Site-directed mutagenesis of the ionizable groups in the active site of Zymomonas mobilis pyruvate decarboxylase: effect on activity and pH dependence.

Pyruvate decarboxylase (PDC, EC 4.1.1.1) is a thiamin diphosphate-dependent enzyme about which there is a large body of structural and functional information. The active site contains several absolutely conserved ionizable groups and all of these appear to be important, as judged by the fact that mutation diminishes or abolishes catalytic activity. Previously we have shown [Schenk, G., Leeper, F.J., England, R., Nixon, P.F. & Duggleby, R.G. (1997) Eur. J. Biochem. 248, 63-71] that the activity is pH-dependent due to changes in kcat/Km while kcat itself is unaffected by pH. The effect on kcat/Km is determined by a group with a pKa of 6.45; the identity of this group has not been determined, although H113 is a possible candidate. Here we mutate five crucial residues in the active site with ionizable side-chains (D27, E50, H113, H114 and E473) in turn, to residues that are nonionizable or should have a substantially altered pKa. Each protein was purified and characterized kinetically. Unexpectedly, the pH-dependence of kcat/Km is largely unaffected in all mutants, ruling out the possibility that any of these five residues is responsible for the observed pKa of 6.45. We conjecture that the kcat/Km profile reflects the protonation of an alcoholate anion intermediate of the catalytic cycle.

Congenital lactic acidosis: evaluation of the properties of the a199t natural variant of human pyruvate dehydrogenase e1alpha by in vitro mutation.

One cause of congenital lactic acidosis is a mutation in the E1 alpha-subunit of the pyruvate dehydrogenase multienzyme complex. Little is known about the consequences of these mutations at the enzymatic level. Here we study the A199T mutation by expressing the protein in Escherichia coli. The specific activity is 25% of normal and the K(m) for pyruvate is elevated by 10-fold. Inhibitors of lactate dehydrogenase might be a useful therapy for patients with such mutations.

Glucose induced IEG expression in the thiamin-deficient rat brain.

Glucose loading of rats made thiamin deficient by dietary deprivation of thiamin and the administration of pyrithiamin (40 microg/100 g, i.p.) precipitates an acute neuropathy, a model of Wernicke's encephalopathy in man (Zimitat and Nixon, Metab. Brain Dis. 1999;14:1-20). Immunohistochemical detection of Fos proteins was used as a marker to identify neuronal populations in the thiamin-deficient rat brain affected by glucose loading. As thiamin deficiency progressed, the extent and intensity of Fos-like immunoreactivity (FLI) in brain structures typically affected by thiamin deficiency (the thalamus, mammillary bodies, inferior colliculus, vestibular nucleus and inferior olives) were markedly increased when compared to thiamin-replete controls. Glucose loading for 1-3 days further increased the intensity of FLI in these same regions, consistent with a dependence of Fos expression on carbohydrate metabolism as well as on thiamin deficiency. The timed acute changes that follow a bolus glucose load administered to thiamin-deficient animals may provide a sequential account of events in the pathogenesis of brain damage in this model of Wernicke's encephalopathy.

Mutagenesis at asp27 of pyruvate decarboxylase from Zymomonas mobilis. Effect on its ability to form acetoin and acetolactate.

Pyruvate decarboxylase (PDC) is one of several enzymes that require thiamin diphosphate (ThDP) and a bivalent cation as essential cofactors. The three-dimensional structure of PDC from Zymomonas mobilis (ZMPDC) shows that Asp27 (D27) is close to ThDP in the active site, and mutagenesis of this residue has suggested that it participates in catalysis. The normal product of the PDC reaction is acetaldehyde but it is known that the enzyme can also form acetoin as a by-product from the hydroxyethyl-ThDP reaction intermediate. This study focuses on the role of D27 in the production of acetoin and a second by-product, acetolactate. D27 in ZMPDC was altered to alanine (D27A) and this mutated protein, the wild-type, and two other previously constructed PDC mutants (D27E and D27N) were expressed and purified. Determination of the kinetic properties of D27A showed that the affinity of D27A for ThDP is decreased 30-fold, while the affinity for Mg2+ and the Michaelis constant for pyruvate were similar to those of the wild-type. The time-courses of their reactions were investigated. Each mutant has greatly reduced ability to produce acetaldehyde and acetoin compared with the wild-type PDC. However, the effect of these mutations on acetaldehyde production is greater than that on acetoin formation. The D27A mutant can also form acetolactate, whereas neither of the other mutants, nor the wild-type PDC, can do so. In addition, acetaldehyde formation and/or release are reversible in wild-type ZMPDC but irreversible for the mutants. The results are explained by a mechanism involving thermodynamic and geometric characteristics of the intermediates in the reaction.

