Identification of GTP as a physiologically relevant inhibitor of Escherichia coli ADP-glucose synthetase

We show that physiological concentrations of GTP can significantly inhibit wild-type Escherichia coli ADP-glucose synthetase (the rate-limiting enzyme of bacterial glycogen synthesis) and that mutant-strain enzymes known to show less inhibition by physiological AMP levels also show less inhibition by physiological levels of GTP. This decreased inhibition by both AMP and GTP can almost totally account for the higher cellular rates of glycogen synthesis observed in the mutant strains. In addition, in metabolic conditions where we have shown that cellular glycogen synthesis increases, cellular GTP levels are known to decrease. Thus, we conclude that GTP inhibition is physiologically relevant.     

Evidence that cyclic AMP stimulates bacterial glycogen synthesis by relieving AMP inhibition of and by increasing the cellular level of ADP-glucose synthetase

Using Escherichia coli mutants that possess an ADP-glucose synthetase (EC 2.7.7.27, the rate-limiting enzyme of bacterial glycogen synthesis) that differs in its inhibition by physiological levels of AMP, evidence was obtained that cyclic AMP stimulates cellular glycogen synthesis during nitrogen starvation by relieving AMP inhibition of this enzyme (without altering the cellular AMP level). Deinhibition for AMP of an enzyme controlled by the adenylate energy charge allows selective release from this control despite the maintenance of a constant cellular energy charge value. It was also shown that an additional increase in rate, not accounted for by AMP deinhibition, was due to an increase in the cellular level of ADP-glucose synthetase.     

Effectors of the stringent response target the active site of Escherichia coli adenylosuccinate synthetase.

Guanosine 5'-diphosphate 3'-diphosphate (ppGpp), a pleiotropic effector of the stringent response, potently inhibits adenylosuccinate synthetase from Escherichia coli as an allosteric effector and/or as a competitive inhibitor with respect to GTP. Crystals of the synthetase grown in the presence of IMP, hadacidin, NO3-, and Mg2+, then soaked with ppGpp, reveal electron density at the GTP pocket which is consistent with guanosine 5'-diphosphate 2':3'-cyclic monophosphate. Unlike ligand complexes of the synthetase involving IMP and GDP, the coordination of Mg2+ in this complex is octahedral with the side chain of Asp13 in the inner sphere of the cation. The cyclic phosphoryl group interacts directly with the side chain of Lys49 and indirectly through bridging water molecules with the side chains of Asn295 and Arg305. The synthetase either directly facilitates the formation of the cyclic nucleotide or scavenges trace amounts of the cyclic nucleotide from solution. Regardless of its mode of generation, the cyclic nucleotide binds far more tightly to the active site than does ppGpp. Conceivably, synthetase activity in vivo during the stringent response may be sensitive to the relative concentrations of several effectors, which together exercise precise control over the de novo synthesis of AMP.     

Protein sequences encoded by the relA and the spoT genes of Escherichia coli are interrelated

relA and spoT are designations for two unlinked Escherichia coli genes whose products function in the synthesis and degradation of guanosine 3',5'-bispyrophosphate during the stringent regulatory response to amino acid deprivation. The RelA protein catalyzes an ATP:GTP 3'-pyrophosphoryl group transfer reaction, and the SpoT protein has a guanosine 3',5'-bispyrophosphate 3'-pyrophosphohydrolyase activity. Both genes have been sequenced recently; the relA gene produces an 84-kDa protein, and the spoT gene is deduced to encode a 79-kDa protein. We report here that the protein sequences of the relA and spoT genes are extensively interrelated.     

Characterization of the relA1 mutation and a comparison of relA1 with new relA null alleles in Escherichia coli

The most widely studied "relaxed" mutant of the relA locus, the relA1 allele, is shown here to consist of an IS2 insertion between the 85th and 86th codons of the otherwise wild-type relA structural gene, which normally encodes a 743-amino acid (84 kDa) protein. The RelA protein is a ribosome-dependent ATP:GTP (GDP) pyrophosphoryltransferase that is activated during the stringent response to amino acid starvation and thereby occasions the accumulation of guanosine 3',5'-bispyrophosphate (ppGpp). We propose that the IS2 insertion functionally splits the RelA protein into two (alpha and beta) peptide fragments which can complement each other in trans to yield residual ppGpp synthetic activity; neither fragment shows this activity when expressed alone. Cell strains with a single copy relA null allele show physiological behavior that is much the same as relA1 mutant strains. Both relA1 and relA null strains accumulate ppGpp during glucose starvation and do not accumulate ppGpp during the stringent response. The presence of ppGpp in verifiable relA null strains is interpreted as unequivocal evidence for an alternate route of ppGpp synthesis that exists in addition to the relA-dependent reaction.     

