Letter


Nature Chemical Biology 2, 324 - 328 (2006)
Published online: 7 May 2006 | Corrected online: 30 May 2006 | doi:10.1038/nchembio788



There is an Erratum (July 2006) associated with this Letter.

The catalytic cycle of a thiamin diphosphate enzyme examined by cryocrystallography

Georg Wille1, Danilo Meyer1, Andrea Steinmetz1, Erik Hinze1, Ralph Golbik1 & Kai Tittmann1


Enzymes that use the cofactor thiamin diphosphate (ThDP, 11), the biologically active form of vitamin B1, are involved in numerous metabolic pathways in all organisms. Although a theory of the cofactor's underlying reaction mechanism has been established over the last five decades1, 2, the three-dimensional structures of most major reaction intermediates of ThDP enzymes have remained elusive. Here, we report the X-ray structures of key intermediates in the oxidative decarboxylation of pyruvate, a central reaction in carbon metabolism catalyzed by the ThDP- and flavin-dependent enzyme pyruvate oxidase (POX)3 from Lactobacillus plantarum. The structures of 2-lactyl-ThDP (LThDP, 22) and its stable phosphonate analog, of 2-hydroxyethyl-ThDP (HEThDP, 33) enamine and of 2-acetyl-ThDP (AcThDP, 44; all shown bound to the enzyme's active site) provide profound insights into the chemical mechanisms and the stereochemical course of thiamin catalysis. These snapshots also suggest a mechanism for a phosphate-linked acyl transfer coupled to electron transfer in a radical reaction of pyruvate oxidase.


The mechanistic understanding of ThDP enzymes has recently increased, a fact largely due to the development of kinetic and spectroscopic tools that detect reaction intermediates4, 5, 6. Of the many available X-ray structures of ThDP enzymes3, 7, 8, 9, only a few show reaction intermediates; these include, for example, an enamine intermediate of transketolase10 and a radical intermediate of pyruvate:ferredoxin oxidoreductase11. Thus, the structures of key intermediates, such as the initial tetrahedral substrate-ThDP complex, have remained elusive. The conformation of this intermediate is important for understanding the reaction mechanism because the orientation of the substrate's leaving group relative to the cofactor is thought to determine the specificity of the reaction and to provide much of its driving force.

Here, we examine the crystal structures of key intermediates of the ThDP- and flavin adenine dinucleotide (FAD)-dependent POX from L. plantarum. This homotetrameric enzyme is of particular mechanistic interest because it combines the electrophilic apparatus of ThDP with the redox device of a flavin3. In the presence of phosphate and oxygen, POX catalyzes the oxidative decarboxylation of pyruvate (55), generating carbon dioxide, hydrogen peroxide and the high-energy metabolite acetylphosphate (66).

Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

The catalytic cycle of POX (Scheme 1) starts with the formation of the reactive C2 carbanion form of ThDP, which attacks pyruvate as a nucleophile, yielding LThDP. After decarboxylation, two electrons are transferred from the resonant enamine/carbanion form of HEThDP to FAD. Kinetic studies have shown that this electron transfer and the phosphorolysis of AcThDP are coupled, a process that probably involves a radical phosphate-HEThDP adduct12. In the absence of phosphate, the enamine is oxidized to AcThDP, which undergoes slow hydrolysis5, 13.

Scheme 1: Proposed mechanism of POX catalysis with major intermediates involved in the oxidative decarboxylation of pyruvate.

Scheme 1 : Proposed mechanism of POX catalysis with major intermediates involved in the oxidative decarboxylation of pyruvate. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

The framed thiamin intermediates are those for which a three-dimensional structure has been determined in this study.

