Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2002 Jan 2;99(2):591–595. doi: 10.1073/pnas.022510999

Snapshot of a key intermediate in enzymatic thiamin catalysis: Crystal structure of the α-carbanion of (α,β-dihydroxyethyl)-thiamin diphosphate in the active site of transketolase from Saccharomyces cerevisiae

Erik Fiedler *, Stina Thorell †, Tatyana Sandalova †, Ralph Golbik *, Stephan König *, Gunter Schneider †,‡
PMCID: PMC117350  PMID: 11773632

Abstract

Kinetic and spectroscopic data indicated that addition of the donor substrate hydroxypyruvate to the thiamin diphosphate (ThDP)-dependent enzyme transketolase (TK) led to the accumulation of the α-carbanion/enamine of (α,β-dihydroxyethyl) ThDP, the key reaction intermediate in enzymatic thiamin catalysis. The three-dimensional structure of this intermediate trapped in the active site of yeast TK was determined to 1.9-Å resolution by using cryocrystallography. The electron density suggests a planar α-carbanion/enamine intermediate having the E-configuration. The reaction intermediate is firmly held in place through direct hydrogen bonds to His-103 and His-481 and an indirect hydrogen bond via a water molecule to His-69. The 4-NH2 group of the amino-pyrimidine ring of ThDP is within 3 Å distance to the α-hydroxy oxygen atom of the dihydroxyethyl moiety but at an angle unfavorable for a strong hydrogen bond. No structural changes occur in TK on formation of the reaction intermediate, suggesting that the active site is poised for catalysis and conformational changes during the enzyme reaction are not very likely. The intermediate is present with high occupancy in both active sites, arguing against previous proposals of half-of-the-sites reactivity in yeast TK.

Thiamin diphosphate (ThDP) is a cofactor in many enzyme-catalyzed reactions where a carbon—carbon bond adjacent to a carbonyl group is to be cleaved. Enzymatic thiamin catalysis proceeds through two carbanion intermediates, the ylide of ThDP, formed after abstraction of the C2 proton of the thiazolium ring, and the 2-α carbanion, formed after nucleophilic attack of the C2 carbon on the substrate (1–4). It is thought that the α-carbanion intermediate is stabilized via the thiazolium ring of ThDP, which acts as an electron sink. A major source of stabilization is the neutral enamine, contributing to the resulting resonance hybrid (Fig. 1). The α-carbanion/enamine intermediate is of central importance in thiamin catalysis because, depending on the nature of its substituent at the α-carbon atom, a variety of enzymatic functions can be derived. Accordingly, the fate of this intermediate differs in the various ThDP-dependent enzymes; for instance, in pyruvate decarboxylase, the α-carbanion will be protonated at the α-carbon position, leading to the expulsion of the product, acetaldehyde (2, 4). In the transketolase (TK) reaction, the carbanion reacts with an acceptor substrate, and the product, a ketose with the carbon chain extended by two carbon atoms, will be released (3, 5). Although it is textbook knowledge that the α-carbanion is a key intermediate common to all ThDP-dependent enzymes, the intermediate has so far been elusive and has escaped unambiguous detection during catalysis.

Figure 1.

Figure 1

Open in a new tab

Formation of the α-carbanion/enamine intermediate by using HPA as donor substrate.

TK is a ubiquitous ThDP-dependent enzyme found in the nonoxidative branch of the pentose phosphate cycle. The enzymes from bacterial, yeast, and mammalian sources are homodimers with a molecular mass of about 74 kDa per subunit (6–8). The ThDP molecule binds in a cleft at the interface between the two subunits and, except for the C2 carbon atom of the thiazolium ring, is completely inaccessible from the solution (9, 10). ThDP is bound in the V-conformation, a feature characteristic of all ThDP-dependent enzymes (3, 11). This conformation positions the 4-amino group of the pyrimidine ring of ThDP in close proximity to the C2 carbon atom of the thiazolium ring of the cofactor, an essential prerequisite for catalysis (3, 4, 9, 12).

