Catalytic acid-base groups in yeast pyruvate decarboxylase. 1. Site-directed mutagenesis and steady-state kinetic studies on the enzyme with the D28A, H114F, H115F, and E477Q substitutions

The roles of four of the active center groups with potential acid-base properties in the region of pH optimum of pyruvate decarboxylase from Saccharomyces cerevisiae have been studied with the substitutions Asp28Ala, His114Phe, His115Phe, and Glu477Gln, introduced by site-directed mutagenesis methods. The steady-state kinetic constants were determined in the pH range of activity for the enzyme. The substitutions result in large changes in k(cat) and k(cat)/S(0.5) (and related terms), indicating that all four groups have a role in transition state stabilization. Furthermore, these results also imply that all four are involved in some manner in stabilizing the rate-limiting transition state(s) both at low substrate (steps starting with substrate binding and culminating in decarboxylation) and at high substrate concentration (steps beginning with decarboxylation and culminating in product release). With the exception of some modest effects, the shapes of neither the bell-shaped k(cat)/S(0.5)-pH (and related functions) plots nor the k(cat)-pH plots are changed by the substitutions. Yet, the fractional activity still remaining after substitutions virtually rules out any of the four residues as being directly responsible for initiating the catalytic process by ionizing the C2H. There is no effect on the C2 H/D exchange rate exhibited by the D28A and E477Q substitutions. These results strongly imply that the base-induced deprotonation at C2 is carried out by the only remaining base, the iminopyrimidine tautomer of the coenzyme, via intramolecular proton abstraction. The first product is released as CO(2) rather than HCO(3)(-) by both wild-type and E477Q and D28A variants, ruling out several mechanistic alternatives.

Catalytic acid-base groups in yeast pyruvate decarboxylase. 3. A steady-state kinetic model consistent with the behavior of both wild-type and variant enzymes at all relevant pH values

The widely quoted kinetic model for the mechanism of yeast pyruvate decarboxylase (YPDC, EC 4.1.1.1), an enzyme subject to substrate activation, is based on data for the wild-type enzyme under optimal experimental conditions. The major feature of the model is the obligatory binding of substrate in the regulatory site prior to substrate binding at the catalytic site. The activated monomer would complete the cycle by irreversible decarboxylation of the substrate and product (acetaldehyde) release. Our recent kinetic studies of YPDC variants substituted at positions D28 and E477 at the active center necessitate some modification of the mechanism. It was found that enzyme without substrate activation apparently is still catalytically competent. Further, substrate-dependent inhibition of D28-substituted variants leads to an enzyme form with nonzero activity at full saturation, requiring a second major branch point in the mechanism. Kinetic data for the E477Q variant suggest that three consecutive substrate binding steps may be needed to release product acetaldehyde, unlikely if YPDC monomer is the minimal catalytic unit with only two binding sites for substrate. A model to account for all kinetic observations involves a functional dimer operating through alternation of active sites. In the context of this mechanism, roles are suggested for the active center acid-base groups D28, E477, H114, and H115. The results underline once more the enormous importance that both aromatic rings of the thiamin diphosphate, rather than only the thiazolium ring, have in catalysis, a fact little appreciated prior to the availability of the 3-dimensional structure of these enzymes.

Catalytic acid-base groups in yeast pyruvate decarboxylase. 2. Insights into the specific roles of D28 and E477 from the rates and stereospecificity of formation of carboligase side products

Yeast pyruvate decarboxylase (YPDC), in addition to forming its metabolic product acetaldehyde, can also carry out carboligase reactions in which the central enamine intermediate reacts with acetaldehyde or pyruvate (instead of the usual proton electrophile), resulting in the formation of acetoin and acetolactate, respectively (typically, 1% of the total reaction). Due to the common mechanism shared by the acetaldehyde-forming and carboligase reactions through decarboxylation, a detailed analysis of the rates and stereochemistry of the carboligase products formed by the E477Q, D28A, and D28N active center YPDC variants was undertaken. While substitution at either position led to an approximately 2-3 orders of magnitude lower catalytic efficiency in acetaldehyde formation, the rate of acetoin formation by the E477Q and D28N variants was higher than that by wild-type enzyme. Comparison of the steady-state data for acetaldehyde and acetoin formation revealed that the rate-limiting step for acetaldehyde formation by the D28A, H114F, H115F, and E477Q variants is a step post-decarboxylation. In contrast to the wild-type YPDC and the E477Q variant, the D28A and D28N variants could synthesize acetolactate as a major product. The lower overall rate of side-product formation by the D28A variant than wild-type enzyme attests to participation of D28 in steps leading up to and including decarboxylation. The results also provide insight into the state of ionization of the side chains examined. (R)-Acetoin is produced by the variants with greater enantiomeric excess than by wild-type YPDC. (S)-Acetolactate is the predominant enantiomer produced by the D28-substituted variants, the same configuration as produced by the related plant acetolactate synthase.

