204 10 Programming Gene Expression by Engineering Transcript Stability Control and Processing in Bacteria
authors attribute to either premature transcript termination due to the cis-
repressor secondary structure or the activities of RNases, such as RNase III, that
cleave double-stranded RNA.
Elsewhere, the space of trans-activating RNA targeting a fixed cis-repressor
RNA, regulating GFP expression in E. coli, has been computationally explored
[108]. In the cis-repressed state, where the level of genetic output was equivalent
to 1–4% of the unrepressed state, activation by one of six designed trans-activat-
ing RNAs increased GFP production 3–11-fold relative to the baseline.
Quantitative reverse transcription polymerase chain reaction (RT-PCR) showed
that the trans-activating RNA–mRNA ratio did not change in an RNase III
knockout strain compared with the wild type. However, the relative genetic out-
put induced by trans-activating RNA more than doubled in the RNase III knock-
out, which, as mentioned, is consistent with the idea that variations in transcript
stability can alter the performance characteristics of these kinds of control
devices and systems.
A riboregulator-like RNA, called an allosteric ribozyme, previously only char-
acterized in vitro [109–111], has recently been used to control translation initia-
tion in vivo [112]. A ribozyme, with the RBS sequestered in its secondary
structure, was designed to autocatalytically cleave itself to expose the RBS and
allow translation initiation. Trans-activating RNAs were designed to bind a com-
plementary sequence within the ribozyme, inhibiting ribozyme cleavage and
exposure of the RBS, leading to 10-fold reductions in relative EGFP expression.
The question of how variations in transcript stability might be in play here has
not been directly investigated.
Overall, the work described here highlights promising approaches for engi-
neering dynamic RNA-based control systems. It is also clear that, to improve
engineering tractability, there is a need to investigate how introducing secondary
structure – whether a cis-repressor RNA, riboswitch, or ribozyme – into the
5 ′ UTR may lead to confounding effects on device outputs stemming from unac-
counted-for RNase binding or premature transcription termination.10.3.5 Experimentally Probing Transcript Stability
The determinants of synthetic transcript stability can be analyzed by experimen-
tally measuring mRNA half-life, through expression studies, by the use of endo-
nuclease gene knockout strains, and with computational RNA folding simulations.
Quantitative gel electrophoresis [21], quantitative PCR, or RNA-seq [1] after the
addition of a transcription-inhibiting antibiotic (e.g., rifampicin) can be done at
intervals to determine average transcript half-life by quantifying transcript abun-
dance as a function of time. A strategy using sRNA to quantify mRNA abun-
dance changes has also been proposed [113]. Comparing measurements from
cells with and without an RNase (via knockout) can lend insight into the RNase
dependence of a phenomenon, though care should be taken to understand the
global impact of an RNase deletion and how that may complicate data interpreta-
tion. Folding simulation tools for calculating minimum free energy (MFE) sec-
ondary structures [114–117] or kinetically driven co-transcriptional [118–120]
folding trajectories can lend insight into whether secondary structure could be