The functional stability of the lacZ transcript is sensitive towards sequence alterations immediately downstream of the ribosome binding site

Occlusion of the Ribosome Binding Site Connects the Translational Initiation Frequency, mRNA Stability and Premature Transcription Termination

Protein production is controlled by ribosome binding to the messenger RNA (mRNA), quantified in part by the binding affinity between the ribosome and the ribosome binding sequence on the mRNA. Using the E. coli lac operon as model, Ringquist et al. (1992) found a more than 1,000-fold difference in protein yield when varying the Shine-Dalgarno sequence and its distance to the translation start site. Their proposed model accounted for this large variation by only a variation in the binding affinity and the subsequent initiation rate. Here we demonstrate that the decrease in protein yield with weaker ribosome binding sites in addition is caused by a decreased mRNA stability, and by an increased rate of premature transcription termination. Using different ribosome binding site sequences of the E. coli lacZ gene, we found that an approximately 100-fold span in protein expression could be subdivided into three mechanisms that each affected expression 3- to 6-fold. Our experiments is consistent with a two-step ribosome initiation model, in which occlusion of the initial part of the mRNA by a ribosome simultaneously protects the mRNA from both premature transcription termination and degradation: The premature termination we suggest is coupled to the absence of occlusion that allows binding of transcription termination factor, possibly Rho. The mRNA stability is explained by occlusion that prevents binding of the degrading enzymes. In our proposed scenario, a mRNA with lower translation initiation rate would at the same time be “hit” by an increased premature termination and a shorter life-time. Our model further suggests that the transcription from most if not all natural promoters is substantially influenced by premature termination.

Dynamic competition between transcription initiation and repression: Role of nonequilibrium steps in cell-to-cell heterogeneity

Transcriptional repression may cause transcriptional noise by a competition between repressor and RNA polymerase binding. Although promoter activity is often governed by a single limiting step, we argue here that the size of the noise strongly depends on whether this step is the initial equilibrium binding or one of the subsequent unidirectional steps. Overall, we show that nonequilibrium steps of transcription initiation systematically increase the cell-to-cell heterogeneity in bacterial populations. In particular, this allows also weak promoters to give substantial transcriptional noise.

Complex degradation processes lead to non-exponential decay patterns and age-dependent decay rates of messenger RNA

Experimental studies on mRNA stability have established several, qualitatively distinct decay patterns for the amount of mRNA within the living cell. Furthermore, a variety of different and complex biochemical pathways for mRNA degradation have been identified. The central aim of this paper is to bring together both the experimental evidence about the decay patterns and the biochemical knowledge about the multi-step nature of mRNA degradation in a coherent mathematical theory. We first introduce a mathematical relationship between the mRNA decay pattern and the lifetime distribution of individual mRNA molecules. This relationship reveals that the mRNA decay patterns at steady state expression level must obey a general convexity condition, which applies to any degradation mechanism. Next, we develop a theory, formulated as a Markov chain model, that recapitulates some aspects of the multi-step nature of mRNA degradation. We apply our theory to experimental data for yeast and explicitly derive the lifetime distribution of the corresponding mRNAs. Thereby, we show how to extract single-molecule properties of an mRNA, such as the age-dependent decay rate and the residual lifetime. Finally, we analyze the decay patterns of the whole translatome of yeast cells and show that yeast mRNAs can be grouped into three broad classes that exhibit three distinct decay patterns. This paper provides both a method to accurately analyze non-exponential mRNA decay patterns and a tool to validate different models of degradation using decay data.

Structured model of genetic control via the lac promoter in Escherichia coli

A model that describes induction of protein synthesis from lac-based promoters has been developed and incorporated into the single-cell model of Escherichia coli with transcriptional and translational modifications. Unlike previous models of lac-based promoters, this model allows a priori prediction of the intracellular parameters controlling transcription from lac-based promoters with only the extracellular levels of substrate and inducer as inputs. Because of the structural detail of the model, it is possible to simulate different genetic constructions for comparison, such as Lacl(q) strains versus wild-type cells, or including lacl on a multicopy plasmid. Expression from lac to tac promoters is predicted to yield 5% and 30% of the total cellular protein, respectively, with a pBR322-type plasmid. The model predicts the experimental observation that the Lacl(q) strain is not as fully induced as the wild-type strains, even at higher inducer concentrations. Additionally, the model predicts the right order of magnitude of protein production from lac and tac promoters when mechanisms for attenuation of transcription at lower translational efficiency are considered. Finally, the model predicts that for high copy number systems ribosomes become limiting in the synthesis of plasmid-encoded proteins. (c) 1994 John Wiley & Sons, Inc.

Ribosomal protein limitations in Escherichia coli under conditions of high translational activity

Details of the mechanism for ribosome synthesis have been incorporated in the single-cell Escherichia coli model, which enable us to predict the amount of protein synthesizing machinery under different environmental conditions. The predictions agree quite well with available experimental data. The model predicts that ribosomal protein limitations are important when the translational apparatus is in high demand. Ribosomal RNA synthesis is induced by an increase in translational activity, which, in turn, stimulates ribosomal protein synthesis. However, as the demand increases still more, the ribosomal protein mRNA must compete with the plasmid mRNA for ribosomes, and the efficiency of translation of ribosomal proteins is reduced. (c) 1994 John Wiley & Sons, Inc.

Genetic manipulation of stationary-phase genes to enhance recombinant protein production in Escherichia coli

Genetic manipulation of the host strain, by which cell physiology could be modulated, was exploited to enhance recombinant protein production in Escherichia coli. The effects of an inactivated stationary-phase gene (rmf or katF) on recombinant protein production in strains with two different expression systems (the pH-inducible and the lac promoters) were investigated. An improvement of recombinant protein production in the katF mutant at low growth rates was observed for both expression systems. A fourfold and a 30% increase in the volumetric recombinant protein activity were observed for the pH-inducible and the lac promoter system, respectively. The effect of the rmf mutation, on the other hand, depends on the expression system. A twofold increase in the volumetric recombinant protein activity was found for the pH-inducible promoter system, but there was no improvement for the lac promoter system. Improvement in culture performance for slow-growing cultures may have an impact on the design strategy of the host/vector system used in fed-batch cultures, where the specific growth rate is usually slow. The information may also be useful for developing optimal host/vector gene expression systems for recombinant protein production.