Consequences of a modified putative substrate-activation site on catalysis by yeast pyruvate decarboxylase

Earlier, it had been proposed in the laboratories at Halle that a cysteine residue is responsible for the hysteretic substrate activation behavior of yeast pyruvate decarboxylase. More recently, this idea has received support in a series of studies from Rutgers with the identification of residue C221 as the site where substrate is bound to transmit the information to H92, to E91, to W412, and finally to the active center thiamin diphosphate. According to steady-state kinetic assays, the C221A/C222A variant is no longer subject to substrate activation yet is still a well-functioning enzyme. Several further experiments are reported on this variant: (1) The variant exhibits lag phases in the product formation progress curves, which can be attributed to a unimolecular step in the pre-steady-state stage of catalysis. (2) The rate of exchange with solvent deuterium of the thiamin diphosphate C2H atom is slowed by a factor of 2 compared to the wild-type enzyme, suggesting that the reduced activity that results from the substitutions some 20 A from the active center is also seen in the first key step of the reaction. (3) The solvent (deuterium oxide) kinetic isotope effect was found to be inverse on V(max)/K(m) (0.62), and small but normal on V(max) (1.26), virtually ruling out residue C221 as being responsible for the inverse effects reported for the wild-type enzyme at low substrate concentrations. The solvent kinetic isotope effects are compared to those on two related enzymes not subject to substrate activation, Zymomonas mobilis pyruvate decarboxylase and benzoylformate decarboxylase.

Cell adhesion molecule expression in the cervix and myometrium during pregnancy and parturition

OBJECTIVE

To determine the expression and localization of cell adhesion molecules intercellular adhesion molecule-1 (ICAM-1), E-selectin, platelet-endothelial cell adhesion molecule (PECAM), and vascular cell adhesion molecule (VCAM) in the cervix and myometrium during pregnancy and labor.

METHODS

Biopsies of myometrium and cervix were obtained from non-pregnant women and from pregnant women before and after onset of spontaneous labor at term. Cell adhesion molecule mRNA expression was quantified using Northern blotting and cell adhesion molecule protein was localized using immunohistochemistry.

RESULTS

ICAM-1 mRNA was upregulated in the cervix (10-fold increase, P <.01) and myometrium (10.5-fold increase, P <.01) during labor. ICAM-1 was localized in the vascular endothelium and in leukocytes in the cervix and myometrium from all three groups of women. VCAM mRNA was upregulated in the cervix (2.5-fold increase, P <.01) during pregnancy and there was no further change during labor. VCAM localized weakly to the vascular endothelium in cervical and myometrial biopsies from pregnant and non-pregnant women. PECAM mRNA was significantly upregulated in myometrium during pregnancy (ninefold increase, P <.01) and did not change with the onset of labor. PECAM localized to the vascular endothelium in all cervical and myometrial biopsies and was identified on leukocytes. There were no significant changes in E-selectin mRNA expression in either tissue with pregnancy or parturition.

CONCLUSION

Cell adhesion molecule expression changes in human cervix and myometrium during pregnancy and parturition. At least part of these changes are attributable to expression by leukocytes infiltrating these tissues.

Spectroscopic detection of transient thiamin diphosphate-bound intermediates on benzoylformate decarboxylase

Thiamin diphosphate (ThDP)-dependent enzymes catalyze a range of transformations, such as decarboxylation and ligation. We report a novel spectroscopic assay for detection of some of the ThDP-bound intermediates produced on benzoylformate decarboxylase. Benzoylformate decarboxylase was mixed with its alternate substrate p-nitrobenzoylformic acid on a rapid-scan stopped-flow instrument, resulting in formation of three absorbing species (lambda(max) in parentheses): I(1) (a transient at 620 nm), I(2) (a transient at 400 nm), and I(3) (a stable absorbance with lambda(max) > 730 nm). Analysis of the kinetics of the two transient species supports a model in which a noncovalent complex of the substrate and the enzyme is converted to the first covalent intermediate I(1); the absorbance corresponding to I(1) is probably a charge-transfer band arising from the interaction of the thiamin diphosphate-p-nitrobenzoylformic acid covalent adduct (2-p-nitromandelylThDP) and the enzyme. The rate of disappearance of I(1) parallels the rate of formation of I(2). Chemical models suggest the lambda(max) of I(2) (near 400 nm) to be appropriate to the enamine, a key intermediate in ThDP-dependent reactions resulting from the decarboxylation of the thiamin diphosphate-p-nitrobenzoylformic acid covalent adduct. Therefore, the rate of disappearance of I(1) and/or the appearance of I(2) directly measure the rate of decarboxylation. A relaxation kinetic treatment of the pre-steady-state kinetic data also revealed a hitherto unreported facet of the mechanism, alternating active-sites reactivity. Parallel studies of the His70Ala BFD active-site variant indicate that it cannot form the complex reported by the charge-transfer band (I(1)) at the level of the wild-type protein.

Interplay of organic and biological chemistry in understanding coenzyme mechanisms: example of thiamin diphosphate-dependent decarboxylations of 2-oxo acids

With the publication of the three-dimensional structures of several thiamin diphosphate-dependent enzymes, the chemical mechanism of their non-oxidative and oxidative decarboxylation reactions is better understood. Chemical models for these reactions serve a useful purpose to help evaluate the additional catalytic rate acceleration provided by the protein component. The ability to generate, and spectroscopically observe, the two key zwitterionic intermediates invoked in such reactions allowed progress to be made in elucidating the rates and mechanisms of the elementary steps leading to and from these intermediates. The need remains to develop chemical models, which accurately reflect the enzyme-bound conformation of this coenzyme.

