Delayed-relaxed response explained by hyperactivation of RelE

Escherichia coli encodes two rel loci, both of which contribute to the control of synthesis of macromolecules during amino acid starvation. The product of relA (ppGpp synthetase I) is responsible for the synthesis of guanosine tetraphosphate, ppGpp, the signal molecule that exerts stringent control of stable RNA synthesis. The second rel locus, relBE, was identified by mutations in relB that confer a so-called 'delayed-relaxed response' characterized by continued RNA synthesis after a lag period of approximately 10 min after the onset of amino acid starvation. We show here that the delayed-relaxed response is a consequence of hyperactivation of RelE. As in wild-type cells, [ppGpp] increased sharply in relB101 relE cells after the onset of starvation, but returned rapidly to the prestarvation level. RelE is a global inhibitor of translation that is neutralized by RelB by direct protein-protein interaction. Lon protease activates RelE during amino acid starvation by degradation of RelB. We found that mutations in relB that conferred the delayed-relaxed phenotype destabilized RelB. Such mutations confer severe RelE-dependent inhibition of translation during amino acid starvation, indicating hyperactivation of RelE. Hyperactivation of RelE during amino acid starvation was shown directly by measurement of RelE-mediated cleavage of tmRNA. The RelE-mediated shutdown of translation terminated amino acid consumption and explains the rapid restoration of the ppGpp level observed in relB mutant cells. Restoration of the prestarvation level of ppGpp, in turn, allows for the resumption of stable RNA synthesis seen during the delayed-relaxed response.     

Turnover as a control of ribonucleic acid accumulation in bacteria undergoing stepdown.

The synthesis of ribosomes was compared in rel+ and rel- strains of Escherichia coli undergoing "stepdown" in growth from glucose medium to one with lactate as principal carbon source. Two strains (CP78 and CP79), isogenic except for rel, showed similar behaviour with respect to (1) the kinetics of labelling total RNA and ribosomes with exogenous uracil, (2) the proportion of newly formed protein that could be bound with nascent rRNA in mature ribosomes, and (3) the rate of induction of enzymically active beta-galactosidase (relative to the rate of ribosome synthesis). It was concluded that, as there was no net accumulation of RNA during stepdown in either strain, rRNA turnover must be occurring at a high rate. The general features of ribosome maturation in rel+ and rel- cells were almost identical with those found in auxotrophic rel+ organisms starved of required amino acids. In both cases, there was a considerable delay in the maturation of new ribosomal particles, owing to a relative shortfall in the rate of synthesis of ribosome-associated proteins. Only about 4-5% of the total protein labelled during stepdown was capable of binding with newly formed rRNA. This compared with 3.5% for rel+ and 0.5% for rel- auxotrophs during amino acid starvation. The turnover rate for newly formed mRNA and rRNA was virtually the same in "stepped-down" rel+ and rel- strains and was similar to that of the same fraction in amino acid-starved rel+ cells. The functional lifetime of mRNA was also identical. It seems that in the rel- strain many of the characteristics typical of the isogenic rel+ strain are displayed under these conditions, at least as regards the speed of ribosome maturation and the induction of beta-galactosidase. Studies on the thermolability of the latter enzyme induced during stepdown indicate that inaccurate translation, which occurs in rel- strains starved for only a few amino acids, is less evident in this situation than in straightforward amino acid deprivation.     

Guanosine polyphosphate concentration and stable RNA synthesis in Bacillus subtilis following suppression of protein synthesis