Effects of deletions at the carboxyl terminus of Zymomonas mobilis pyruvate decarboxylase on the kinetic properties and substrate specificity.

The three-dimensional structure of Zymomonas mobilis pyruvate decarboxylase shows that the carboxyl-terminal region of the protein occludes the active site. This observation is consistent with earlier suggestions that the active site is inaccessible to solvent during catalysis. However, the carboxyl-terminal region must move aside to allow entry of the substrate, and again to permit the products to leave. Here we have examined the role of the carboxyl terminus by making 15 variants of the enzyme with serial deletions. The activity is largely unaffected by removal of up to seven residues but deletion of the next two, R561 and S560, results in a drastic loss of activity. Five of these deletion mutants were purified and fully characterized and showed progressive decreases in activity, in the ability to discriminate between pyruvate and larger substrates, and in cofactor affinity. Several substitution mutants at residues R561 and S560 were prepared, purified, and fully characterized. The results indicate important roles for the side-chain of R561 and the backbone atoms of S560. It is suggested that the carboxyl-terminal region of pyruvate decarboxylase is needed to lock in the cofactors and for the proper closure of the active site that is required for discrimination between substrates and for decarboxylation to occur.

Glucose loading precipitates acute encephalopathy in thiamin-deficient rats.

A rat model of glucose-precipitated Wernicke's encephalopathy (WE) has been developed in which glucose loading (10 g/kg, i.p.) of ataxic thiamin-deficient (TD) rats induced episodes of gross neurological dysfunction and sometimes death. The acute effects of a glucose load on the neurological state of thiamin-replete control and TD rats were assessed by scoring of clinical observations and performance measured on a moving belt (MB) apparatus at 30 min intervals for 2 hr after the challenge. Glucose loading or saline treatment (2.5 mL, i.p.) had no significant behavioural or clinical consequences when administered to controls or rats fed TD diet for <21 days. Glucose loading of ataxic rats fed TD diet for 28-35 days precipitated episodes of gross ataxia and signs of advanced neurological dysfunction (e.g. loss of righting reflex and hyperexcitability) leading to significant increases in the Ataxia (p<0.05) and Advanced Sign (p<0.05) scores within 2 hr after the challenge. Simultaneously, the performance of these animals on the MB decreased 10-fold. Regular glucose challenges significantly increased the rate of progression of disease in TD rats when compared with untreated TD rats. This model may be useful for the further investigation of the pathogenesis of WE at the molecular level.

Aspartate-27 and glutamate-473 are involved in catalysis by Zymomonas mobilis pyruvate decarboxylase.

Zymomonas mobilis pyruvate decarboxylase (EC 4.1.1.1) was subjected to site-directed mutagenesis at two acidic residues near the thiamin diphosphate cofactor in the active site. Asp-27 was changed to Glu or Asn, and Glu-473 was mutated to Asp (E473D) or Gln (E473Q). Each mutant protein was purified to near-homogeneity, and the kinetic and cofactor-binding properties were compared with those of the wild-type protein. Despite the very conservative nature of these alterations, all mutants had a very low, but measurable, specific activity ranging from 0.025% (E473Q) to 0.173% (E473D) of the wild type. With the exception of E473Q, the mutants showed small decreases in the affinity for thiamin diphosphate, and binding of the second cofactor (Mg2+) was also weakened somewhat. With E473Q, both cofactors seemed to be very tightly bound so that they were not removed by the treatment that was effective for the wild-type enzyme and other mutant forms. All mutants showed minor changes in the Km for substrate, but these alterations did not account for the low activities. These low specific activities, accompanied by little change in the Km for pyruvate, are consistent with a quantitative model of the catalytic cycle in which the main effect of the mutations is to slow the decarboxylation step with a minor change in the rate constant for pyruvate binding.

