Fig. 1

Analysis of MRE regulatory landscape at single-nucleotide resolution. a MRE reporter library diagram. Values indicate the proportion of nucleotides at each position in the MRE (shaded squares = nucleotides complementary to miR-17). b MRE regulatory landscape analysis pipeline. c Impact of MRE variants on transcript abundance. Bar graph shows relative contribution of each nucleotide to MRE function, as determined by high-throughput sequencing (n = 3 biological replicates, mean + s.d.; dashed line = expression of a perfectly complementary MRE, solid line = expression of a non-targeted MRE- Cel-miR-67). Heat-map displays the effect of each possible mismatch by position and reflects the mean of three replicates (complementary bases are displayed in black). d The impact of di-nucleotide substitutions on reporter expression (mean of 3 biological replicates; colour scale is the same as in c; grey box = seed region). e Schematic representation of the two major pathways underlying miRNA-mediated repression. f Polysome profiles generated by sucrose gradient fractionation. Blue trace denotes the spatial distribution of RNA across the gradient as monitored by UV absorbance. Analysed fractions (monosomes and heavy polysomes) are shaded in grey (representative of two biological replicates). g Correlation between translational efficiency and transcript stability for all single nucleotide miR-17 MRE variants. P value indicates that the slope of a linear regression model (black diagonal line) significantly differs from 0 (n = 69 variants). h Linear regression comparing expression measured by high-throughput sequencing with expression measured by RT-qPCR in HEK-293T cells transfected with each of the 15 MRE variants in the validation set (P < 0.0001, slope differs from 0, n = 17 variants). Expression was calculated by the ∆∆CT method using iBlue as a reference gene. i Linear regression comparing variant expression measured by high-throughput sequencing and flow cytometry (flow cytometry expression was calculated by normalising ECFP levels to iBlue levels on a single cell basis and taking the mean of that value for each MRE variant) (P < 0.0001, slope significantly differs from 0, n = 17 variants). Source data are provided as a Source Data file