Evidence for new factors in the coordinate regulation of energy metabolism in Escherichia coli. Effects of hypoxia, chloramphenicol succinate, and 2,4-dinitrophenol on glucose utilization, glycogen synthesis, adenylate energy charge, and hexose phosphates during the first two periods of nitrogen starvation.

We studied the effects of decreased aeration, chloramphenicol succinate, and 2,4-dinitrophenol on the cellular rates of glycogen synthesis and glucose utilization and on the cellular concentrations of adenine nucleotides, glucose 6-phosphate, fructose 1,6-diphosphate, and phosphoenolpyruvate during the first two periods of nitrogen starvation of Escherichia coli W4597(K). A quantitative relationship between the changes in the rates and the accompanying changes in the hexose phosphates is demonstrated. However, the relationship for glycogen synthesis is different in different sets of metabolic conditions. We suggest that this difference reflects a change in the steady state level of a previously unknown effector of ADP-glucose synthetase (glucose 1-phosphate adenylyltransferase, EC 2.7.7.27) the rate-limiting enzyme of bacterial glycogen synthesis. We show that the properties of the hypothetical in vivo effector are consistent with the inhibitory effects of ppGpp (guanosine 3'-diphosphate 5'-diphosphate) and pppGpp (guanosine 3'-diphosphate 5'-triphosphate) on this enzyme in vitro. In addition, tetracycline, an inhibitor of the synthesis of these nucleotides, apparently prevents the change in the quantitative relationship. The relationship between glucose utilization and the hexose phosphates is altered at the transition to Period II of nitrogen starvation. We propose that this change reflects the alteration of the cellular steady state level of an unknown effector of the glucose phosphotransferase system. In contrast to the ATP-hexose phosphate system of shared regulatory effects, the specific effects of the unknown effectors allow the rates of glucose utilization and glycogen synthesis to be altered independently of each other and independently of changes in the rate of glycolysis. This independence allows a greater latitude of response for the individual pathways in more severe metabolic stress or in accommodating the metabolic changes necessary for long term survival.     

Regulation of bacterial glycogen synthesis. Stimulation of glycogen synthesis by endogenous and exogenous cyclic adenosine 3':5'-monophosphate in Escherichia coli and the requirement for a functional CRP gene

In Escherichia coli cya mutants, deficient in adenylate cyclase (EC 4.6.1.1), basal cellular rates of glycogen synthesis were lower and the relative increases produced by exogenous cyclic adenosine 3',5'-monophosphate during growth on glucose were greater than in their respective parent strains. These observations provide strong evidence that endogenous cyclic AMP is one of the key regulators of glycogen synthesis in growing E. coli. In crp mutants, deficient in cyclic AMP receptor protein (CRP), the basal cellular rates of glycogen synthesis were much lower than in their respective parent strains. Stimulation of glycogen synthesis by exogenous cyclic AMP was markedly attenuated in the three crp mutants. Thus, stimulation of glycogen synthesis by either endogenous or exogenous cyclic AMP appears to require CRP. Functional CRP appeared to be required for all three responses observed after cyclic AMP addition: an abrupt step-up in the cellular rate of glycogen synthesis, a continuing exponential increase in rate, and a stimulation of the rate during a subsequent nitrogen starvation. To account for these responses, we derived a mathematical model in which the cyclic AMP-CRP complex regulates the differential rate of synthesis of an enzyme metabolizing an effector of the rate-limiting enzyme of glycogen synthesis.     

