Expression of the gene for Bacillus subtilis aspartokinase II in Escherichia coli

The gene coding for the subunits of aspartokinase II from Bacillus subtilis has been identified in a B. subtilis DNA library and cloned in a bacterial plasmid (Bondaryk, R. P., and Paulus, H. (1984) J. Biol. Chem. 259, 585-591). The introduction of a plasmid carrying the aspartokinase II gene into an auxotrophic Escherichia coli strain lacking all three aspartokinases restored its ability to grow in the absence of L-lysine, L-threonine, and L-methionine. The B. subtilis aspartokinase gene could thus be functionally expressed in E. coli and substitute for the E. coli aspartokinases. Measurement of aspartokinase levels in extracts of aspartokinaseless E. coli transformed with the B. subtilis aspartokinase II gene revealed an enzyme level comparable to that in a genetically derepressed B. subtilis strain. In spite of the high level of aspartokinase, the growth of the transformed E. coli strain was severely inhibited by the addition of L-lysine but could be restored by also adding L-homoserine. This apparently paradoxical sensitivity to lysine was due to the allosteric inhibition of B. subtilis aspartokinase II by that amino acid, a property which was also observed in extracts of the transformed E. coli strain. The synthesis and degradation of the aspartokinase II subunits were measured by labeling experiments in E. coli transformed with the B. subtilis aspartokinase II gene. In contrast to exponentially growing cells of B. subtilis which contained equimolar amounts of the aspartokinase alpha and beta subunits, the transformed E. coli strain contained a 3-fold molar excess of beta subunit. Pulse-chase experiments showed that the disproportionate level of beta subunit was not due to more rapid turnover of alpha subunit, both subunits being quite stable, but presumably to a more rapid rate of synthesis. After the addition of rifampicin, the synthesis of alpha subunit declined much more rapidly than that of beta subunit, indicating that the two subunits were translated independently from mRNA species that differ in functional stability. In conjunction with the results described in the preceding paper which demonstrated that the aspartokinase subunits are encoded by a single DNA sequence, these observations imply that the alpha and beta subunits of B. subtilis aspartokinase II are the products of in-phase overlapping genes.     

Comparison of the three aspartokinase isozymes in Bacillus subtilis Marburg and 168.

The levels of two aspartokinase isozymes, a lysine-sensitive enzyme and an aspartokinase that is inhibited synergistically by lysine plus threonine, differ strikingly in different strains of Bacillus subtilis. In derivatives of B. subtilis 168 growing in minimal medium, the predominant isozyme is the lysine-sensitive aspartokinase. In B. subtilis ATCC 6051, the Marburg strain, the level of the lysine-sensitive aspartokinase is much lower during growth in minimal medium, and the major aspartokinase activity is the lysine-plus-threonine-sensitive isozyme. Molecular cloning and nucleotide sequence determination of the genes for the lysine-sensitive isozymes from the two B. subtilis strains and their upstream control regions showed these genes to be identical. Evidence that the lysine-sensitive aspartokinase, referred to as aspartokinase II, is distinct from the threonine-plus-lysine-sensitive aspartokinase comes from the observation that disruption of the aspartokinase II gene by recombinational insertion had no effect on the latter. Mutants were obtained from the aspartokinase II-negative strain that also lacked the threonine-plus-lysine-sensitive aspartokinase, which will be referred to as aspartokinase III. Aspartokinase II could be selectively restored to these mutants by transformation with plasmids carrying the aspartokinase II gene. Study of the growth properties of the various mutant strains showed that the loss of either aspartokinase II or aspartokinase III had no effect on growth in minimal medium but that the loss of both enzymes interfered with growth unless the medium was supplemented with the three major end products of the aspartate pathway. It appears, therefore, that aspartokinase I alone cannot provide adequate supplies of precursors for the synthesis of lysine, threonine, and methionine by exponentially growing cells.     