Full size image (28 KB)

Using a chemical quenching and 1H NMR spectroscopic approach5, we confirmed the nature of covalent ThDP intermediates in POX catalysis (Fig. 1). At steady state, decarboxylation of LThDP is rate limiting for catalysis, as evidenced by the extensive accumulation of this intermediate at the active site (Fig. 1b). Catalysis can also be blocked at the HEThDP carbanion/enamine stage: under anaerobic conditions and when POX's flavin is fully reduced, a reaction between POX and pyruvate cannot cause an electron transfer from HEThDP to the reduced flavin (Fig. 1c). Similarly, catalysis stops at the AcThDP stage if the substrate phosphate is missing (Fig. 1d). On the basis of these studies, we soaked crystals of POX with substrates under specific conditions so that either LThDP, HEThDP or AcThDP accumulated at the active site. Testing revealed that crystals of POX variant F479W are best suited to meet all requirements. The three-dimensional structures of POX with the different intermediates were determined by X-ray crystallography (Supplementary Table 1 online). Next, we reacted POX with the pyruvate analog methylacetylphosphonate14 (MAP, 77). This compound forms a covalent bond with C2 of ThDP similar to that of pyruvate, but the adjacent Calpha-P bond cannot be cleaved enzymatically. Therefore, phosphono-LThDP (PLThDP, 88) is a stable analog of LThDP.

Figure 1: Detection and quantification of covalent ThDP intermediates in POX catalysis by 1H NMR spectroscopy after acid-quench isolation according to.

Figure 1 : Detection and quantification of covalent ThDP intermediates in POX catalysis by 1H NMR spectroscopy after acid-quench isolation according to.

(a) Upfield section of the 1H NMR spectra of the chemically synthesized intermediates LThDP, HEThDP and AcThDP and of unsubstituted ThDP. The signals shown correspond to the C6'-H resonances (aminopyrimidine ring) of the intermediates and the C2-H resonance (thiazolium ring) of ThDP. Isolated AcThDP exists in an equilibrium of three forms: the keto form, the internal hydrate and the tricyclic carbinolamine13. Thus it yields three C6'-H peaks in the NMR spectrum. (b) Intermediate distribution of POX F479W at the true steady state at pH 6.0 and 25 °C. (c) Intermediate distribution of POX after reaction of the reduced enzyme with pyruvate. (d) Intermediate distribution of POX after reaction of the enzyme with pyruvate in the absence of the substrate phosphate (previously published in ref. 12). (e) Control experiment in the absence of the substrate pyruvate. The exclusive accumulation of LThDP at steady state in b is unique to the POX variant F479W, whereas under the conditions specified for c and d similar intermediate distributions are observed for the wild-type enzyme12 (shown here) and the variant (Supplementary Fig. 4).

Full size image (70 KB)

The structure of the POX variant F479W at 2.03 Å without pyruvate is identical to that of the wild-type enzyme3. At the active site (Fig. 2), the thiamin and flavin cofactors are separated by approximately 7 Å. The cavity between the two cofactors is sandwiched by the side chains of Phe121 and Trp479. The side chains of Gln122, Glu483, Val394 and Ile480 and the main chain atoms of Gly35 and Ser36 are also located near the active site. The presence of numerous hydrophobic side chains suggests that the active site has a low polarity, a conclusion consistent with studies on a related thiamin enzyme15.


Our NMR studies indicate that the LThDP intermediate accumulates at the active site of POX at steady state. The X-ray crystallographic snapshot of POX at steady state with a resolution of 1.97 Å reveals additional electron density close to C2 of ThDP (Fig. 3a). The well-defined shape is consistent with a 2-lactyl moiety bound to ThDP, especially in view of the tetrahedral Calpha carbon. The carboxylate group of LThDP is oriented perpendicular to the thiazolium ring of ThDP and is located in a hydrophobic pocket between the side chains of Trp479 and Ile480. It forms hydrogen bonds with the carboxyl group of Glu483 and the amide nitrogen of Gly35. Even though the substrate soak was short (approx2 s), a distance Fourier** (Fo - Fc) map showed some negative electron density near the carboxylate group, indicating that a fraction of the carboxylate had already split off and linearized to form CO2 (Supplementary Fig. 1 online*). The pyruvate-derived methyl and hydroxyl groups of LThDP are positioned on the opposite side of the thiazolium plane, close to the side chains of Val394 and Gln122 and the 4'-amino group of the pyrimidine part of ThDP (Supplementary Fig. 2 online*). We propose that the methyl group points toward Val394 and the hydroxyl group forms hydrogen bonds with Gln122 and the 4'-amino group. This assignment is corroborated by kinetic studies on various ThDP enzymes, in which the 4'-amino group of ThDP was shown to protonate the carbonyl oxygen of pyruvate5. Our proposal is also consistent with the structure of the (alpha,beta-dihydroxyethyl)-ThDP intermediate in transketolase, for which the corresponding moieties can be definitely assigned10. No major conformational changes of active site residues or of either cofactor can be seen upon covalent binding of pyruvate (Supplementary Fig. 2)*. However, there is a slight distortion of the thiazolium part of LThDP. Notably, the C2-Calpha bond is out of the aromatic ring plane by a few degrees, indicating strain that might help to drive the decarboxylation reaction (Supplementary Fig. 3 online).