Killer and protective ribosomes

In prokaryotes, translation influences mRNA decay. The breakdown of most Escherichia coli mRNAs is initiated by RNase E, a 5'-dependent endonuclease. Some mRNAs are protected by ribosomes even if these are located far upstream of cleavage sites ("protection at a distance"), whereas others require direct shielding of these sites. I argue that these situations reflect different modes of interaction of RNase E with mRNAs. Protection at a distance is most impressive in Bacilli, where ribosomes can protect kilobases of unstable downstream sequences. I propose that this protection reflects the role in mRNA decay of RNase J1, a 5'-->3' exonuclease with no E. coli equivalent. Finally, recent years have shown that besides their protective role, ribosomes can also cleave their mRNA under circumstances that cause ribosome stalling. The endonuclease associated with this "killing" activity, which has a eukaryotic counterpart ("no-go decay"), is not characterized; it may be borne by the distressed ribosome itself.

In vivo dynamics of intracistronic transcriptional polarity

Transcriptional polarity occurs in Escherichia coli when cryptic Rho-dependent transcription terminators become activated as a consequence of reduced translation. Increased spacing between RNA polymerase and the leading ribosome allows the transcription termination factor Rho to bind to mRNA, migrate to the RNA polymerase, and induce termination. Transcriptional polarity results in decreased synthesis of inefficiently translated mRNAs and, therefore, in decreased expression not only of downstream genes in the same operon (intercistronic polarity) but also of the cistron in which termination occurs (intracistronic polarity). To quantitatively measure the effect of different levels of translation on intracistronic transcription termination, the polarity-prone lacZ reporter gene was fused to a range of mutated ribosome binding sites, repressed to different degrees by local RNA structure. The results show that polarity gradually increases with decreasing frequency of translational initiation, as expected. Closer analysis, with the help of a newly developed kinetic model, reveals that efficient intracistronic termination requires very low translational initiation frequencies. This finding is unexpected because Rho is a relatively small protein that binds rapidly to its RNA target, but it appears to be true also for other examples of transcriptional polarity reported in the literature. The conclusion must be that polarity is more complex than just an increased exposure of the Rho binding site as the spacing between the polymerase and the leading ribosome becomes larger. Biological consequences and possible mechanisms are discussed.

Messenger RNA Decay

This chapter discusses several topics relating to the mechanisms of mRNA decay. These topics include the following: important physical properties of mRNA molecules that can alter their stability; methods for determining mRNA half-lives; the genetics and biochemistry of proteins and enzymes involved in mRNA decay; posttranscriptional modification of mRNAs; the cellular location of the mRNA decay apparatus; regulation of mRNA decay; the relationships among mRNA decay, tRNA maturation, and ribosomal RNA processing; and biochemical models for mRNA decay. Escherichia coli has multiple pathways for ensuring the effective decay of mRNAs and mRNA decay is closely linked to the cell's overall RNA metabolism. Finally, the chapter highlights important unanswered questions regarding both the mechanism and importance of mRNA decay.

Control of functional mRNA stability in bacteria: multiple mechanisms of nucleolytic and non-nucleolytic inactivation

Messenger RNA in bacteria may be inactivated by several parallel mechanisms acting independently on different target sites. For any species of mRNA the overall rate of inactivation is determined by the sum of the contributions from the different mechanisms. Transcripts may be inactivated directly by endonucleolytic attack or by processive nucleolytic degradation, which may proceed in the 3'-5' direction and probably also in the 5'-3' direction. Moreover, the functional lifetime of many mRNAs may be determined by processes that are not nucleolytic, such as the binding of translational repressors or the formation of secondary structures which prevent initiation of translation. These non-nucleolytic processes may also determine the chemical stability as chemical degradation frequently appears to be closely coupled to functional inactivation. The relative importance of the different mechanisms in the inactivation of bulk cellular mRNA, as well as the general prospects for engineering of stable mRNAs are discussed.

Single-cell analysis of lambda immunity regulation

We have examined expression of the lambdacI operon in single cells via a rex Colon, two colons gfp substitution. Although average fluorescence agreed with expectations for expression of lambda-repressor, fluorescence fluctuated greatly from cell-to-cell. Fluctuations in repressor concentration are not predicted by previous models and are tolerated in part by a regulatory response to DNA damage.

Resistance to the peptidyl transferase inhibitor tiamulin caused by mutation of ribosomal protein l3

The antibiotic tiamulin targets the 50S subunit of the bacterial ribosome and interacts at the peptidyl transferase center. Tiamulin-resistant Escherichia coli mutants were isolated in order to elucidate mechanisms of resistance to the drug. No mutations in the rRNA were selected as resistance determinants using a strain expressing only a plasmid-encoded rRNA operon. Selection in a strain with all seven chromosomal rRNA operons yielded a mutant with an A445G mutation in the gene coding for ribosomal protein L3, resulting in an Asn149Asp alteration. Complementation experiments and sequencing of transductants demonstrate that the mutation is responsible for the resistance phenotype. Chemical footprinting experiments show a reduced binding of tiamulin to mutant ribosomes. It is inferred that the L3 mutation, which points into the peptidyl transferase cleft, causes tiamulin resistance by alteration of the drug-binding site. This is the first report of a mechanism of resistance to tiamulin unveiled in molecular detail.

The cryptic adenine deaminase gene of Escherichia coli. Silencing by the nucleoid-associated DNA-binding protein, H-NS, and activation by insertion elements

In Escherichia coli there are two pathways for conversion of adenine into guanine nucleotides, both involving the intermediary formation of IMP. The major pathway involves conversion of adenine into hypoxanthine in three steps via adenosine and inosine, with subsequent phosphoribosylation of hypoxanthine to IMP. The minor pathway involves formation of ATP, which is converted via the histidine pathway to the purine intermediate 5-amino-4-imidazolecarboxamide ribonucleotide and, subsequently, to IMP. Here we describe E. coli mutants, in which a third pathway for conversion of adenine to IMP has been activated. This pathway was shown to involve direct deamination of adenine to hypoxanthine by a manganese-dependent adenine deaminase encoded by a cryptic gene, yicP, which we propose be renamed ade. Insertion elements, located from -145 to +13 bp relative to the transcription start site, activated the ade gene as did unlinked mutations in the hns gene, encoding the histone-like protein H-NS. Gene fusion analysis indicated that ade transcription is repressed more than 10-fold by H-NS and that a region of 231 bp including the ade promoter is sufficient for this regulation. The activating insertion elements essentially eliminated the H-NS-mediated silencing, and stimulated ade gene expression 2-3-fold independently of the H-NS protein.