Amount of guanosine-5'-triphosphate, 3'-diphosphate (pppGpp) and guanosine-5'-diphosphate, 3'-diphosphate (ppGpp) in the cells of b. subtilis increased several times during starvation for lysine or after treatment with serine hydroxamate (analog of serine) or norvaline (analog of leucine), or in the presence of trimethoprim, which induced deficiency of methionine and leucine. In exponentially growing cells the concentration of pppGpp was found to be 10-20 pmol/A600. When serine hydroxamate or trimethoprim were added, concentration of pppGpp increased to 500-800 pmol/A600 and then slowly diminished. Elimination of lysine or addition to the culture medium of norvaline caused slight transitory accumulation of pppGpp (150 pmol/A600). The amount of another nucleotide ppGpp was always 2-3 times lower than one of pppGpp. Accumulation of (p)ppGpp in rel+ cells was accompanied by cessation of stable RNA synthesis. Under conditions described above rel- cells continued RNA synthesis and did not accumulate (p)ppGpp. In the rel+ cells treated with serine hydroxamate synthesis of stable RNA resumed and the amount of (p)ppGpp decreased after addition of serine or tetracycline and chloramphenicol. The half-life period for pppGpp in the presence of chloramphenicol was determined to be 30-40 seconds. Thus, during aminoacyl-tRNA deficiency rel+ cells of B. subtilis accumulate (p)ppGpp, which are believed to participate in negative regulation of RNA synthesis. Slight accumulation of pppGpp without concomitant inhibition of stable RNA synthesis was observed after treatment of growing cells with chloramphenicol.     

Role of guanosine tetraphosphate in gene expression and the survival of glucose or seryl-tRNA starved cells of Escherichia coli K12

The concentration of guanosine 3′,5′-bispyrophosphate (ppGpp) increases in bacteria in response to amino acid or carbon/energy source starvation. An Escherichia coli K12 ΔrelAΔspoT mutant lacking the ability to synthesize ppGpp lost viability at an increased rate during both glucose and seryl-tRNA starvation. Also, the deleterious effect of chloramphenicol on starved wild-type cells could be overcome by inducing expression of RelA from a plasmid carrying the relA gene transcribed from a tac promoter, prior to starvation and chloramphenicol treatment. As demonstrated by two dimensional gel electrophoresis, this induction of the RelA protein resulted in global alterations in gene expression including increased synthesis of some rpoS-dependent proteins. The ΔrelAΔspoT mutant maintained high expression of several ribosomal proteins during starvation and appeared to exhibit significantly decreased translational fidelity, as demonstrated by an unusual heterogeneity in the isoelectric point of several proteins and the failure to express higher molecular weight proteins during starvation. Moreover, both rpoS-dependent and independent genes failed to exhibit increased expression in the mutant. It is suggested that the deleterious effects on the cells of the relA, spoT deletions are not due solely to the inability of these cells to induce the sigma factor σ^s, but also to deficiencies in translational fidelity and failure to exert classical stringent regulation.     

Role of peptide chain elongation factor G in guanosine 5'-diphosphate 3'-diphosphate synthesis.

In a wild-type strain (relA+) of Escherichia coli, starvation of amino acid led to an immediate cessation of the synthesis of stable ribonucleic acids, together with the accumulation of an unusual nucleotide, guanosine 5'-diphosphate 3'-diphosphate, commonly known as ppGpp. This compound also accumulated during heat shock. When temperature-sensitive protein synthesis elongation factor G (EF-G) was introduced into E. coli NF859, a relA+ strain, the synthesis of ppGpp was reduced to approximately one-half that of wild-type EF-G+ cells at a nonpermissive temperature of 40 degrees C. Furthermore, fusidic acid, an inhibitor of protein synthesis which specifically inactivates EF-G, prevented any accumulation of ppGpp during the heat shock. We suggest that a functional EF-G protein is necessary for ppGpp accumulation under temperature shift conditions, possibly by mediating changes in the function of another protein, the relA gene product. However, EF-G is probably not required for the synthesis of ppGpp during the stringent response, since its inactivation did not prevent ppGpp accumulation during amino acid starvation.     

Synthesis of beta, beta'-subunits of RNA-polymerase in E. coli cells starved for an essential amino acid]

The differential rate of RNA-polymerase beta,beta'-subunits synthesis in E. coli rel+ and rel- cells starved for an essential amino acid, i. e. leucine, is decreased. Inhibition of the rpo BC genes expression proceeds faster in the rel A strain. Guanosine tetraphosphate within the concentration range of 0,2--0,6 mM specifically inhibits lambda drifd18 DNA directed beta,beta'-subunits synthesis in vitro. rpo BC genes are believed to be subject to a weak "stringent control".     

DNA synthesis in vitro in human fibroblast preparations

With several pairs of rel+ and rel- strains of Escherichia coli, the effects of amino acid starvation on the intracellular concentration of K+ and the rate of uptake of 42K+ were investigated. In the early phase of the experiments, the intracellular concentration of K+ was estimated by the conventional method in which the cell volume per A660 value of the culture was assumed to be constant, being not influenced by the variation of growth condition and strain. Apparently, the K+ concentration of rel+ cells was kept almost constant, while that of rel- cells increased about 1.5-fold 2 h after the exposure to amino acid starvation. Unexpectedly, however, the above assumption was found not to be valid in the present study. The cell volume per A660 changed only slightly in CP78 (rel+) cells, while it increased markedly in CP79 (rel-) cells after the exposure to amino acid starvation. Reestimation of the K+ concentrations based on the estimated respective values of cell volumes per A660 revealed no significant difference between both strains. After all, the above apparent phenomenon was found to be due to the fact that the increase in cell volume of the rel+ cells was arrested upon amino acid starvation whereas that in the rel- cells was not. The 42K+ uptake by the rel+ cells was depressed upon amino acid starvation, whereas that by the rel- cells increased. Some regulatory mechanism was suggested to operate in both strains to keep their K+ concentrations constant. When intracellular concentration of a metabolite is to be determined, importance of measurement of cell volume under the respective conditions, without assuming the constancy of the cell volume per A660 of the culture, was pointed out.     

On the control of ribosomal protein biosynthesis in Escherichia coli. I. Studies on ribosomal protein biosynthesis in amino acid-starved cells.

The rate of individual ribosomal protein synthesis relative to total protein synthesis has been determined in Escherichia coli rel+ and rel- cells, under valyltRNA deprivation. These strains have a temperature-sensitive valyl-tRNA synthetase. Starvation was obtained following transfer to the cells to non-permissive temperature. Ribosomal proteins were obtained by treatment of either total lysates of freeze-thawed lysozyme spheroplasts or ammonium sulphate precipitate of ribosomes, with acetic acid. Differential labelling of the ribosomal proteins was observed in both strains: proteins from the rel+ strain appear more labelled than those from the rel- strain, the rate of labelling of individual proteins being about the same in both strains. Moreover ribosomal proteins were found as stable during starvation as total protein. It is thus concluded that in starving cells individual ribosomal proteins are not synthesized at equal rates. This indicates that the synthesis of ribosomal proteins is not only under the control of the rel gene.     

Charging levels of four tRNA species in Escherichia coli Rel(+) and Rel(-) strains during amino acid starvation: a simple model for the effect of ppGpp on translational accuracy

Escherichia coli strains mutated in the relA gene lack the ability to produce ppGpp during amino acid starvation. One consequence of this deficiency is a tenfold increase in misincorporation at starved codons compared to the wild-type. Previous work had shown that the charging levels of tRNAs were the same in Rel(+) and Rel(-) strains and reduced, at most, two- to fivefold in both strains during starvation. The present reinvestigation of the charging levels of tRNA(2)(Arg), tRNA(1)(Thr), tRNA(1)(Leu) and tRNA(His) during starvation of isogenic Rel(+) and Rel(-) strains showed that starvation reduced charging levels tenfold to 40-fold. This reduction corresponds much better with the decreased rate of protein synthesis during starvation than that reported earlier. The determination of the charging levels of tRNA(2)(Arg) and tRNA(1)(Thr) during starvation were accurate enough to demonstrate that charging levels were at least fivefold lower in the Rel(-) strain compared to the Rel(+) strain. Together with other data from the literature, these new data suggest a simple model in which mis-incorporation increases as the substrate availability decreases and that ppGpp has no direct effect on enhancing translational accuracy at the ribosome.     

On the control of ribosomal protein biosynthesis in Escherichia coli. II. Studies during recovery from amino acid starvation.

The rate of synthesis of ribosomal proteins relative to that of total protein was measured at various times during recovery from arginine starvation in isogenic re+ and rel- strains of Escherichia coli K 12. Total ribosomal proteins are preferentially synthesized early during recovery. Higher rates of synthesis are obtained in the rel+ strain than in the rel- strain. Differential rates of synthesis of individual ribosomal proteins are observed at the various times studied. The rate of synthesis of individual proteins increases with time up to maximum values then the rates come down to values similar to those found in exponentially growing cells. The time of restart of synthesis of each protein has been estimated (1) by the time at which the maximum value is reached, and (2) by measuring the rate of synthesis at early time (3 min). Most ribosomal proteins behave similarlly in rel- and rel+ strains. Proteins have been listed from highly labelled (early proteins) to poorly labelled (late proteins). The significance of the order of restart is considered.     

Correlation between RNA synthesis and basal level guanosine 5'-diphosphate 3'-diphosphate in relaxed mutants of Escherichia coli

Through the use of a new nucleotide extraction procedure, we had previously shown that relaxed mutants of Escherichia coli exhibit a unique response to amino acid starvation (Lagosky, P. A., and Chang, F. N. (1980) J. Bacteriol. 144, 499-508). The basal level amounts of guanosine 5'-diphosphate 3'-diphosphate (ppGpp) in both relA and phenotypically relaxed relA+ rplK (relC) strains were shown to decrease at the onset of amino acid limitation and to remain severely depressed throughout the course of the starvation. Upon resupplementation of amino acid-starved relaxed mutants, the production of ppGpp resumes and results in the temporary overaccumulation of this nucleotide beyond its original basal level amount. We now show that the basal level ppGpp content of relaxed bacteria, as well as its subsequent fluctuations in response to amino acid starvation, is inversely correlated with the initial rates of RNA synthesis in these strains. The ability of ppGpp to control the rate of protein synthesis in relA mutants was also examined. It was observed that ppGpp had no apparent direct effect on the initial rates of protein synthesis in relA mutants. The constant inverse correlation which exists between ppGpp content in relA mutants, and their rates of RNa synthesis provide evidence which indicates that basal level ppGpp synthesis has definite physiological significance. It also suggests that the synthesis of basal level ppGpp might be an absolute requirement needed for normal bacterial growth.     

Continued Expression of the Ribonucleic Acid Control Gene During Inhibition of Escherichia coli Ribonucleic Acid and Protein Synthesis

The effect of the ribonucleic acid (RNA) control (RC) gene on the biosynthesis of viral RNA has been examined in an RC(str) and an RC(rel) host infected with R17 RNA bacteriophage under conditions in which host RNA and protein synthesis were inhibited by the addition of rifampicin. Methionine and isoleucine starvation depressed viral RNA biosynthesis in an RC(str) host but not in an RC(rel) host. However, histidine starvation had little effect on viral RNA and protein synthesis in both RC(str) and RC(rel) cells, although it had a marked effect on host protein and RNA synthesis in an RC(str) host. Chloramphenicol relieved the effect of amino acid starvation on viral RNA synthesis in an RC(str) host. It is concluded that stringent control of viral RNA biosynthesis does not require the continued biosynthesis of the RC gene product (RNA or protein) and that a preformed RC gene product can regulate the biosynthesis of the exogenous RNA. It is suggested that the amino acid dependence of viral RNA biosynthesis is due to its obligatory coupling with the translation of the viral coat protein which lacks histidine. It may be inferred that the amino acid requirement of bacterial RNA is due to its coupling with the translation of a host-specific protein (other than the RC gene product) which requires a full complement of amino acids. Since chloramphenicol is known to permit ribosome movement in the absence of protein synthesis, it is suggested that ribosome movement along the nascent RNA chain is a sufficient condition for the continuation of RNA synthesis.     

Regulation of Chromosome Replication in Bacillus subtilis: Effects of Amino Acid Starvation in Strain W23

Amino acid starvation allows limited synthesis of deoxyribonucleic acid (DNA) in Bacillus subtilis strain W23. DNA synthesis increased by about 30% after leucine starvation and by about 60% after histidine starvation. Genetic analysis on the DNA synthesized after amino acid starvation showed that all genetic markers examined have replicated, regardless of which amino acid was starved for. Initially, all markers replicated equally, but upon further replication, the thr cysB and the argA to lys regions replicated ahead of their neighboring, proximal regions. This could indicate that preferred stopping sites exist in these regions or additional sites from which replication can originate reside there. The results suggest that chromosome replication continues from those sites where it had stopped during amino acid starvation.     

Characterization of the stringent and relaxed responses of Streptococcus equisimilis.

The 739-codon rel(Seq) gene of Streptococcus equisimilis H46A is bifunctional, encoding a strong guanosine 3',5'-bis(diphosphate) 3'-pyrophosphohydrolase (ppGppase) and a weaker ribosome-independent ATP:GTP 3'-pyrophosphoryltransferase [(p)ppGpp synthetase]. To analyze the function of this gene, (p)ppGpp accumulation patterns as well as protein and RNA synthesis were compared during amino acid deprivation and glucose exhaustion between the wild type and an insertion mutant carrying a rel(Seq) gene disrupted at codon 216. We found that under normal conditions, both strains contained basal levels of (p)ppGpp. Amino acid deprivation imposed by pseudomonic acid or isoleucine hydroxamate triggered a rel(Seq)-dependent stringent response characterized by rapid (p)ppGpp accumulation at the expense of GTP and abrupt cessation of net RNA accumulation in the wild type but not in the mutant. Tetracycline added to block (p)ppGpp synthesis caused the accumulated (p)ppGpp to degrade rapidly, with a concomitant increase of the GTP pool (decay constant of ppGpp, approximately 0.7 min(-1)). Simultaneous addition of pseudomonic acid and tetracycline to mimic a relaxed response caused wild-type RNA synthesis to proceed at rates approximating those seen under either condition in the mutant. Glucose exhaustion provoked the (p)ppGpp accumulation response in both the wild type and the rel(Seq) insertion mutant, consistent with the block of net RNA accumulation in both strains. Although the source of (p)ppGpp synthesis during glucose exhaustion remains to be determined, these findings reinforce the idea entertained previously that rel(Seq) fulfils functions that reside separately in the paralogous reL4 and spoT genes of Escherichia coli. Analysis of (p)ppGpp accumulation patterns was complicated by finding an unknown phosphorylated compound that comigrated with ppGpp under two standard thin-layer chromatography conditions. Unlike ppGpp, this compound did not adsorb to charcoal and did not accumulate appreciably during isoleucine deprivation. Like ppGpp, the unknown compound did accumulate during energy source starvation.     

A mutation in Escherichia coli that mimics diauxie lag.

We have described a mutant of $\text{E.}\text{coli}$ (2S142) which shows a specific inhibition of stable RNA synthesis at 42°. The temperature sensitive lesion differs from the stringent response to amino acid starvation in that the shut off of rRNA synthesis is not associated with an inhibition of protein synthesis. The decay of ppGpp is slow at 42° with little or no pppGpp detectable. This slow decay rate is not observed in the parental strain, D10, or in 2S142 at 30°. Neither 2S142 or D10 are spoT, nor does the temperature sensitive lesion map near the spoT locus. Thus, the effect of the temperature sensitive lesion on ppGpp metabolism and rRNA synthesis seems to resemble a carbon source downshift (diauxie lag) rather than a stringent response to amino acid starvation.     

Responses to multiple-nutrient starvation in marine Vibrio sp. strain CCUG 15956.

The response of marine Vibrio sp. strain S14 (CCUG 15956) to long-term (48-h) multiple-nutrient starvation (i.e., starvation for glucose, amino acids, ammonium, and phosphate simultaneously) can be described as a three-phase process. The first phase, defined as the stringent control phase, encompasses an accumulation of guanosine 5'-diphosphate 3'-diphosphate (ppGpp) and decreases in RNA and protein synthesis during the first 40 min. In the second phase, there is a temporary increase in the rates of RNA and protein synthesis between 1 and 3 h paralleling a decrease in the ppGpp pool. The third phase includes gradual decline in macromolecular synthesis after 3 h. Using two-dimensional gel electrophoresis of pulse-labeled proteins, a total of 66 proteins were identified as starvation inducible (Sti), temporally expressed throughout the three phases of starvation. The inhibition of protein synthesis during the first phase of starvation partly disrupted the subsequent temporally ordered synthesis of starvation proteins and prevented the expression of some late starvation proteins. It was also found that the early temporal class of starvation proteins, which included the majority of the Sti proteins, was the most essential for long-term survival. Vibrio sp. strain S14 cultures prestarved (1 h) for glucose, amino acids, ammonium, or phosphate as well as cultures exposed (1 h) to CdCl2 exhibited enhanced survival during the subsequent multiple-nutrient starvation in the presence of chloramphenicol or rifampin, while heat or the addition of cyclic AMP or nalidixic acid prior to starvation had no effect. It was demonstrated that amino acid starvation and CdCl2 exposure, which induced the stringent response, were the most effective in conferring enhanced survival. A few Sti proteins were common to all starvation conditions. In addition, the total number of proteins induced by multiple-nutrient starvation significantly exceeded the sum of those induced by starvation for each of the individual nutrients.     