Role of glutamate 91 in information transfer during substrate activation of yeast pyruvate decarboxylase

Oligonucleotide-directed site-specific mutagenesis was carried out on pyruvate decarboxylase (EC 4.1.1.1) from Saccharomyces cerevisiae at E91, located on the putative substrate activation pathway and linking the alpha and gamma domains of the enzyme. While C221 on the beta domain is the residue at which substrate activation is triggered [Baburina, I., et al. (1994) Biochemistry 33, 5630-5635; Baburina, I., et al. (1996) Biochemistry 35, 10249-10255], that information, via the substrate bound at C221, is transmitted to H92 on the alpha domain, across the domain divide from C221 [Baburina, I. , et al. (1998) Biochemistry 37, 1235-1244], thence to E91 on the alpha domain, and then on to W412 on the gamma domain [Li, H., and Jordan, F. (1999) Biochemistry 38, 10004-10012] and to the active site thiamin diphosphate located at the interface of the alpha and gamma domains [Arjunan, D., et al. (1996) J. Mol. Biol. 256, 590-600]. Substitution at E91 with Q, D, or A led to modest reductions in the specific activity (4-, 5-, and 30-fold), as well as in both the turnover number and the catalytic efficiency, in that order. Interestingly, the Hill coefficient was only slightly reduced for the E91D variant, but cooperativity was virtually abolished for the E91Q and E91A variants. The thermal stability of the variants was reduced in the following order: wild type > E91Q > E91D > E91A; circular dichroism and fluorescence experiments also demonstrated that the tertiary structure of the enzyme was affected by these substitutions. The variants could be purified as apoenzymes, demonstrating their impaired ability to bind thiamin diphosphate. Apparently, the charge at residue 91 is quite important for maintaining optimal cooperativity. To maintain strong domain-domain interactions, the length of the side chain at position 91 with hydrogen bonding potential to W412 is sufficient.

Effects of substitution of tryptophan 412 in the substrate activation pathway of yeast pyruvate decarboxylase

Oligonucleotide-directed site-specific mutagenesis was carried out on pyruvate decarboxylase (EC 4.1.1.1) from Saccharomyces cerevisiae at W412, located on the putative substrate activation pathway and linking E91 on the alpha domain with W412 on the gamma domain of the enzyme. While C221 on the beta domain is the residue at which substrate activation is triggered [Baburina, I., et al. (1994) Biochemistry 33, 5630-5635; Baburina, I., et al. (1996) Biochemistry 35, 10249-10255], that information, via the substrate bound at C221, is transmitted to H92 on the alpha domain, across the domain divide from C221 [Baburina, I., et al. (1998) Biochemistry 37, 1235-1244; Baburina, I., et al. (1998) Biochemistry 37, 1245-1255], thence to E91 on the alpha domain [Li, H., and Jordan, F. (1999) Biochemistry 38, 9992-10003], and then on to W412 on the gamma domain and to the active site thiamin diphosphate located at the interface of the alpha and gamma domains [Arjunan, D., et al. (1996) J. Mol. Biol. 256, 590-600]. Substitution at W412 with F and A was carried out, resulting in active enzymes with specific activities about 4- and 10-fold lower than that of the wild-type enzyme. Even though W412 interacts with E91 and H115 via a main chain hydrogen bond donor and acceptor, respectively, there is clear evidence for the importance of the indole side chain of W412 from a variety of experiments: thermostability, fluorescence quenching, and the binding constants of the thiamin diphosphate, and circular dichroism spectroscopy, in addition to conventional steady-state kinetic measurements. While the substrate activation is still prominent in the W412F variant, its level is very much reduced in the W412A variant, signaling that the size of the side chain is also important in positioning the amino acids surrounding the active center to achieve substrate activation. The fluorescence studies demonstrate that W412 is a relatively minor contributor to the well-documented fluorescence of apopyruvate decarboxylase in its native state. The information about the W412 variants provides strong additional support for the putative substrate activation pathway from C221 --> H92 --> E91 --> W412 --> G413 --> thiamin diphosphate. The accumulating evidence for the central role of the beta domain in stabilizing the overall structure is summarized.

Remarkable stabilization of zwitterionic intermediates may account for a billion-fold rate acceleration by thiamin diphosphate-dependent decarboxylases

When the E91D variant of apo-yeast pyruvate decarboxylase (EC 4.1.1. 1) is exposed to C2alpha-hydroxybenzylthiamin diphosphate, this putative intermediate is partitioned on the enzyme between release of the benzaldehyde product (as evidenced by regeneration of active enzyme) and dissociation of the proton at C2alpha to form the enamine-C2alpha-carbanion intermediate. While the pKa (the negative log of the acid dissociation constant) for this dissociation is approximately 15.4 in water, formation of the enamine at pH 6.0 on the enzyme indicates a >9 unit pKa suppression by the enzyme environment. The dramatic stabilization of this zwitterionic enamine intermediate at the active center is sufficient to account for as much as a 10(9)-fold rate acceleration on the enzyme. This "solvent" effect could be useful for achieving the bulk of the rate acceleration provided by the protein over and above that afforded by the coenzyme on all thiamin diphosphate-dependent 2-oxo acid decarboxylases.