Cloning and structure of the gene for the subunits of aspartokinase II from Bacillus subtilis

A library of Bacillus subtilis DNA in lambda Charon 4A (Ferrari, E., Henner, D.J., and Hoch, J.A. (1981) J. Bacteriol. 146, 430-432) was screened by an immunological procedure for DNA sequences encoding aspartokinase II of B. subtilis, an enzyme composed of two nonidentical subunits arranged in an alpha 2 beta 2 structure (Moir, D., and Paulus, H. (1977a) J. Biol. Chem. 252, 4648-4654). A recombinant bacteriophage was identified that harbored an 18-kilobase B. subtilis DNA fragment containing the coding sequences for both aspartokinase subunits. The coding sequence for aspartokinase II was subcloned into bacterial plasmids. In response to transformation with the recombinant plasmids, Escherichia coli produced two polypeptides immunologically related to B. subtilis aspartokinase II with molecular weights (43,000 and 17,000) indistinguishable from those found in enzyme produced in B. subtilis. Peptide mapping by partial proteolysis confirmed the identity of the polypeptides produced by the transformed E. coli cells with the B. subtilis aspartokinase II subunits. The size of the cloned B. subtilis DNA fragment could be reduced to 2.9 kilobases by cleavage with PstI restriction endonuclease without affecting its ability to direct the synthesis of complete aspartokinase II subunits, irrespective of its orientation in the plasmid vector. Further subdivision by cleavage with BamHI restriction endonuclease resulted in the production of truncated aspartokinase subunits, each shortened by the same extent. This suggested that a single DNA sequence encoded both aspartokinase subunits and provided an explanation for the earlier observation that the smaller beta subunit of aspartokinase II was highly homologous or identical with the carboxyl-terminal portion of the alpha subunit (Moir, D., and Paulus, H. (1977b) J. Biol. Chem. 252, 4655-4661). A map of the gene for B. subtilis aspartokinase II is proposed in which the coding sequence for the smaller beta subunit overlaps in the same reading frame the promoter-distal portion of the coding sequence for the alpha subunit.     

Aspartokinase III, a new isozyme in Bacillus subtilis 168.

A previously undetected Bacillus subtilis aspartokinase isozyme, which we have called aspartokinase III, has been characterized. The new isozyme was most readily detected in extracts of cells grown with lysine, which repressed aspartokinase II and induced aspartokinase III, or in extracts of strain VS11, a mutant lacking aspartokinase II. Antibodies against aspartokinase II did not cross-react with aspartokinase III. Aspartokinases II and III coeluted on gel filtration chromatography at Mr 120,000, which accounts for the previous inability to detect it. Aspartokinase III was induced by lysine and repressed by threonine. It was synergistically inhibited by lysine and threonine. Aspartokinase III activity, like aspartokinase II activity, declined rapidly in B. subtilis cells that were starved for glucose. In contrast, the specific activity of aspartokinase I, the diaminopimelic acid-inhibitable isozyme, was constant under all growth conditions examined.     

Mechanism of expression of the overlapping genes of Bacillus subtilis aspartokinase II

The mechanism of expression of the overlapping genes that encode the alpha and beta subunits of aspartokinase II of Bacillus subtilis was studied by specific mutagenesis of the cloned coding sequence. Escherichia coli or B. subtilis VB31 (aspartokinase II-deficient), transformed with plasmids carrying either a deletion of the translation start site and about one-half of the coding region for the larger alpha subunit or a frameshift mutation early in the alpha subunit coding region, produced the smaller beta subunit in the absence of alpha subunit synthesis, indicating that beta subunit is not derived from alpha subunit and that its synthesis does not depend on the alpha subunit translation initiation site. The beta subunit translation start site was identified by oligonucleotide-directed mutagenesis of the putative translation start codon. Modification of the nucleotide sequence encoding methionine residue 247 of the alpha subunit from ATG to either TTA or AAT (but not GTG) abolished beta subunit synthesis but had no effect on the production of alpha subunit. This observation is consistent with peptide chain initiation by N-formylmethionine, which specifically requires an ATG or GTG sequence, and indicates that translation of the beta subunit starts at a site corresponding to Met247 of the alpha subunit. Initial studies on the function of the aspartokinase II subunits, using E. coli as a heterologous host, showed that beta subunit was not essential for the expression of the catalytic function of aspartokinase, measured in vitro and in vivo, nor for its allosteric regulation by L-lysine. Whether the beta subunit has a function specific to B. subtilis needs to be explored in a homologous expression system.     