The RihA, RihB, and RihC ribonucleoside hydrolases of Escherichia coli. Substrate specificity, gene expression, and regulation

Pyrimidine-requiring cdd mutants of Escherichia coli deficient in cytidine deaminase utilize cytidine as a pyrimidine source by an alternative pathway. This has been presumed to involve phosphorylation of cytidine to CMP by cytidine/uridine kinase and subsequent hydrolysis of CMP to cytosine and ribose 5-phosphate by a putative CMP hydrolase. Here we show that cytidine, in cdd strains, is converted directly to cytosine and ribose by a ribonucleoside hydrolase encoded by the previously uncharacterized gene ybeK, which we have renamed rihA. The RihA enzyme is homologous to the products of two unlinked genes, yeiK and yaaF, which have been renamed rihB and rihC, respectively. The RihB enzyme was shown to be a pyrimidine-specific ribonucleoside hydrolase like RihA, whereas RihC hydrolyzed both pyrimidine and purine ribonucleosides. The physiological function of the ribonucleoside hydrolases in wild-type E. coli strains is enigmatic, as their activities are paralleled by the phosphorolytic activities of the nucleoside phosphorylases, and a triple mutant lacking all three hydrolytic activities grew normally. Furthermore, enzyme assays and lacZ gene fusion analysis indicated that rihB was essentially silent unless activated by mutation, whereas rihA and rihC were poorly expressed in glucose medium due to catabolite repression.

Control of copper homeostasis in Escherichia coli by a P-type ATPase, CopA, and a MerR-like transcriptional activator, CopR

We have isolated and characterized a copper sensitive Escherichia coli mutant that is deficient in the copper transporting P-type ATPase encoded by the copA gene (previously ybaR). Measurements of uptake and efflux of 64Cu by wild-type and mutant cells implicated the CopA protein in copper efflux from the cytoplasm, and further demonstrated that cell-associated copper in intact E. coli cells is distributed between two kinetically distinguishable pools, the ratio of which was dramatically disturbed by the copA mutation. Using a copA-lacZ gene fusion the copA promoter was found to be specifically induced by copper, and this induction was shown to be dependent on a MerR-like transcriptional activator encoded by a previously uncharacterized gene, copR (previously ybbI). In the copA deficient background the copA-lacZ fusion was super induced to very high levels even in the absence of copper addition to the medium, and this induction was dependent on CopR. These results indicated that the cytoplasmic copper concentration was dramatically increased in the copA mutant, in agreement with the 64Cu uptake experiments. Moreover, they implied, that the copper concentration in wild type cells is determined primarily by the CopA efflux pump, while copper is taken up by an essentially constitutive mechanism.

Cell differentiation by interaction of two HMG-box proteins: Mat1-Mc activates M cell-specific genes in S.pombe by recruiting the ubiquitous transcription factor Ste11 to weak binding sites

The Schizosaccharomyces pombe mfm1 gene is expressed in an M cell-specific fashion. This regulation requires two HMG-box proteins: the ubiquitous Ste11 transcription factor and the M cell-controlling protein Mat1-Mc. Here we report that the mfm1 promoter contains a single, weak Stell-binding site (a so-called TR-box) that can confer M-specificity on a heterologous promoter when present in eight copies. In vitro, both Mat1-Mc and Ste11 can bind this box with approximately the same affinity. The Mat1-Mc protein caused a dramatic increase in the DNA-binding of Ste11 to this box, under conditions where we could not detect Mat1-Mc in the resulting protein-DNA complex. When we changed a single base in the mfm1 TR-box, such that it resembled those boxes found in ubiquitously expressed genes, Ste11 binding was enhanced, and in vivo the mfm1 gene also became expressed in P cells where Mat1-Mc is absent. These findings suggest that M-specificity results from Mat1-Mc-mediated Ste11 binding to weak TR-boxes. We have also defined a novel motif (termed M-box), adjacent to the mfm1 TR-box, to which Mat1-Mc binds strongly. A DNA fragment containing both the TR- and the M-box allowed the formation of a complex containing both Ste11 and Mat1-Mc. A single copy of this fragment was sufficient to activate a heterologous promoter in an M-specific fashion, suggesting that these two boxes act in a synergistic manner.

Identification of ribosome binding sites in Escherichia coli using neural network models

This study investigated the use of neural networks in the identification of Escherichia coli ribosome binding sites. The recognition of these sites based on primary sequence data is difficult due to the multiple determinants that define them. Additionally, secondary structure plays a significant role in the determination of the site and this information is difficult to include in the models. Efforts to solve this problem have so far yielded poor results. A new compilation of E. coli ribosome binding sites was generated for this study. Feedforward backpropagation networks were applied to their identification. Perceptrons were also applied, since they have been the previous best method since 1982. Evaluation of performance for all the neural networks and perceptrons was determined by ROC analysis. The neural network provided significant improvement in the recognition of these sites when compared with the previous best method, finding less than half the number of false positives when both models were adjusted to find an equal number of actual sites. The best neural network used an input window of 101 nucleotides and a single hidden layer of 9 units. Both the neural network and the perceptron trained on the new compilation performed better than the original perceptron published by Stormo et al. in 1982.

Control of mRNA processing and decay in prokaryotes

Post-transcriptional mechanisms operate in regulation of gene expression in bacteria, the amount of a given gene product being also dependent on the inactivation rate of its own message. Moreover, segmental differences in mRNA stability of polycistronic transcripts may be responsible for differential expression of genes clustered in operons. Given the absence of 5' to 3' exoribonucleolytic activities in prokaryotes, both endoribonucleases and 3' to 5' exoribonucleases are involved in chemical decay of mRNA. As the 3' to 5' exoribonucleolytic activities are readily blocked by stem-loop structures which are usual at the 3' ends of bacterial messages, the rate of decay is primarily determined by the rate of the first endonucleolytic cleavage within the transcripts, after which the resulting mRNA intermediates are degraded by the 3' to 5' exoribonucleases. Consequently, the stability of a given transcript is determined by the accessibility of suitable target sites to endonucleolytic activities. A considerable number of bacterial messages decay with a net 5' to 3' directionality. Two different alternative models have been proposed to explain such a finding, the first invoking the presence of functional coupling between degradation and the movement of the ribosomes along the transcripts, the second one implying the existence of a 5' to 3' processive '5' binding nuclease'. The different systems by which these two current models of mRNA decay have been tested will be presented with particular emphasis on polycistronic transcripts.

