SARS-CoV-2 5’-UTR Is an Unexpectedly Efficient Promoter of Translation

Secondary structure and minimum free energy of A. 4X repeat of Hbb 5’-UTR and SARS-CoV-2 5’-UTR (bp 1-265).B. Adjusted Minimum Free Energy for human 5’-UTRs with GC Content of ∼50% and Length between 150-280bp. Hbb (62% GC) and CoV-2 (47% GC) marked in blue and red respectively. C. Luciferase assay from pGL3-Hbb-fLuc and pGL3-CoV-2 5’-UTR-fLuc transfected HEK293T cells. Results of two experiments (3 biological replicates each) shown as mean±SEM. Statistical testing was performed using Welch’s T-test.

m6A enhances SARS-CoV-2 5’-UTR Initiated Translation

A. Experimental design for m6A-RNA Immunoprecipitation (ME-RIP) and RT-qPCR for SARS-CoV-2 5’-UTR. B. ME-RIP-RT-qPCR enrichment of CoV-2 5’-UTR from HEK293T cells transfected with pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc or pGL3-CoV-2-5’-UTRC75G (SL1-TSS)-fLuc treated with and without METTL3 siRNA knockdown. Results of two experiments (3 biological replicates each) normalized to input fractions shown as geometric mean; error bars are geometric SD. Statistical testing was performed using Tukey’s HSD under two-way ANOVA. C. Sequences of putative m6A site in SARS-CoV2 WT and engineered m6A mutant variants and predicted DG for respective 5’UTRs (1-265). D. Luciferase assay of pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc or pGL3-CoV-2-5’-UTRC75G (SL1-TSS)-fLuc or pGL3-CoV-2-5’-UTRC75G+G61C (SL1-TSS)-fLuc transfected HEK293T with or without METTL3 siRNA knockdown. Results from three experiments (3 biological replicates each) shown as mean ± SEM. Statistical testing was performed using Tukey’s HSD under two-way ANOVA E. Luciferase assay of pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc or pGL3-CoV-2-5’-UTRC75G (SL1-TSS)-fLuc HEK293T treated with increasing concentrations of METTL3 Inhibitor STM2457. Results from three experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing was performed using Bonferroni-Šidák correction under two-way ANOVA F. Luciferase assay of pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc, pGL3-CoV-2-5’-UTRC75G+G61C (SL1-TSS)-fLuc, pGL3-CoV-2-5’-UTR (SL1-5)-fLuc, and pGL3-CoV-2-5’-UTRC75G+G61C (SL1-5) -fLuc. Results from three experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing was performed using Tukey’s HSD under two-way ANOVA. G. Luciferase assay of pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc or pGL3-CoV-25’-UTRC75G (SL1-TSS)-fLuc or pGL3-CoV-2-5’-UTRC75G+G61C (SL1-TSS)-fLuc transfected HEK293T with or without METTL3 and/or DF1/2 siRNA knockdown. Results from three experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing was performed using Tukey’s HSD under two-way ANOVA.

m6A Inhibition Impairs SARS-CoV-2 5’-UTR Promoted Translation Initiation

Ribosome profiling of HEK293T cells transfected with A. pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc or B. pGL3-CoV-2-5’-UTR (SL1-TSS)C75G-fLuc with METTL3 siRNA knockdown. Relative distribution (%mRNA) of fLuc (top) and hprt (bottom) determined by RT-qPCR in fractions 5-19 from C. pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc D. pGL3-CoV-2-5’-UTRC75G (SL1-TSS)-fLuc transfected HEK293T cells. Results from three experiments (3 biological replicates each) shown as mean±SEM. Statistical testing performed using two-way ANOVA with Šídák’s multiple comparisons test. Representative semi-quantitative low cycle number PCR for fLuc from E. pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc F. pGL3-CoV-2-5’-UTRC75G (SL1-TSS)-fLuc transfected HEK293T cells.