Autoregulation of yeast pyruvate decarboxylase gene expression requires the enzyme but not its catalytic activity

In the yeast, Saccharomyces cerevisiae, pyruvate decarboxylase (Pdc) is encoded by the two isogenes PDC1 and PDC5. Deletion of the more strongly expressed PDC1 gene stimulates the promoter activity of both PDC1 and PDC5, a phenomenon called Pdc autoregulation. Hence, pdc1Delta strains have high Pdc specific activity and can grow on glucose medium. In this work we have characterized the mutant alleles pdc1-8 and pdc1-14, which cause strongly diminished Pdc activity and an inability to grow on glucose. Both mutant alleles are expressed as detectable proteins, each of which differs from the wild-type by a single amino acid. The cloned pdc1-8 and pdc1-14 alleles, as well as the in-vitro-generated pdc1-51 (Glu51Ala) allele, repressed expression of PDC5 and diminished Pdc specific activity. Thus, the repressive effect of Pdc1p on PDC5 expression seems to be independent of its catalytic activity. A pdc1-8 mutant was used to isolate spontaneous suppressor mutations, which allowed expression of PDC5. All three mutants characterized had additional mutations within the pdc1-8 allele. Two of these mutations resulted in a premature translational stop conferring phenotypes virtually indistinguishable from those of a pdc1Delta mutation. The third mutation, pdc1-803, led to a deletion of two amino acids adjacent to the pdc1-8 mutation. The alleles pdc1-8 and pdc1-803 were expressed in Escherichia coli and purified to homogeneity. In the crude extract, both proteins had 10% residual activity, which was lost during purification, probably due to dissociation of the cofactor thiamin diphosphate (ThDP). The defect in pdc1-8 (Asp291Asn) and the two amino acids deleted in pdc1-803 (Ser296 and Phe297) are located within a flexible loop in the beta domain. This domain appears to determine the relative orientation of the alpha and gamma domains, which bind ThDP. Alterations in this loop may also affect the conformational change upon substrate binding. The mutation in pdc1-14 (Ser455Phe) is located within the ThDP fold and is likely to affect binding and/or orientation of the cofactor in the protein. We suggest that autoregulation is triggered by a certain conformation of Pdc1p and that the mutations in pdc1-8 and pdc1-14 may lock Pdc1p in vivo in a conformational state which leads to repression of PDC5.

Is a hydrophobic amino acid required to maintain the reactive V conformation of thiamin at the active center of thiamin diphosphate-requiring enzymes? Experimental and computational studies of isoleucine 415 of yeast pyruvate decarboxylase

The residue I415 in pyruvate decarboxylase from Saccharomyces cerevisiae was substituted with a variety of uncharged side chains of varying steric requirements to test the hypothesis that this residue is responsible for supporting the V coenzyme conformation reported for this enzyme [Arjunan et al. (1996) J. Mol. Biol. 256, 590-600]. Changing the isoleucine to valine and threonine decreased the kcat value and shifted the kcat-pH profile to more alkaline values progressively, indicating that the residue at position 415 not only is important for providing the optimal transition state stabilization but also ensures correct alignment of the ionizable groups participating in catalysis. Substitutions to methionine (the residue used in pyruvate oxidase for this purpose) or leucine (the corresponding residue in transketolase) led to greatly diminished kcat values, showing that for each thiamin diphosphate-dependent enzyme an optimal hydrophobic side chain evolved to occupy this key position. Computational studies were carried out on the wild-type enzyme and the I415V, I415G, and I415A variants in both the absence and the presence of pyruvate covalently bound to C2 of the thiazolium ring (the latter is a model for the decarboxylation transition state) to determine whether the size of the side chain is critically required to maintain the V conformation. Briefly, there are sufficient conformational constraints from the binding of the diphosphate side chain and three conserved hydrogen bonds to the 4'-aminopyrimidine ring to enforce the V conformation, even in the absence of a large side chain at position 415. There appears to be increased coenzyme flexibility on substitution of Ile415 to Gly in the absence compared with the presence of bound pyruvate, suggesting that entropy contributes to the rate acceleration. The additional CH3 group in Ile compared to Val also provides increased hydrophobicity at the active center, likely contributing to the rate acceleration. The computational studies suggest that direct proton transfer to the 4'-imino nitrogen from the thiazolium C2H is eminently plausible.

Regulation of thiamin diphosphate-dependent 2-oxo acid decarboxylases by substrate and thiamin diphosphate.Mg(II) - evidence for tertiary and quaternary interactions

The regulatory mechanism of substrate activation in yeast pyruvate decarboxylase is triggered by the interaction of pyruvic acid with C221 located on the beta domain at >20 A from the thiamin diphosphate (ThDP). To trace the putative information transfer pathway, substitutions were made at H92 on the alpha domain, across the domain divide from C221, at E91, next to H92 and hydrogen bonded to W412, the latter being intimately involved in the coenzyme binding locus. Additional substitutions were made at D28, E51, H114, H115, I415 and E477, all near the active center. The pH-dependent steady-state kinetic parameters, including the Hill coefficient, provide useful insight to this effort. In addition to C221, the residues H92, E91, E51 and H114 and H115 together appear to have a critical impact on the Hill coefficient, providing a pathway for information transfer. To study the activation by ThDP.Mg(II), variants at G231 (of the conserved GDG triplet) and at N258 and C259 (all three being part of the putative ThDP fold) of the E1 component of the Escherichia coli pyruvate dehydrogenase multienzyme complex were studied. Kinetic and spectroscopic evidence suggests that the Mg(II) ligands are very important to activation of the enzymes by cofactors.

