Identification and characterization of two new types of bacterial L-serine dehydratases and assessment of the function of the ACT domain.

Two new types of bacterial Fe-S L-serine dehydratases have been identified. These join two previously recognized enzyme types, for a total of four, that are distinguished on the basis of domain arrangement and amino acid sequence. A Type 3 enzyme from Amphibacillus xylanus (axLSD) and a Type 4 enzyme from Heliscomenobacter hydrossis (hhLSD) were cloned, expressed, purified, and characterized. Like the Type 1 enzyme from Bacillus subtilis (bsLSD), axLSD required a monovalent cation, preferably potassium, for activity. However, the hhLSD was without activity even after reconstitution of the iron-sulfur center by a process that successfully restored activity to oxygen-inactivated axLSD. This and other characteristics suggest that this Type 4 protein may be a pseudoenzyme. The oxygen sensitivity of axLSD was greater than other L-serine dehydratases so far studied and suggested that there may be significant conformational differences among the four types resulting in widely different solvent accessibility of the Fe-S clusters in these enzymes. The role of the ACT domain in these enzymes was explored by deleting it from bsLSD. Although there was an effect on the kinetic parameters, this domain was not responsible for the cation requirement nor did its removal have a significant effect on oxygen sensitivity.

Some biochemical and histochemical properties of human liver serine dehydratase.

In rat, serine dehydratase (SDH) is abundant in the liver and known to be a gluconeogenic enzyme, while there is little information about the biochemical property of human liver serine dehydratase because of its low content and difficulty in obtaining fresh materials. To circumvent these problems, we purified recombinant enzyme from Escherichia coli, and compared some properties between human and rat liver serine dehydratases. Edman degradation showed that the N-terminal sequence of about 75% of human serine dehydratase starts from MetSTART-Met2-Ser3- and the rest from Ser3-, whereas the N-terminus of rat enzyme begins from the second codon of MetSTART-Ala2-. The heterogeneity of the purified preparation was totally confirmed by mass spectrometry. Accordingly, this observation in part fails to follow the general rule that the first Met is not removed when the side chain of the penultimate amino acid is bulky such as Met, Arg, Lys, etc. There existed the obvious differences in the local structures between the two enzymes as revealed by limited-proteolysis experiments using trypsin and Staphylococcus aureus V8 protease. The most prominent difference was found histochemically: expression of rat liver serine dehydratase is confined to the periportal region in which many enzymes involved in gluconeogenesis and urea cycle are known to coexist, whereas human liver serine dehydratase resides predominantly in the perivenous region. These findings provide an additional support to the previous notion suggested by physiological experiments that contribution of serine dehydratase to gluconeogenesis is negligible or little in human liver.

Evidence for a dimeric structure of rat liver serine dehydratase.

Rat liver serine dehydratase (SDH) is known to be involved in gluconeogenesis. It has long been believed to be a dimeric protein with the subunit molecular weight (M(r)) of 34,000. Recently, sheep liver SDH was reported to be a monomer with a M(r) of 38,000. The native M(r) of rat SDH was only determined by the ultracentrifugation method more than three decades ago, and that of sheep SDH was done by the method of gel chromatography. The primary to quaternary structures of a given enzyme in a specific mammalian organ are usually conserved among various species. The aim of the present investigation is to clarify the structural differences between rat and sheep SDHs. First, we found that the amino acid composition reported for sheep SDH was statistically similar to that of rat SDH. Second, immunoblot analysis using anti-rat SDH IgG as the probe showed the size of sheep SDH to be a M(r) of 30,500, whereas that of SDH was about M(r) of 35,000. On the other hand, the native size of rat SDH was assessed by two methods: (1) the laser light scattering method demonstrated that rat SDH had a M(r) of 66,800, consistent with the previous value (M(r)=64,000); (2) cross-linking experiments of the purified rat SDH with dimethyl suberimidate revealed the existence of a dimeric form by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The present results clearly confirm that rat SDH is a dimer, and suggest that sheep SDH is similar to rat SDH immunologically, but with a molecular weight 7500 smaller than reported previously.