Identification of the catalytic glutamate in the E1 component of human pyruvate dehydrogenase.

The pyruvate dehydrogenase complex catalyzes the conversion of pyruvate to acetyl-CoA. The first component (E1) converts pyruvate to bound acetaldehyde using thiamine diphosphate (ThDP) and Mg2+ as cofactors. There is no 3D structure of E1 available but those of other ThDP-dependent enzymes show some similarities including a glutamate residue that assists in ThDP activation. Eukaryotic E1 has an alpha2beta2 structure and the conserved Glu89 of the beta-subunit was identified as a possible catalytic residue by sequence alignment. Human E1 was expressed in Escherichia coli and purified. Mutating Glu89 to glutamine, aspartate and alanine markedly reduces catalytic activity and the affinity for ThDP, consistent with a role as the catalytic glutamate.

Heterologous expression of human transketolase.

Transketolase belongs to the family of thiamin diphosphate dependent enzymes. The aim of this study was to establish a bacterial expression system for human transketolase in order to investigate the functional characteristics of mammalian transketolases. The level of recombinant human enzyme expressed in Escherichia coli was modest. Purification of recombinant transketolase and separation from the E. coli enzyme has been greatly simplified by means of a non-cleavable hexa-histidine tag. The highest specific activity was 13.5 U/mg and the K(m) values were 0.27 +/- 0.02 and 0.51 +/- 0.05 mM for the substrates D-xylulose 5-phosphate and D-ribose 5-phosphate, respectively. Binding of cofactors to the apoenzyme showed the expected hysteresis. Without preincubation, the K(m) values for thiamin diphosphate and for Mg2+ were, respectively, 4.1 +/- 0.8 and 2.5 +/- 0.4 microM, but after 1 h of preincubation these values were 85 +/- 16 nM and 0.74 +/- 0.23 microM. The kinetic constants are similar to those of the native enzyme purified from human erythrocytes. Despite the modest expression level the reported system is well suited to a variety of functional studies.

Changes in the hippocampus induced by glucose in thiamin deficient rats detected by MRI.

Using T2-weighted Magnetic Resonance Imaging (MRI) in a pyrithiamin-treated, thiamin deficient (TD) rat model of Wernicke's encephalopathy (WE), we have observed hyperintensity in the thalamus, hypothalamus, collicular bodies and hippocampus which was enhanced 40 min after a glucose load. Hyperintensity was not evident in these structures in thiamin replete rats receiving glucose nor was it enhanced in TD rats administered 2-deoxyglucose. Residual hyperintensity was still evident in the hippocampus as long as 30 days after thiamin administration and was increased by repeat glucose challenge at that time. These data indicate that the hippocampus is as vulnerable as the thalamus to some persistent pathological change when glucose is metabolised in a state of thiamin deficiency.

Properties and functions of the thiamin diphosphate dependent enzyme transketolase.

This review highlights recent research on the properties and functions of the enzyme transketolase, which requires thiamin diphosphate and a divalent metal ion for its activity. The transketolase-catalysed reaction is part of the pentose phosphate pathway, where transketolase appears to control the non-oxidative branch of this pathway, although the overall flux of labelled substrates remains controversial. Yeast transketolase is one of several thiamin diphosphate dependent enzymes whose three-dimensional structures have been determined. Together with mutational analysis these structural data have led to detailed understanding of thiamin diphosphate catalysed reactions. In the homodimer transketolase the two catalytic sites, where dihydroxyethyl groups are transferred from ketose donors to aldose acceptors, are formed at the interface between the two subunits, where the thiazole and pyrimidine rings of thiamin diphosphate are bound. Transketolase is ubiquitous and more than 30 full-length sequences are known. The encoded protein sequences contain two motifs of high homology; one common to all thiamin diphosphate-dependent enzymes and the other a unique transketolase motif. All characterised transketolases have similar kinetic and physical properties, but the mammalian enzymes are more selective in substrate utilisation than the nonmammalian representatives. Since products of the transketolase-catalysed reaction serve as precursors for a number of synthetic compounds this enzyme has been exploited for industrial applications. Putative mutant forms of transketolase, once believed to predispose to disease, have not stood up to scrutiny. However, a modification of transketolase is a marker for Alzheimer's disease, and transketolase activity in erythrocytes is a measure of thiamin nutrition. The cornea contains a particularly high transketolase concentration, consistent with the proposal that pentose phosphate pathway activity has a role in the removal of light-generated radicals.

