological replicates of each strain and the mean values were normalized to the mean activity measured in WT grown with glucose as carbon source. **, P-value <0.01 from student’s t-test; ns, not significant. (G) Measurements of mitochondrial membrane potential. WT cells or transformants of the scd6Δedc3Δ mutant containing the EDC3 plasmid pLfz614-7 or empty vector were cultured in SC-Ura to mid-log phase. TMRM (500 nM) was added and incubated for 30 min before samples were collected and washed once with deionized water. ΔΨm was determined by measuring TMRM fluorescence intensity using flow cytometry. Data are presented in arbitrary fluorescence intensity units per OD600. 2-way ANOVA was used for statistical analysis and data are given as mean values ± SD (n=3) (****p<0.0001).

Eliminating decapping activators or decapping enzyme confers similar changes in polar metabolites.

(A) Principal component analysis of the levels of 147 metabolites in biological replicates of each strain. (B-C) Hierarchical clustering analysis of log2 changes in all 147 metabolites analyzed (B) or the 46 metabolites up-regulated in any two of the four mutants (C) conferred by the indicated mutations vs. WT, including the Spearman coefficients (ρ) and P values for the indicated correlations. (D) Results of pathway analysis of the 46 up-regulated metabolites described in (C) conducted at https://www.metaboanalyst.ca/MetaboAnalyst/Secure/pathway. Red ovals depict groups of metabolites significantly enriched among the set of 46 compounds, with P-value <0.05. (E-F) Log2 changes in levels of TCA cycle intermediates (E) or amino acids (F) conferred by the indicated mutations vs. WT.

Elimination of decapping activators or decapping enzyme up-regulates respiration, increasing flux from glucose into TCA cycle intermediates and proportion of ATP produced by Ox. Phos.

(A) Three biological replicates of cells of each indicated genotype were cultured in YP with 2% unlabeled glucose, shifted to YP with 1% unlabeled glucose for 20 min, and pulsed with 13C6 labelled glucose (at final concentration of 1%) for 8 min, followed by extraction of metabolites and quantification of the indicated TCA cycle intermediates by mass spectrometry. In the diagrams, green circles signify the labelled carbon atom and the notation M+1, M+2, etc. indicate the mass increase in the molecules due to the labelled carbon. The depicted labelling pattern of metabolites reflects on cycle of the TCA cycle, resulting in mass additions of M+1 and M+2; however, across multiple cycles, a broader range of metabolite species with different mass additions will emerge. The metabolite signal intensities in all samples are expressed relative to that determined for the first replicate of the WT strain. (B-C) Measurements of ATP levels and proportions of total ATP impaired by azide inhibition of ETC activity. Cells cultured in YPD were treated or untreated with sodium azide for 30 min prior to harvesting. ATP levels were determined in extracts and normalized to OD600 units of cells for three biological replicates each of treated and untreated cell aliquots. Mean values for each strain are plotted in (B) and relative fractions of ATP in untreated samples retained following azide treatment (ATP_untreated)-(ATP_Azide)/ATP_untreated), normalized to the values determined for WT are plotted in (C) with results from a student’s t-test indicated with asterisks: **, P<0.005; *, P<0.05.

Yeast strains employed.

Plasmids employed.

Combining scd6Δ and edc3Δ mutations confer synthetic reductions in cell growth and polysome assembly.

(A) Serial dilutions of WT strain HFY114, edc3Δ strain FZY862, scd6Δ strain SYY2352, and scd6Δedc3Δ strain FZY858 transformed with empty URA3 CEN vector YCplac33 or derivatives of this vector containing EDC3 (pLfz614-7) or SCD6 (pLfz615-5) were spotted on synthetic complete plates minus uracil (SC-Ura) and incubated at the indicated temperatures. (B) Polysome profiles of the strains in (A) but lacking plasmids cultured in YPD medium at 30°C to log-phase growth and treated with cycloheximide prior to harvesting to block run-off of elongating ribosomes during cell lysis. Cell extracts were resolved by sedimentation through 10-50% sucrose gradients, and gradients were scanned at 260 nm to yield the indicated tracings. Average polysome/monosome ratios (P/M) from 2 biological replicates are shown, with means +/-SEMs.

Reproducibility among biological replicates of RNA-Seq, Ribo-Seq, ERCC-normalized RNA-Seq, Rpb1 ChIP-seq, and TMT-MS data.