Partial purification of an iron-dependent L-serine dehydratase from Clostridium sticklandii.

An oxygen-sensitive and highly unstable L-serine dehydratase was partially purified from the Gram-positive anaerobe Clostridium sticklandii. The final active preparation contained five proteins of 27, 30, 44.5, 46, and 58 kDa as judged by SDS-PAGE. The N-terminal sequence of the 30 kDa subunit showed some similarity to the alpha-subunits of the iron-containing L-serine dehydratases from Clostridium propionicum and Peptostreptococcus asaccharolyticus. Oxygen-inactivated L-serine dehydratase from C. sticklandii was reactivated by incubation with Fe2+ under reducing conditions. Furthermore, the enzyme was inactivated by iron-chelating substances like phenanthroline and EDTA. Pyridoxal-5-phosphate (PLP) did not stimulate the activity, and known inhibitors of PLP-containing enzymes such as NaBH4 had no effect on the activity of L-serine dehydratase from C. sticklandii.

Cofactor identification of threonine-serine dehydratase from sheep liver.

L-Threonine-serine dehydratase catalyzes the conversion of L-threonine and L-serine to alpha-ketobutyric acid and pyruvate, respectively. The enzyme has been purified to homogeneity from extracts of sheep liver. In the past, various cofactors have been suggested for threonine dehydratase from both prokaryotic and eukaryotic tissue. While some direct evidence for the presence of pyriodoxal 5'-phosphate in impure preparations is present in the literature no direct evidence for the cofactor in homogeneous dehydrogenase from mammalian tissue has been reported. The threonine dehydratase of sheep liver has been obtained in a homogeneous form and a spectral study provides clear evidence for the presence of pyridoxal 5'-phosphate. Both the physical properties of homogeneous threonine dehydratase and a study of spectral properties of its cofactor are reported in this communication.

L-serine and L-threonine dehydratase from Clostridium propionicum. Two enzymes with different prosthetic groups.

L-Serine dehydratase from the Gram-positive bacterium Peptostreptococcus asaccharolyticus is novel in the group of enzymes deaminating 2-hydroxyamino acids in that it is an iron-sulfur protein and lacks pyridoxal phosphate [Grabowski, R. and Buckel, W. (1991) Eur. J. Biochem. 199, 89-94]. It was proposed that this type of L-serine dehydratase is widespread among bacteria but has escaped intensive characterization due to its oxygen lability. Here, we present evidence that another Gram-positive bacterium, Clostridium propionicum, contains both an iron-sulfur-dependent L-serine dehydratase and a pyridoxal-phosphate-dependent L-threonine dehydratase. These findings support the notion that two independent mechanisms exist for the deamination of 2-hydroxyamino acids. L-Threonine dehydratase was purified 400-fold to apparent homogeneity and revealed as being a tetramer of identical subunits (m = 39 kDa). The purified enzyme exhibited a specific activity of 5 mu kat/mg protein and a Km for L-threonine of 7.7 mM. L-Serine (Km = 380 mM) was also deaminated, the V/Km ratio, however, being 118-fold lower than the one for L-threonine. L-Threonine dehydratase was inactivated by borohydride, hydroxylamine and phenylhydrazine, all known inactivators of pyridoxal-phosphate-containing enzymes. Incubation with NaB3H4 specifically labelled the enzyme. Activity of the phenylhydrazine-inactivated enzyme could be restored by pyridoxal phosphate. L-Serine dehydratase was also purified 400-fold, but its extreme instability did not permit purification to homogeneity. The enzyme was specific for L-serine (Km = 5 mM) and was inhibited by L-cysteine (Ki = 0.5 mM) and D-serine (Ki = 8 mM). Activity was insensitive towards borohydride, hydroxylamine and phenylhydrazine but was rapidly lost upon exposure to air. Fe2+ specifically reactivated the enzyme. L-Serine dehydratase was composed of two different subunits (alpha, m = 30 kDa; beta, m = 26 kDa), their apparent molecular masses being similar to the ones of the two subunits of the iron-sulfur-dependent enzyme from P. asaccharolyticus. Moreover, the N-terminal sequences of the small subunits from these two organisms were found to be 47% identical. In addition, 38% identity with the N-terminus of one of the two L-serine dehydratases of Escherichia coli was detected.

