This paper was a fascinating combination of theory and application. It began with the finding that a gene we were trying to express in E. coli displayed very poor overall expression, while also displaying an unusual artefact in the DNA sequencing chromatogram. We deduced that the artefact owed to formation of secondary structure in oligonucleotides through intrachain hydrogen bonding akin to that seen in tRNA, and concluded that similar structure(s) could also form in messenger RNA resulting in the prevention of ribosome binding and translation. We also found that the literature of the seventies and early eighties referred to this artefact as a ‘compression artefact’, i.e., it had long been recognized that secondary structure forms in oligonucleotides generated and analysed during DNA sequencing, with high temperatures (60 deg C) and formamide used to break such secondary structures during sequencing, especially after the advent of automated DNA sequencers. We had just been lucky that our DNA sequencing chromatogram had displayed such an anomaly. What was remarkable was that we managed to analyse and break the secondary structure through the introduction of three silent mutations which raised the expression levels of the gene by a factor of more than 300-fold. We believe that to discover secondary structures in RNA, therefore, one must conduct automated DNA sequencing of the DNA encoding the RNA under sub-optimal conditions by allowing compression artefacts to form through the lowering of column oven temperatures and exclusion of formamide. We show that this method of sequencing DNA under suboptimal conditions works, and that it produces actual information regarding sites of secondary structure formation in RNA which hold up translation, and which can be subjected to silent mutations that result in increased efficiency of translation of mRNA. The method potentially works even better than any mRNA secondary structure-prediction computer program.

Click to Read in Detail