192 10 Programming Gene Expression by Engineering Transcript Stability Control and Processing in Bacteria
they are 5′-PPP- or 5′-OH-terminated RNAs), thought to occur when RNase E
binds directly to unpaired regions in the coding sequence [26].
Once RNase E is associated with the RNA, either through 5′-P binding or
direct entry, it can scan the transcript and catalyze cleavage at the initial target
site [27, 28], setting in motion the recruitment of the other members of the
degradosome and further degradation by 3′ → 5 ′ exoribonucleases and subse-
quent rounds of RNase E activity [29]. PNPase, a 3′ → 5 ′ exoribonuclease, binds
to polyadenylated 3′ mRNA ends [30]. RhlB is an adenosine triphosphate (ATP)-
dependent helicase implicated in preparing RNA for RNase E and PNPase cleav-
age by removing secondary structure. When inhibited, RhlB no longer enhanced
PNPase-mediated degradation in an ATP-dependent way [31], and when RhlB
was deleted, lacZ mRNA was stabilized in a ribosome-free context by impaired
RNase E cleavage at the 5′ end [32]. Enolase is the least well-understood member
of the degradosome and is thought to have a role in metabolism-related tran-
script degradation [33].
Additional means of initiating degradation occur through RNase III cleavage
and RNase G cleavage. RNase III is thought to primarily bind and cleave second-
ary structures [34], often in the context of rRNA maturation and decay [35].
RNase G, an RNase E homolog, is usually involved in 9S rRNA maturation but in
a small number of cases initiates mRNA decay as well [36].
While this is not an exhaustive account of the mechanisms related to RNA
decay in E. coli, the aforementioned mechanisms are responsible for the majority
of messenger RNA decay [3] and are the most salient for programming variations
in gene expression levels.10.1.3 The Effects of Translation on Transcript Stability
The development of a complete mechanistic understanding of RNA degradation
has been complicated by the effects that ribosomes and translation have on tran-
script stability (Figure 10.2a) [37]. For instance, ribosome binding has been found
to attenuate RNase E cleavage in several studies. Incubation of the ompA mRNA
with increasing molar excesses of 30S ribosomal subunits substantially reduced
RNase E cleavage in the 5′ untranslated region (UTR) [38]. lacZ transcript half-
life was correlated with β-galactosidase enzyme activity (a proxy for translation
efficiency) when changes were made to the RBS [39], suggesting that ribosome
occupancy positively influences transcript half-life. Taken together, these results
support a simple steric hindrance model where the presence of ribosomes on an
mRNA inhibits RNase E binding and cleavage [37].
Moreover, because translation is co-transcriptional in bacteria, the transcrip-
tion rate can influence susceptibility to RNase E cleavage by determining the
length of exposed transcript. If transcription outpaces the rate of ribosome bind-
ing and translation initiation, much of the transcript, including potential RNase E
binding sites, will be exposed. A study with the lacZ gene and mutant T7 bacte-
riophage polymerases in E. coli showed an inverse correlation between
β-galactosidase activity and the rate of T7 RNA transcription, a trend that was
RNase E dependent [40]. Experiments using premature stop codons to render