Structure-function relationships and flexible tetramer assembly in pyruvate decarboxylase revealed by analysis of crystal structures

The crystal structures of pyruvate decarboxylase from the yeast Saccharomyces uvarum and Saccharomyces cerevisiae have been determined at 2.4 and 2.3 A resolution, respectively. These structures provide details about the protein fold and domain assembly within subunits, about subunit assembly to form dimers and about dimer assembly to form tetramers. They also provide a clear picture of the active site centered on the thiamin diphosphate cofactor, and have allowed amino acids critical for catalysis and involved in stabilization of the unusual cofactor conformation to be identified. The structural information has enabled identification of the site of allosteric activation to be centered on Cys-221, and suggests that a six residue segment leading from the regulatory site to the catalytic site may be involved in transmission of a binding signal. The importance of several amino acids within this segment in the regulatory process, as well as some involved in stabilizing and activating the cofactor has been confirmed by analyzing the behavior of recombinant enzymes with single point mutations introduced at these sites. Additional structures have been determined for pyruvate decarboxylase in multiple crystal forms, some of which were obtained from crystals grown with known allosteric activators present in the media. Currently four distinct types of tetramers have been observed, with each showing a different mode of association of dimers to form the tetramers. In some of the cases involving the presence of allosteric activators drastic changes in the mode of dimer assembly to form tetramers is seen.

Interdomain information transfer during substrate activation of yeast pyruvate decarboxylase: the interaction between cysteine 221 and histidine 92

Oligonucleotide-directed site-specific mutagenesis was carried out on pyruvate decarboxylase (EC 4.1.1.1) from Saccharomyces cerevisiae at two cysteines on the beta domain (221 and 222) and at H92 on the alpha domain, across the domain divide from C221. While C221 has been shown to provide the trigger for substrate activation [Baburina, I., et al. (1994) Biochemistry 33, 5630-5635], the information must be transmitted from the substrate bound at this site [Arjunan, D., et al. (1996) J. Mol. Biol. 256, 590-600] to the active center thiamin diphosphate located at the interface of the alpha and gamma domains. Substitution at H92 with G, A, or C leads to great reduction of the Hill coefficient (from 2.0 in the wild-type enzyme to 1.2-1.3), while substitution for Lys affords an active enzyme with a Hill coefficient of 1.5-1.6. Iodoacetate at 10 mM reduced the Hill coefficient from 2.0 to 1.1, while also causing significant inactivation of the enzyme, presumably by carboxymethylation of C221. 1,3-Dibromoacetone, a potential cross-linker when added to the H92C/C222S variant at 0.1 mM, abolished substrate activation while reducing the activity only by 30%. Therefore, 1,3-dibromoacetone may cross-link C92 and C221. It was concluded that H92 is on the information transfer pathway during the substrate activation process and the interaction between C221 on the beta domain and H92 on the alpha domain is required for substrate activation. Extensive pH studies of the steady-state kinetic constants provide support for the interaction of C221 and H92 and the transmission of regulatory information to the active center via this pathway and pKaS for the two groups. This important interaction between the C221-bound pyruvate and His92 probably has both electrostatic and steric components.

Reactivity at the substrate activation site of yeast pyruvate decarboxylase: inhibition by distortion of domain interactions

