Fig. 4: Hairpin and RG4 structures synergistically promote translation.

a Schematic of the 5′UTR. Domain 1 is upstream of the 5′UTR and the position for hairpin insertion. Domain 2 is the potential G-quadruplex sequence (PQS) location. Domain 3 is downstream of PQS. Domain 4 begins at the ribosome binding site (RBS) and ends at +9 of translation codons. The predicted mRNA folding energies are provided in Supplementary Table 1. b Translation efficiencies of non-template constructs containing different hairpins (hp) and RNA G-quadruplex (RG4) in the 5′UTR. The constructs are grouped according to hairpin stem length, ranging from short (left) to long (right). The 0 hp data set was copied from Fig. 3c and normalized to 10 hp-control. The x-axis of each group indicates the RG4 sequence in domain 2, arranged in increasing total loop length. Translation efficiencies for all constructs are normalized to the 10 hp-control. Data are represented as mean ± SEM of n = 3–16. Raw data points and exact mean value are provided as a Source Data file. c The correlation between translation efficiency and total loop length. The data are grouped according to  hairpin stem length, and each data point is normalized to the control within the group. Strong correlations are observed between translation efficiency and loop length for all hairpin structures, indicating that dependence on domain 2 RG4 remains regardless of domain 1. The regression line was plotted with 95% line confidence band. d Heat map of the mean translation efficiencies from (b). The x-axis is RG4 arranged in order of longer loop lengths, and the y-axis is the length of hairpin stem. The color represents the mean translation efficiency and is scaled from 0 (white) to 12 (dark red). The color trend demonstrates the synergistic dependence of hairpin and RG4 size on translation efficiencies.