Evidence that the iron-sulfur cluster of Bacillus subtilis glutamine phosphoribosylpyrophosphate amidotransferase determines stability of the enzyme to degradation in vivo

Bacillus subtilis glutamine P-Rib-PP amidotransferase contains a [4Fe-4S] cluster which is essential for activity. The enzyme also undergoes removal of 11 NH2-terminal residues from the primary translation product in vivo to form the active enzyme. It has been proposed that oxidative inactivation of the FeS cluster in vivo is the first step in degradation of the enzyme in starving cells. Four mutants of amidotransferases that alter cysteinyl ligands to the FeS cluster or residues adjacent to them have been prepared by site-directed mutagenesis, expressed in Escherichia coli, and characterized (Makaroff, C. A., Paluh, J. L., and Zalkin, H. (1986) J. Biol. Chem. 261, 11416-11423). These mutations were integrated into the B. subtilis chromosome in place of the normal purF gene. Inactivation and degradation in vivo of wild type and mutant amidotransferases were characterized in these integrants. Mutants FeS1 (C448S) and FeS2 (C451S) failed to form active enzyme, assemble FeS clusters, or undergo NH2-terminal processing. The immunochemically cross-reactive protein produced by both mutants was degraded rapidly (t1/2 = 16 min) in exponentially growing cells. In contrast the wild type enzyme was stable in growing cells, and activity and cross-reactive protein were lost from glucose-starved cells with a t1/2 of 57 min. Mutant FeS3 (F394V) contained an FeS cluster and was processed normally, but had only about 40% of normal specific activity. The FeS3 enzyme was also inactivated by reaction with O2 in vitro about twice as fast as the wild type. The amidotransferase produced by the FeS3 integrant was stable in growing cells but was inactivated and degraded in glucose-starved cells more rapidly (t1/2 = 35 min) than the wild type enzyme. Mutant FeS4 (C451S, D442C) also contained an FeS cluster and was processed; the enzyme had about 50% of wild type-specific activity and reacted with O2 in vitro at the same rate as the wild type. Inactivation and degradation of the FeS4 mutant in vivo in glucose-starved cells proceeded at a rate (t1/2 = 45 min) that was somewhat faster than normal. The correlation between absence of an FeS cluster or enhanced lability of the cluster to O2 and increased degradation rates in vivo supports the conclusions that stability of the enzyme in vivo requires an intact FeS cluster and that O2-dependent inactivation is the rate-determining step in degradation of the enzyme. The fact that mutant FeS3 was processed normally but degraded rapidly argues against a role for NH2-terminal processing in controlling degradation rates.     

Role of Clp protease subunits in degradation of carbon starvation proteins in Escherichia coli.

When deprived of a carbon source, Escherichia coli induces the synthesis of a group of carbon starvation proteins. The degradation of proteins labeled during starvation was found to be an energy-dependent process which was inhibited by the addition of KCN and accelerated when cells were resupplied with a carbon source. The degradation of the starvation proteins did not require the ATP-dependent Lon protease or the energy-independent proteases protease I, protease IV, OmpT, and DegP. During starvation, mutants lacking either the ClpA or ClpP subunit of the ATP-dependent Clp protease showed a partial reduction in the degradation of starvation proteins. Strains lacking ClpP failed to increase degradation of starvation proteins when glucose was added to starving cells. The clpP mutants showed a competitive disadvantage compared with wild-type cells when exposed to repeated cycles of carbon starvation and growth. Surprisingly, the glucose-stimulated, ClpP-dependent degradation of starvation proteins did not require either the ClpA or ClpB protein. The patterns of synthesis of starvation proteins were similar in clpP+ and clpP cells. The clpP mutants had reduced rates of degradation of certain starvation proteins in the membrane fraction when a carbon source was resupplied to the starved cells.     

Inactivation of Aspartic Transcarbamylase in Sporulating Bacillus subtilis: Demonstration of a Requirement for Metabolic Energy

The aspartic transcarbamylase (ATCase) activity of Bacillus subtilis cells disappears rapidly from stationary-phase cells prior to sporulation. ATCase activity does not appear in the culture fluid during the stationary phase; hence the enzyme appears to be inactivated in the cells. The enzyme is inactivated normally in two different mutants lacking proteases; the activity is very stable in crude extracts of cells or in the culture fluid. These results suggest that ATCase is not inactivated by the general proteolysis that occurs in sporulating bacteria. The inactivation of ATCase can be completely inhibited after it has begun by oxygen starvation or addition of fluoroacetate. Inhibitors of oxidative phosphorylation and electron transport also interrupt the inactivation of ATCase. The inactivation of ATCase is very slow in two mutant strains that are deficient in enzymes of tricarboxylic acid cycle. Addition of gluconate to stationary cultures of the mutant strains, which is known to restore depleted adenosine 5'-triphosphate pools in these bacteria, also restores inactivation of ATCase. These experiments support the conclusion that the generation of metabolic energy is necessary for the inactivation of ATCase in stationary cells. ATCase activity is stable in growing cells in which ATCase synthesis is repressed by addition of uracil; the enzyme is inactivated normally, however, when such cells cease growing.     