TK catalyses ketol transfer between ketose and aldose sugars. The first half of the reaction cycle consists of the cleavage of the donor substrate and release of the first product, an aldose, and the formation of a covalently bound intermediate, the α-carbanion/enamine of α,β-dihydroxyethyl-ThDP (DHEThDP). In the second half of the reaction, the two-carbon unit is transferred to the acceptor substrate, leading to the formation of a ketose with an extended carbon skeleton. A peculiarity of TK is the acceptance of β-hydroxypyruvate (HPA) as donor substrate, a feature that has been exploited for the organic synthesis of novel ketoses by using TK as biocatalyst (13, 14). Formation of the α-carbanion/enamine of DHEThDP starting from HPA is in practice irreversible, because it is coupled to the decarboxylation of the substrate (Fig. 1). In the absence of a suitable acceptor substrate, the intermediate decomposes into enzyme–ThDP and glycolaldehyde (15), albeit at a rather low rate. This suggested that it would be possible to establish experimental conditions that would lead to significant steady-state levels of this intermediate in the active site of TK, making its detection by biophysical methods possible. Here we describe the three-dimensional structure of the α-carbanion/enamine of DHEThDP at 1.9-Å resolution, trapped by flash-freezing in crystals of TK. The study, which provides, to our knowledge, the first crystallographic evidence of a covalent intermediate in enzymatic ThDP catalysis, reveals groups on the enzyme responsible for stabilization of the reaction intermediate and which may participate in catalysis.

Materials and Methods

Materials.

ThDP and HPA were obtained from Sigma–Aldrich; polyethylene glycol 6000 and ethylene glycol were from Merck. All other chemicals were of the highest purity available.

Protein Expression and Purification.

H402 (tkl1∷HIS3 derivative of W303–1A) yeast cells carrying the wild-type plasmid pTKL1 (6) were cultured in a synthetic medium (with galactose and without leucine) as described earlier (16). TK was purified according to the protocol of Wikner et al. (17), with modifications as described in ref. 15.

Circular Dichroism (CD) Measurements.

To investigate the stability of the reaction intermediate, near-UV CD spectroscopy was used. The CD spectra of TK were recorded on a Aviv 62 CD (Aviv Associates, Lakewood, NJ) spectrophotometer under conditions similar to the crystallization experiments (3.4 mg TK/ml in 100 mM glycyl-glycine/1 mM ThDP/5 mM CaCl2, pH 7.9/40 mM HPA). The temperature was set to 4°C, and spectra were recorded after incubation times of 30 s, 9 min, 15 min, and 30 min by using cuvettes with an optical path length of 1 cm.

Kinetics of Formation of the TK-HPA Adduct.

The formation of the TK-intermediate complex formed through reaction with HPA was followed directly by the associated absorbance change at 300 nm (18) by using a stopped-flow spectrophotometer (SX.18MV, Applied Photophysics, Surrey, U.K.) at 25°C. The two syringes had the same volume, resulting in a 1:2 dilution of the solutions. One syringe contained 50 mM glycyl-glycine, 2.5 mM CaCl2, 0.1 mM ThDP, and 13.4 μM TK, pH 7.6. The other syringe contained the same mixture but, instead of TK, 4 mM HPA.

Crystallography and Data Collection.

TK was crystallized by the vapor diffusion technique as described previously (19) by using polyethylene glycol (PEG) 6000 (13–17% wt/vol) as precipitant in the presence of 5 mM ThDP and 5 mM CaCl2 (drop volume, 15 μl). The crystals grew within 6 days at 4°C. The reaction was started by incubating crystals in the cryo-protecting solution [50 mM glycyl-glycine buffer, containing 5 mM ThDP, 5 mM CaCl2, 20% (wt/vol) PEG 6000, 20% (vol/vol) ethylene glycol pH 7.9] at 4°C, including the donor substrate HPA (40 mM). The reaction was stopped by transfer of crystals into a nitrogen stream at 110 K at various time intervals. X-ray data were collected at beamline BW7B, European Biological Molecular Laboratory outstation, Deutsches Elektronen Synchrotron Hamburg, by using crystals incubated for 30 s and 30 min, respectively. All data sets were processed with denzo and scalepack (20); the statistics are given in Table 1.

Table 1.