Figure 3: The four reaction intermediates of the cofactor ThDP in pyruvate oxidase.

Figure 3 : The four reaction intermediates of the cofactor ThDP in pyruvate oxidase.

Each panel shows the thiazolium ring of ThDP with the modification at its C2 carbon atom, proximate amino acid residues and water molecules in the protein structure. The diphosphate moiety of ThDP has been omitted for clarity, and the electron density was contoured at the indicated level in a 2FoFc map. (a) LThDP (1.0sigma). (b) PLThDP, a stable analog of LThDP (1.8sigma). (c) HEThDP in the enamine form with a nearby phosphate ion (1.5sigma). (d) AcThDP in the keto form (1.8sigma). Amino acids labeled with an asterisk are contributed from the neighboring subunit.

Full size image (94 KB)

The stable analog PLThDP adopts a conformation at the active site of POX similar to that of LThDP (Fig. 3b). The phosphonolactyl moiety forms the same interactions as are observed for the native intermediate and also shows some out-of-plane distortion. The negatively charged phosphonate part, which mimics the carboxylate group, is oriented perpendicular to the thiazolium ring of ThDP. The additional methoxy moiety points toward the exit of the active site without any steric clashes. In a study on the E1 component of pyruvate dehydrogenase published in parallel to this work, the PLThDP was shown to adopt a similar conformation16.

The three-dimensional structures of LThDP and PLThDP suggest the ways that ThDP enzymes (i) selectively destabilize C-C bonds and (ii) facilitate decarboxylation. In the structure of LThDP, the carboxylate moiety is above the plane of the thiazolium system. This orientation makes the positively charged thiazolium ring an optimal electron sink for the electron pair released upon decarboxylation because the formed p orbital conjugates with the thiazolium pi-system. This principle is termed "maximum overlap mechanism"2 and is thought to determine the specificity of other cofactor-dependent reactions—for example, those in pyridoxal phosphate enzymes17. We suggest that this stereochemical control exists in all ThDP enzymes. The presence of the carboxylate leaving group in a presumably low-polarity microenvironment15 is indicative of a nonpolar transition state during decarboxylation that resembles the uncharged HEThDP enamine–carbon dioxide pair. This stabilization of the transition state by the protein results in an enzymatic decarboxylation that is seven orders of magnitude faster than the nonenzymatic reaction12, 18. The observed distortion of LThDP and PLThDP might also drive the decarboxylation reaction.