The role of His113 and His114 in pyruvate decarboxylase from Zymomonas mobilis.

Pyruvate decarboxylase (PDC) is one of several enzymes that require thiamin diphosphate (ThDP) and a divalent cation as essential cofactors. Recently, the three-dimensional structures of the enzyme from two yeasts have been determined. While these structures shed light on the binding of the cofactors and the reaction mechanism, the interactions between the substrate pyruvate and the enzyme remain unclear. We have used PDC from Zymomonas mobilis as a model for these enzymes in order to study substrate binding. The recombinant enzyme was expressed in Escherichia coli. High yield, simplicity of purification, high stability and simple kinetics make this model well suited for these studies. Activity measurements in the pH range between 5.8 and 8.5 indicated that a His residue may be involved in substrate binding. Analysis of an alignment of all known PDC protein sequences showed two invariant His residues (His113 and His114) which, according to the crystal structure of yeast PDC, are in the vicinity of the active site. Here we demonstrate that replacement of His114 by Gln does not have a great effect on cofactor and substrate binding. However, the k(cat) is decreased indicating that His114 may assist in catalysis. In contrast, substitution of His113 by Gln renders the enzyme completely inactive. This mutant has decreased affinity for both cofactors, as revealed by measurements of tryptophan fluorescence quenching. However, this decreased affinity is insufficient to account for the complete loss of activity. Despite its inability to support overall catalysis, this [Gln113]PDC mutant is capable of releasing acetaldehyde from 2-(1-hydroxyethyl)thiamin diphosphate supplied exogenously. It is proposed that upon substrate binding, His113 is placed close to C2 of the thiazole ring. Subsequent deprotonation of this atom leads to a conformational change that allows a flexible loop (residues 105-112) that precedes His113 to close over the active site. Hence, replacement of His113 by another residue interferes with this closure of the active site and thus disrupts the catalytic process.

Molecular evolutionary analysis of the thiamine-diphosphate-dependent enzyme, transketolase.

Members of the transketolase group of thiamine-diphosphate-dependent enzymes from 17 different organisms including mammals, yeast, bacteria, and plants have been used for phylogenetic reconstruction. Alignment of the amino acid and DNA sequences for 21 transketolase enzymes and one putative transketolase reveals a number of highly conserved regions and invariant residues that are of predicted importance for enzyme activity, based on the crystal structure of yeast transketolase. One particular sequence of 36 residues has some similarities to the nucleotide-binding motif and we designate it as the transketolase motif. We report further evidence that the recP protein from Streptococcus pneumoniae might be a transketolase and we list a number of invariant residues which might be involved in substrate binding. Phylogenies derived from the nucleotide and the amino acid sequences by various methods show a conventional clustering for mammalian, plant, and gram-negative bacterial transketolases. The branching order of the gram-positive bacteria could not be inferred reliably. The formaldehyde transketolase (sometimes known as dihydroxyacetone synthase) of the yeast Hansenula polymorpha appears to be orthologous to the mammalian enzymes but paralogous to the other yeast transketolases. The occurrence of more than one transketolase gene in some organisms is consistent with several gene duplications. The high degree of similarity in functionally important residues and the fact that the same kinetic mechanism is applicable to all characterized transketolase enzymes is consistent with the proposition that they are all derived from one common ancestral gene. Transketolase appears to be an ancient enzyme that has evolved slowly and might serve as a model for a molecular clock, at least within the mammalian clade.