Use of gene fusions of the structural gene sdaA to purify L-serine deaminase 1 from Escherichia coli K-12.

The purification by affinity chromatography of beta-galactosidase from strains carrying sdaA/lacZ gene fusions results in the copurification of L-serine deaminase 1. We conclude that sdaA is the structural gene for the latter enzyme. The purified L-serine deaminase 1 obtained after collagenase treatment of an sdaA-collagen-lacZ fusion differs from the native enzyme by the addition of several amino acids at the C-terminal. Like the enzyme in crude extracts, this purified enzyme is catalytically inactive, and is activated by incubation with iron and dithiothreitol.

The iron-sulfur-cluster-containing L-serine dehydratase from Peptostreptococcus asaccharolyticus. Stereochemistry of the deamination of L-threonine.

The stereochemistry of the deamination of L-threonine to 2-oxobutyrate, catalyzed by purified L-serine dehydratase of Peptostreptococcus asaccharolyticus, was elucidated. For this purpose the enzyme reaction was carried out with unlabelled L-threonine in 2H2O and in 3HOH, as well as with L-[3-3H]threonine in unlabelled water. Isotopically labelled 2-oxobutyrate thus formed was directly reduced in a coupled reaction with L- or D-lactate dehydrogenase and NADH. The (2R)- or (2S)-2-hydroxybutyrate species obtained were then subjected to configurational analyses of their labelled methylene group. The results from 1H-NMR spectroscopy and, after degradation of 2-hydroxybutyrate to propionate, the transcarboxylase assay consistently indicated that the deamination of L-threonine catalyzed by L-serine dehydratase of P. asaccharolyticus proceeds with inversion and retention in a 2:1 ratio. This partial racemization is the first ever to be observed for a reaction catalyzed by serine dehydratase, therefore confirming the distinction of the L-serine dehydratase of P. asaccharolyticus as an iron-sulfur protein from those dehydratases dependent on pyridoxal phosphate. For the latter enzymes exclusively, retention has been reported.

Purification and properties of an iron-sulfur-containing and pyridoxal-phosphate-independent L-serine dehydratase from Peptostreptococcus asaccharolyticus.

L-Serine dehydratase with a specific activity of 15 nkat/mg protein was present in the anaerobic eubacterium Peptostreptococcus asaccharolyticus grown either on L-glutamate or L-serine. The enzyme was highly specific for L-serine with the lowest Km = 0.8 mM ever reported for an L-serine dehydratase. L-Threonine (Km = 22 mM) was the only other substrate. V/Km for L-serine was 500 times higher than that for L-threonine. L-Cysteine was the best inhibitor (Ki = 0.3 mM, competitive towards L-serine). The enzyme was purified 400-fold to homogeneity under anaerobic conditions (specific activity 6 mukat/mg). PAGE in the presence of SDS revealed two subunits with similar intensities (alpha, 30 kDa; beta, 25 kDa). The molecular mass of the native enzyme was estimated as 200 +/- 20 kDa (gel filtration) and 180 kDa (gradient PAGE). In the absence of oxygen the enzyme was moderately stable even in the presence of sodium borohydride or phenylhydrazine (5 mM each). However, by exposure to air the activity was lost, especially when the latter agent was added. The enzyme was reactivated by ferrous ion under anaerobic conditions. The inability of several nucleophilic agents to inactivate the enzyme indicated the absence of pyridoxal phosphate. This was confirmed by a microbiological determination of pyridoxal phosphate. However, the enzyme contained 3.8 +/- 0.2 mol Fe and 5.6 +/- 0.3 mol inorganic sulfur/mol heterodimer (55 kDa) indicating the presence of an [Fe-S] center. The enzyme was successfully applied to measure L-serine concentrations in bacterial media and in human sera.