Regulation of hepatic fatty acid synthetase in the obese-hyperglycemic mutant mouse.

Regulation of fatty acid synthetase has been studied in the obese-hyperglycemic mouse and compared with regulation in non obese, littermate control animals. The mechanisms underlying the regulatory changes were defined by immunochemical techniques. Several major conclusions are justified from the data obtained: (1) Although the hepatic specific activity of fatty acid synthetase is higher in obese than in non obese animals pair-fed chow, no difference in hepatic activities is apparent in animals pair-fed the fat-free diet; (2) The higher enzymatic activity in obese animals fed chow is related to a higher content of enzyme, and this higher content is associated with a higher rate of enzyme synthesis; (3) The decrease in hepatic synthetase activity with starvation is distinctly more striking in non obese than in obese animals, and the changes in activity reflect changes in content of enzyme; (4) With starvation there is a decrease in synthesis of enzyme in obese and non obese animals, but only in non obese animals is there also a marked increase in the rate of synthetase degradation (t1/2 = 24 h during starvation, t1/2 = 76 h during normalfeeding); (5) Refeeding starved mice a fat-free diet results in a more striking increase in hepatic synthetase activity in non obese than in obese animals; (6) Administration of triiodothyronine causes a more marked increase in hepatic synthetase activity in non obese than in obese animals. The data thus define a variety of differences in regulation of hepatic fatty acid synthetase in mutant and normal animals. The roles of enzyme synthesis and degradation in the etiology of these differences are defined, and possible mechanisms underlying regulation of synthetase synthesis and degradation in normal mammalian liver are suggested by the observations.     

Aspartokinase Activity and the Developmental Cycle of Myxococcus xanthus

The relationship between aspartokinase activity and fruiting body formation in Myxococcus xanthus was investigated. Two required amino acids, methionine and isoleucine, which stimulated the enzyme in vitro also inhibited fruiting body formation when added to 0.1% Casitone agar. Threonine, a potent feedback inhibitor of the aspartokinase, completely reversed the effects of methionine and isoleucine both on enzyme activity and fruiting body formation. A mutant, M. xanthus FB-S, which had the unusual property of forming fruiting bodies on 1.0% Casitone agar, also exhibited an altered regulation of aspartokinase activity. Spermidine, which is a strong stimulator of the enzyme in vitro, interfered with the developmental cycle of both M. xanthus FB and FS-S. During glycerol induction of myxospores the level of aspartokinase dropped more than 75% during the first hour. These data indicate a strong correlation between aspartokinase activity and the induction of the developmental cycle in M. xanthus. It is suggested that the decrease in aspartokinase activity results in diaminopimelic acid starvation, blockage of cell wall growth, and subsequent induction of the developmental cycle.     

Role of protein degradation in the survival of carbon-starved Escherichia coli and Salmonella typhimurium.

When an Escherichia coli K-12 culture was starved for glucose, 50% of the cells lost viability in about 6 days. When a K-12 mutant lacking five distinct peptidase activities, CM89, was starved in the same manner, viability was lost much more rapidly; 50% of the cells lost viability in about 2 days, whereas a parent strain lacking only one peptidase activity lost 50% viability in about 4 days. Compared with the wild-type strain and with its parent strain CM17, CM89 was defective in both protein degradation and protein synthesis during carbon starvation. Similar results were obtained with glucose-starved Salmonella typhimurium LT2 and LT2-derived mutants lacking various peptidase activities. An S. typhimurium mutant lacking four peptidases, TN852, which was deficient in both protein degradation and synthesis during carbon starvation (Yen et al., J. Mol. Biol. 143:21-33, 1980), was roughly one-third as stable as the isogenic wild type. Isogenic S. typhimurium strains that lacked various combinations of three of four peptidases and that displayed protein degradation and synthesis rates intermediate between those of LT2 and TN852 (Yen et al., J. Mol. Biol. 143:21-33, 1980) displayed corresponding stabilities during carbon starvation. These results point to a role for protein degradation in the survival of bacteria during starvation for carbon.     