The role of residues glutamate-50 and phenylalanine-496 in Zymomonas mobilis pyruvate decarboxylase.

Several enzymes require thiamine diphosphate (ThDP) as an essential cofactor, and we have used one of these, pyruvate decarboxylase (PDC; EC 4.1.1.1) from Zymomonas mobilis, as a model for this group of enzymes. It is well suited for this purpose because of its stability, ease of purification, homotetrameric subunit structure and simple kinetic properties. Crystallographic analyses of three ThDP-dependent enzymes [Müller, Lindqvist, Furey, Schulz, Jordan and Schneider (1993) Structure 1, 95-103] have suggested that an invariant glutamate participates in catalysis. In order to evaluate the role of this residue, identified in PDC from Zymomonas mobilis as Glu-50, it has been altered to glutamine and aspartate by site-directed mutagenesis of the cloned gene. The mutant proteins were expressed in Escherichia coli. Here we demonstrate that substitution with aspartate yields an enzyme with 3% of the activity of the wild-type, but with normal kinetics for pyruvate. Replacement of Glu-50 with glutamine yields an enzyme with only 0.5% of the catalytic activity of the wild-type enzyme. Each of these mutant enzymes has a decreased affinity for both ThDP and Mg2+. It has been reported that the binding of cofactors to apoPDC quenches the intrinsic tryptophan fluorescence [Diefenbach and Duggleby (1991) Biochem. J. 276, 439-445] and we have identified the residue responsible as Trp-487 [Diefenbach, Candy, Mattick and Duggleby (1992) FEBS Lett. 296, 95-98]. Although this residue is some distance from the cofactor binding site, it lies in the dimer interface, and the proposal has been put forward [Dyda, Furey, Swaminathan, Sax, Farrenkopf and Jordan (1993) Biochemistry 32, 6165-6170] that alteration of ring stacking with Phe-496 of the adjacent subunit is the mechanism of fluorescence quenching when cofactors bind. The closely related enzyme indolepyruvate decarboxylase (from Enterobacter cloacae) has a leucine residue at the position corresponding to Phe-496 but shows fluorescence quenching properties that are similar to those of PDC. This suggests that the fluorescence quenching is due to some perturbation of the local environment of Trp-487 rather than to a specific interaction with Phe-496. This latter hypothesis is supported by our data: mutation of this phenylalanine to leucine, isoleucine or histidine in PDC does not eliminate the fluorescence quenching upon addition of cofactors.

MRI demonstration of impairment of the blood-CSF barrier by glucose administration to the thiamin-deficient rat brain.

Contrast-enhanced T1-weighted spin-echo magnetic resonance imaging (MRI) has demonstrated that Gd-diethylenetriaminepentaacetate (Gd-DTPA), which normally does not cross the blood-brain or blood-CSF barriers, does so approximately 40 min after administration of glucose to a vitamin B1 deficient rat. The period of the onset of this blood-CSF or blood-brain barrier dysfunction coincides with our previous observations of accumulation of glutamate or glutamate derivatives following an equivalent glucose load under identical conditions of thiamin deficiency, consistent with a relationship between these two observations. The dysfunction was reversed when a thiamin deficient animal was made thiamin replete.

Effect of chronic alcohol ingestion on hepatic folate distribution in the rat.