Contribution of a conserved arginine near the active site of Escherichia coli D-serine dehydratase to cofactor affinity and catalytic activity.

We have employed site-directed mutagenesis to investigate the contribution of a conserved arginyl residue to the catalytic activity and cofactor affinity of D-serine dehydratase, a model pyridoxal 5'-phosphate (vitamin B6) enzyme. Replacement of R-120 in the active site peptide of D-serine dehydratase by L decreased the affinity of the enzyme for pyridoxal 5'-phosphate by 20-fold and reduced turnover by 5-8-fold. kappa cat displayed modified substrate alpha-deuterium isotope effects and altered dependence on both temperature and pH. Analysis of the pH rate profiles of DSD and the R-120----L variant indicated that R-120 interacts electrostatically with catalytically essential ionizable groups at the active site of wild type D-serine dehydratase. The decrease in cofactor affinity observed for DSD(R120L) was not accompanied by significant perturbations in the UV, CD, or 31P NMR spectrum of the holoenzyme, suggesting that the contribution of R-120 to pyridoxal phosphate affinity may be indirect or else involve an interaction with a cofactor functional group other than the 5'-phosphoryl moiety. The properties of two other site-directed variants of D-serine dehydratase indicated that the pyridoxal 5'-phosphate:K-118 Schiff base was indifferent to a small change in the shape of the side chain at position 117 (I-117----L), whereas replacement of K-118 by H resulted in undetectable levels of enzyme. A poor ability to bind cofactor may have rendered DSD(K118H) susceptible to intracellular proteolysis.

Occurrence of a catabolic L-serine (L-threonine) deaminase in Saccharomyces cerevisiae.

Saccharomyces cerevisiae mutants lacking the anabolic L-threonine deaminase, the ilv1- mutants, have been found to exhibit a normal ability to grow, without auxotrophy towards isoleucine, on L-threonine of L-serine as only nitrogen nutrient. Starting from a strain carrying a ilv1- mutation, a new mutation affecting the ability to utilize L-threonine as nitrogen source was selected. This mutation, which also impairs the ability to utilize L-serine, has been denominated cha-, for 'catabolism of hydroxyamino acids' and was found to result in the lack of a catabolic L-serine (L-threonine) deaminase. This enzyme which, unlike the anabolic threonine deaminase, is more active towards serine than towards threonine, differs from the latter enzyme by a number of biochemical and regulatory properties. Whereas the anabolic enzyme is an allosteric enzyme sensitive to feedback inhibition by isoleucine, the catabolic enzyme exhibits Michaelian kinetics: no control of its activity has been detected. Its synthesis is induced by L-serine and L-threonine. These two enzymes, which thus can be easily differentiated by means of their regulations, display a limited ability to compensate for one another's absence and appear to play clearly distinct roles under normal physiological conditions.

In vitro studies on L-serine deaminase activity of Escherichia coli K12.

Extracts of Escherichia coli K12 contain an enzyme which deaminates L-serine. This serine deaminase appears to be a soluble enzyme and is inhibited by substrate analogues, metal ions, and chelators. The activity, which is very unstable in vitro, is protected, and in some cases, even activated by substrate, substrate analogues, and by ferrous ion. The enzyme has proved unstable in all attempts at purification. It resembles closely the L-serine deaminase activity in other microorganisms, but is very different from the mammalian enzyme. As judged by comparison with organisms in which this enzyme serves as part of the principal carbon-handling pathway, L-serine deaminase activity is present in E. coli extracts in physiologically significant amounts.

Comparative studies on threonine and serine dehydratases in rat liver.