Some regulatory properties of glycogen phosphorylase from Streptococcus salivarius

Purified (200-fold) glycogen phosphorylase (EC 2.4.1.1) of Streptococcus salivarius was activated by AMP and NaF when assayed both in the direction of synthesis and in the direction of phosphorolysis. Activation by NaF + AMP was greater than the sum of their individual effects. In the direction of synthesis, the K_m for AMP was 0.25 mm and was decreased to 0.125 mm in the presence of NaF. The K_m for NaF was 0.49 m and was decreased to 0.40 m in the presence of AMP. Glycogen phosphorolysis was similarly affected by AMP and NaF, except that above a concentration of 2 mm AMP was inhibitory. The effects of AMP and NaF were reversible since preincubation with these compounds, followed by dialysis, restored activity almost to the control values although some inhibition of enzyme activity was noted with the samples preincubated with NaF. The presence of both NaF and AMP had no effect on the K_m values for glucose-1-P and glycogen in the direction of synthesis, but increased the V of the enzyme.When assayed in the absence of AMP and NaF in the direction of synthesis, the enzyme was slightly inhibited by glucose and glucose-6-P, and activated by P-enolpyruvate and ADP-glucose. In the presence of AMP and NaF, the enzyme was inhibited by glucose, glucose-6-P and ADP-glucose, but was activated by P-enolpyruvate. Fructose-1,6-P₂ had no effect on the enzyme. The enzyme was further activated in the absence of AMP and NaF by adenosine, ATP, GMP, cyclic AMP and ADP, and was slightly inhibited by GTP and GDP. In the presence of AMP and NaF, however, these compounds, with the exception of adenosine, either did not show any effect or were slightly inhibitory. Adenosine was slightly stimulatory with NaF + AMP, but not with AMP alone. In the direction of phosphorolysis, the enzyme was inhibited by glucose and ADP-glucose, and activated by P-enolpyruvate, fructose-1,6-P₂ and ATP, both in the presence and absence of AMP + NaF.     

Glycogen metabolism and cyclic AMP levels in isolated islets of lean and genetically obese mice.

The levels of glycogen and cyclic AMP, incorporation of glucose into glycogen and activities of glycogen synthetase and phosphorylase were determined in pancreatic islets isolated from genetically obese mice and their lean litter-mates. Islets from obese mice had elevated glycogen levels, increased phosphorylase activity and an increased amount of glycogen synthetase in the physiologically more effective I-form, indicating an increased turnover of glycogen. There was no significant difference in cyclic AMP levels between islets of lean and obese mice, but inhibition of phosphodiesterase or stimulation of adenyl cyclase increased cyclic AMP levels more in obese than in lean mouse islets, indicating a more rapid turnover of cyclic AMP in the former. It is suggested that cyclic AMP stimulated phosphorolytic breakdown of glycogen may be one of the mechanisms responsible for the increased insulin secretory response to glucose observed in islets from genetically obese mice.     

Analysis of the relA gene product of Escherichia coli.

The relA gene product, ATP: GTP 3'-pyrophosphotransferase (stringent factor) has been isolated in homogeneous form from an Escherichia coli strain polyploid for this gene at a yield of 1 mg/100 g cells and at a specific activity in a ribosome-activated assay at 37 degrees C of 120 mumol guanosine pentaphosphate formed min-1 mg protein-1. The specific activity in a methanol-activated assay at 25 degrees C was found to be 4 mumol guanosine pentaphosphate formed min-1 mg protein-1. These values are about 100 times higher than reported by others. Our further studies of this enzyme led to the following results. Antibodies raised against this enzyme inhibit the ribosome-activated synthesis of guanosine tetraphosphate and pentaphosphate but have no effect on the much slower synthesis, detected in the absence of ribosomes. The amount of stringent factor in the relA+ strain CP78 is estimated to about 1 copy per 200 ribosomes. The amount of antibody-binding material in CP79 (relA) is at least 5 times lower.     

Influence of amino acid starvation on guanosine 5''-diphosphate 3''-diphosphate basal-level synthesis in Escherichia coli.