(A) Correlation matrix showing Spearman correlation coefficients calculated for pair-wise comparisons of numbers of RPKM-normalized RNA-Seq reads for all expressed genes among all 8 RNA-Seq libraries generated for 2 biological replicates (_1,_2) of the indicated genotypes. The correlation coefficients between replicates are ≥0.98 and for all comparisons >0.95 with P-values of ≈0. For this and all similar plots below, the eccentricity of the ellipses are scaled parametrically to the correlation value between the two samples. (B) Same as (A) but for Ribo-Seq data, with correlation coefficients between replicates >0.995 and for all comparisons >0.96 with P-values of ≈0. (C) Same as (A) but for ERCC-normalized RNA-Seq data, with correlation coefficients between replicates >0.97 and for all comparisons >0.94 with P-values of ≈0. (D) Same as (A) but for Rpb1 ChIP-Seq data comparing 3 normalized occupancies averaged across the CDS of all 5770 expressed genes, with correlation coefficients between replicates >0.98 and for all comparisons >0.91 with P-values of ≈0. (E) Same as (A) but for TMT-MS data, comparing log2 cyclic Loess-normalized exclusive MS1 intensities for all expressed proteins with correlation coefficients between replicates >0.96 and for all comparisons >0.95 with P-values of ≈0. (F) Correlation between changes in abundance of RNA vs. RPFs for 5835 expressed transcripts in the scd6Δedc3Δ double mutant compared to WT, calculated from RPKM-normalized RNA-Seq and Ribo-Seq reads obtained after combining data from biological replicates, with the Pearson correlation coefficient (r) and P-value of the correlation indicated.

Functional redundancy between Scd6 and Edc3 in controlling mRNA abundance and mobilizing the ESR.

(A-B) Venn diagrams of overlaps between all mRNA_up (A) and all mRNA_dn (B) groups identified in scd6Δ, edc3Δ, or scd6Δedc3Δ mutants vs. WT, with fold-enrichments and P values from the hypergeometric distribution indicated for overlapping sets. (C) Gene browser image showing the mRNA (top 16 tracks) and RPF (bottom 16 tracks) reads measured by ribosome profiling for two biological replicates of WT and the indicated mutants for CIT1, with fold-changes in mRNA or RPFs between mutant and WT indicated to the right of each track. (D-E) Notched box-plot analyses of log2 changes in mRNA abundance in mutant vs. WT for 585 rESR (C) and 283 iESR (D) mRNAs conferred by scd6Δ, edc3Δ, or scd6Δedc3Δ mutations and those observed for the slowest-growing yeast deletion mutants (M) analyzed previously (9). Each box depicts the interquartile range containing 50% of the data, intersected by the median; the notch indicates a 95% confidence interval (CI) around the median. Median changes (un-logged) are shown at the bottom.

Reproducibility among biological replicates of Rpb1 ChIP-seq data.

Log2 values of relative Rpb1 occupancies averaged over the CDSs for three replicates of WT or scd6Δedc3Δ cells for the mRNAs up-or down-regulated by scd6Δedc3Δ vs. WT, including ESR transcripts. (B) Notched box-plots showing log2 changes in absolute mRNA abundance from ERCC spike-in normalized RNA-Seq (left) or relative mRNA abundance determined by DESeq2 analysis of RNA-Seq results (right) in scd6Δedc3Δ vs. WT cells for all mRNAs or the 554 or 526 non-ESR mRNAs up-or down-regulated, respectively, by scd6Δedc3Δ vs. WT.

Average median codon optimality scores and average median TE values in WT cells for mRNAs repressed in abundance by Scd6/Edc3 or Dhh1/Pat1.

(A) Notched box-plots of log2ΔmRNA between scd6Δedc3Δ (left, cyan) or dhh1Δ (right, orange) vs. WT for 6 bins of all non-ESR mRNAs sorted according to stAI values, designated by the median stAI value for each bin, or for all non-ESR mRNAs (All). (B) tAI, stAI, and average CSC values for the 591 non-iESR mRNAs derepressed in abundance by the scd6Δ, edc3Δ, or scd6Δedc3Δ mutations, the 1018 non-ESR mRNAs up-regulated by the dhh1Δ, pat1Δ, or pat1Δdhh1Δ mutations, or all 5802 non-ESR mRNAs. (C) Log2 values of TE determined in WT cells for the same groups of non-iESR mRNAs derepressed by the scd6Δ, edc3Δ, or scd6Δedc3Δ mutations or by the dhh1Δ, pat1Δ, or pat1Δdhh1Δ mutations analyzed in (A).

Properties of mRNAs translationally repressed by Scd6/Edc3 or Dhh1/Pat1.

