高通量分析信使核糖核酸的二級結構對轉譯起始的影響

Abstract

合成生物學致力於組裝來自於不同生物體的元件以建造嶄新的基因或代謝途徑。要達成這樣的目的，準確調控連接各個迴路部分的蛋白質表現量非常關鍵。先前的研究已建立根據編碼區上游的信使核醣核酸的序列來預測蛋白質的轉譯速率的熱力學模型，雖然這些模型有考慮到信使核醣核酸的折疊能量，卻忽略其確切的結構型態對轉譯的影響。為了解信使核糖核酸二級結構對轉譯的影響，我做出綠色螢光蛋白的Shine-Dalgarno序列周邊9個核苷酸的所有排列可能的質體。將這些質體送進大腸桿菌之後，我用流式細胞術和次世代定序去量化各個序列種類的轉譯速率。我的結果顯示之前宣稱具有準確預測能力的熱力學模型實際上預測很差。具有髮夾結構將Shine-Dalgarno序列遮蔽的信使核糖核酸能顯著抑制轉譯。帶有一致周邊序列但不同的Shine-Dalgarno序列所造成的二級結構效應也不相同。藉由釐清被現有模型忽視的要素，我的研究可增進我們對細菌轉譯的根本認知、改善計算預測的準確性並加速達到合成生物學的目標。

Synthetic biology aims for building novel genetic or metabolic pathways by assembling elements from different organisms. To this end, precise control of the protein expression levels connecting each part of the circuit is crucial. Prior works have established thermodynamics models that predict translation rate of a protein based on the mRNA sequence upstream of the coding region. Although these models consider the mRNA folding energy, they neglect the influence of exact mRNA conformation on ribosome binding. To identify the influence of mRNA secondary structure on translation, I applied saturated mutagenesis to the 9 nucleotides flanking the Shine-Dalgarno (SD) sequence of a green fluorescent protein (GFP) gene on the plasmids. Upon transformation into Escherichia coli, I used fluorescence-activated cell sorting and next-generation sequencing to profile the translation rate of each sequence variant. My results show that the thermodynamic model, formerly claimed to be accurate, actually makes poor prediction. Particularly, mRNA hairpins which sequester the SD sequence inhibit translation significantly. The flanking sequences also have different effects on different SD sequence background. By identifying factors ignored by the current model, my study may advance our fundamental understanding of protein translation in bacteria, improve the accuracy of computational prediction, and facilitate the goal of synthetic biology.