The essential role of mRNA degradation in understanding and engineering E. coli metabolism.

Metabolic engineering strategies are crucial for the development of bacterial cell factories with improved performance. Until now, optimal metabolic networks have been designed based on systems biology approaches integrating large-scale data on the steady-state concentrations of mRNA, protein and metabolites, sometimes with dynamic data on fluxes, but rarely with any information on mRNA degradation. In this review, we compile growing evidence that mRNA degradation is a key regulatory level in E. coli that metabolic engineering strategies should take into account. We first discuss how mRNA degradation interacts with transcription and translation, two other gene expression processes, to balance transcription regulation and remove poorly translated mRNAs. The many reciprocal interactions between mRNA degradation and metabolism are also highlighted: metabolic activity can be controlled by changes in mRNA degradation and in return, the activity of the mRNA degradation machinery is controlled by metabolic factors. The mathematical models of the crosstalk between mRNA degradation dynamics and other cellular processes are presented and discussed with a view towards novel mRNA degradation-based metabolic engineering strategies. We show finally that mRNA degradation-based strategies have already successfully been applied to improve heterologous protein synthesis. Overall, this review underlines how important mRNA degradation is in regulating E. coli metabolism and identifies mRNA degradation as a key target for innovative metabolic engineering strategies in biotechnology.

mRNA degradation. A tale of poly(A) and multiprotein machines.

The Escherichia coli RNA degradosome is a multiprotein complex containing an endoribonuclease, polynucleotide phosphorylase and a DEAD-box RNA helicase. A related complex has been described in the spinach chloroplast. The exosome and the mtEXO complex have recently been described in yeast and it is likely that related complexes also exist in animal cells. This research suggests the widespread existence of sophisticated machines for the efficient degradation of messenger RNA. The DEAD-box helicase in the degradosome can unwind regions of RNA structure that interfere with 3'-5' degradation. The polyadenylation of RNA 3' ends is also known to promote degradation by creating a 'toehold' for the degradation machinery. Much remains to be learned about the regulation of mRNA stability. The complexity of the degradation process, both in the eubacteria and in the eukaryotes, suggests that many steps are possible points of control.

Interaction of RNA polymerase with lacUV5 promoter DNA during mRNA initiation and elongation. Footprinting, methylation, and rifampicin-sensitivity changes accompanying transcription initiation.

We have used enzymatic and chemical probes to follow the movement of Escherichia coli RNA polymerase along lacUV5 promoter DNA during transcription initiation. The RNA polymerase does not escape from the promoter but remains tightly bound during the synthesis of the initial bases of the transcript. This initial phase of RNA synthesis involves the reiterative synthesis and release of RNA chains up to ten bases long via the RNA polymerase cycling reaction and the enzyme remains sensitive to rifampicin inhibition. When longer chains are made, promoter-specific binding is disrupted and the enzyme forms a rifampicin-resistant elongation complex with downstream DNA sequences. This elongation complex covers less than half as much DNA and lacks the DNase I-hypersensitive sites and the base-specific contacts that characterize promoter-bound RNA polymerase. These results lead us to suggest that lacUV5 mRNA synthesis is primed by a promoter-bound enzyme complex that synthesizes the initial nine or ten bases in the mRNA chain. Subsequently, when a chain of ten bases, or slightly longer, is made, contacts with promoter DNA are irreversibly disrupted, sigma subunit is lost, and a "true" elongation complex is formed.

Nucleolytic inactivation and degradation of the RNase III processed pnp message encoding polynucleotide phosphorylase of Escherichia coli.

