9.3 Mechanisms 183of individual mutants. The Gallivan and the Hartig groups, among others, have
taken this approach by evaluating hundreds to thousands of riboswitch clones
by reporter gene assay [13–15]. Although more laborious and costly compared
with genetic selection, screening of individual clones provides quantitative
characteristics of every mutant evaluated (ON and OFF expression levels).
More recently, fluorescence‐activated cell sorting (FACS) has been used to fur-
ther increase throughput [16, 17].
9.2.3 Rational Design
Despite the extensive research on the natural riboswitch mechanisms and struc-
tures, rational or computational design of synthetic bacterial riboswitches has
been few and far between. An earlier example by Suess et al. highlighted the
potential of rationally engineering ligand‐induced structural shift to regulate
bacterial gene expression [18]. More recently, computationally driven designs of
bacterial RNA switches based on ribozymes [19] and transcriptional regulation
[20] have emerged. However, some level of experimental feedback is still expected
to be essential due to the complexity of parameters that influence the perfor-
mance of these RNA devices in living cells.
9.3 Mechanisms
9.3.1 Translational Regulation
Translational regulation by bacterial riboswitches involves a change in the
local structure of the ribosome binding site (RBS) upon ligand binding. RBS
within a stable structure generally hinders ribosome access and results in
repressed translation. Engineering such ligand‐induced structural changes,
however, is not trivial, and screening or selection is often used in the process.
In some cases, naive randomization of the nucleotides peripheral to the RBS
followed by screening or selection was sufficient for isolation of suitable
expression platforms [9, 13]. In other cases, riboswitch libraries were carefully
designed to predispose the riboswitch mutants to undergo a specific structural
shift. An example of the latter is shown in Figure 9.1a where the RBS was stra-
tegically placed to form a putative stem at the base of the aptamer upon ligand
binding so that the riboswitch negatively responds to the aptamer ligand [21].
In another strategy, the RBS was placed so that it becomes accessible only
when the hammerhead ribozyme self‐cleaves, and the aptamer was inserted
in one of the stem‐loops of the ribozyme to control its activity (Figure 9.1b)
[14, 15, 22]. In this strategy, because the translation efficiency is directly coupled
to the ribozyme activity, small molecule response is actually engineered at the
level of the aptamer–ribozyme hybrid, or aptazyme. The Hartig group has
exploited the aptazyme strategy further by adapting them to control other
translational components such as tRNA [23] and rRNA [24] to construct small
molecule‐responsive RNA switches in E. coli.