We observed that the synthesis of basal-level guanosine 5'-diphosphate 3'-diphosphate (ppGpp) in both relA mutants and relA+ relC strains of Escherichia coli decreased in response to amino acid limitation and that this was accompanied by an increase in ribonucleic acid (RNA) synthesis. Addition of the required amino acid to starved cultures of relaxed bacteria resulted in the resumption of ppGpp synthesis and a concomitant decrease in RNA production. Our results indicate that relA mutants retain a stringent factor-independent ribosomal mechanism for basal-level ppGpp synthesis. They also suggest that in relA+ bacteria, stringent factor-mediated ppGpp synthesis and the production of basal-level ppGpp are mutually exclusive. These findings substantiate the hypothesis that there are two functionally discrete mechanisms for ppGpp synthesis in E. coli. Through these studies we have also obtained new evidence which indicates that ppGpp serves as a modulator of RNA synthesis during balanced growth as well as under conditions of nutritional downshift and starvation.     

Nonsense and insertion mutants in the relA gene of E. coli: cloning relA.

We have made use of lysogens of a specialized transducing bacteriophage, lambdapyrG+ relA+, to select nonsense (relAnon) and insertion (relAins) mutations in the relA gene. Three independent relAnon mutants were isolated on the phage. In all three, the relaxed phenotype was suppressed by supD, supE, supF or sup6. Three independent relAins mutants were isolated, all containing an insertion element (probably IS2) in an apparently identical location in the relA gene. Polyacrylamide gel electrophoretic analysis of peptides synthesized by the phages in ultraviolet lightkilled host cells revealed that no stringent factor was coded for by either the relAins or relAnon phages (the latter in a sup+ cell); stringent factor was detected when the relAnon phages were used in a similar experiment with supD or supE host cells. The relAnon and relAins mutations could be crossed in haploid form in the E. coli chromosome. These recombinants grew with a normal doubling time, had a ppGpp pool which was between 70 and 100% compared with the classical relA strain, and underwent a normal carbon source shift-down. A restriction endonuclease map of the pyrG relA region of the specialized transducing phage is presented in which the position of the insertion element (recognized by a novel Hind III-cut site) defines the position of the relA gene. This position was verified by an analysis of the structure of five plasmids formed by cloning portions of the region in the pBR322 cloning vehicle. Our results indicate that the relA gene is not an essential cellular function, that there might be a second mechanism for the synthesis of basal level ppGpp in the cell and that the sole function of the relA gene is apparently the high level ppGpp synthesis triggered in response to deacylated tRNA.     

Biosynthesis of bacterial glycogen. Use of site-directed mutagenesis to probe the role of tyrosine 114 in the catalytic mechanism of ADP-glucose synthetase from Escherichia coli

Previous covalent modification studies showed that tyrosine 114 of Escherichia coli ADP-glucose synthetase is involved in substrate binding (Lee, Y. M., and Preiss, J. (1986) J. Biol. Chem. 261, 1058-1064). We have prepared, via site-directed mutagenesis, an E. coli ADP-glucose synthetase variant (Phe114) containing a Tyr114 to Phe substitution in order to test whether the phenolic hydroxyl group plays a critical role in catalysis. Kinetic characterization of Phe114 ADP-glucose synthetase indicates that the Tyr114 hydroxyl is not obligatory for the enzyme catalysis. However, the variant enzyme showed altered properties. It showed a decreased apparent affinity for the substrates. The variant enzyme showed less than 2-fold activation by 5 mM fructose 1,6-bisphosphate in the ADP-glucose synthesis direction. In contrast, in the pyrophosphorolysis direction, the mutant enzyme showed about a 30-fold activation by 5 mM fructose 1,6-bisphosphate. The variant enzyme is heat-labile compared to wild type enzyme. It lost about 60% enzyme activity on incubation at 65 degrees C for 5 min in the presence of 30 mM Pi. The wild type enzyme is stable under these conditions. The results indicate that tyrosine 114 is involved directly or indirectly in enzyme catalysis, but is not obligatory for the enzyme catalysis. Conversion of Tyr114 to Phe also alters the regulatory properties of the enzyme with respect to activation by fructose-1,6-P2 and inhibition by AMP.     

A Novel Dual Allosteric Activation Mechanism of Escherichia coli ADP-Glucose Pyrophosphorylase: The Role of Pyruvate

Fructose-1,6-bisphosphate activates ADP-glucose pyrophosphorylase and the synthesis of glycogen in Escherichia coli. Here, we show that although pyruvate is a weak activator by itself, it synergically enhances the fructose-1,6-bisphosphate activation. They increase the enzyme affinity for each other, and the combination increases V _max, substrate apparent affinity, and decreases AMP inhibition. Our results indicate that there are two distinct interacting allosteric sites for activation. Hence, pyruvate modulates E. coli glycogen metabolism by orchestrating a functional network of allosteric regulators. We postulate that this novel dual activator mechanism increases the evolvability of ADP-glucose pyrophosphorylase and its related metabolic control.     