m6A Methylation Promotes SARS-CoV-2 5’-UTR Accessibility

A. Schematic of SARS-CoV-2-5’UTR segments used in this study. B. Luciferase assay of pGL3-CoV-2-5’-UTR (SL1-3)-fLuc, pGL3-CoV-2-5’-UTR (SL1-4.5)-fLuc, pGL3-CoV-2-5’-UTR (SL1-TSS)-fLuc, pGL3-CoV-2-5’-UTR (SL1-5)-fLuc, with or without METTL3 siRNA knockdown or mutation at A74T+T62A or C75G+G61C. Results from three experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing was performed using Tukey’s HSD under two-way ANOVA. C. Calculated minimum free energy structure for SARS-CoV2 (SL1-5) RNA under native (37°C) or denaturing (55°C) conditions. D. RT-qPCR of CoV-2-5’-UTR (SL1-3)-fLuc, CoV-2-5’-UTR (SL1-4.5)-fLuc, CoV-2-5’-UTR (SL1-TSS)-fLuc, CoV-2-5’-UTR (SL1-5)-fLuc cDNA generated under denaturing (55 °C, top) and native (37 °C, bottom) with or without METTL3 siRNA knockdown or m6A mutation, normalized to hprt mRNA level. Results of three experiments (3 biological replicates each) normalized to CoV-2-5’-UTR (SL1-5)-fLuc and shown as geometric mean; error bars are geometric SD. Statistical testing was performed using Tukey’s HSD under two-way ANOVA.

SARS-CoV-2 5’-UTR is Highly Conserved Among SARS-CoV-2 Variants

A. Genome browser tracks showing Nextstrain recurrent mutations in SARS-CoV-2 5’-UTR (top) and SARS-CoV-2 S Glycoprotein (bottom). Non-synonymous mutations in protein coding regions are shown in red, and synonymous mutations are shown in green. B. Lineage-specific SARS-CoV-2 5-’UTR mutations for all WHO variants of interest and variants of concern mapped to predicted secondary structure. C. Lineage-specific multiple alignment of consensus sequences to reference SARS-CoV-2 genome (Wuhan-Hu-1; NC_045512.2) for all WHO variants of interest and variants of concern. Polymorphism sites are highlighted in green, with mutant allele highlighted in red.

m6A Methylation Does not Alter mRNA levels of pGL3-SARS-CoV-2 5’UTR-fLuc

A. qPCR measurement of HBB-fLuc and CoV-2 (SL1-TSS)-fLuc mRNA in transfected HEK293T cells. Results from 2 experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing performed using Welch’s T-Test. B. qPCR measurement of CoV-2 (SL1-TSS)-fLuc mRNA with or without METTL3 siRNA or C75G/G61C mutations in transfected HEK293T cells. Results from 2 experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing performed using two-way ANOVA with Šidák’s multiple comparison test. C. qPCR measurement of CoV-2 (SL1-TSS)-fLuc mRNA with or without C75G mutations in HEK293T cells treated with STM2457. Results from a single experiment (3 biological replicates) shown. D. qPCR measurement of CoV-2 (SL1-TSS)-fLuc or CoV-2 (SL1-5)-fLuc with or without METTL3 siRNA or C75G+G61C or A74T+T62A mutation as indicated in transfected HEK293T cells. Results from 3 experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing performed using two-way ANOVA with Tukey’s HSD. E. qPCR measurement of CoV-2 (SL1-TSS)-fLuc with or without METTL3 and DF1/2 siRNA or C75G+G61C or A74T+T62A mutation as indicated in transfected HEK293T Cells. Results from 2 experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing performed using two-way ANOVA with Tukey’s HSD. F. qPCR measurement of CoV-2 (SL1-3)-fLuc, CoV-2 (SL1-4.5)-fLuc, CoV-2 (SL1-TSS)-fLuc, or CoV-2 (SL1-5)-fLuc with or without METTL3 siRNA in transfected HEK293T cells. Results from 3 experiments (3 biological replicates each) shown as geometric mean; error bars are geometric SD. Statistical testing performed using two-way ANOVA with Šidák’s multiple comparison test. n.s. indicates p-value > 0.05. For all panels hprt was used as internal normalization control.

m6A Methylation Destabilizes SARS-CoV-2 5’-UTR Secondary Structure

A. m6A-dependent minimum free energy secondary structure prediction of stem loop 3 using RNAstructure with m6A alphabet. B. m6A-dependent minimum free energy secondary structure prediction of stem loops 2+3 using RNAstructure with m6A alphabet. C. Minimum free energy secondary structure prediction of acta2, cox8a, and hprt 5’-UTRs at 37°C and 55°C. D. cDNA abundance of acta2, cox8a, and hprt mRNA from HEK293T cells with or without METTL3 siRNA knockdown reverse transcribed under native (37°C) or denaturing (55°C) conditions normalized to hprt. Multiple t-test with Welch correction, and multiple testing adjustment with Holm-Šídák method