An efficient Shine-Dalgarno sequence but not translation is necessary for lacZ mRNA stability in Escherichia coli

The 5' ends of many bacterial transcripts are important in determining mRNA stability. A series of Shine-Dalgarno (SD) sequence changes showed that the complementarity of the SD sequence to the anti-SD sequence of 16S rRNA correlates with lacZ mRNA stability in Escherichia coli. Several initiation codon changes showed that an efficient initiation codon is not necessary to maintain lacZ mRNA stability. A stop codon in the 10th codon of lacZ increased mRNA stability. Therefore, ribosomal binding via the SD sequence but not translation of the coding region is necessary to maintain lacZ mRNA stability.

Limited differential mRNA inactivation in the atp (unc) operon of Escherichia coli

Individual subunits of ATP synthase, encoded by the eight genes of the atp operon (atpA through atpH), have been found to be synthesized at a 10-fold range in molar amounts (D.L. Foster and R.H. Fillingame, J. Biol. Chem. 257:2009-2015, 1982; K. von Meyenburg, B.B. Jorgensen, J. Nielsen, F.G. Hansen, and O. Michelsen. Tokai J. Exp. Clin. Med. 7:23-31, 1982). We have determined the functional half-lives at 30 degrees C of mRNAs transcribed from these genes either during constitutive expression in a partial diploid strain or after induced expression from a plasmid. Accurate decay kinetics of the relative mRNA levels were determined by monitoring the rates of synthesis of the individual ATP synthase subunits by radioactive pulse labeling at different times after blocking transcription initiation with rifampin. The mRNA transcribed from the atp operon was found to be inactivated about twice as fast as the bulk mRNA in E. coli. Exceptions are the mRNA from the promoter-proximal atpB gene, which was inactivated about three times as fast as the bulk mRNA, and atpC mRNA, the inactivation rate of which was comparable to that of the bulk mRNA. These moderate differences in the kinetics of functional decay explain only a minor part of the differences in expression levels of the atp genes. We conclude, therefore, that the individual atp mRNAs must be translated with widely different efficiencies. The present analysis further revealed that mRNA degradation is sensitive to heat shock; i.e., after incubation at 39 degrees C for 5 min followed by a shift back to 30 degrees C, the decay rate of the bulk mRNA was decreased by 30%.

Isolation and characterization of mutants with impaired regulation of rpsA, the gene encoding ribosomal protein S1 of Escherichia coli

In order to select mutants that would help to characterize the post-transcriptional regulation of rpsA, we constructed a strain in which the growth rate on lactose minimal medium is determined by the amount of an rpsA-lacZ' alpha-fragment fusion protein produced, even when this is encoded by a high-copy-number plasmid. In the parental strain, synthesis of the fusion protein is repressed by a wild-type rpsA gene, present on a compatible plasmid. Twenty-eight spontaneous and independent mutants, all of them mapping in the rpsA leader region, were isolated as strains that showed higher growth rates, on lactose medium, due to increased synthesis of the rpsA-lacZ' fusion protein. Among these mutants only three sequence changes were found, mapping 9, 10 and 27 bases upstream of the rpsA start codon. At both the -9 and -10 positions an A to G transition and at -27 a C to G transversion all resulted in a sequence with better complementarity to the 3' end of 16S rRNA. We also isolated two mutations mapping in the plasmid-encoded rpsA structural gene: an ochre nonsense mutation in codon 15 of the rpsA gene and a frameshift mutation, deleting the T residue at position +1186. To facilitate the in vitro assay of alpha-fragment activity we also constructed a strain that overproduces the alpha-acceptor fragment four-fold relative to a strain that is diploid for this lacZ delta M15 allele.

Construction and use of lambda PL promoter vectors for direct cloning and high level expression of PCR amplified DNA coding sequences

A set of plasmid vectors which allow single-step cloning and expression of PCR-amplified DNA coding sequences has been constructed. The vectors contain the phage lambda PL promoter, a synthetic translation initiation region (TIR), and convenient cloning sites. The cloning sites provide all or part of an AUG translation initiation codon and facilitate the precise fusion of target DNA sequences to vector transcriptional and translational signals. The vectors were constructed with synthetic TIRs because there is evidence which suggests that the efficiency of the phage lambda cII gene TIR present in the parental vector depends strongly on information contained within the cII N-terminal coding sequence. Bovine brain 14-3-3 eta chain cDNA was PCR-amplified and used to demonstrate the expression capacity of the newly constructed vectors. A significant increase in expression of 14-3-3 protein was observed when synthetic TIRs were used in the place of the cII TIR. Expression levels vary from 15% to 48% of total cell protein. The effects of a reported translational enhancer from phage T7 on expression of the 14-3-3 protein are also discussed. The vectors should be generally useful for high level heterologous protein expression in Escherichia coli.

Interdependence of translation, transcription and mRNA degradation in the lacZ gene

We have constructed a collection of Escherichia coli strains which differ by point mutations in the ribosome binding site (RBS) that drives the translation of the lacZ gene. These mutations affect the Shine-Dalgarno sequence or the initiation codon, or create secondary structures that sequester these elements, and result in a 200-fold variation in beta-galactosidase expression. Surprisingly, these variations of expression are paralleled by nearly equivalent changes in the lacZ mRNA level. The ratio of the beta-galactosidase expression to the mRNA level reflects the average spacing between translating ribosomes: hence, paradoxically, mutations that affect translation initiation do not correspondingly change this spacing. Further analysis of the mRNA level variations shows that they originate from two independent mechanisms. When beta-galactosidase expression exceeds a threshold corresponding roughly to one translation event per transcript, the variations in the efficiency of translation initiation affect largely the chemical and functional lifetimes of the mRNA. We further show that the rate-limiting step in the chemical decay process is an RNase E-dependent cleavage, which is outcompeted by translation initiation. Below this expression threshold, the mRNA lifetime levels out and strain-to-strain variations in mRNA level arise solely from polarity effects. We suggest that, in this activity range, most mRNA molecules that escape polarity are crossed by a single ribosome, and hence are identical from the viewpoint of degradation. Altogether, the tight couplings between translation initiation on one hand, polarity and/or mRNA degradation on the other, result in translation initiation events being closely spaced in time even from inefficient RBS, at the expense of the mRNA level. Finally, we evocate the possible beneficial consequences of a coupling between translation, transcription and mRNA degradation, for the management of cellular resources.