The decarboxylation of LThDP leads to the formation of the HEThDP carbanion/enamine. Our previous NMR studies have shown that catalysis stops at this stage if the flavin in POX has already been reduced, as two electrons cannot be transferred after binding and decarboxylation of pyruvate. We determined the X-ray structure of the HEThDP intermediate in POX at 2.29 Å. Close to the reactive C2 of ThDP, we observed a well-defined electron density that can be modeled as a hydroxyethyl group (Fig. 3c). Our assignment of the methyl and hydroxyl groups is based on the same premises as listed for that of LThDP above. Unlike the tetrahedral LThDP and PLThDP, the HEThDP is planar, indicating that the intermediate is present in the enamine form with sp2 hybridization of Calpha. The stabilization of the enamine is notable: to prevent protonation, the enzyme must lower the pKa of Calpha-H by 11–12 units as compared to a pKa of 17–18 for the free compound19. This finding is in accordance with recent studies on related thiamin enzymes10, 15, 20. Both thermodynamic and kinetic factors seem to contribute to this stabilization in POX; the protonated enamine carries a positive net charge, so that the low-polarity microenvironment stabilizes the enamine thermodynamically. Furthermore, POX lacks the His-Asp-Glu proton relay found in the related pyruvate decarboxylase5, for which rapid protonation of the carbanion/enamine is part of catalysis. Close to the HEThDP enamine intermediate, we found additional electron density that is too high to originate from a water molecule but is consistent with a tetrahedral oxoanion such as sulfate or the substrate phosphate. As both compounds are components of the crystallization mixture, we cannot determine which is responsible for this additional electron density. For the model we have chosen phosphate, as it is a substrate of the reaction whereas sulfate is not even a competitive inhibitor. Furthermore, when only sulfate was used in the crystallization setup, we found no such well-defined electron density in the structure of POX that contained AcThDP (see below). The phosphate forms multiple contacts with the main chain atoms of Gly34 and Ser35, with the side chains of Gln122 and (via a water molecule) Glu483, with the hydroxyethyl moiety and with the 4'-amino group of HEThDP. Kinetic studies on POX variants further corroborate our assignment because the binding and reaction of phosphate are mediated by these functional groups (unpublished data). The observation of a phosphate bound before electron transfer requires a revision of the earlier mechanistic assumption that phosphate entered the catalytic cycle after electron transfer, that is, at the AcThDP stage3. The crystal structure supports our proposal of a coupling between electron transfer and phosphorolysis12. According to this proposal, one electron is transferred from the enamine to the flavin, resulting in a kinetically stabilized HEThDP radical–flavin semiquinone pair. Considering the Bürgi-Dunitz trajectory, the phosphate in the crystal structure is perfectly positioned for a nucleophilic attack on the long-lived HEThDP radical. The presumably low redox potential of the anionic phosphate-radical adduct facilitates the transfer of the second electron to the flavin, leading to an forward commitment of acyl transfer.

When POX reacts with pyruvate and oxygen in the absence of phosphate, the reaction generates acetate instead of acetylphosphate. The rate-limiting step of this alternative pathway is the hydrolysis of the AcThDP intermediate. We determined the X-ray structure of POX with AcThDP trapped at the active site at a resolution of 2.16 Å (Fig. 3d). A well-defined electron density with a planar conformation is visible close to the reactive C2 of ThDP; it is similar in shape to that of the HEThDP enamine. Apparently, in POX the AcThDP intermediate is stabilized in the reactive keto form. Although free AcThDP is also present as a hydrate (formed by nucleophilic attack of OH- on the acetyl group) and as a tricyclic carbinolamine (with a covalent bond between the carbonyl carbon and the neighboring 4'-NH2 of the pyrimidine part)13, we found no evidence for either of them in the POX crystal structure. The absence of the carbinolamine shows that the rigid V conformation of the cofactor at the active site prevents the orbital overlap required for the cyclization reaction to occur.

In all enzyme structures obtained from pyruvate-soaked crystals, we observed well-defined electron densities consistent with a pyruvate molecule in a surface loop close to the entrance of the active site (Fig. 4). This observation suggests that POX can hold an additional substrate molecule in a 'queuing position' for the next turnover while catalysis takes place simultaneously at the active site. It remains to be studied kinetically whether this auxiliary binding site is an artifact produced by the high pyruvate concentration in the soaking experiments and irrelevant for catalysis or whether it indeed affects substrate binding at the active site.

Figure 4: Exo-site binding cavity of the substrate pyruvate in the structure of the LThDP-containing enzyme.

Figure 4 : Exo-site binding cavity of the substrate pyruvate in the structure of the LThDP-containing enzyme.

(a) The amino acid residues comprising the binding pocket and the two nearest water molecules are shown. The electron density assigned to pyruvate is contoured at 1.5sigma in a 2FoFc map. (b) Surface of the enzyme. The funnel leading to the active site can be seen at the lower right with the lactyl moiety bound to the C2 carbon of ThDP. The auxiliary pyruvate binding site on the left is separated from the active site by about 23 Å.