Data collection and refinement statistics

Data sets
0.5 min 30 min
Space group P212121 P212121
Cell dimensions, Å 74.3 × 113.2 × 159.4 73.8 × 113.1 × 159.2
Resolution, Å 2.37 1.86
Completeness, % 93.6 92.9
Total no. of reflections 178,260 333,802
Unique reflections 51,947 104,364
I//σ 8.7 15.9
R, % 22.1 19.8
Rfree, % 25.7 22.6
B factors, Å2
Overall 22.6 21.3
DHEThDP 23.8 20.2
Solvent 23.5 27.7
Number of solvent
Molecules 370 854
Number of atoms 10,822 11,330
rms bond lengths 0.008 0.006
rms bond angles 1.42 1.28
Ramachandran plot
Percentage of nonglycine residues in favorable regions 87.9 88.0
Additionally allowed regions 12.1 12.0
Open in a new tab

Crystallographic Model Building and Refinement.

The model of holo-TK, refined to 2.0-Å resolution (10), was used as the source of initial phase information for the calculation of the electron density maps. Inspection of electron density maps and model building were carried out by using the program o (21). Refinement was done with cns (22). The refinement procedure started with simulated annealing, and the following cycles consisted of positional and B factor refinement. Five percent of the diffraction data was set aside to monitor the progress by means of the free R factor. Tight noncrystallographic symmetry restraints were imposed for all residues of the two polypeptide chains in the asymmetric unit. The protocol consisted of iterative rounds of refinement and model examination/rebuilding with o (21), until the Rfree value had converged. The final model contains amino acids 3–680 for each subunit of the dimer, two DHEThDP molecules, two Ca2+ ions, and 854 (370 in the lower-resolution data set) water molecules. The model was analyzed with procheck (23); the statistics of the refinement and model parameters are given in Table 1. Structural comparisons and superpositions were made with the program o (21) by using default parameters.

Results

CD Measurements and Kinetics of HPA Binding to TK.

The binding of the cofactor ThDP and the donor substrate HPA to TK can be monitored by near-UV CD spectroscopy (24, 25). Holo-TK is characterized by a negative band in the CD spectrum at 320 nm. Inversion of this band is observed on addition of donor substrate. The spectrum of holo-TK can be restored by addition of an acceptor substrate, indicating that the spectroscopic changes might be related to catalytic intermediates. This signal, proposed to correspond to the formation of the α-carbanion/enamine intermediate of (α,β-dihydroxyethyl)-ThDP (24), was used as an analytical tool to follow the stability of the intermediate. As illustrated in Fig. 2, the CD signal at 320 nm is stable in solution for at least 30 min under crystallization conditions. From the progress curves of the stopped-flow experiments (data not shown), a rate constant of 44.6 mM−1⋅s−1 was calculated for the formation of the TK–HPA complex and a rate constant of 0.0026 s−1 for release of glycolaldehyde. Decomposition of DHEThDP is thus the rate-limiting step under these conditions.

Figure 2.

Figure 2

Open in a new tab

Time dependence of the CD spectrum of holo-TK incubated with HPA (Top to Bottom, after 30-s incubation, 9-min incubation, 15-min incubation, 30-min incubation, holo-TK without HPA). The reaction mixture contained 1 mM ThDP, 5 mM CaCl2, 40 mM HPA, and 3.4 mg/ml of TK in 100 mM glycyl-glycine, pH 7.9. The optical pathway of the cuvette was 1 cm and the temperature 4°C.

Electron Density Maps and Overall Structure.

After reaction initiation with HPA, two data sets of presumptive TK-intermediate complexes were collected. The structures of these complexes were solved by using difference Fourier methods and were refined. The resulting electron density maps are of very good quality, and the refinement statistics are given in Table 1. The R factors and stereochemistry are as expected for models at the given resolution. The overall structure of the polypeptide chain in these complexes is very similar to the structure of holo-TK. Superposition of the 678 Cα atoms of the TK subunit results in a rms deviation of 0.3 Å, and there are no local deviations larger than 0.5 Å, even for side chains. Thus, there is no significant structural change in the structures described here that would have indicated a conformational transition on formation of the intermediate.

Structure of the DHEThDP Intermediate in the Active Site of TK.