Decreasing transcription elongation rate in Escherichia coli exposed to amino acid starvation

The time required for transcription of the lacZ gene in Escherichia coli was determined during exponential growth and under conditions, when the bacterium was exposed to partial isoleucine starvation. To do this, RNA was extracted from the cells at 10 s intervals following induction and quantified by Northern hybridization with probes complementary to either the beginning or the end of the lacZ mRNA. The time lag between inducer addition and the appearance of a hybridization signal at the 'late' probe represents the transit time for RNA polymerase on the lacZ gene, and this parameter and the known length of the transcribed sequence were used to calculate the lacZ mRNA chain growth-rate. The transcription elongation rate was c. 43 nucleotides s-1 during exponential growth and decreased abruptly to c. 20 nucleotides s-1 in a relA+ strain after the onset of isoleucine starvation, when massive concentrations of guanosine tetraphosphate (ppGpp) accumulated in the cells. The starvation condition did not affect initiation of transcription at the lac-promoter, but a substantial fraction of the initiated lacZ mRNA chains was never completed. For the rel+ strain the polarity was moderate, since c. 25% of the initiated lacZ mRNA' chains were continued into full-length mRNAs, but for the relA strain the polarity was so strong that no completed lacZ mRNA could be detected. The protein chain elongation rates decreased from 13 amino acids (aa) s-1 in the unperturbed growth phase to approximately 6 as s-1, when the cells starved for isoleucine. In combination, these results suggest that ppGpp plays a major role in maintaining the coupling between transcription and translation during the downshift by inhibiting mRNA chain elongation. The implications of this result for the control of stable RNA synthesis during the stringent response are discussed.

The Tn10-encoded tetracycline resistance mRNA contains a translational silencer in the 5' nontranslated region

We performed a mutational analysis of the left half of Tn10-encoded tet operator O2, located in the 5' nontranslated region of the mRNA for the resistance protein TetA, and determined the importance of that region for translation efficiency and mRNA stability. Transcriptional fusions of 17 mutants to lacZ expressed the same amounts of beta-galactosidase, while translational fusions varied 35-fold in expression efficiency. The mRNA half-lives varied 24-fold, with 9.6 min for the most highly expressed mRNA and 0.4 min for the least efficiently expressed mRNA. Toeprint experiments were performed to distinguish whether these mutations define a determinant of mRNA stability or influence translation initiation. The highly expressed mRNA was 24-fold more efficient in forming the initiation complex in vitro than the low-expression mutant. It was concluded that this sequence, albeit located upstream of the ribosome-binding sequence, is an important determinant for efficient initiation of translation. Secondary-structure calculations of the mRNAs revealed no correlation of the potential to form double strands masking the ribosome-binding sequence with expression efficiency.

In the Escherichia coli lacZ gene the spacing between the translating ribosomes is insensitive to the efficiency of translation initiation

We have constructed a series of 44 Escherichia coli strains in which the chromosomal region corresponding to the Ribosome Binding Site (RBS) of the lacZ gene, has been replaced by small DNA fragments harboring either RBSs from other genes, or artificial RBSs. The beta-galactosidase expression from these strains ranges from 1 to 130 per cent of that of the parental strain. Using this collection, we demonstrate here that strain-to-strain variations in expression are paralleled by nearly equivalent variations in lacZ mRNA content. We propose that, in this system, polarity and mRNA stability are tightly coupled to translation initiation, so that changes in RBS efficiency are detected mainly as changes in mRNA concentration rather than in the spacing between translating ribosomes. In addition, we show that the mRNA sequence immediately downstream from the initiator codon influences per se the lifetime of the lacZ mRNA. We discuss the mechanism of the interdependence between translation, transcription and degradation in this gene, and speculate about the general role of this interdependence in determining the expression of bacterial genes.

Chloramphenicol-induced stabilization of cat messenger RNA in Bacillus subtilis

The expression of the chloramphenicol-inducible chloramphenicol-acetyltransferase gene (cat), encoded on Staphylococcus aureus plasmid pUB112, is regulated via a translational attenuation mechanism. Ribosomes, which are arrested by chloramphenicol during synthesis of a short leader peptide, activate catmRNA translation by opening a 5'-located stem-loop structure, thus setting free the cat ribosome-binding site. We have determined the 5' and 3' ends of catmRNA and analysed its stability in Bacillus subtilis. In the absence of the antibiotic, the half-life of catmRNA is shorter than 0.5 min; it is enhanced to about 8 min by sub-inhibitory concentrations of the drug. No decay intermediates of catmRNA could be detected, indicating a very fast degradation after an initial rate-limiting step. ochre nonsense mutations in the 5' region of the cat structural gene, which eliminate catmRNA translation, did not affect its chloramphenicol-induced stabilization. Mutations in the leader-peptide coding region, which abolish ribosome stalling and, therefore, cat gene induction, also eliminate catmRNA stabilization. We conclude that catmRNA is stabilized on induction by a chloramphenicol-arrested ribosome, which physically protects a nuclease-sensitive target site in the 5' region of catmRNA against exo- or endonucleolytic initiation of degradation. This protection is analogous to ermA and ermC mRNA and seems to reflect a general mechanism for stabilization of mRNA derived from inducible antibiotic resistance genes in B. subtilis.