Full size image (35 KB)

The three-dimensional structures of the intermediates LThDP, HEThDP enamine and AcThDP at the active site of POX provide important insights into the molecular mechanisms of thiamin enzymes. The stereochemical control of decarboxylation is accomplished by the perpendicular orientation of the carboxylate leaving group that has long been proposed but appears to be shown here for the first time. The nonpolar nature of the active site also guarantees a substantial stabilization of the enamine formed after decarboxylation. The redox interplay of the thiamin and flavin cofactors results in a thiamin radical that is sufficiently stable for nucleophilic attack by a phosphate, thereby coupling electron transfer and phosphorolysis. No large structural changes occur during the enzymatic reaction, indicating that the active site is poised for catalysis.

Top

Methods

Protein production, reaction intermediate analysis and chemical synthesis.

We produced pyruvate oxidase recombinantly in Escherichia coli as described previously21. The analysis of the reaction intermediates LThDP, HEThDP, AcThDP and unsubstituted ThDP in POX catalysis using rapid-quenched flow–1H NMR spectroscopy was carried out as detailed in ref. 5. These studies of intermediates show that decarboxylation of LThDP is completely rate limiting only in POX F479W, the latter having an active site occupancy of LThDP greater than or equal to90% (Fig. 1b). At the steady state of wild-type POX the majority of active sites contain LThDP, but approximately 25% of the active sites are occupied by the HEThDP enamine12 (Supplementary Fig. 4 online). Hence, an assignment of electron density originating from the covalent intermediates in the variant is more reliable than in wild-type POX. The pyruvate analog MAP was synthesized according to ref. 14. The reaction of POX with 100 mM MAP for 90 s at 25 °C in 0.2 M potassium phosphate buffer (pH 6.0) leads to the formation of PLThDP as detected by 1H NMR spectroscopy after acid-quench isolation (data not shown).

X-ray crystallography.

We grew the crystals of POX variant F479W in hanging drops using the vapor diffusion method against a reservoir solution containing 1.3 M ammonium sulfate3. Before freezing a crystal in liquid nitrogen, we incubated it in a cryosolution with 1.3 M ammonium sulfate, 115 mM phosphate buffer (pH 5.8; MES buffer for AcThDP), 25% (v/v) glycerol and further additives specific for the intended reaction intermediate as follows: (i) underivatized ThDP: no further additive; (ii) LThDP: 200 mM pyruvate, very short soak for approx2 s; (iii) PLThDP: 100 mM methylacetylphosphonate, soaking for approx90 s; (iv) HEThDP enamine: 200 mM pyruvate, soaking for approx90 s until bleaching of the crystal's bright yellow color indicates fully reduced FAD (anaerobic conditions via enzymatic oxygen consumption) and thus the generation of HEThDP enamine; (v) AcThDP: initial rinse (approx120 s) in cryosolution with MES buffer instead of phosphate buffer to remove the phosphate, second soak (approx15 s) in cryosolution with MES buffer + 200 mM pyruvate, + 10 mM ferricyanide (to ensure fully oxidized FAD).

We then flash froze the crystals in liquid nitrogen and collected datasets at a cryogenic temperature at beamlines X13 (structure without substrate) and BW7B (all others) of the Doppel-Ring Synchrotron (DORIS) radiation facility at the European Molecular Biology Laboratory outstation Deutsches Elektronen-Synchrotron, Hamburg. We processed the datasets with Denzo/Scalepack22 and obtained initial phases through molecular replacement with Molrep23 using the POX wild-type structure (PDB code 1POW) as a search model. We conducted several cycles of model building and refinement with Coot24 and Refmac23, respectively. We prepared crystallographic figures with PyMOL (DeLano Scientific, http://www.pymol.org).

We observed no differences between the two active sites in the crystallographic asymmetric unit. For the discussion, we arbitrarily chose the site where Glu483 belongs to subunit A.

Database accession numbers.