The most significant feature in the initial Fo − Fc difference electron density maps calculated with both data sets is a strong positive electron density extending from the C2 carbon atom of the thiazolium ring of ThDP into the active site (Fig. 3), indicating a covalent adduct at this carbon atom. A model of the α,β dihydroxyethyl moiety of DHEThDP could be fitted straightforwardly into this difference electron density, and the resulting maps after refinement clearly indicate successful trapping of the α-carbanion/enamine intermediate in the active site of TK (Fig. 3). After refinement, the B factors for the atoms of the α,β dihydroxyethyl moiety (22.8 Å2) are very similar to those for ThDP (19.9 Å2) and surrounding active-site residues, indicating stoichiometric formation of the intermediate. The results are very similar for the data sets collected 30 sec and 30 min after reaction initiation. In the following, we will describe the structure of the intermediate as observed in the 30-min data set because of its higher resolution.

Figure 3.

Figure 3

Open in a new tab

Electron density maps for the bound α-carbanion/enamine of DHEThDP in the active site of TK. (a) Initial Fo − Fc difference electron density map, contoured at 3.5σ (green) and 6.0σ (red), data set t = 30 sec. (b) Refined 2Fo − Fc difference electron density map, contoured at 1.0σ, t = 30 sec. The atoms of the dihydroxyethyl moiety were not included in the structure factor calculation. (c) Initial Fo − Fc difference electron density map, contoured at 3.5σ (green) and 6.0σ (red), data set t = 30 min. (d) Refined 2Fo − Fc difference electron density map, contoured at 1.0σ, t = 30 min. The atoms of the dihydroxyethyl moiety were not included in the structure factor calculation.

The crystallographic data show that the atoms of the thiazolium ring, as well as the Cα, Cβ, and the Cα oxygen atoms of the dihydroxyethyl moiety, are all in one plane. The electron density fits best to a planar structure of DHEThDP, i.e., sp2 hybridization of the α carbon atom, suggesting a predominant enamine character of the intermediate (Fig. 1). Attempts to model the intermediate with sp3 geometry at the α carbon atom are not consistent with the electron density and also result in a small but significant increase of the free R factor by 0.4%. The carbanion/enamine has the E-configuration, as predicted from molecular modeling of decarboxylation of pyruvate in pyruvate decarboxylase (26). The α,β dihydroxyethyl moiety is held in place through a number of hydrogen bonds made to surrounding amino acids (Fig. 4). The β-hydroxyl oxygen interacts through hydrogen bonds with the side chain of His-103 and a water molecule that, in turn, is anchored to the protein by hydrogen bonds to the side chain of His-69 and main-chain atoms of residues His-69 and Gly-116. The α-hydroxyl oxygen forms a hydrogen bond to the side chain of His-481. It is also within close distance (2.9 Å) of the 4-NH2 group of the aminopyrimidine ring of ThDP, however at an angle (74° between the Cα carbon, the oxygen at the Cα and the 4-amino nitrogen) not favorable for the formation of a strong hydrogen bond. None of the side chains interacting with the α,β dihydroxyethyl moiety of DHEThDP changes its position compared with the structure of holo-TK, and they are thus poised already in the absence of the donor substrate to interact with the α-carbanion intermediate. The interactions of the ThDP moiety with residues in the cofactor-binding site also remain unchanged. Furthermore, the formation of the covalent intermediate does not result in changes in conformation of the cofactor itself (Fig. 5).

Figure 4.

Figure 4

Open in a new tab

Stereoview of the surroundings of the reaction intermediate DHEThDP in the active site of TK. Hydrogen bonds (distances <3.1 Å) to neighboring residues are indicated by dashed lines.

Figure 5.

Figure 5

Open in a new tab

Superposition of the ThDP molecule bound in holoTK with the covalent reaction intermediate DHEThDP in the TK-intermediate complex. The superposition is based on 678 Cα atoms of the TK subunit.

Discussion

The kinetic and spectroscopic data indicated that, in the absence of the acceptor substrate, the α-carbanion/enamine intermediate is relatively stable in the active site of TK. The use of the donor substrate HPA, coupled with an in practice irreversible decarboxylation step, led to the accumulation of this intermediate and allowed determination of its three-dimensional structure by using cryocrystallography. A large body of indirect evidence accumulated over the last decades has suggested the existence of covalent intermediates during catalysis by ThDP-dependent enzymes, and the present study provides the first direct observation, to our knowledge, of such a covalent reaction intermediate in the active site of a ThDP-dependent enzyme.

