10.2 Synthetic Control of Transcript Stability 197to almost 20 min [81]. Similarly, 3′ hairpin introduction has been shown to
inhibit degradation [70] and increased the penicillinase (penP) transcript half-
life threefold in both E. coli and Bacillus subtilis [71].
These results demonstrated that transcript secondary structures inhibiting
RNase E and exoribonuclease binding are useful tools for varying mRNA half-life
in a static manner. Despite these successes, however, it has been difficult to con-
trol transcript stability in a quantitatively predictable manner through secondary
structure engineering [72, 82]. In principle, it should be possible to develop more
explicit design rules if the relationships between secondary structure folding
kinetics, stability, and RNase E binding occlusion can be further developed.
Cambray et al. combined experiment with kinetic RNA folding simulation analy-
sis of a large number of transcriptional terminators to derive heuristics for relat-
ing sequence and structural features to termination efficiency [83]. Similarly, by
testing the effect of different hairpin structures on mRNA stabilities in multiple
transcript contexts, it may be possible to identify rules to understand how
the sequence and structure of a given hairpin affects half-life. Furthermore, as
RNase III [34] or helicase activity [32] may mitigate the stabilizing effects of sec-
ondary structure, more study of RNA sequence and structure interactions with
these enzymes should lead to better genetic design predictability.
10.2.3 Noncoding RNA-Mediated
ncRNA has been used for TSC in two related forms, namely, sRNA and asRNA.
Both sRNA and asRNA act via an antisense mechanism and base-pair with a
region – usually the 5′ UTR – of a target mRNA (Figure 10.2b). In one well-known
example, asRNA was derived from the RNA-IN/RNA-OUT system from the
insertion sequence IS10 in E. coli [84]. There, the RNA-IN antisense hairpin
binds to the RNA-OUT portion of the target mRNA. Although engineered sRNA
mechanisms have originated from distinct naturally occurring ncRNA systems,
both asRNA [85] and sRNA target [74, 86] base pairing has been shown to be
Hfq-mediated in vitro, so it is likely that asRNA and sRNA functions are mecha-
nistically similar.
Substantial progress has been made in the past few years toward developing
sRNA and noncoding trans-RNA as avenues for controlling gene expression. In
2011, Man et al. developed initial design principles for creating novel sRNAs
with Hfq-binding sites and regions targeting enhanced green fluorescent protein
(EGFP) and a native E. coli gene [10]. They tested 16 such sRNAs and reported
relative expression knockdown levels ranging from 6% to 71%. Furthermore, they
showed sRNA-dependent target mRNA half-life reduction and used a tempera-
ture-sensitive RNase E mutant to establish the RNase E dependence of target
transcript level reduction. Surprisingly, the reduction in gene expression was
unaffected by the presence or absence of RNase E, suggesting that sRNA binding
alone was sufficient to reduce translation and that TSC does not play a dominant
role in this system (see also [60]).
Sharma et al. randomized the antisense seed portion of the E. coli sRNA
Spot42 [73] to screen for sRNAs that downregulate a natively targeted gene.
After a single round of screening, sRNAs were identified that downregulate a