The residue C221 on pyruvate decarboxylase (EC. 4.1.1.1) from Saccharomyces cerevisiae has been shown to be the site where the substrate activation cascade is triggered [Baburina et al. (1994) Biochemistry 33, 5630-5635] and is located on the beta domain [Arjunan et al. (1996) J. Mol. Biol. 256, 590], while the active-center thiamin diphosphate is located > 20 A away, at the interface of the alpha and gamma domains. The reactivity of all three exposed cysteines (152, 221, and 222) was examined under the influence of known activators and inhibitors. Protein chemical methods, in conjunction with [1-14C] and [3-3H] analogues of the mechanism-based inhibitor p-ClC6H4CH=CHCOCOOH, demonstrated that the holoenzyme bound approximately 2-3 atoms of tritium/atom of C-14. However, when the labeled enzyme was subjected to trypsinization, followed by sequencing of the labeled peptide, only the tritium label was in evidence at C221, with a stoichiometry of 2 atoms of tritium/tetrameric holoenzyme. Apparently, the product of decarboxylation bonded to the enzyme survived the limited proteolysis and sequencing, but the bound 2-oxoacid was released during the protocol. Surprisingly, the C221S or C222A variants, although they still possess 20-30% specific activity compared to the wild-type enzyme, could still be inhibited by the XC6H4CH=CHCOCOOH class of inhibitors/substrate analogues, as well as by the product of decarboxylation from such compounds, cinnamaldehydes. Other potential nucleophilic sites for the inhibitor [C152 (the third exposed cysteine), residues D28, H114, H115, and E477 at the active center and H92 at the regulatory site] were also substituted by a nonnucleophilic side chain. All variants were still subject to inhibition by p-ClC6H4CH=CHCOCOOH, the active-center variants being inactivated even faster than the wild-type enzyme, suggesting that the active center is involved in the inactivation process. It appears that C221 is one of only two sites of interaction with such compounds (perhaps the result of a Michael addition across the C=C bond), yet the bound [1-14C]-labeled inhibitor could no longer be detected after peptide mapping at this site or at the catalytic site. Upon combining the tritiated inhibitor with [2-14C]-thiamin diphosphate, no evidence could be found for a thiamin-inhibitor-protein ternary complex, suggesting that the thiamin-bound enamine intermediate did not react further with the protein. It is likely that the second form of inhibition is at the active center, with the inhibitor cofactor-bound, which would have been released during the proteolytic protocol. Among other known activators, ketomalonate was found to react at C221 only. Glyoxalic acid, a mechanism-based inhibitor, on the other hand, could react at both the regulatory and the catalytic center. The high reactivity of C221 is consistent with it being in the thiolate form at the optimal pH of the enzyme [forming a Cys221S(-) + HHis92 ion pair; see Baburina et al. (1996) Biochemistry 35, 10249-10255, and Baburina et al. (1998) Biochemistry 37, 1235-1244]. Several additional compounds were tested as potential regulatory site-directed reagents: iodoacetate, 1,3-dibromoacetone, and 1-bromo-2-butanone. All three compounds reduced the Hill coefficient and hence appear to react at C221. It was concluded that either substitution of C221 by a nonnucleophilic residue or large groups attached to C221 in the wild-type enzyme lead to a distortion of domain interactions, interactions which are required for both optimal activity and substrate activation.

D,L-S-methyllipoic acid methyl ester, a kinetically viable model for S-protonated lipoic acid as the oxidizing agent in reductive acyl transfers catalyzed by the 2-oxoacid dehydrogenase multienzyme complexes

D,L-S(6,8)-Methyllipoic acid methyl ester triflate salt (D,L-S-methyllipoic acid methyl ester) was synthesized as a model for S-protonated lipoic acid, suggested to be the active form of lipoic acid in the reductive acylation catalyzed by the E1 and E2 enzymes of the 2-oxoacid dehydrogenase multienzyme complexes by a previous model [Chiu, C. C., Chung, A., Barletta, G., and Jordan, F. (1996) J. Am. Chem. Soc. 118, 11026-11029]. While in that earlier study lipoic acid could only trap only the enamine/C2 alpha-carbanion intermediate in an intramolecular model, and with the assistance of mercury compound to shift the equilibrium to the products, D,L-S-methyllipoic acid methyl ester could trap the enamine derived from 2-alpha-methoxybenzyl-3,4,5-trimethylthiazolium salt in an intermolecular reaction in the absence of a mercury compound, and with a rate constant of 6.6 x 10(4) M-1 S-1. A tetrahedral adduct at the C2 alpha-position formed between the enamine and D,L-S-methyllipoic acid methyl ester was isolated and characterized. The reaction likely takes place by two-electron nucleophilic attack, since no evidence was found for C2 alpha-linked homodimers, expected from a free-radical mechanism. The results suggest that, in the reductive acyl transfer, there is nucleophilic attack by the enamine at one of the sulfur atoms of the lipoic acid [probably at S8, according to Frey, P. A., Flournoy, D. S., Gruys, K., and Yang, Y. S. (1989) Ann. N.Y. Acad. Sci. 373, 21-35], while there is concomitant electrophilic catalysis by a proton juxtaposed at S6 via a general acid catalyst located on the E1 enzyme. Oxidation of the enamine derived from C2 alpha-hydroxybenzyl-3,4,5-trimethylthiazolium salt by D,L-S-methyllipoic acid methyl ester was also deduced on the basis of the formation of 2-benzoylthiazolium ion as a major product; however, the tetrahedral intermediate could not be detected. Oxidation of the enamine by D,L-S-methyllipoic acid methyl ester can take place with either an ether or an alcohol at the C2 alpha position of the enamine.

Systematic study of the six cysteines of the E1 subunit of the pyruvate dehydrogenase multienzyme complex from Escherichia coli: none is essential for activity