The reaction intermediate is held in place by several hydrogen bonds. The interaction with the invariant residue His-103 is important for two reasons. It is a key feature for the recognition of the β-hydroxyl group, i.e., the discrimination between hydroxypryuvate and pyruvate. Indeed, mutagenesis data for this amino acid indicated a major role in binding of the donor substrate but not the acceptor substrate (27). Furthermore, the His-103→Ala mutant is severely impaired in the formation of the α-carbanion intermediate, consistent with the stabilizing interaction seen in the crystal structure of this intermediate. The hydrogen bond of the α-hydroxyl oxygen to the side chain of His-481 also contributes to interactions with the donor substrate, because its replacement with glutamine, serine, or alanine led to significantly increased Km values for the donor substrate but only moderate decrease in kcat and no changes for the Km values for the acceptor substrate (28). This situation appears to be different in mammalian TK, where this residue is replaced by glutamine. Site-directed mutagenesis of this glutamine residue in human TK led to only minor changes in the kinetic parameters (29).

In TK, the carbanion/enamine acts as nucleophile in the second step of the reaction. The enzyme therefore has to avoid protonation of the carbanion/enamine, different from the situation, for instance, in pyruvate decarboxylase, where protonation of the carbanion/enamine is part of the catalytic mechanism. These mechanistic differences are reflected in the active-site topology of the two enzymes. In pyruvate decarboxylase, there is a residue (Glu-477) suitably positioned to participate in proton transfer (26), whereas in TK, no such amino acid is found close to the Cα atom of the intermediate.

A key interaction of central importance to catalysis is the interaction of the 4-NH2 group of the aminopyrimidine ring with the oxygen at the α-carbon position. The essential function of this group had early been demonstrated by showing that deamino-ThDP is catalytically inactive (12, 30). Although in the structure of the α-carbanion/enamine intermediate, the geometry is not favorable for a hydrogen bond because of the bond angle of 74°, this situation will be different in the steps immediately preceding and subsequent to the formation of DHEThDP. During the initial nucleophilic attack of the C2 carbanion on the carbonyl carbon of the donor substrate, a negative charge will develop at the carbonyl oxygen atom, and the hybridization at the carbonyl carbon atom will change from sp2 to sp3. This electronic rearrangement leads to a geometry in the transition state and in the first covalent adduct that is more favorable for hydrogen bond formation between the α-hydroxyl oxygen and the 4-NH2 group. At this stage, the 4-NH2 group can act in proton transfer to this oxygen atom in a manner similar to that described for the deprotonation of the C2 carbon of the thiazolium ring of ThDP (31). In this way, the negative charge of the oxygen atom at the α-carbon is compensated. The structure of the α-carbanion/enamine intermediate described here thus supports mechanistic proposals pinpointing the importance of the 4-NH2 group of ThDP in proton transfer steps during catalysis (3, 4, 12). The conformation of the carbanion/enamine intermediate with respect to the thiazolium ring, the Cα, Cβ, and the Cα oxygen atoms of the α,β dihydroxyethyl moiety observed in TK is very likely common to other ThDP-dependent enzymes. For instance, it agrees well with the conformation of the hydroxyethyl ThDP in pyruvate decarboxylase derived from modeling studies (26). The mechanistic conclusions concerning the 4-NH2 group of ThDP, therefore, may hold also for other ThDP-dependent enzymes.

The reaction was initiated by soaking of the small donor substrate into crystals of the holoenzyme. Despite the time required for diffusion of the substrate into the active site of the enzyme, the intermediate is observed with high occupancy already in the 30-sec data set, indicating that the catalytic steps leading to the α-carbanion must occur at significant rates also in the crystal. The lack of any electron density for the carboxyl group of HPA suggests further that the decarboxylation step is fast compared with the decomposition of the α-carbanion into ThDP and glycolaldehyde. This finding in turn supports the proposal (24) that the spectroscopic changes in the absorption and CD spectra around 300–320 nm are because of the presence of the α-carbanion/enamine of D 72 ΕThDP at the active site of the enzyme.

The reaction intermediate accumulates to similar high occupancies in the two active sites of TK, as indicated by the electron density maps and the B factors for the α,β dihydroxyethyl moiety of the bound DHEThDP. The active sites in the dimer are thus independent of each other, and our results are not compatible with proposals of half-of-the-site reactivity of TK (32).

