The action of alpha-ketoglutarate dehydrogenase on 4-chloronitrosobenzene: evidence for species-dependent differences in active site properties

The reaction of bovine (Bos taurus) and porcine (Sus scrufa) cardiac alpha-ketoglutarate dehydrogenase complex (alpha-KGD) with 4-chloronitrosobenzene (I) was shown to produce a hydroxamic acid (IV) and a product due to a Bamberger rearrangement as previously shown for Escherichia coli alpha-KGD. The conversion of I into an active site-bound electrophile was general among the three alpha-KGD enzymes tested, but quantitative differences in products and kinetics were shown. The reaction of I was specific for the resolved alpha-ketoglutarate decarboxylase subunit.     

Functional nonequivalence of transketolase active centers

Transketolase (TK, EC 2.2.1.1), the key enzyme of the non-oxidative branch of pentose phosphate pathway of hydrocarbon transformation, plays an important role in a system of substrate rearrangement between pentose shunt and glycolysis, acting as a reversible link between the two metabolic pathways. In addition, it supplies precursors for biosyntheses of nucleotides, aromatic amino acids, and vitamins. In plants, the enzyme plays a central role in the Calvin cycle. TK catalyzes interconversion of sugar phosphates. Thiamine diphosphate (TDP) and bivalent cations serve as its cofactors. Being a typical TDP-dependent enzyme, TK is the least complex representative of this group of enzymes, and this accounts for its use as a model in studies of their structure and mechanism of action. TK is readily crystallized, this being the reason why the first crystal X-ray structure analysis of TDP-dependent enzymes was performed with a TK sample. Both the general structure of TK and the structures of its active centers have been studied in detail. In this article, we review experimental evidence of functional nonequivalence of the two active centers of TK, which are known to be identical by crystal X-ray structure analysis.     

The carboxyl group in the active center of transketolase

Transketolase from baker's yeast is rapidly inactivated in the presence of 1-ethyl-3 (3'-dimethylaminopropyl)-carbodiimide or Woodward's reagent K. In both cases the kinetics of inactivation is biphasic, which agrees with the presence of two active centers in the enzyme molecule differing in their sensitivity to the inhibitors. There is some evidence that inactivation of transketolase is due to modification of carboxyl groups of enzyme. Complete inactivation is achieved by modification of one carboxyl per active site of the enzyme. The experimental results suggest that the carboxyl group is essential for the enzymatic activity of transketolase.     

Functional carboxylic group in the active center of transketolase

Baker's yeast transketolase is rapidly inactivated in the presence of carboxylic group modifiers, i.e., 1-ethyl-3(3'-dimethylaminopropyl)-carbodiimide or Woodward's reagent K. This inactivation is due to modification of the carboxylic group in the enzyme active center. The essential groups localized in the two active centers of transketolase differ in the rate of modification; accordingly, the inactivation kinetics appears as biphasic. A complete loss of the enzyme activity occurs as a result of modification of one carboxylic group per enzyme active center. The pKa value of modifiable groups is equal to about 6.5. This modification decreases by two orders of magnitude the affinity of the substrate for the active center. The carboxylic groups are not directly involved in the interaction with the substrates; their modification does not significantly affect the coenzyme binding. It is supposed that these groups are responsible for the deprotonation of the second carbon in the thiamine pyrophosphate thiazolium ring.     

The functional identity of the active centres of transketolase.

Direct determination of the number of catalytically active molecules of the coenzyme in holotransketolase (sedoheptulose-7-phosphate:D-glyceraldehyde-3-phosphate glycoaldehydetransferase, EC 2.2.1.1) has corroborated our previous data indicating that in the native enzyme there are two active centres. They have been provided to be functionally identical. It has been shown that the decrease in the specific activity of transketolase during its storage is due to inactivation of one of the active centres, having a lower affinity for the coenzyme. The second active centre retains thereby its full catalytic activity.     

Binding of the coenzyme and formation of the transketolase active center

Transketolase (TK) is a homodimer, the simplest representative of thiamine diphosphate (ThDP)-dependent enzymes. It was first ThDP-dependent enzymes the crystal structure of which has been solved and revealed the general fold for this class of enzymes and the interactions of the non-covalently bound coenzyme ThDP with the protein component. Transketolase is a convenient model to study the structure(s) of the active center and the mechanism of action of ThDP-dependent enzymes. This review summarizes the results of studies on the kinetics of the interaction of ThDP with TK from Saccharomyces cerevisiae as well as the generation of the catalytically active form of the coenzyme within the holoenzyme and formation of the enzyme's active center.     