Variants of the Escherichia coli 1-lip pyruvate dehydrogenase multienzyme complex (1-lip PDHc) with the C259N and C259S substitutions in the putative thiamin diphosphate-(ThDP-) binding motif of the pyruvate dehydrogenase component (E1, EC 1.2.4.1) were characterized. Single substitutions were made at the five remaining cysteines of the E1 component, creating the C120A, C575A, C610A, C654A, and C770S variants to test the hypothesis that the activity loss that accompanies exposure of the enzyme to fluoropyruvate, bromopyruvate, and 2-oxo-3-butynoic acid is the result of the modification of approximately one cysteine residue per E1 monomer. Surprisingly, all single cysteine E1 variants could be reconstituted with E2-E3 subcomplex and showed PDHc activity ranging from 74% to 96% that of the parental enzyme. The specific activities of C259N and C259S variants of 1-lip PDHc were 58% and 27% relative to that of the parental 1-lip PDHc. All five single cysteine E1 variants, along with the C259N and C259S variants of 1-lip PDHc, could also (1) be inactivated with fluoropyruvate and 2-oxo-3-butynoic acid, (2) were subject to inactivation by the monoclonal antibody 18A9 reported from one of our laboratories, and (3) were subject to regulation by pyruvate and acetyl-CoA. It was therefore concluded that none of the six cysteine residues is essential for the activity of the E1 component or of the complex. When tested with the putative transition-state analogue, thiamin 2-thiothiazolone diphosphate, all but the C259S and C259N variants were very potently inhibited, the stoichiometry for parental E1 being about 1.6 mol of inhibitor/mol of E1 subunit. The C259S and C259N E1 variants required at least 25-fold greater inhibitor concentration to achieve the same level of inhibition. C259 is located in the putative thiamin diphosphate-binding motif of the enzyme [more exactly, it is adjacent to a ligand to the Mg(II) ion]. It is therefore concluded that thiamin 2-thiothiazolone diphosphate is not a transition-state analogue; rather, it is a potent inhibitor of the complex because of a specific interaction with the C259 residue.

Low barrier hydrogen bond is absent in the catalytic triads in the ground state but Is present in a transition-state complex in the prolyl oligopeptidase family of serine proteases

High frequency proton NMR spectra for two members of the prolyl oligopeptidase class of serine proteases, prolyl oligopeptidase and oligopeptidase B, showed that resonances corresponding to the active center histidine Ndelta1H and Nepsilon2H generally observed in this region, are absent in these enzymes. However, for both enzymes, as well as with the H652A and H652Q active center variants of oligopeptidase B, there are two resonances observed in this region that could be assigned to two protonated histidines with a noncatalytic function. The results indicate that these two histidines participate in strong hydrogen bonds. The absence of resonances pertinent to the active center histidine resonances suggests the absence of a low barrier hydrogen bond between the Asp and His in these two enzymes in their ground states. Addition of the peptide boronic acid t-butoxycarbonyl-(D)Val-Leu-(L)boroArg to oligopeptidase B resulted in potent, slow binding inhibition of the enzyme and the appearance of a new resonance at 15.8 ppm, whose chemical shift is appropriate for a tetrahedral boronate complex and a low barrier hydrogen bond. The results demonstrate important dissimilarities between the active centers of the prolyl oligopeptidase class of serine proteases and the pancreatic and subtilisin classes both in the ground state and in the transition-state analog complexes.

2-Oxo-3-alkynoic acids, universal mechanism-based inactivators of thiamin diphosphate-dependent decarboxylases: synthesis and evidence for potent inactivation of the pyruvate dehydrogenase multienzyme complex

A new class of compounds, the 2-oxo-3-alkynoic acids with a phenyl substituent at carbon 4 was reported by the authors as potent irreversible and mechanism-based inhibitors of the thiamin diphosphate- (ThDP-) dependent enzyme pyruvate decarboxylase [Chiu, C.-F., & Jordan, F. (1994) J. Org. Chem. 59, 5763-5766]. The method has been successfully extended to the synthesis of the 4-, 5-, and 7-carbon aliphatic members of this family of compounds. These three compounds were then tested on three ThDP-dependent pyruvate decarboxylases: the Escherichia coli pyruvate dehydrogenase multienzyme complex (PDHc) and its E1 (ThDP-dependent) component, pyruvate oxidase (POX, phosphorylating; from Lactobacillus plantarum),and pyruvate decarboxylase (PDC) from Saccharomycescerevisiae. All three enzymes were irreversibly inhibited by the new compounds. The 4-carbon acid is the best substrate-analog inactivator known to date for PDHc, more potent than either fluoropyruvate or bromopyruvate. The following conclusions were drawn from extensive studies with PDHc: (a) The kinetics of inactivation of PDH complexes and of resolved E1 by 2-oxo-3-alkynoic acids is time- and concentration-dependent. (b) The 4-carbon acid has a Ki 2 orders of magnitude stronger than the 5-carbon acid, clearly demonstrating the substrate specificity of PDHc. (c) The rate of inactivation of PDH complexes and of resolved E1 by 2-oxo-3-alkynoic acids is enhanced by the addition of ThDP and MgCl2. (d) Pyruvate completely protects E1 and partially protects PDHc from inactivation by 2-oxo-3-butynoic acid. (e) E1 but not E2-E3 is the target of inactivation by 2-oxo-3-butynoic acid. (f) Inactivation of E1 by 2-oxo-3-butynoic acid is accompanied by modification of 1.3 cysteines/E1 monomer. The order of reactivity with the 4-carbon acid was PDHc > POX > PDC. While the order of reactivity with PDHc and POX was 2-oxo-3-butynoic acid > 2-oxo-3-pentynoic acid > 2-oxo-3-heptynoic acid, the order of reactivity was reversed with PDC.

