Evaluation of synthetic post-transcription regulatory sequences reveals design principle to enhance mRNA stability and translation efficiency

Figures

• 

  Fig. 1. Expression regulation of GFP by PTRCs with different structures. The effect of TSS with varied stem length (A) and varied loop size (B) on protein expression. (C) The impact of the insertion of linear NCS on protein expression. (D) The effect of different PTRC on 15-min post-transcription mRNA survival rate. Data are mean ± s.d. from n = 3 biologically independent replicates. Statistical significance was assessed by unpaired two-tailed Student's t-test: ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; ns, not significant.

• 

  Fig. 2. The PTRC library construction and FlowSeq experimental scheme, as well as the collections of cell pools. (A) The random assembled DNA fragments were ligated upstream of a reporter gene for transformation into E. coli, thereby generating a transformed plasmid library. GFP, serving as a reporter gene, was used to assess the regulatory efficacy of the post-transcriptional regulatory assembly. The cells containing randomly assembled variants were sorted into three bins based on their fluorescence intensity. The DNA sequence of PTRC region and the corresponding reverse transcription PCR product of transcripts were subjected to high throughput sequencing. (B) Distribution of PTRCs across Bins. (C) The weighted fluorescence intensity (GFPScore) of each post-transcriptional regulatory component is calculated based on the distribution of each variant among three bins and the GFP intensity of each bin; the inset is the variants showed bias towards a specific bin, for instance, if more than 40 % reads of a variant was in bin1, this sequence was grouped into Low group. (D) Based on the relative abundance of transcript of each PTRC variant, all PTRC variants were classified into R-High,…

• 

  Fig. 3. The distribution of each part in groups with varied protein level and transcript abundance level. (A) The distribution of each part in the High, Med and Low groups (varied protein level). (B) The distribution of each part in the R-High, R-Med, and R-Low groups (varied mRNA level). Each ternary map shows all 576 PTRC variants as individual dots; the corresponding color code is given beneath the stacked bars on the left.

• 

  Fig. 4. Comparison of the effects of each part on relative abundance of transcript and GFPScore.

• 

  Fig. 5. Parts occurring in 576 PTRCs were grouped to resolve their regulatory levels. (A) All PTRCs in the library were quantified and ranked in quartiles according to GFPScore and relative abundance of transcript. The upper and lower quartiles for GFPScore were set at 21337 and 14942, respectively, while those for relative abundance of transcript were set at 3.19 and 0.82, respectively. Totally four typical PRTC variants were obtained, including: high mRNA abundance (H-H), high GFP expression and low mRNA abundance (H-L), low GFP expression and high mRNA abundance (L-H), and low GFP expression and low mRNA abundance (L-L). (B) The top 10 % of PTRCs in H-H, L-L, H-L and L-H group. (C) The frequency of each part appeared in H-H, L-L, H-L or L-H group. (D) The coefficient of variation of each part is calculated to quantify its background dependence, and a random forest model is constructed, with GFPScore and the relative abundance of transcripts regulated by each part as input data, and feature importance (FI) as output data.

• 

  Fig. 6. Regulation of nrk and sam2 genes using selected PTRCs. (A). Ten PTRCs were selected to regulate the nrk gene to improve yield and transcript abundance. (B). Nine PTRCs were selected to regulate the sam2 gene to improve yield and transcript abundance.