An investigation of the carboxyl group function in the active center of transketolase

Transketolase from baker's yeast is rapidly inactivated in the presence of 1-ethyl-3 (3'-dimethylaminopropyl)-carbodiimide. pKa of the modified carboxyl groups is approximately 6.5. An investigation of the initial steps of enzymatic catalysis monitored by a changes in the circular dichroism spectra and in an oxidation reaction with ferricyanide made it possible to conclude that the modification interferes with the donor substrate attachment to the enzyme. Evidence obtained was suggesting that the carboxyl group of the active center facilitates dissociation of a proton from the carbon atom in the second position of the thiamine pyrophosphate thiazolium ring.     

The nonmicrosomal production of N-(4-chlorophenyl)-glycolhydroxamic acid from 4-chloronitrosobenzene by rat liver homogenates

The incubation of 4-chloronitrosobenzene (4-CNB) with subcellular fractions of rat liver resulted in the formation of a previously unknown type of hydroxamic acid metabolite for mammals. This new metabolite, N-(4-chlorophenyl)glycolhydroxamic acid (Gl-CHA), is most likely formed through the action of liver transketolase on the substrate 4-CNB. Gl-CHA was produced only by the 10 000g and 105 000g supernatant fractions, and required glucose-6-phosphate as an energy source. No hydroxamic acid metabolites were produced in detectable quantities by the microsomal fraction of the rat liver homogenate. Gl-CHA was positively identified by isolation and comparison to an authentic sample of Gl-CHA. Authentic Gl-CHA was prepared by the condensation of 4-chlorophenylhydroxylamine with glycolic acid in the presence of dicylohexylcarbodiimide. The highest observed conversion of 4-CNB to Gl-CHA was 18%, which occurred at the lowest concentration of 4-CNB incubated with the 105 000g supernatant. Gl-CHA was not produced by C-hydroxylation of the corresponding acetyl-derived hydroxamic acid, since none of the subcellular fractions of rat liver would effect this conversion. The incubation of 4-chloroaniline under identical conditions failed to result in the production of Gl-CHA; however, such an observation is probably not important to the possibility that Gl-CHA might be a significant metabolite in vivo.     

Evidence for the mechanism of hydroxylation by 4-hydroxyphenylpyruvate dioxygenase and hydroxymandelate synthase from intermediate partitioning in active site variants

4-Hydroxyphenylpyruvate dioxygenase (HPPD) and hydroxymandelate synthase (HMS) each catalyze similar complex dioxygenation reactions using the substrates 4-hydroxyphenylpyruvate (HPP) and dioxygen. The reactions differ in that HPPD hydroxylates at the ring C1 and HMS at the benzylic position. The HPPD reaction is more complex in that hydroxylation at C1 instigates a 1,2-shift of an aceto substituent. Despite that multiple intermediates have been observed to accumulate in single turnover reactions of both enzymes, neither enzyme exhibits significant accumulation of the hydroxylating intermediate. In this study we employ a product analysis method based on the extents of intermediate partitioning with HPP deuterium substitutions to measure the kinetic isotope effects for hydroxylation. These data suggest that, when forming the native product homogentisate, the wild-type form of HPPD produces a ring epoxide as the immediate product of hydroxylation but that the variant HPPDs tended to also show the intermediacy of a benzylic cation for this step. Similarly, the kinetic isotope effects for the other major product observed, quinolacetic acid, showed that either pathway is possible. HMS variants show small normal kinetic isotope effects that indicate displacement of the deuteron in the hydroxylation step. The relatively small magnitude of this value argues best for a hydrogen atom abstraction/rebound mechanism. These data are the first definitive evidence for the nature of the hydroxylation reactions of HPPD and HMS.     