Benzyloxycarbonylprolylprolinal, a transition-state analogue for prolyl oligopeptidase, forms a tetrahedral adduct with catalytic serine, not a reactive cysteine

N-Benzyloxycarbonyl-l-prolyl-l-[1-13C]prolinal was synthesized starting with reduction of l-[1-13C]Pro to l-[1-13C]prolinol, followed by coupling with N-benzyloxycarbonyl-l-Pro to N-benzyloxycarbonyl-l-Pro-l-[1-13C]prolinol (Z-Pro-[1-13C]prolinol), and finally oxidation of the alcohol to the aldehyde with dimethyl sulphoxide. While the 13C NMR chemical shift of the aldehyde carbon is 202 p.p.m., that of the aldehyde hydrate is between 91.6 and 91.8 p.p.m., that of the dithiothreitol adduct is between 74.8 and 75.0 p. p.m., and that in the presence of the serine protease prolyl oligopeptidase is at 92.3 p.p.m.. The linewidth of the latter is 114 Hz, roughly consistent with the molecular mass of 80 kDa reported for the enzyme. Inverse detection experiments gave a 1H resonance at 5.29 p.p.m. with a linewidth of 80 Hz, also consistent with the expected chemical shift and linewidth for a hemiacetal bound to such a large enzyme, while the free hydrate gave resonances at 5.18 and 5. 25 p.p.m., with very much narrower linewidths. It is concluded that Z-Pro-prolinal, a putative transition-state analogue for prolyl oligopeptidase, forms a tetrahedral complex with the enzyme at its catalytic serine, rather than at a neighbouring cysteine that was found to be highly reactive according to chemical modification studies.

Effect of substitutions in the thiamin diphosphate-magnesium fold on the activation of the pyruvate dehydrogenase complex from Escherichia coli by cofactors and substrate

The homotropic regulation of the Escherichia coli pyruvate dehydrogenase multienzyme complex (PDHc) by its coenzyme thiamin diphosphate and its substrate pyruvate was re-examined with complexes containing three and one lipoyl domains per E2 chain, and several variants of the latter, containing substitutions in the putative thiamin diphosphate fold of E1 (G231A, G231S, C259S, C259N, and N258Q). It was found that all of the E1 variants had significantly reduced specific activities, as reported elsewhere (Russell, G. C., Machado, R. S., and Guest, J. R. (1992) Biochem. J. 287, 611-619). In addition, extensive kinetic studies were performed in an attempt to determine the effects of the amino acid substitutions on the Hill coefficients with respect to thiamin diphosphate and pyruvate. All but one of the variants were incapable of being saturated with thiamin diphosphate, even at concentrations > 5 mM. Most importantly, the striking activation lag phase lasting for many seconds in the parental complexes containing three and one lipoyl domains per E2 chain was totally eliminated in the variants. Furthermore, activation by the coenzyme was localized to the E1 subunit, because resolved E1 exhibits virtually the same behavior during the activation lag phase as does the complex. In the parental complexes two distinct lag phases could be resolved, the duration of both decreases with increasing ThDP concentration. A mechanism that is consistent with all of the kinetic data on the parental complexes involves rapid equilibration of the first ThDP with the E1 dimer, followed by a slow conformational equilibration, that in turn is followed by slow addition of the second ThDP to form the fully activated dimer. When the diphosphate site is badly impaired, the binding affinity is very much reduced, this perhaps eliminates the slow step leading to the activated dimer form of the E1.

Solution structure of vasoactive intestinal polypeptide (11-28)-NH2, a fragment with analgesic properties

An 18-residue-long fragment of vasoactive intestinal polypeptide [VIP(11-28)-NH2] that is known to be analgesic was synthesized by solid-phase t-Boc methodology on a 4-methylbenzhydrylamine resin. Circular dichroism spectroscopy gave evidence that the peptid acquires about 60% helical structure in 50/50 methanol/phosphate buffer, pH 6.0, and 65% (+/-5%) helicity in 80/20 methanol/phosphate buffer pH 7.0, A 2.0 mM solution of VIP (11-28) NH2 in 80% methanol, 20% phosphate buffer pH 7.0 was subjected to 2-dimensional nuclear magnetic resonance (NMR) studies The NMR results suggested formation of an extended helical structure extending from residue 11 to 27 essentially the same region found to be helical in a VIP(1-28)-NH2 and log. This finding suggests that the sequence required for analgesia assumes a helical structure at the receptor.

Evidence for intramolecular processing of prosubtilisin sequestered on a solid support

Subtilisin E is synthesized in Bacillus subtilis as a preprosubtilisin. The prepeptide is removed by a signal peptidase, and the propeptide is cleaved from the mature protein by the catalytic domain of subtilisin itself in an autocatalytic fashion. A six residue histidine-tag was attached to the C terminus of prosubtilisin and mature subtilisin to enable immobilization on a metal chelating resin. Guanidine-HC1 denatured histidine-tagged subtilisin and prosubtilisin were immobilized on Co2+ charged Talon resin, then renatured by dialysis of the resin against renaturation buffer. Refolding of the immobilized prosubtilisin resulted in its quantitative autoprocessing and the formation of active enzyme. Mature subtilisin on the other hand refolded into an active conformation with very low efficiency, and at the same concentration the steady-state rate attained was at least a 1000 times lower than that from prosubtilisin. The results give very strong support for an intramolecular autoprocessing pathway for prosubtilisin, in addition to an intermolecular one demonstrated before. The results also demonstrate rather convincingly the very much higher yield of active enzyme refolded from prosubtilisin than from mature protein under sequestered unimolecular conditions.