1. Two isoenzymes of serine (threonine) dehydratase can be separated by DEAE cellulose chromatography. 2. Each of these isoenzymes possesses different S:T (serine dehydratase activity: threonine dehydratase activity) ratio and kinetic properties. 3. In early postnatal development, the S:T ratio shows marked variation. 4. The hepatic S:T ratio also changes with dietary manipulations. Furthermore L-alanine inhibits enzymic activities towards serine and threonine to different extent. 5. The results are consistent with the hypothesis postulating the presence of two distinct proteins each having a different activity ratio towards serine and threonine.

[Extraction of crystalline L-threonine(serine)-dehydratase from rat liver]. / Poluchenie kristallicheskoi L-treonin(serin)-degidratazy iz pecheni krys.

Isolation on a preparative scale of crystalline pyridoxal phosphate-dependent threonine dehydratase (responsible for threonine deamination) from rat liver tissue is described. The enzyme was purified by stepwise salting out with (NH4)2SO4, two precipitations with acetone, gel filtration through Sephadex G-25, chromatography on DEAE cellulose, repricipitation with ammonium sulfate and crystallization. The ratio of threonine to serine dehydratase activities was altered only slightly through all the steps of the purification procedure. Both enzymes proved to be similar in their chromatographic properties; this suggests that a single enzyme is responsible for dehydrative deamination of both hydroxyamino acids in rat liver tissue. Stability of the enzyme preparations was distinctly increased after DEAE cellulose chromatography. The yield of crystalline threonine (serine) dehydratase was about 3%; the enzyme was purified 1500-1800-fold.

A D-serine dehydratase acting also on L-serine from Klebsiella pneumoniae.

D-Serine dehydratase [EC 4.2.1.14] was purified from a strain of Klebsiella pneumoniae 140-fold from crude extract with a yield of 5%. This enzyme catalyzed formation of pyruvate and ammonia not only from D-serine but also from L-serine, and also catalyzed the formation of alpha-ketobutyrate and ammonia from D-threonine. Km values for D-serine, L-serine, and D-threonine were 2.8 mM, 20 mM, and 3.6 mM, respectively. Km for pyridoxal 5'-phosphate was 2.5 micron. The molecular weight was estimated to be 46,000 by Sephadex G-150 gel filtration and 40,000 by SDS-polyacrylamide gel electrophoresis. This enzyme was inducible by D-serine. Induction by casamino acids appeared to depend on the presence of D-serine.

L-serine dehydratase from Arthrobacter globiformis.

1. L-Serine dehydratase (EC 4.2.1.13) was purified 970-fold from glycine-grown Arthrobacter globiformis to a final specific activity of 660micronmol of pyruvate formed/min per mg of protein. 2. The enzyme is specific for L-serine; D-serine, L-threonine and L-cysteine are not attacked. 3. The time-course of pyruvate formation by the purified enzyme, in common with enzyme in crude extracts and throughout the purification, is non-linear. The reaction rate increases progressively for several minutes before becoming constant. The enzyme is activated by preincubation with L-serine and a linear time-course is then obtained. 4. The substrate-saturation curve for L-serine is sigmoid. The value of [S]0.5 varies with protein concentration, from 6.5mM at 23microng/ml to 20mM at 0.23microng/ml. The Hill coefficient remains constant at 2.9.5 The enzyme shows a non-specific requirement for a univalent or bivalent cation. Half-maximal activity is produced by 1.0mM-MgCl2 or by 22.5mM-KCl. 6. L-Cysteine and D-serine act as competitive inhibitors of L-serine dehydratase, with Ki values of 1.2 and 4.9mM respectively. L-Cysteine, at higher concentrations, also causes a slowly developing irreversible inhibition of the enzyme. 7. Inhibition by HgCl2 (5micronM)can be partially reversed in its initial phase by 1mM-L-cysteine, but after 10 min it becomes irreversible. 8. In contrast with the situation in all cell-free preparations, toluene-treated cells of A. globiformis form pyruvate from L-serine at a constant rate from the initiation of the reaction, show a hyperbolic substrate-saturation curve with an apparent Km of 7mM and do not require a cation for activity.