The refined model and structure factors have been deposited in the Research Collaboratory for Structural Biology (http://www.rcsb.org) under the following accession numbers: 2EZ4 (POX variant F479W), 2EZ8 (POX variant F479W with 2-lactyl-ThDP), 2EZ9 (POX variant F479W with 2-phosphonolactyl-ThDP), 2EZT (POX variant F479W with 2-hydroxyethyl-ThDP enamine) and 2EZU (POX variant F479W with 2-acetyl-ThDP).

Note: Supplementary information is available on the Nature Chemical Biology website.

* supp info PDFs 1 and 2 replaced; compound numbering fixed; notes added to HTML

Top

Acknowledgments

We gratefully acknowledge access to synchrotron radiation beamlines X13 and BW7B at the European Molecular Biology Laboratory outstation, Deutsches Elektronen-Synchrotron, Hamburg. We thank R. Schowen, R. Kluger, S. Ghisla, F. Jordan and in particular G. Hübner for many stimulating discussions.

Competing interests statement:

The authors declare no competing financial interests.

Received 2 January 2006; Accepted 4 April 2006; Published online 7 May 2006; Corrected 30 May 2006.

Top

References

  1. Schowen, R.L. Thiamin-dependent enzymes. in Comprehensive Biological Catalysis Vol. 2 (ed. Sinnot, M.) 217–266 (Academic, London, 1998).
  2. Kluger, R. Thiamin diphosphate: a mechanistic update on enzymic and nonenzymic catalysis of decarboxylation. Chem. Rev. 87, 863–876 (1987). | Article | ISI | ChemPort |
  3. Muller, Y.A. & Schulz, G.E. Structure of the thiamine- and flavin-dependent enzyme pyruvate oxidase. Science 259, 965–967 (1993). | PubMed | ISI | ChemPort |
  4. Kern, D. et al. How thiamine diphosphate is activated in enzymes. Science 275, 67–70 (1997). | Article | PubMed | ISI | ChemPort |
  5. Tittmann, K. et al. NMR analysis of covalent intermediates in thiamin diphosphate enzymes. Biochemistry 42, 7885–7891 (2003). | Article | PubMed | ISI | ChemPort |
  6. Nemeria, N. et al. Tetrahedral intermediates in thiamin diphosphate-dependent decarboxylations exist as a 1',4'-imino tautomeric form of the coenzyme, unlike the Michaelis complex or the free coenzyme. Biochemistry 43, 6565–6575 (2004). | Article | PubMed | ISI | ChemPort |
  7. Lindqvist, Y., Schneider, G., Ermler, U. & Sundstrom, M. Three-dimensional structure of transketolase, a thiamine diphosphate-dependent enzyme, at 2.5 A resolution. EMBO J. 11, 2373–2379 (1992). | PubMed | ISI | ChemPort |
  8. Arjunan, P. et al. Crystal structure of the thiamin diphosphate-dependent enzyme pyruvate decarboxylase from the yeast Saccharomyces cerevisiae at 2.3 A resolution. J. Mol. Biol. 256, 590–600 (1996). | Article | PubMed | ISI | ChemPort |
  9. Arjunan, P. et al. Structure of the pyruvate dehydrogenase multienzyme complex E1 component from Escherichia coli at 1.85 A resolution. Biochemistry 41, 5213–5221 (2002). | Article | PubMed | ISI | ChemPort |
  10. Fiedler, E. et al. Snapshot of a key intermediate in enzymatic thiamin catalysis: crystal structure of the alpha-carbanion of (alpha,beta-dihydroxyethyl)-thiamin diphosphate in the active site of transketolase from Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 99, 591–595 (2002). | Article | PubMed | ChemPort |
  11. Chabriere, E. et al. Crystal structure of the free radical intermediate of pyruvate:ferredoxin oxidoreductase. Science 294, 2559–2563 (2001). | Article | PubMed | ISI | ChemPort |
  12. Tittmann, K. et al. Radical phosphate transfer mechanism for the thiamin diphosphate- and FAD-dependent pyruvate oxidase from Lactobacillus plantarum. Kinetic coupling of intercofactor electron transfer with phosphate transfer to acetyl-thiamin diphosphate via a transient FAD semiquinone/hydroxyethyl-ThDP radical pair. Biochemistry 44, 13291–13303 (2005). | Article | PubMed | ISI | ChemPort |