The two cleavages made by RNase III in the transcripts of the pnp gene of Escherichia coli, 80 nucleotides upstream of the coding sequence of polynucleotide phosphorylase, were previously demonstrated to trigger the rapid degradation of the pnp messenger. In this paper, we demonstrate that the 5' end of the RNase III processed pnp mRNA is attacked by ribonucleases more efficiently than the rest of the molecule. Several 5' extremities resulting from cleavages occurring in the first 500 nucleotides of the pnp transcript have been identified. Three of them referred to as X, Y and W occur in the wild-type strain at the beginning of the coding sequence of the pnp mRNA. The mRNA appears to be cleaved more efficiently at the X site, proximal to the initiation codon, than at sites Y and W located downstream. In vitro, the maturation at X is catalysed by RNase E but not by RNase III. Accumulation of RNA processed at X in RNase E deficient strains leads us to postulate that X is a high affinity primary site which is slowly cleaved by the residual activity of thermosensitive RNase E at non-permissive temperature and that secondary sites located downstream are processed less efficiently than X. Taken together, our results suggest that in wild-type E. coli the degradation of the RNase III processed mRNA is mediated by RNase E.

Bacterial poly(A) polymerase: an enzyme that modulates RNA stability.

We have constructed a strain that overexpresses E coli poly(A) polymerase (PAP I). The recombinant protein was soluble, and a partially purified extract had high levels of poly(A) polymerising activity. An antiserum raised against the overexpressed PAP I has permitted two types of analysis: the identification of other E coli proteins that may interact with PAP I, and the search for PAP I-like proteins in other bacteria. Immunoprecipitation experiments suggest that PAP I is associated with a 48-kDa protein. This protein remains to be identified. Western blotting using the antiserum against E coli PAP I revealed related proteins in a variety of Gram-negative bacteria and in B subtilis. A comparison of the E coli protein with putative poly(A) polymerases recently identified in H influenza and B subtilis showed highly conserved sequences in the amino terminal and central portions of the proteins that may be important for enzyme activity.

The Escherichia coli RNA degradosome: structure, function and relationship in other ribonucleolytic multienzyme complexes.

mRNA instability is an intrinsic property that permits timely changes in gene expression by limiting the lifetime of a transcript. The RNase e of Escherichia coli is a single-strand-specific endo-nuclease involved in the processing of rRNA and the degradation of mRNA. A nucleolytic multi-enzyme complex now known as the RNA degradosome was discovered during the purification and characterization of RNase E. Two other components are a 3' exoribonuclease (polynucleotide phosphorylase, PNPase) and a DEAD-box RNA helicase (RNA helicase B, RhlB). RNase E is a large multidomain protein with N-terminal ribonucleolytic activity, an RNA-binding domain and a C-terminal "scaffold" that binds PNPase, enolase and RhlB. RhlB by itself has little activity but is strongly stimiulated by its interaction with RNase E. RhlB in vitro can facilitate the degradation of structured RNA by PNPase. Since the discovery of the RNA degradosome in E. coli, related complexes have been described in other organisms.

Cycling of ribonucleic acid polymerase to produce oligonucleotides during initiation in vitro at the lac UV5 promoter.

High-resolution gel electrophoresis has been used to detect and quantitate promoter-specific oligonucleotides produced during initiation of transcription in vitro at the lactose operon (lac) UV5 promoter. The resolved products are RNA species of various lengths which correspond to the initial lac mRNA sequence. Quantitation shows that many oligonucleotides can be formed per preinitiation complex, including species as long as hexanucleotide. Synthesis occurs without dissociation of the enzyme, as evidenced by levels of synthesis in the presence of heparin, a selective inhibitor of free RNA polymerase. Thus, RNA polymerase cycles at this promoter in vitro producing oligonucleotides reiteratively. In general, the yield of oligonucleotides decreases when the total concentration of all four substrates is increased or when a missing nucleoside triphosphate substrate is added. Nevertheless, oligonucleotide synthesis persists under all conditions tested. Strikingly, the dinucleotide always represents 50% of the total of all oligonucleotides, even when conditions are manipulated to cause a 100-fold variation in this total. This shows that, after formation of the first phosphodiester bond at the lac UV5 promoter, dissociation of the dinucleotide is as likely as formation of the second phosphodiester bond. As discussed above, after release of a small RNA, RNA polymerase may then begin another RNA chain, which is again subject to premature release. These considerations lead to a model in which RNA polymerase cycles to produce oligonucleotides during initiation of transcription at the lac UV5 promoter in vitro. Production of a long RNA transcript is then essentially an escape from this cycling reaction. The drug rifampicin, which drastically inhibits escape to produce RNA, limits, but dose not prevent, the cycling reaction.