The steps from binding of donor substrate to the formation of the α-carbanion intermediate can occur within the crystal lattice, and we do not observe any large conformational changes in the enzyme during these catalytic steps, which represent the first half of the TK reaction. Structural changes on formation of the enzyme intermediate, which had been postulated on the basis of molecular dynamics simulations (33), are thus not supported by the present crystal structure analysis. The second part of the reaction, transfer of the glycolaldehyde moiety onto an acceptor substrate, is in principle the reversal of the first catalytic steps. Together with the observation that binding of the acceptor substrate erythrose-4-phosphate to holoTK does not induce any large-scale conformational changes in the enzyme (34), it appears not likely that catalysis by TK is associated with any significant structural changes. Conformational transitions that might be invoked to occur transiently, i.e., escape detection by crystallography, must also be limited in size because the crystal lattice is not disturbed on addition of the donor substrate HPA.

Acknowledgments

We gratefully acknowledge access to synchrotron radiation at beamline BW7b, European Molecular Biology Laboratory outstation, Deutsches Elektronen Synchrotron, Hamburg. We thank C. Enroth for assistance at the beamline and H. Lilie for help with CD measurements. E.F. was supported by grants of the Graduiertenförderung of Sachsen-Anhalt. This work was supported by the Deutsche Forschungsgemeinschaft, Fonds der Chemischen Industrie, and the Swedish Research Council.

Abbreviations

DHEThDP

(α,β-dihydroxyethyl)-thiamin diphosphate

HPA

β-hydroxypyruvate

ThDP

thiamin diphosphate

CD

circular dichroism

TK

transketolase

Footnotes

This paper was submitted directly (Track II) to the PNAS office.

Data deposition: The atomic coordinates and structure factor amplitudes have been deposited in the Protein Data Bank, www.rcsb.org (PDB ID codes 1gpu and r1gpusf, respectively).