  13. Gruys, K.J., Halkides, C.J. & Frey, P.A. Synthesis and properties of 2-acetylthiamin pyrophosphate: an enzymatic reaction intermediate. Biochemistry 26, 7575–7585 (1987). | Article | PubMed | ISI | ChemPort |
  14. Kluger, R. & Pike, D.C. Active site generated analogs of reactive intermediates in enzymic reactions. Potent inhibition of pyruvate dehydrogenase by a phosphonate analog of pyruvate. J. Am. Chem. Soc. 99, 4505–4506 (1977).
  15. Jordan, F., Li, H. & Brown, A. Remarkable stabilization of zwitterionic intermediates may account for a billion-fold rate acceleration by thiamin diphosphate-dependent decarboxylases. Biochemistry 38, 6369–6373 (1999). | Article | PubMed | ISI | ChemPort |
  16. Arjunan, P. et al. A thiamin-bound, pre-decarboxylation reaction intermediate analogue in the pyruvate dehydrogenase E1 subunit induces large-scale disorder-to-order transformations in the enzyme and reveals novel structural features in the covalently bound adduct. J. Biol. Chem. (in the press) (2006).
  17. Dunathan, H.C. Stereochemical aspects of pyridoxal phosphate catalysis. Adv. Enzymol. 35, 79–134 (1971). | PubMed | ISI | ChemPort |
  18. Kluger, R. & Smyth, T. Thiamin-catalyzed decarboxylation of pyruvate. Synthesis and reactivity analysis of the central, elusive intermediate, alpha-lactylthiamin. J. Am. Chem. Soc. 103, 884–888 (1981). | Article | ISI | ChemPort |
  19. Jordan, F. & Nemeria, N. Experimental observation of thiamin diphosphate-bound intermediates on enzymes and mechanistic information derived from these observations. Bioorg. Chem. 33, 190–215 (2005). | Article | PubMed | ISI | ChemPort |
  20. Zhang, S. et al. Evidence for dramatic rate acceleration of an C-H bond ionization rate in thiamin diphosphate enzymes by the protein environment. Biochemistry 44, 2237–2243 (2005). | Article | PubMed | ISI | ChemPort |
  21. Wille, G. et al. The role of Val-265 for flavin adenine dinucleotide (FAD) binding in pyruvate oxidase: FTIR, kinetic, and crystallographic studies on the enzyme variant V265A. Biochemistry 44, 5086–5094 (2005). | Article | PubMed | ISI | ChemPort |
  22. Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276A, 307–326 (1997).
  23. Collaborative Computational Project Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D50, 760–763 (1994).
  24. Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D 60, 2126–2132 (2004).
  1. Institut für Biochemie, Martin-Luther-Universität Halle-Wittenberg, Kurt-Mothes-Str. 3, 06120 Halle/Saale, Germany.

Correspondence to: Kai Tittmann1 e-mail: kai.tittmann@biochemtech.uni-halle.de

Correspondence to: Georg Wille1 e-mail: georg.wille@biochemtech.uni-halle.de

*Note: In the supplementary information initially published online to accompany this letter, the legends of Supplementary Figures 1 and 2 are incorrect. The legends contain incorrect citations to other figures; Supplementary Figure 1 should cite Fig. 3a and Supplementary Figure 2 should cite Supplementary Fig. 3. These errors have been corrected online. Also, in the version of this article initially published online, compounds in the Compound Data Index are listed in the wrong order; thus the numbered compounds in the article link to the wrong structures. This error has been corrected in the HTML version of the article.

**Note: In the version of this article initially published, there was an error in the text of the second page. In line 15 of the first column, the text should read "a difference Fourier" rather than "a distance Fourier." The error has been corrected in the PDF version of the article.

MORE ARTICLES LIKE THIS

These links to content published by NPG are automatically generated.

NEWS AND VIEWS

Snapshots of three intermediates at the active site of pyruvate oxidase

Nature Chemical Biology News and Views (01 Jun 2006)

Adenosine triphosphate and thiamine cross paths

Nature Chemical Biology News and Views (01 Apr 2007)