Poly(A) polymerase I of Escherichia coli: characterization of the catalytic domain, an RNA binding site and regions for the interaction with proteins involved in mRNA degradation.

Poly(A) polymerase I (PAP I) of Escherichia coli is a member of the nucleotidyltransferase (Ntr) superfamily that includes the eukaryotic PAPs and all the known tRNA CCA-adding enzymes. Five highly conserved aspartic acids in the putative catalytic site of PAP I were changed to either alanine or proline, demonstrating their importance for polymerase activity. A glycine that is absolutely conserved in all Ntrs was also changed yielding a novel mutant protein in which ATP was wastefully hydrolysed in a primer-independent reaction. This is the first work to characterize the catalytic site of a eubacterial PAP and, despite the conservation of certain sequences, we predict that the overall architecture of the eukaryotic and eubacterial active sites is likely to be different. Binding sites for RNase E, a component of the RNA degradosome, and RNA were mapped by North-western and Far-western blotting using truncated forms of PAP I. Additional protein-protein interactions were detected between PAP I and CsdA, RhlE and SrmB, suggesting an unexpected connection between PAP I and these E. coli DEAD box RNA helicases. These results show that the functional organization of PAP I is similar to the eukaryotic PAPs with an N-terminal catalytic domain, a C-terminal RNA binding domain and sites for the interaction with other protein factors.

Productive and abortive initiation of transcription in vitro at the lac UV5 promoter.

The rates of productive and abortive initiation of transcription in vitro at the lac UV5 promoter have been determined and compared to values determined for phage lambda and T7 promoters. The rate constants for productive initiation of lac transcript are consistently lower over a range of low to moderate concentration of initiating nucleoside triphosphate (ATP). Abortive initiation of lac dinucleoside tetraphosphate is also slower at low to moderate concentrations of ATP. These data demonstrate the existence of significant differences in initiation rate among promoters. We suggest that these differences may be a consequence of the initial mRNA sequences and extents of RNA polymerase cycling during initiation of promoter-specific transcription.

Escherichia coli RNase E has a role in the decay of bacteriophage T4 mRNA.

Bacteriophage T4 mRNAs are markedly stabilized, both chemically and functionally, in an Escherichia coli strain deficient in the RNA-processing endonuclease RNase E. The functional stability of total T4 messages increased 6-fold; we were unable to detect a T4 message whose functional stability was not increased. There was a 4-fold increase in the chemical stability of total T4 RNA. The degree of chemical stabilization of six specific T4 mRNAs examined varied from a maximum of 28-fold to a minimum of 1.5-fold. In the RNase E-deficient strain, several minutes delay and a slower rate of progeny production led to a reduction in final phage yield of approximately 50%. Although the effect of the rne temperature-sensitive mutation could be indirect, the simplest interpretation of our results is that RNase E acts directly in the degradation of many T4 mRNAs.

Gene 32 transcription and mRNA processing in T4-related bacteriophages.

We have analysed transcription and mRNA processing for the gene 32 region of five phages related to T4. Two different organizations of gene 32 proximal promoters were found. In T4 and M1, middle- and late-mode promoters are separated by 50 nucleotides and located within an upstream open reading frame. In T2, K3, Ac3, and Ox2, the 626bp T4 sequence that includes these promoters is replaced by a 59bp sequence containing overlapping middle and late promoters. The RNase E-dependent processing of the g32 mRNAs is conserved in all of the phages. The processing site immediately upstream of g32 in T4 and M1 has been replaced in the other phages by a different sequence that is also cleaved by RNase E. The remarkable conservation of these regulatory features, despite the sequence divergences, suggests that they play an important role in the control of gene expression.