The relation between translation and mRNA degradation in the lacZ gene

The technique of gene fusion, in which the gene of interest, severed from its 3' end, is in-phase fused to a reporter gene--usually lacZ--is widely used to study translational regulation in Escherichia coli. Implicit in these approaches is the assumption that the activity of the ribosome binding site (RBS) fused in-phase with lacZ, does not per se modify the steady-state level of the lacZ mRNA. Herein, we have tested this hypothesis, using a model system in which the RBS of the lamB gene is fused to lacZ. Several point mutations affecting translation initiation have been formerly characterized in this RBS, and we used Northern blots to study their effect upon the lacZ mRNA pattern. Two series of constructs were assayed: in the first one, a 51-bp fragment centered around the lamB initiator codon, was inserted in front of lacZ within the natural lactose operon, whereas in the second the lacZ gene was fused to the genuine malK-lamB operon just downstream from the lamB RBS. We observed that in the first series, the concentration and average molecular weight of the lacZ mRNA dropped sharply as the efficiency of the RBS decreased. This apparently arose from a decreased stability of the message, since the mRNA patterns are equalized when the endonuclease RNase E is inactivated. We suggest that in this case the rate limiting step in the decay process is an RNase E cleavage that is outcompeted by translation.(ABSTRACT TRUNCATED AT 250 WORDS)

Induction kinetics of RNA and proteins in exponentially growing organisms

A mathematical model of the induction kinetics of RNAs and proteins in exponentially growing organisms is derived, and the cellular concentrations of the induced macromolecules at a given time after induction are related to three parameters: the fraction of the synthesis of these macromolecules in total synthesis, the half life of the inducible macromolecules, and the generation time of the organisms. The model predicts that the concentrations of the inducible macromolecules reach one half of the maximum induction level within one generation time after the onset of the induction. The model also predicts that induction curves of proteins are parabolic when their mRNAs are short-lived, but sigmoid when they are stable. Observed induction curves of beta-galactosidase in Escherichia coli cells fit in the theoretical induction curves.

Lack of a 5' non-coding region in Tn1721 encoded tetR mRNA is associated with a low efficiency of translation and a short half-life in Escherichia coli

The repressor-encoding tetR gene from Tn1721 is expressed with a very low efficiency. Its mRNA lacks an untranslated leader sequence. We have constructed protein fusions with the lacZ gene which contain between 14 and 157 5' nucleotides from the tetR gene. Since they are all expressed with similar efficiencies we conclude that the sequence information for initiation of translation is contained within the first 14 bases of the tetR coding region. These fusion transcripts are about 20-fold less efficiently translated than the wild type lacZ transcript. A toeprint analysis confirms that the initiation complex is indistinguishable from those formed by regular transcripts with 5' untranslated regions but occurs in a very low amount in vitro. Thus, the absence of a 5' leader causes a poor rate of translation initiation. The half-lives of tetR and tetR-lacZ mRNAs are about 30 seconds, which is 3-times lower than that of the wt lacZ mRNA. Inactivation of the ams/rne locus in E. coli stabilizes the tetR transcript more than ten-fold. The influence of translation on the tetR half-life is discussed.

Codon bias and gene expression

The frequencies with which individual synonymous codons are used to code their cognate amino acids is quite variable from genome to genome and within genomes, from gene to gene. One particularly well documented codon bias is that associated with highly expressed genes in bacteria as well as in yeast; this is the so-called major codon bias. Here, it is suggested that the major codon bias is not an arrangement for regulating individual gene expression. Instead, the data suggest that this codon bias, which is correlated with a corresponding bias of tRNA abundance, is a global arrangement for optimizing the growth efficiency of cells. On the practical side, it is suggested that heterologous gene expression is not as sensitive to codon bias as previously thought, but that it is quite sensitive to other characteristics of the heterologous gene.

Unusual codon bias occurring within insertion sequences in Escherichia coli

The large open reading frames of insertion sequences from Escherichia coli were examined for their spatial pattern of codon usage bias and distribution of rarely used codons. There is a bias in codon usage that is generally lower toward the terminal ends of the coding regions, which is reflected in the occurrence of an excess of nonpreferred codons in the 3' portions of the coding regions as compared with the 5' portions. In contrast, typical chromosomal genes have a lower codon usage bias toward the 5' ends of the coding regions. These results imply that the selective forces reflected in codon usage bias may differ according to position within the coding sequence. In addition, these constraints apparently differ in important ways between genes contained in insertion sequences and those in the chromosome.

Multiple determinants of functional mRNA stability: sequence alterations at either end of the lacZ gene affect the rate of mRNA inactivation

The Escherichia coli lacZ gene was used as a model system to identify specific sequence elements affecting mRNA stability. Various insertions and substitutions at the ribosome-binding site increased or decreased the rate of mRNA inactivation by up to fourfold. Deletion of a dyad symmetry, which may give rise to a very stable secondary structure in the mRNA immediately downstream of the gene, decreased the functional stability of the lacZ message. The magnitude of the latter effect was strongly dependent on the sequences at the ribosome-binding site, ranging from practically no effect for the most labile transcripts to a threefold decrease in stability for the most stable one. The results suggest that the wild-type lacZ message is inactivated predominantly by attacks near the ribosome-binding site, presumably in part because the putative secondary structure downstream of the gene protects against 3'-exonucleolytic attack. Taken together, the data for all of the modified variants of lacZ were shown to be quantitatively compatible with a general model of mRNA inactivation involving multiple independent target sites.

The importance of the 5'-region in regulating the stability of sdh mRNA in Bacillus subtilis

The decay of the polycistronic Bacillus subtilis sdh mRNA was analysed using probes specific for each of the component cistrons, sdhC, sdhA and sdhB. In exponentially growing cells, the entire sdh mRNA seems to decay with an 'all or nothing' mechanism and with a uniform half-life of 2-3 min for all cistrons. In stationary-phase cells, the half-life of the 5'-part had dropped to about 0.6 min whereas that of the 3'-part was about 1.2 min. Decay of sdh mRNA was also measured in exponentially growing cells containing a 'down-mutation' in the ribosomal binding site preceding sdhC which decreases the expression of sdhC by about 90%. The mutation has a moderate effect on expression of the downstream cistron sdhA. In this mutant, the half-life of the 5'-part of sdh mRNA was about 0.5 min (i.e. the same as in stationary phase wild-type cells) and the half-life of the 3'-part about 1.3 min. Also, analysis of the decay of an sdh-cat fusion transcript revealed that the sdh (5') part decayed more rapidly than the cat part and this difference was more pronounced in stationary-phase cells compared to exponentially growing cells. The results of these experiments demonstrate the importance of the 5'-segment of sdh mRNA in controlling the stability of the transcript under different growth conditions.

