198 10 Programming Gene Expression by Engineering Transcript Stability Control and Processing in Bacteria
natively targeted gene with 45–145-fold repression – an improvement over the
27-fold repression of the native sRNA. After two rounds of screening for sRNAs
against a gene with no native sRNA, sRNAs were identified that could repress the
relative level of gene expression 23–85-fold.
Most recently, Ishikawa et al. [74], Park et al. [53], and Na et al. [75] have taken
systematic approaches to uncover design principles for sRNAs that are effective
at repressing gene expression. Ishikawa et al. studied the SgrS sRNA in E. coli
using mutational analysis and Northern blotting to elucidate the Hfq-binding
motif of that sRNA. This motif was incorporated into artificial sRNA against
three mRNA targets that showed orthogonal Hfq-dependent knockdown via
Northern blotting. The authors speculate that any mRNA can be effectively
targeted by designing sRNAs with at least 14 nucleotides (nt) of sequence
complementarity to the RBS and a cis Hfq-binding motif located within 10 nt. Na
et al. screened native sRNA scaffolds and potential mRNA target binding sites
around the SD region and found that a MicC sRNA scaffold with a binding site
spanning the first 21 nt (not including the SD region) of the target mRNA was
particularly effective. Using that insight, sRNAs were developed to target native
genes in a microbial platform engineered to produce l-tyrosine and cadaverine.
In both cases, the authors were able to employ sRNA-mediated genetic repres-
sion to divert metabolic flux and increase product formation in the engineered
system. Collectively, these sRNA design studies suggest that an Hfq-binding,
scaffold-based sRNA platform may provide a means of downregulating gene
expression predictably, as binding energy of the antisense region is strongly
correlated with repression capacity [75].
In addition to sRNA, at least two studies have utilized IS10-based asRNA
against the 5′ UTR. The first study built a model from 529 possible combina-
tions of 23 sense and antisense pairs (termed RNA-IN and RNA-OUT), which
was then used to forward-engineer new regulators [76]. A second study built
upon the RNA-IN and RNA-OUT system by adding a theophylline aptamer-
based domain upstream of the RNA-IN asRNA, which functions similarly to a
riboswitch in that gene expression is controlled through structures modulat-
ing ribosomal access to the RBS. Several designed mutants were screened to
find an aptamer–RNA-IN pseudoknot interaction that impaired the RNA-IN
asRNA’s ability to bind its RNA-OUT partner when the aptamer domain was
not bound to its ligand [87]. This provides another means of dynamic control
and, given the similarities between asRNA and sRNA, indicates that sRNA
could be engineered for dynamic control by appending ligand-binding aptamer
domains.10.2.4 Model-Driven Transcript Stability Control for Metabolic
Pathway Engineering
Ribozyme-catalyzed phosphodiester bond cleavage can affect mRNA half-life in
primarily two ways (Figure 10.3b). Depending on the sequence context, and
whether the target site is within a 5′ or 3′ UTR, cleavage may remove or alter
secondary structures that influence RNase E or ribosome docking, resulting
in differences in transcript stability, and, potentially, levels of gene expression