Specificity of Escherichia coli endoribonuclease RNase E: in vivo and in vitro analysis of mutants in a bacteriophage T4 mRNA processing site.

Endoribonuclease RNase E has an important role in the processing and degradation of bacteriophage T4 and Escherichia coli mRNAs. We have undertaken a mutational analysis of the -71 RNase E processing site of T4 gene 32. A series of mutations were introduced into a synthetic T4 sequence cloned on a plasmid, and their effects on processing were analyzed in vivo. The same mutations were transferred into T4 by homologous recombination. In both the plasmid and the phage contexts the processing of the transcripts was similarly affected by the mutations. Partially purified RNase E has also been used to ascertain the effect of these mutations on RNase E processing in vitro. The hierarchy of the efficiency of processing of the various mutant transcripts was the same in vivo and in vitro. These results and an analysis of all of the known putative RNase E sites suggest a consensus sequence RAUUW (R = A or G; W = A or U) at the cleavage site. Modifications of the stem-loop structure downstream of the -71 site indicate that a secondary structure is required for RNase E processing. Processing by RNase E was apparently inhibited by sequences that sequester the site in secondary structure.

mRNA degradation in procaryotes.

The fast turnover of mRNA permits rapid changes in the pattern of gene expression. In procaryotes, many enzymes involved in mRNA degradation have been identified and some of these endo- and exo-ribonucleases are now being intensively studied. Some of the structural features of mRNA that influence decay rates have also recently been defined. Although important components of the decay pathway are still elusive, a coherent and simple model for mRNA decay has emerged in the last few years.

Transcription and messenger RNA processing upstream of bacteriophage T4 gene 32.

Bacteriophage T4 gene 32 lies at the 3' end of a complex transcription unit which includes genes 33, 59, and several open reading frames. In the course of an infection, four major transcripts are synthesized from this unit: two overlapping polycistronic transcripts about 3800 and 2800 nucleotides in length, and two monocistronic gene 32 transcripts about 1150 and 1100 nucleotides in length. These transcripts are made at different times in infection and the polycistronic transcripts have segmental differences in stability. Messenger RNA processing yields a 1025 nucleotide monocistronic gene 32 transcript, and a 135 nucleotide transcript containing part of the gene 59 coding sequence. Processing depends on Escherichia coli encoded ribonuclease E. This pattern of transcription and processing leads to the synthesis of gene 32 mRNA throughout infection, whereas transcripts encoding the upstream genes are present only early in infection. The 3800 nucleotide polycistronic transcript initiates at a promoter that does not require T4 encoded factors for activity. However, full-length synthesis of this transcript depends on the T4 mot gene product. The region upstream of gene 32 also contains four E. coli-like promoters that are active on chimeric plasmids in uninfected cells, but inactive in bacteriophage T4. The location of these cryptic T4 promoters is intriguing in that they lie near the 5' ends of open reading frame B, gene 59 and gene 32. They could play a role in phage development under particular conditions of growth or in bacterial hosts other than those examined here.

The Bacillus subtilis nucleotidyltransferase is a tRNA CCA-adding enzyme.

There has been increased interest in bacterial polyadenylation with the recent demonstration that 3' poly(A) tails are involved in RNA degradation. Poly(A) polymerase I (PAP I) of Escherichia coli is a member of the nucleotidyltransferase (Ntr) family that includes the functionally related tRNA CCA-adding enzymes. Thirty members of the Ntr family were detected in a search of the current database of eubacterial genomic sequences. Gram-negative organisms from the beta and gamma subdivisions of the purple bacteria have two genes encoding putative Ntr proteins, and it was possible to predict their activities as either PAP or CCA adding by sequence comparisons with the E. coli homologues. Prediction of the functions of proteins encoded by the genes from more distantly related bacteria was not reliable. The Bacillus subtilis papS gene encodes a protein that was predicted to have PAP activity. We have overexpressed and characterized this protein, demonstrating that it is a tRNA nucleotidyltransferase. We suggest that the papS gene should be renamed cca, following the notation for its E. coli counterpart. The available evidence indicates that cca is the only gene encoding an Ntr protein, despite previous suggestions that B. subtilis has a PAP similar to E. coli PAP I. Thus, the activity involved in RNA 3' polyadenylation in the gram-positive bacteria apparently resides in an enzyme distinct from its counterpart in gram-negative bacteria.