Transcription of single-copy hybrid lacZ genes by T7 RNA polymerase in Escherichia coli: mRNA synthesis and degradation can be uncoupled from translation

In Escherichia coli transcription of individual genes generally requires concomitant translation, and thus the decay of mRNAs cannot be studied without the complication of translation. Here we have used T7 RNA polymerase to transcribe in vivo lacZ genes carrying ribosome binding sites of variable efficiency. We show that neither cell viability nor growth rate is affected by the T7-driven transcription of these genes, provided that they are present as single chromosomal copy. Furthermore, transcription is now completely uncoupled from translation, allowing large amounts of even completely untranslated mRNAs to be synthesized. Taking advantage of these features, we discuss the influence of the frequency of translation upon the processing and degradation of the lac message.

The ompA 5' untranslated RNA segment functions in Escherichia coli as a growth-rate-regulated mRNA stabilizer whose activity is unrelated to translational efficiency

The 5' untranslated region (UTR) of the long-lived Escherichia coli ompA message can function in vivo as an mRNA stabilizer. Substitution of this ompA mRNA segment for the corresponding segment of the labile bla gene transcripts prolongs their lifetime by a factor of 6. We show here that the function of this ompA mRNA stabilizer requires the presence of a 115-nucleotide ompA RNA segment that lies upstream of the ribosome-binding site. Although deletion of this segment reduced the half-life of the ompA transcript by a factor of 5, its absence had almost no effect on the translational efficiency of ompA mRNA. Like the ompA transcript, but unlike bla mRNA, hybrid ompA-bla messages containing the complete ompA 5' UTR were significantly less stable under conditions of slow bacterial growth. We conclude that the stabilizing activity of the ompA 5' UTR is growth rate regulated and that the mechanism of mRNA stabilization by this RNA segment is not related to the spacing between translating ribosomes.

Escherichia coli ribosomal protein L10 is rapidly degraded when synthesized in excess of ribosomal protein L7/L12

In Escherichia coli the genes encoding ribosomal proteins L10 and L7/12, rplJ and rplL, respectively, are cotranscribed and subject to translational coupling. Synthesis of both proteins is coordinately regulated at the translational level by binding of L10 or a complex of L10 and L7/L12 to a single target in the mRNA leader region upstream of rplJ. Unexpectedly, small deletions that inactivated the ribosome-binding site of the rplL gene carried on multicopy plasmids exerted a negative effect on expression of the upstream rplJ gene. This effect could be overcome by overproduction of L7/L12 in trans from another plasmid. This apparent stimulation resulted from stabilization of the overproduced L10 protein by L7/L12, presumably because free L10, in contrast to L10 complexed with L7/L12, is subject to rapid proteolytic decay. The contribution of this decay mechanism to the regulation of the rplJL operon is evaluated.

A regulatory element within a gene of a ribosomal protein operon of Escherichia coli negatively controls expression by decreasing the translational efficiency

The trmD operon of Escherichia coli consists of the genes for the ribosomal protein (r-protein) S16, a 21 kDa protein (21K) of unknown function, the tRNA(m1G37)methyltransferase (TrmD), and r-protein L19, in this order. Previously we have shown that the steady-state amount of the two r-proteins exceeds that of the 21K and TrmD proteins 12- and 40-fold, respectively, and that this differential expression is solely explained by translational regulation. Here we have constructed translational gene fusions of the trmD operon and lacZ. The expression of a lacZ fusion containing the first 18 codons of the 21K protein gene is 15-fold higher than the expression of fusions containing 49 or 72 codons of the gene. This suggests that sequences between the 18th and the 49th codon may act as a negative element controlling the expression of the 21K protein gene. Evidence is presented which demonstrates that this regulation is achieved by reducing the efficiency of translation.

Codon contexts from weakly expressed genes reduce expression in vivo

Nucleotides that neighbor codons in Escherichia coli genes are highly non-random. Furthermore, these context biases are stronger and extend farther from the codon in weakly expressed than in highly expressed genes. We therefore suggested that codon contexts are selected to reduce gene expression levels. We now compare the expression levels of lacZ genes containing two specific coding sequences (context inserts). One context insert represents contexts seen in weakly expressed genes (low variant); the other represents contexts seen in highly expressed genes (high variant). The two variants have identical nucleotide and codon compositions, and encode the same protein. A permutation of four nucleotides, which changes eight codon:codon interfaces of 1043, comprises the only difference between the high and low context variant genes. In three different lacZ mRNAs, the low variant was expressed at a level significantly below that of the high variant. This context effect depends entirely on translation of the contexts in the correct frame; its magnitude depends in part on the placement of other features (e.g. transcriptional pauses and terminators, or perhaps other slow codons or contexts) in the mRNAs. Changing the ribosome density on the message by changing the ribosome binding site distinguishes between dropoff, interference and polarity, three fundamentally different types of models for the context effect. The expression difference between context variants is eliminated by both increases and decreases in the ribosome initiation frequency, as uniquely predicted by the polarity model. In fact, data from all constructions are accommodated by a model in which slow translation of the low context insert increases rho-dependent transcriptional termination within the test gene. The data suggest that the rates of translational initiation and elongation are poised with respect to the rate of transcriptional elongation so that all are influential in setting the expression level of wild-type lacZ. We conclude that context-induced polarity will exist in genes wherever low and reproducible gene product levels have been selected.

Differentially expressed trmD ribosomal protein operon of Escherichia coli is transcribed as a single polycistronic mRNA species

The trmD operon is a four-cistron operon in which the first and fourth genes encode ribosomal proteins S16 (rpsP) and L19 (rplS), respectively. The second gene encodes a 21,000 Mr polypeptide of unknown function and the third gene (trmD) encodes the enzyme tRNA(m1G37)methyltransferase, which catalyzes the formation of 1-methylguanosine (m1G) next to the 3' end of the anticodon (position 37) of some tRNAs in Escherichia coli. Here we show under all regulatory conditions studied, transcription initiates at one unique site, and the entire operon is transcribed into one polycistronic mRNA. Between the promoter and the first gene, rpsP, an attenuator-like structure is found (delta G = -18 kcal; 1 cal = 4.184 J), followed by four uridine residues. This structure is functional in vitro, and terminates more than two-thirds of the transcripts. The different parts of the trmD operon mRNA decay at a uniform rate. The stability of the trmD mRNA is not reduced with decreasing growth rate, which is in contrast to what has been found for other ribosomal protein mRNAs. Furthermore, earlier experiments have shown the existence of differential expression as well as non-co-ordinate regulation within the operon. Our results are consistent with the regulation of the trmD operon being due to some mechanism(s) operating at the post-transcriptional level, and do not involve differential degradation of different mRNA segments, internal promoters or internal terminators.

Codon usage determines translation rate in Escherichia coli

We wish to determine whether differences in translation rate are correlated with differences in codon usage or with differences in mRNA secondary structure. We therefore inserted a small DNA fragment in the lacZ gene either directly or flanked by a few frame-shifting bases, leaving the reading frame of the lacZ gene unchanged. The fragment was chosen to have "infrequent" codons in one reading frame and "common" codons in the other. The insert in these constructs does not seem to give mRNAs that are able to form extensive secondary structures. The translation time for these modified lacZ mRNAs was measured with a reproducibility better than plus or minus one second. We found that the mRNA with infrequent codons inserted has an approximately three-seconds longer translation time than the one with common codons. In another set of experiments we constructed two almost identical lacZ genes in which the lacZ mRNAs have the potential to generate stem structures with stabilities of about -75 kcal/mol. In this way we could investigate the influence of mRNA structure on translation rate. This type of modified gene was generated in two reading frames with either common or infrequent codons similar to our first experiments. We find that the yield of protein from these mRNAs is reduced, probably due to the action in vivo of an RNase. Nevertheless, the data do not indicate that there is any effect of mRNA secondary structure on translation rate. In contrast, our data persuade us that there is a difference in translation rate between infrequent codons and common codons that is of the order of sixfold.

Long-range translational coupling in the rplJL-rpoBC operon of Escherichia coli

In Escherichia coli the genes encoding ribosomal proteins L10 and L7/L12, rplJ and rplL, are cotranscribed, and translation of both cistrons is regulated by binding of L10 or a complex of L10 and L7/L12 to a single target in the mRNA leader region. Co-ordinated regulation is assured by some kind of translational coupling, the mechanism of which was investigated here by deletion analysis of plasmids carrying either the intact rplL gene or rplL-lacZ gene fusions. Unless the rplL ribosome binding site was modified by deletion, efficient initiation of translation required translation of a region located more than 500 nucleotides upstream on the transcript within the rplJ cistron. It is proposed that the wild-type rplL ribosome binding site is blocked by long-range RNA base-pairing to this region, when translation of the rplJ sequence is inhibited.

Mechanisms of mRNA decay in bacteria: a perspective

Messenger RNA decay plays an important role in prokaryotic gene expression. The disparate stabilities of bacterial messages in vivo are a consequence of their differential susceptibility to degradation by cellular endoribonucleases and 3' -exoribonucleases, which in turn results from differences in mRNA sequence and structure. RNase II and polynucleotide phosphorylase, the major bacterial exonucleases involved in mRNA turnover, rapidly degrade single-stranded RNA from the 3' end, but are impeded by 3' stem-loop structures. At present, the identify and substrate specificity of the endonucleases that control mRNA decay rates are relatively poorly defined. Ribosomes and antisense RNA also can influence the stability of transcripts with which they associate. Differences in mRNA stability can contribute to differential expression of genes within polycistronic operons and to modulation of gene expression in response to changes in bacterial growth conditions.

The differential stability of the Escherichia coli ompA and bla mRNA at various growth rates is not correlated to the efficiency of translation

Using two monocistronic gene transcripts, bla and ompA, we have studied the relationship between mRNA stability and translational efficiency. It was found that changes in the ompA mRNA stability are not correlated with an alteration in translational efficiency. In addition, at slow bacterial growth rates, the ompA transcript is translated ten times more efficiently than the bla messenger although the stability of the two transcripts is about equal. At rapid bacterial growth rate, chloramphenicol slightly stabilises both the bla and ompA transcripts without affecting their characteristic difference in half-life. Thus, control of mRNA stability seems not necessarily to be mediated either by the efficiency of loading ribosomes on a transcript, or by the arrest or slowing down of translating ribosomes.

What constitutes the signal for the initiation of protein synthesis on Escherichia coli mRNAs?

Small DNA fragments (60 to 80 nucleotides), randomly obtained from a collection of 14 catabolic, biosynthetic or regulatory Escherichia coli genes, have been shot-gun cloned in place of the lacZ ribosome binding site. A total of 47 recombinants showing substantial beta-galactosidase synthesis (at least 1/30th of the wild-type) were isolated, and their newly acquired translational starts were characterized. Of these, 46 were found to carry a ribosome binding site from one of the original genes, and only one, a non-natural start. Moreover, 12 out of the 14 natural starts were found. The two that were not found are the only ones lacking a Shine-Dalgarno element. So, real starts are generally active in the lac mRNA, whereas the many sites (approx. 100 in this gene collection) that carry a Shine-Dalgarno element followed by AUG or GUG but are located in intra- or intergenic regions, or on non-transcribed strands, are inactive. I conclude that: (1) these "false" starts, being strongly discriminated against in the lac message, are presumably also inactive in their original mRNAs; (2) the discriminating information, being portable from one mRNA to another, must be contained within a small DNA region surrounding the starts. Indeed, I further show that it generally lies within a sequence of about 35 nucleotides bracketing real starts; and (3) this information must have a larger effect on initiation than the exact structure of the mRNA, because the discrimination persists despite a complete change of this structure. Previous statistical analysis has shown that real starts differ from false starts in having a non-random sequence composition from nucleotides -20 to +15 with respect to the start. To uncover whether these biases constitute the discriminating information or simply reflect coding constraints, translational starts were randomly searched in eukaryotic, largely non-coding, DNA. These "eukaryotic" starts all have an in-phase AUG or GUG, preceded by a typical Shine-Dalgarno sequence; outside these elements, the initiator region is strikingly rich in A, and poor in C. These biases match those found around real starts, demonstrating that they are indeed part of the initiation signal. Finally, I describe a simple procedure for introducing any DNA fragment in place of the lac operator site on the E. coli chromosome.