The mechanism by which ethanol impairs folate metabolism remains uncertain. In the present study, we used our new technique (affinity/HPLC) for folate analysis to study the effect of chronic alcohol ingestion on the content and distribution of folates in livers. Twelve male Sprague-Dawley rats (180 g) were divided into two groups, and fed for 4 weeks with Lieber-DeCarli semi-liquid isocaloric diets, with and without 5% ethanol. Livers were extracted in boiling, pH 9.3 borate buffers containing ascorbate/dithioerythritol. Folates in the supernatant fractions were purified by affinity chromatography and analyzed using ion pair high performance liquid chromatography. The data obtained showed that hepatic folate distribution in alcohol-treated rats differed from that of control animals in two ways. Livers from the ethanol-fed rats, when compared with those from control rats, exhibited increases in the percent concentrations of methylated tetrahydrofolates (21.46 +/- 2.21 vs 14.8 +/- 1.23), decreases in the percent concentrations of formylated tetrahydrofolates (25.62 +/- 4.02 vs 46.18 +/- 2.65) and higher concentrations of unsubstituted tetrahydrofolates (52.91 +/- 3.84 vs 38.88 +/- 2.50). In addition, alcohol ingestion was associated with longer glutamate chains of the folate molecules, characterized by lower relative concentrations of pentaglutamyl folates (29 vs 48%), and higher relative concentrations of hexa- and heptaglutamyl folates (55 vs 46% and 15 vs 6%) when compared with controls. The data are discussed in relation to the possibility that alcohol exerts its effect through: (1) inhibition of B12-dependent methyl transfer from methyltetrahydrofolate to homocysteine; (2) diversion of formylated tetrahydrofolates toward serine synthesis; and (3) interaction of acetaldehyde with tetrahydrofolates, thereby interfering with folate coenzyme metabolism.

Reconstitution of holotransketolase is by a thiamin-diphosphate-magnesium complex.

When human erythrocyte apo-transketolase is mixed with cofactors and substrates, the progress curve exhibits a lag phase. Elimination of the lag phase requires the presence of saturating concentrations of cofactors, thiamin diphosphate and Mg2+. The most simple explanation of the observed hysteretic transition is that the slow binding of a Mg(2+)-thiamin-diphosphate species precedes slow isomerisation of the enzyme to the active form. Although the hysteretic transition involves more than one process, it does not involve the dissociation-association of enzyme subunits. The best estimate of the apparent Km, 1.59 +/- 0.23 microM, for the binding of Mg(2+)-thiamin diphosphate to transketolase was obtained in the presence of a high non-inhibitory concentration of magnesium and varied concentrations of thiamin diphosphate. Thus the reconstitution of the human enzyme differs from the yeast enzyme, which undergoes a rate-limiting dimerisation during reconstitution.

Application of high field localised in vivo 1H MRS to study biochemical changes in the thiamin deficient rat brain under glucose load.

In vivo, volume-selected 1H NMR spectroscopy employing the SPACE technique was used to monitor biochemical changes in the thiamin deficient rat brain in response to glucose loading. The concentrations of brain N-acetylaspartate, glutamate/glutamine/gamma-aminobutyric acid, lactate and glucose differed significantly from those of control animals. The results are consistent with a metabolic block at the reaction catalyzed by the thiamin dependent enzyme alpha-keto glutarate dehydrogenase soon after the onset of neurological symptoms of thiamin deficiency, and a further block at pyruvate dehydrogenase arising late in the course of thiamin deficiency.

High control coefficient of transketolase in the nonoxidative pentose phosphate pathway of human erythrocytes: NMR, antibody, and computer simulation studies.

The degree of control exerted by transketolase over metabolite flux in the nonoxidative pentose phosphate pathway in human erythrocytes was investigated using transketolase antiserum to modulate the activity of that enzyme. 31P NMR enabled the simultaneous measurement of the levels of pentose phosphate pathway metabolites following incubation of hemolysates with ribose 5-phosphate. The variations in metabolic flux which occurred as the transketolase activity of hemolysate samples was altered indicated that a high degree of control was exerted by transketolase. Investigations using transaldolase-depleted hemolysates showed that transaldolase exhibits a lesser degree of control over pathway flux. Experimental data were compared with simulations generated by a computer model encompassing the reactions of the classical nonoxidative pentose phosphate pathway. The sensitivity coefficients (also called "control strengths" or "flux-control coefficients") calculated from the computer simulations were 0.74 and 0.03 for transketolase and transaldolase, respectively.

Inhibition of transketolase and pyruvate decarboxylase by omeprazole.

Omeprazole inhibited two thiamin diphosphate-dependent enzymes, pyruvate decarboxylase (EC 4.1.1.1, PDC) from Zymomonas mobilis and transketolase (EC 2.2.1.1, TK) from human erythrocytes. Inhibition of PDC was competitive with the coenzyme with a Ki value of 42 +/- 3 microM, whereas inhibition of TK was complex.