References

  • 1.Breslow R. J Am Chem Soc. 1958;80:3719–3726. [Google Scholar]
  • 2.Kluger R. Chem Rev. 1987;87:863–876. [Google Scholar]
  • 3.Schneider G, Lindqvist Y. Bioorg Chem. 1993;21:109–117. [Google Scholar]
  • 4.Schellenberger A. Biochem Biophys Acta. 1998;1385:177–186. doi: 10.1016/s0167-4838(98)00067-3. [DOI] [PubMed] [Google Scholar]
  • 5.Gubler C J. In: Biochemistry and Physiology of Thiamin Diphosphate Dependent Enzymes. Bisswanger H, Ulrich J, editors. Weinheim, Germany: VCH; 1991. pp. 311–321. [Google Scholar]
  • 6.Sundström M, Lindqvist Y, Schneider G, Hellman U, Ronne H. J Biol Chem. 1993;268:24346–24352. [PubMed] [Google Scholar]
  • 7.Sprenger G A, Schörken U, Sprenger G, Sahm H. Eur J Biochem. 1995;230:525–532. doi: 10.1111/j.1432-1033.1995.0525h.x. [DOI] [PubMed] [Google Scholar]
  • 8.McCool B A, Plonk S G, Martin P R, Singleton C. J Biol Chem. 1993;268:1397–1404. [PubMed] [Google Scholar]
  • 9.Lindqvist Y, Schneider G, Ermler U, Sundström M. EMBO J. 1992;11:2373–2379. doi: 10.1002/j.1460-2075.1992.tb05301.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Nikkola M, Lindqvist Y, Schneider G. J Mol Biol. 1994;238:387–404. doi: 10.1006/jmbi.1994.1299. [DOI] [PubMed] [Google Scholar]
  • 11.Muller Y, Lindqvist Y, Furey W, Schulz G E, Jordan F, Schneider G. Structure (London) 1993;1:95–103. doi: 10.1016/0969-2126(93)90025-c. [DOI] [PubMed] [Google Scholar]
  • 12.Schellenberger A. Chem Ber. 1990;123:1489–1494. [Google Scholar]
  • 13.Kobori Y, Myles DC, Whitesides G M. J Org Chem. 1992;57:5899–5907. [Google Scholar]
  • 14.Hecquet L, Bolte J, Demuynck C. Tetrahedron. 1994;50:8677–8684. [Google Scholar]
  • 15.Fiedler E, Golbik R, Schneider G, Tittmann K, Neef H, König S, Hübner G. J Biol Chem. 2001;276:16051–16058. doi: 10.1074/jbc.M007936200. [DOI] [PubMed] [Google Scholar]
  • 16.Erhart E, Hollenberg C P. J Bacteriol. 1983;156:625–635. doi: 10.1128/jb.156.2.625-635.1983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Wikner C, Meshalkina L, Nilsson U, Nikkola M, Lindqvist Y, Sundström M, Schneider G. J Biol Chem. 1994;269:32144–32150. [PubMed] [Google Scholar]
  • 18.Kochetov G A, Usmanov R A, Mevkh A T. Biochem Biophys Res Commun. 1973;54:1619–1626. doi: 10.1016/0006-291x(73)91172-8. [DOI] [PubMed] [Google Scholar]
  • 19.Schneider G, Sundström M, Lindqvist Y. J Biol Chem. 1989;264:21619–21620. [PubMed] [Google Scholar]
  • 20.Otwinowski Z. In: Proceedings of the CCP4 Study Weekend. Lawyer L, Isaacs N, Bailey S, editors. Warrington, U.K.: Daresbury Laboratory; 1993. pp. 56–62. [Google Scholar]
  • 21.Jones T A, Zou J-Y, Cowan S W, Kjeldgaard M. Acta Crystallogr D. 1991;47:110–119. doi: 10.1107/s0108767390010224. [DOI] [PubMed] [Google Scholar]
  • 22.Brünger A T, Adams P D, Clore G M, DeLano W L, Gros P, Grosse-Kunstleve R W, Jiang J-S, Kuszewski J, Nilges M, Pannu N S, et al. Acta Crystallogr D. 1998;54:905–921. doi: 10.1107/s0907444998003254. [DOI] [PubMed] [Google Scholar]
  • 23.Laskowski R A, MacArthur M W, Moss D S, Thornton J M. J Appl Crystallogr. 1993;26:283–291. [Google Scholar]
  • 24.Kochetov G A, Usmanov R A, Merzlov V P. FEBS Lett. 1970;9:265–266. doi: 10.1016/0014-5793(70)80372-6. [DOI] [PubMed] [Google Scholar]
  • 25.Heinrich C P, Noack K, Wiss O. Biochem Biophys Res Commun. 1971;44:275–279. doi: 10.1016/0006-291x(71)90595-x. [DOI] [PubMed] [Google Scholar]
  • 26.Lobell M, Crout D H G. J Am Chem Soc. 1996;118:1867–1873. [Google Scholar]
  • 27.Wikner C, Meshalkina L, Nilsson U, Bäckström S, Lindqvist Y, Schneider G. Eur J Biochem. 1995;233:750–755. doi: 10.1111/j.1432-1033.1995.750_3.x. [DOI] [PubMed] [Google Scholar]
  • 28.Wikner C, Nilsson U, Meshalkina L, Udekwu C, Lindqvist L, Schneider G. Biochemistry. 1997;36:11643–11649. doi: 10.1021/bi971606b. [DOI] [PubMed] [Google Scholar]
  • 29.Singleton C K, Wang J L, Shan L, Martin P R. Biochemistry. 1996;35:15865–15869. doi: 10.1021/bi9616920. [DOI] [PubMed] [Google Scholar]
  • 30.Golbik R, Neef H, Hübner G, König S, Seliger B, Meshalkina M, Kochetov G A, Schellenberger A. Bioorg Chem. 1991;19:10–17. [Google Scholar]
  • 31.Kern D, Kern G, Neef H, Tittmann K, Killenberg-Jabs M, Wikner C, Schneider G, Hübner G. Science. 1997;275:67–70. doi: 10.1126/science.275.5296.67. [DOI] [PubMed] [Google Scholar]
  • 32.Kovina M V, Kochetov G A. FEBS Lett. 1998;440:81–84. doi: 10.1016/s0014-5793(98)01423-9. [DOI] [PubMed] [Google Scholar]
  • 33.Kovina M V, Tikhonova O V, Soloveva O N, Bykova I A, Ivanov A S, Kochetov G A. Biochem Biophys Res Commun. 2000;275:968–972. doi: 10.1006/bbrc.2000.3412. [DOI] [PubMed] [Google Scholar]
  • 34.Nilsson U, Meshalkina L, Lindqvist Y, Schneider G. J Biol Chem. 1997;272:1864–1869. doi: 10.1074/jbc.272.3.1864. [DOI] [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES