Figures and data
Enrichment of UTR variants relative to open reading frames.
(A) Metagene analysis of native human UTRs (blue line) or disease-relevant variants reported in ClinVar (orange line). Note the enrichment of UTR variants near the coding region. (B) Metagene analysis of all native human variants (blue line) or disease-relevant variants reported in ClinVar and HGMD (orange line) in relation to uORFs. Note the distinct enrichment of disease-relevant variants at the uORF start site.
Massively parallel polysome assays.
(A) UTR fragments were synthesized in bulk and ligated with EGFP reporters for expression in human cells. The reporter-expressing cell lysate was fractionated into monosome, light (2-5 ribosomes) and heavy (6+ ribosomes) polysome-containing fractions, and the UTR library was then amplified for amplicon sequencing. Mutations that altered relative distribution among fractions were defined as significant polysome-shifting mutations. (B) Examples of UTR pairs that significantly differed in polysome distribution. (C) Enrichment of pathogenic variants among HC-sig polysome-shifting UTR variants.
Validation of mutant effect on translation by luciferase assays.
(A) Selected UTRs were ligated with firefly luciferase reporters and co-transfected with Renilla luciferase into human cells. The firefly luminescence was first normalized to the Renilla luminescence, and further normalized to the respective RNA expression levels to determine translation efficiency. The log2 value of mut over WT translation efficiency was plotted ascendingly. (B) Translation efficiency of eight full-length UTR pairs was examined in the same manner as in (A). UTR pairs are color-coded according to their gene origin (A&B). The individual assays were repeated three times and the differences were accessed by a two-tailed Student t-test: *: p < 0.05, **: p < 0.01, ***: p < 0.001, ****: p < 0.0001 (A&B). (C) Comparison of the effect of mutations on the translation efficiency of library (short) and full-length forms of UTRs.
Identification of polysome-shifting RBP-binding motifs.
(A) Motifs significantly more frequently mutated by HC-sig variants. (B) Two 5’ UTR-enriched motifs from (A), i.e., GTTTTG and GAAGAAG, were inserted into a 5’ UTR of firefly luciferase to determine their luciferase activity and normalized translation efficiency. Mutant forms of these two motifs, GTTAAC and TCATGAC, were generated for comparison. (C) Similar to the 5’ UTR design in (B), the 3’ UTR-enriched motifs CTTTTC and AGGGGGA were examined against their mutant forms, i.e., CATAGT and AGCTGTA. The individual assays were repeated four times. Statistical significance was determined by a two-tailed Student t-test: **: p < 0.01, ***: p < 0.001, ****: p < 0.0001. (D) Upper: Relative SRSF2 transcript levels, as determined by RT-qPCR. Lower: Relative luciferase activity with various 5’ UTR motifs, normalized to scrambled shRNA infection.
5’ UTR Polysome-shifting mutations elicit structural changes.
(A) Significantly weaker folding energy of 5’ UTR polysome-shifting variants. The left panel is 5’ UTR ΔG and the right panel is 3’ UTR ΔG. (B) Structural change induced by a CCR5 5’ UTR single nucleotide change, as determined by RNAfold. (C) Significant differences in ΔΔG values for 5’ UTRs that contain upstream ATG (uATG).
Elastic model distinguishes polysome-shifting UTR variants.
(A) Selected features from elastic regression explaining the likelihood of a polysome-shifting UTR variant. A positive coefficient indicates a feature that promote polysome shifting, while a negative coefficient suppresses it. “mut” represents a feature of the mutant UTR, “ref” represents a feature of WT UTR, and “diff” indicates the difference (delta) associated with the variant. (B) Receiver operating characteristic (ROC) curve for the prediction of polysome-shifting variants on the test set of UTR variants.
5’ UTR mutations cause translation inhibition by generating upstream ATG
(A) Luciferase activity of eight UTRs with or without a uATG in the 5’ UTR. Error bars represent standard errors. (B) Constructs of the mutant IRF6 5’ UTRs. Mutant (mut) IRF6 generates an in-frame luciferase protein from an upstream ATG, whereas the uORF and luciferase are coded in different frames for the out-of-frame mutant (mut out-of-frame). (C) Western blots showing a reduction in luciferase protein levels following 5’ UTR mutations. (D) Luciferase assays showing a reduction in translation efficiency due to 5’ UTR mutations. RLU: relative luminescence units. The individual assays were repeated three times. Statistical significance was determined by a two-tailed Student t-test against WT: ***: p < 0.001. (E) Design of guided ADAR-mediated editing to remove a uORF. (F) Editing efficiency achieved by guided ADAR, as determined by Sanger sequencing. (G) Restoration of translation efficiency by means of guided ADAR editing, as determined by luciferase assays.