Sequence studies on D-serine dehydratase of Escherichia coli. Primary structure of the tryptic phosphopyridoxyl peptide and of the N-terminus.

An improved procedure for large-scale production of crystalline D-serine dehydratase (EC 4.2.1.14) from Escherichia coli is described. The N-terminal sequence of the enzyme (Mr 45500) was determined in a solid-phase sequencer as Met-Glu-Asn-Ala-Lys-Met-Asn-Ser-Leu-Ile-Ala-Gln-Tyr-Pro-Leu-Val-Lys-Asp-Leu-Val-Ala-LEU-Lys. Four of the first five N-terminal residues are homologeous with tryptophanase, another pyridoxal-phosphate (P-Pxy) enzyme that catalyzes alpha,beta-elimination reactions. After borohydride reduction and tryptic digestion of the enzyme, a peptide was isolated showing the sequence Lys-Asp-Ser-His-Leu-Pro-Ile-Ser-Gly-Ser-Ile-Lys(P-Pxy)-Ala-Arg. No clear homology of this portion of the enzyme with tryptophanase or another pyridoxal-phosphate enzyme was observed.

[Guinea pig liver cytosol has an L-threonine deaminase activity distinct from L-serine deaminase]. / Mise en évidence dans le cytosol du foie de cobaye d'une activité L-thréonine désaminasique distinct de la L-sérine désaminase

Using sodium sulfate precipitation, "Sephadex G200" gel filtration and polyacrylamide gel electrophoresis, a L-threonine desaminase was demonstrated in the Guinea-Pig liver cytosol. This enzyme was separated from the guinea pig liver L-serine desaminase possessing an auxiliary activity on L-threonine substrate described by us in a previous work. The optimals for pH (7,1) and temperature (+ 55 degrees C) and the apparent molecular weight (134,000 + 20,000) were established.

Immunochemical studies of serine dehydratase and ornithine aminotransferase regulation in rat liver in vivo.

Previous studies of serine dehydratase (EC 4.2.1.13) and ornithine aminotransferase (EC 2.6.1.13) adaptation in rat liver showed that in rats on a high protein diet, glucocorticoid administration increased serine dehydratase activity while simultaneously reducing the activity of ornithine aminotransferase. The present study examines the role of enzyme synthesis in the expression of these and other dissimilar adaptive characteristics of the two enzymes. Both enzymes were purified to crystallinity and used to prepare specific antibodies. Changes in the rate of synthesis of each enzyme during adaptation were then measured immunochemically. In rats fed ad libitum, the synthetic rates for both enzymes exhibited circadian rhythm, although enzyme levels remained relatively constant. The circadian cycle for ornithine aminotransferase synthesis was in phase with the cycles for body weight and relative liver weight (maxima at 9 a.m., minima at 9 p.m.) but was approximately 12 hours out of phase with the cycle for serine dehydratase synthesis. 9alpha-Fluoro-11beta, 21-dihydroxy-16alpha, 17alpha-isopted at 9 a.m., increased serine dehydratase synthesis and simultaneously decreased the synthesis of ornithine aminotransferase. When triamcinolone was injected at 9 p.m., however, serine dehydratase synthesis was not stimulated, although the reduction of ornithine aminotransferase synthesis was still produced. These results suggest that: (a) circadian cycling of synthesis may be a general phenomenon in enzyme regulation even though for enzymes with relatively long half-lives, such cycling may not be reflected as fluctuations in enzyme levels; (b) such circadian rhythmicity may also involve cyclic changes in the responsiveness of the enzyme-forming system to regulatory stimuli; (c) whereas the adaptive behavior of serine dehydratase typifies that of amino acid-catabolizing enzymes in general, the responses of ornithine aminotransferase denote a functional association of this enzyme with anabolic processes. On this basis, the possibility that ornithine aminotransferase plays a pivotal role in the regulation of urea cycle activity and nitrogen balance is discussed.