190 10 Programming Gene Expression by Engineering Transcript Stability Control and Processing in Bacteria
could, therefore, increase the speed with which synthetic biological systems can
be created for applications in basic science; for the production of renewable
chemicals, fuels, and materials for global health; and for the development of new
therapeutic agents. Even when transcript stability is not an explicit aspect of a
given genetic control device (e.g., ribosome binding site (RBS) control of transla-
tion initiation), unknown or poorly characterized effects on transcript degrada-
tion may affect genetic device outputs. It is therefore important to regard
transcript stability through two lenses: as a “tuning knob” for predictably con-
trolling gene expression dynamics and as a confounding factor if unaccounted
for in genetic device design.
In this chapter, we describe current understanding of transcript stability and
processing for designing and engineering genetic expression devices with
predictable functions. In Section 10.1, we consider the machinery that controls
transcript stability within bacteria, with specific focus on Escherichia coli.
Section 10.2 examines efforts to utilize this machinery for controlling gene
expression dynamics. In Section 10.3, we consider ways of managing transcript
stability to reduce unintentional and confounding effects. Section 10.4 details
possible strategies for controlling transcript stability and points to future
research directions in computation and wet-lab experimentation that may lead
to design technologies for rapidly engineering genetic devices. The final section,
Section 10.5, will provide a summary of the chapter.10.1.2 The RNA Degradation Process in E. coli
RNA is degraded through multistep pathways that can begin as soon as a
transcript has been synthesized by an RNA polymerase. Degradation of mRNA
typically begins with a rate-limiting, RNase E-mediated phosphodiester bond
cleavage event. RNase E cleavage is followed by subsequent rounds of 3′ → 5 ′
degradation (E. coli has no known 5′ → 3 ′ exoribonuclease) [4], carried out in
concert with the degradosome, a collection of four enzymes – RNase E, RhlB,
PNPase (polynucleotide phosphorylase), and enolase [3, 12–14] – that localizes
to the membrane [15, 16].
RNase E [17], an endoribonuclease and a rate-limiting cleavage enzyme, is
thought to bind and process transcripts via two mechanisms (Figure 10.1), the
first of which is 5′ entry at a monophosphorylated end. It was discovered to
prefer substrates with unpaired 5′ ends in vivo [18], and early in vitro analysis of
RNase E activity showed a manyfold reduction in cleavage rate when three dif-
ferent RNAs were 5′ triphosphorylated (5′-PPP) instead of 5′ monophosphoryl-
ated (5′-P) [19]. The structure of the RNase E domain was later shown to have a
binding pocket that cannot accommodate substrates larger than a 5′-P [20],
explaining the selectivity toward 5′-P- versus 5′-PPP-terminated transcripts and
the increased half-life of 5′ hydroxyl (5′-OH)-terminated transcripts [21]. As
transcripts are synthesized natively with 5′-PPP, it was hypothesized, and later
shown, that 5′-P RNAs are created in cells through the removal of the gamma-
and beta-phosphate from 5′-PPP RNAs [21]. This conversion, which creates the
direct substrates for RNase E cleavage, was found to be catalyzed by RppH, an
RNA pyrophosphohydrolase [22].