(A) Notched box-plots of log2ΔmRNA values between the indicated mutants vs. WT for the translationally up-regulated mRNAs belonging to the specified sectors of the diagram in Fig. 5A. (B) Notched box-plots of log2ΔTE values between the indicated mutants vs. WT for the mRNAs belonging to the 591 non-iESR mRNAs derepressed in abundance by scd6Δ, edc3Δ, or scd6Δedc3Δ vs. WT or the 1018 non-ESR mRNAs up-regulated by dhh1Δ, pat1Δ, or pat1Δdhh1Δ vs. WT. (C) Box-plot of log2 values of TE determined in WT cells for the translationally up-regulated mRNAs belonging to the specified sectors of the diagram in Fig. 5A, the 54 mRNAs showing >2-fold TE increases conferred by both scd6Δedc3Δ and pat1Δdhh1Δ mutations vs. WT, or for all mRNAs. (D) Box-plot of ratios of capped to total mRNA abundance in TPMs (Relative C/T) in WT cells for for the translationally up-regulated mRNAs belonging to the specified sectors of the diagram in Fig. 5A, or for all mRNAs.

mRNAs repressed in abundance or translation by Scd6/Edc3 or Dhh1/Pat1 are enriched for common functional categories, including Ox. Phos. proteins and cell wall components.

(A-B) Functional categories showing enrichment for genes encoding the 591 non-iESR mRNAs derepressed in abundance by scd6Δ, edc3Δ, or scd6Δedc3Δ vs. WT or the 1018 non-ESR mRNAs up-regulated by dhh1Δ, pat1Δ, or pat1Δdhh1Δ vs. WT, using color-coding to indicate related functions or cellular components. (C-D) Functional categories showing enrichment for genes encoding the 853 mRNAs with RPFs up-regulated by scd6Δ, edc3Δ, or scd6Δedc3Δ vs. WT or the 1350 mRNAs with RPFs up-regulated by dhh1Δ, pat1Δ, or pat1Δdhh1Δ vs. WT, conducted using the Web-based tool FunSpec and applying the Bonferroni correction and P<0.05 cutoff.

Scd6/Edc3 post-transcriptionally repress carbon-catabolite-repressed (CCR) genes in rich medium.

(A) Log2 changes in mRNA, RPFs, or TE conferred by the indicated double mutations vs. WT for 106 genes subject to carbon catabolite repression or activated by transcription factors Adr1 or Cat8. (B) Notched box-plots showing log2ΔRPFs or log2ΔRpb1 relative occupancies averaged over the CDSs in scd6Δedc3Δ vs. WT cells for the same Ox-Phos. (left) or CCR genes (right) analyzed in Figures 6A and S7A, respectively. (C) Expression of the ADH2-lacZ reporter on plasmid pLGADH2, containing the entire ADH2 5’ non-coding region, in the WT strain grown on SC-Ura medium containing either 2% glucose or 3% glycerol/2% ethanol as carbon sources (i), or in WT and the indicated mutant strains on SC-Ura cultured with 2% glucose (ii). For (ii), mean values of β-galactosidase activity measured for 3 biological replicates of each strain were normalized to the mean value measured for WT cells. **, P-value <0.01 from student’s t-test; ns, not significant. (D) Log2 changes in RPFs for 93 genes encoding amino acid biosynthetic enzymes in the indicated mutants vs. WT.

Synthetic genetic derepression of four Dhh1 target mRNAs on combining scd6Δ and edc3Δ mutations.

RNA-seq results obtained here for the four Dhh1 target mRNAs, two of which were designated as Pat1 targets as well, identified by He et al. (2022), in the scd6Δ, edc3Δ, and scd6Δedc3Δ mutants versus WT. The log2ΔmRNA values for each of the three mutants vs. WT are listed together with the FDR values and color-coded according to derepression ratios.

Hypothetical model to explain concerted repression of mRNA abundance or translation of particular mRNAs by decapping activators Scd6, Edc3, Dhh1, and Pat1.

A complex of Dcp1:Dcp2 containing Scd6/Edc3, Pat1/Dhh1, or different combinations of these factors, is recruited to target mRNAs, possibly by a sequence-specific RNA-binding protein (RBP) binding to the 3’UTR. In agreement with a recent proposal (10), the recruitment of Dhh1 is enhanced by its redundant interactions with Scd6 or Edc3, which interact with the same segment of the Dcp2 CTT, whereas Pat1 is recruited independently to a distinct region of the CTT. One outcome of association of the decapping complex with an mRNA is activation of decapping with attendant 5’-3’ degradation by Xrn1 occurring without detectable repression of translation (i). A different outcome is translational repression by decapping wherein degradation by Xrn1 is inefficient and decapped intermediates accumulate that cannot bind the cap-binding initiation factors eIF4E and eIF4G (depicted as subunits of eIF4F excluded from the m7G mRNA cap) and thus persist as translationally inert isoforms (ii). Alternatively, decapping may not occur and the decapping complex competes with the cap-binding initiation factors to selectively inhibit translation initiation (iii), a fate that could be favored by sequestration of the transcript in RNA granules in a manner facilitated by the decapping activator proteins (iv).