Polyadenylation promotes degradation of 3'-structured RNA by the Escherichia coli mRNA degradosome in vitro.

Polyadenylation contributes to the destabilization of bacterial mRNA. We have investigated the role of polyadenylation in the degradation of RNA by the purified Escherichia coli degradosome in vitro. RNA molecules with 3'-ends incorporated into a stable stem-loop structure could not readily be degraded by purified polynucleotide phosphorylase or by the degradosome, even though the degradosome contains active RhlB helicase which normally facilitates degradation of structured RNA. The exoribonucleolytic activity of the degradosome was due to polynucleotide phosphorylase, rather than the recently reported exonucleolytic activity exhibited by a purified fragment of RNase E (Huang, H., Liao, J., and Cohen, S. N. (1998) Nature 391, 99-102). Addition of a 3'-poly(A) tail stimulated degradation by the degradosome. As few as 5 adenosine residues were sufficient to achieve this stimulation, and generic sequences were equally effective. The data show that the degradosome requires a single-stranded "toehold" 3' to a secondary structure to recognize and degrade the RNA molecule efficiently; polyadenylation can provide this single-stranded 3'-end. Significantly, oligo(G) and oligo(U) tails were unable to stimulate degradation; for oligo(G), at least, this is probably due to the formation of a G quartet structure which makes the 3'-end inaccessible. The inaccessibility of 3'-oligo(U) sequences is likely to have a role in stabilization of RNA molecules generated by Rho-independent terminators.

RNase II levels change according to the growth conditions: characterization of gmr, a new Escherichia coli gene involved in the modulation of RNase II.

In Escherichia coli, ribonucleases are effectors that rapidly modulate the levels of mRNAs for adaptation to a changing environment. Factors involved in the regulation of these ribonucleases can be relevant for mRNA stability. RNase II is one of the main ribonucleases responsible for exonucleolytic activity in E. coli extracts. We have identified and characterized a new E. coli gene, which was named gmr (gene modulating RNase II). The results demonstrate that a deletion of gmr can be associated with changes in RNase II levels and activity. Western analysis and exoribonuclease activity assays showed a threefold increase in RNase II in the gmr deletion strain. Gmr does not affect RNase II mRNA, but modulates RNase II at the level of protein stability. RNase II protein turnover is slower in the gmr deletion strain. We also show that RNase II levels change in different media, and that this regulation is abolished in a strain lacking gmr. The data presented here show that the regulation of ribonucleolytic activity can depend on growth conditions, and this regulation can be mediated by factors that are not RNases.

Copurification of E. coli RNAase E and PNPase: evidence for a specific association between two enzymes important in RNA processing and degradation.

Ribonuclease E (RNAase E) was isolated in a complex that also contained polynucleotide phosphorylase (PNPase). Besides copurification, evidence for an association of these enzymes comes from sedimentation and immunoprecipitation experiments. Highly purified RNAase E correctly processed E. coli 5S ribosomal RNA, bacteriophage T4 gene 32 mRNA and E. coli ompA mRNA at sites known to depend on the rne gene for cleavage in vivo. The difference between previous smaller estimates of the size of RNAase E and that reported here apparently is due to the sensitivity of the enzyme to proteolysis during purification. The discovery of a specific association between RNAase E and PNPase raises the intriguing possibility that these enzymes act cooperatively in the processing and degradation of RNA.