Global analysis of the carbon starvation response of a marine Vibrio species with disruptions in genes homologous to relA and spoT.

The stringent control response, which involves a rapid accumulation of ppGpp, is triggered if the marine Vibrio sp. strain S14 is subjected to carbon and energy starvation. By means of high-resolution two-dimensional gel electrophoresis analysis, we addressed the role of the major ppGpp-synthesizing enzyme (RelA) in the regulation of the carbon starvation response of Vibrio sp. strain S14. The finding that a large number of the carbon starvation-induced proteins were underexpressed in the Vibrio sp. S14 relA mutant strain after the onset of glucose starvation suggests that a rapid accumulation of ppGpp is required for induction of many of the carbon starvation-induced proteins. However, it was also found that a majority of the carbon starvation-induced proteins were significantly less induced if the stringent control response was provoked by amino acid starvation. We therefore also addressed the notion that a carbon starvation-specific signal transduction pathway, complementary to the stringent control, may exist in Vibrio sp. strain S14. It was found that a majority of the proteins that were underexpressed in the relA mutant strain were also underexpressed in the Vibrio sp. S14 spoT mutant strain (csrS1). Interestingly, a large proportion of these underexpressed proteins were found to belong to a group of proteins that are not, or significantly less, induced by starvation conditions that do not promote starvation survival. On the basis of these observations and the finding that the csrS1 strain survives poorly but accumulates ppGpp in a fashion similar to the wild type during carbon and energy source starvation, the gene product of the csrS gene is suggested to be responsible for the mediation of a signal which is complementary to ppGpp and essential for the successful development of the starvation- and stress-resistant cell. This conclusion was also supported by experiments in which changes in phenotypic characteristics known to be induced during carbon starvation were studied. The starvation induction of the high-affinity glucose uptake system was found to be dependent on the csrS gene but not relA, and the synthesis of carbon starvation-specific periplasmic space proteins was dependent, at different times of starvation, on both the relA and the csrS gene products.     

Mutational analysis of the Escherichia coli spoT gene identifies distinct but overlapping regions involved in ppGpp synthesis and degradation

The spoT gene of Escherichia coli encodes a guanosine 3′,5′-bis(diphosphate) 3′-pyrophosphohydrolase (ppGppase) as well as an apparent guanosine 3′,5′-bis(diphosphate) synthetase (designated PSII). To determine the regions of the SpoT protein that are required for these two competing activities, we analysed plasmid-borne deletion mutations for their ability to complement chromosomal mutations defective in each activity. We found that a region containing the first 203 amino acids of the 702-amino-acid SpoT protein was sufficient for ppGppase activity while an overlapping region containing residues 67–374 was sufficient for PSII activity. These data indicate that the catalytic sites involved in the two activities are separate but closely linked in the primary sequence of the SpoT protein. A ppGppase-defective Δ1–58 deletion mutant strain failed to synthesize ppGpp in response to nutrient limitation, also supporting the notion that PSII activity from wild-type SpoT does not increase in response to nutrient limitation. Using a strain lacking PSII activity but retaining ppGppase activity, we determined the contribution of the RelA protein (ppGpp synthetase I, PSI) to ppGpp synthesis following glucose starvation. We found that the RelA protein activity accounts for the initial burst of ppGpp synthesis at the onset of glucose starvation but that this source of synthesis is absent when amino acids are present during glucose starvation.