Inhibition of transketolase by p-hydroxyphenylpyruvate

The effect of p-hydroxyphenylpyruvate, a natural analogue of transketolase substrate, on the catalytic activity of the enzyme was investigated. p-Hydroxyphenylpyruvate proved to be a reversible and competitive inhibitor of transketolase with respect to substrate; it was also able to displace thiamine diphosphate from holotransketolase. The data suggest that p-hydroxyphenylpyruvate participates in the regulation of tyrosine biosynthesis by influencing the catalytic activity of transketolase.     

Induction of the soxRS regulon of Escherichia coli by glycolaldehyde

The ability of short-chain sugars to cause oxidative stress has been examined using glycolaldehyde as the simplest sugar. Short-chain sugars autoxidize in air, producing superoxide and alpha,beta-dicarbonyls. In Escherichia coli the soxRS regulon mediates an oxidative stress response, which protects the cell against both superoxide-generating agents and nitric oxide. In superoxide dismutase-deficient E. coli mutants, glycolaldehyde induces fumarase C and nitroreductase A, which are regulated as members of the soxRS regulon. A mutational defect in soxRS eliminates that induction. This establishes that glycolaldehyde can cause induction of this defensive regulon. This effect of glycolaldehyde was oxygen-dependent, was not shown by glyoxal, and was not seen in the superoxide dismutase-replete parental strain, and it was abolished by a cell-permeable SOD mimetic. All of these suggest that superoxide radicals produced by the oxidation of glycolaldehyde played a key role in the induction.     

The number of active sites in a molecule of transketolase

It has been demonstrated that the previously described changes in the optical properties of apotransketolase interacting with thyamine pyrophosphate are associated only with the catalytically active molecules of coenzyme being bound. Titration of apoenzyme by TPP has shown a molecule of transketolase to have two active centres.     

Inhibition of transketolase by N-acetylimidazole

Transketolase from baker's yeast is rapidly inactivated in the presence of N-acetylimidazole. According to kinetic data, acetylation of one amino acid residue of the protein per active site is sufficient for TK* inactivation. The holoenzyme is inhibited more slowly than is apotransketolase. The presence of a tyrosine residue in the enzyme's active site, essential for activity, is suggested.     

Nonequivalence of transketolase active centers with respect to acceptor substrate binding

The interaction of transketolase with its acceptor substrate, ribose 5-phosphate, has been studied. The active centers of the enzyme were shown to be functionally nonequivalent with respect to ribose 5-phosphate binding. Under the conditions where only one out of the two active centers of transketolase is functional, their affinities for ribose 5-phosphate are identical. The phenomenon of nonequivalence becomes apparent when the substrate interacts with one of the two active centers. As a consequence of such interaction, the affinity of the second active center for ribose 5-phosphate decreases.     

Biotransformation of β‐hydroxypyruvate and glycolaldehyde to l‐erythrulose by Pichia pastoris strain GS115 overexpressing native transketolase

Transketolase is a proven biocatalytic tool for asymmetric carbon‐carbon bond formation, both as a purified enzyme and within bacterial whole‐cell biocatalysts. The performance of Pichia pastoris as a host for transketolase whole‐cell biocatalysis was investigated using a transketolase‐overexpressing strain to catalyze formation of l ‐erythrulose from β‐hydroxypyruvic acid and glycolaldehyde substrates. Pichia pastoris transketolase coding sequence from the locus PAS_chr1‐4_0150 was subcloned downstream of the methanol‐inducible AOX1 promoter in a plasmid for transformation of strain GS115, generating strain TK150. Whole and disrupted TK150 cells from shake flasks achieved 62% and 65% conversion, respectively, under optimal pH and methanol induction conditions. In a 300 μL reaction, TK150 samples from a 1L fed‐batch fermentation achieved a maximum l ‐erythrulose space time yield (STY) of 46.58 g L^?1 h^?1, specific activity of 155 U , product yield on substrate (Y_p/s) of 0.52 mol mol^?1 and product yield on catalyst (Y_p/x) of 2.23g . We have successfully exploited the rapid growth and high biomass characteristics of Pichia pastoris in whole cell biocatalysis. At high cell density, the engineered TK150 Pichia pastoris strain tolerated high concentrations of substrate and product to achieve high STY of the chiral sugar l ‐erythrulose. © 2017 The Authors Biotechnology Progress published by Wiley Periodicals, Inc. on behalf of American Institute of Chemical Engineers Biotechnol. Prog., 34:99–106, 2018