Regulation of lysine- and lysine-plus-threonine-inhibitable aspartokinases in Bacillus brevis.

Further studies on the expression of the two aspartokinase activities in Bacillus bovis are presented. Aspartokinase I (previously shown to be inhibited and repressed by lysine) was found to be repressed by diaminopimelate in the wild-type strain. However, in a mutant unable to convert diaminopimelate to lysine, starvation for lysine resulted in an increase in aspartokinase I activity. Thus, lysine itself or an immediate metabolite was the true effector of repression. Aspartokinase II (previously shown to be inhibited by lysine plus threonine) was repressed by threonine. Studies with the parent strain and auxotrophs inidicated that only threonine or an immediate metabolite of threonine was involved in this repression. Methionine and isoleucine were not effectors of any of the detected aspartokinase activities. Apart from inhibition and repression controls, a third as yet undefined regulatory mechanism operated to decrease the levels of both aspartokinases as growth declined, even in mutants in which repression control was absent. In thiosine-resistant, lysine-excreting mutants with elevated levels of aspartokinase, the increase in activity could always be attributed to one enzyme or the other, never both. The existence of separate structural genes for each aspartokinase is therefore suggested.     

A Disease-causing Mutation in the Active Site of Serine Palmitoyltransferase Causes Catalytic Promiscuity

The autosomal dominant peripheral sensory neuropathy HSAN1 results from mutations in the LCB1 subunit of serine palmitoyltransferase (SPT). Serum from patients and transgenic mice expressing a disease-causing mutation (C133W) contain elevated levels of 1-deoxysphinganine (1-deoxySa), which presumably arise from inappropriate condensation of alanine with palmitoyl-CoA. Mutant heterodimeric SPT is catalytically inactive. However, mutant heterotrimeric SPT has ∼10–20% of wild-type activity and supports growth of yeast cells lacking endogenous SPT. In addition, long chain base profiling revealed the synthesis of significantly more 1-deoxySa in yeast and mammalian cells expressing the heterotrimeric mutant enzyme than in cells expressing wild-type enzyme. Wild-type and mutant enzymes had similar affinities for serine. Surprisingly, the enzymes also had similar affinities for alanine, indicating that the major affect of the C133W mutation is to enhance activation of alanine for condensation with the acyl-CoA substrate. In vivo synthesis of 1-deoxySa by the mutant enzyme was proportional to the ratio of alanine to serine in the growth media, suggesting that this ratio can be used to modulate the relative synthesis of sphinganine and 1-deoxySa. By expressing SPT as a single-chain fusion protein to ensure stoichiometric expression of all three subunits, we showed that GADD153, a marker for endoplasmic reticulum stress, was significantly elevated in cells expressing mutant heterotrimers. GADD153 was also elevated in cells treated with 1-deoxySa. Taken together, these data indicate that the HSAN1 mutations perturb the active site of SPT resulting in a gain of function that is responsible for the HSAN1 phenotype.     

Active lipoprotein precursors in the Gram-positive eubacterium Lactococcus lactis

Lipid-modified proteins play important roles at the interface between eubacterial cells and their environment. The importance of lipoprotein processing by signal peptidase II (SPase II) is underscored by the fact that this enzyme is essential for viability of the Gram-negative eubacterium Escherichia coli. In contrast, SPase II is not essential for growth and viability of the Gram-positive eubacterium Bacillus subtilis. This could be due to alternative amino-terminal lipoprotein processing, which was shown previously to occur in SPase II mutants of B. subtilis. Alternatively, uncleaved lipoprotein precursors might be functional. To explore further the importance of lipoprotein processing in Gram-positive eubacteria, an SPase II mutant strain of Lactococcus lactis was constructed. Although some of the 39 (predicted) lactococcal lipoproteins, such as PrtM and OppA, are essential for growth in milk, the growth of SPase II mutant L. lactis cells in this medium was not affected. Furthermore, the activity of the strictly PrtM-dependent extracellular protease PrtP, which is required for casein degradation, was not impaired in the absence of SPase II. Importantly, no alternative processing of pre-PrtM and pre-OppA was observed in cells lacking SPase II. Taken together, these findings show for the first time that authentic lipoprotein precursors retain biological activity.