Induction of autolysis in nongrowing Escherichia coli

Unless relaxation of the stringent response is achieved, all nongrowing bacteria rapidly develop resistance to autolysis induced by a variety of agents, including all classes of cell wall synthesis inhibitors. We now describe inhibitors of cell wall synthesis which were unusual in that they could continue to effectively induce autolysis in relA+ Escherichia coli even after prolonged amino acid starvation. The process of cell wall degradation seems to be catalyzed by similar hydrolytic enzymes in nongrowing and growing cells, yet the activity of these new agents capable of inducing autolysis in the nongrowing relA+ cells did not involve relaxation of RNA or peptidoglycan synthesis. We propose that the suppression of autolysis characteristic of nongrowing cells can be bypassed by a novel mechanism of autolytic triggering which is independent of the relA locus.     

Glucose upshift of carbon-starved marine Vibrio sp. strain S14 causes amino acid starvation and induction of the stringent response.

The physiological status of carbon-starved cells of the marine Vibrio sp. strain S14 has been investigated by the analysis of their immediate response to carbon and energy sources. During the first minute after glucose addition to 48-h-starved cells, the pools of ATP and GTP increased rapidly, and the [ATP]/[ADP] ratio reached the level typical for growing cells within 4 min. The total rates of RNA and protein synthesis increased initially but were inhibited 4 to 5 min after glucose addition by the induction of the stringent response. A mutation in the relA gene abolished stringent control during the recovery and significantly prolonged the lag phase, before the starved cells regrew, after the addition of a single source of carbon. However, both the wild-type and the relA cells regrew without a significant lag phase when given glucose supplemented with amino acids. On the basis of these results, it is suggested that carbon-starved cells are deficient in amino acid biosynthesis and that ppGpp and the stringent response are involved in overcoming this deficiency, presumably by depressing the synthesis of amino acid biosynthetic enzymes. Furthermore, the data suggest that the starved cells primarily are starved for energy, and evidence is presented that the step-up in the rate of protein synthesis after refeeding is partially dependent on de novo RNA synthesis.     

Insights into Glycogen Metabolism in Chemolithoautotrophic Bacteria from Distinctive Kinetic and Regulatory Properties of ADP-Glucose Pyrophosphorylase from Nitrosomonas europaea

Nitrosomonas europaea is a chemolithoautotroph that obtains energy by oxidizing ammonia in the presence of oxygen and fixes CO₂ via the Benson-Calvin cycle. Despite its environmental and evolutionary importance, very little is known about the regulation and metabolism of glycogen, a source of carbon and energy storage. Here, we cloned and heterologously expressed the genes coding for two major putative enzymes of the glycogen synthetic pathway in N. europaea, ADP-glucose pyrophosphorylase and glycogen synthase. In other bacteria, ADP-glucose pyrophosphorylase catalyzes the regulatory step of the synthetic pathway and glycogen synthase elongates the polymer. In starch synthesis in plants, homologous enzymes play similar roles. We purified to homogeneity the recombinant ADP-glucose pyrophosphorylase from N. europaea and characterized its kinetic, regulatory, and oligomeric properties. The enzyme was allosterically activated by pyruvate, oxaloacetate, and phosphoenolpyruvate and inhibited by AMP. It had a broad thermal and pH stability and used different divalent metal ions as cofactors. Depending on the cofactor, the enzyme was able to accept different nucleotides and sugar phosphates as alternative substrates. However, characterization of the recombinant glycogen synthase showed that only ADP-Glc elongates the polysaccharide, indicating that ATP and glucose-1-phosphate are the physiological substrates of the ADP-glucose pyrophosphorylase. The distinctive properties with respect to selectivity for substrates and activators of the ADP-glucose pyrophosphorylase were in good agreement with the metabolic routes operating in N. europaea, indicating an evolutionary adaptation. These unique properties place the enzyme in a category of its own within the family, highlighting the unique regulation in these organisms.