Present address: Biozentrum, Universität Basel, Departement Biophysikalische Chemie, Klingelbergstrasse 70, 4056 Basel, Switzerland.
Pyruvate decarboxylase from Kluyveromyces lactis
An enzyme with an extraordinary substrate activation behaviour
Article first published online: 4 AUG 2009
European Journal of Biochemistry
Volume 269, Issue 13, pages 3256–3263, July 2002
How to Cite
Krieger, F., Spinka, M., Golbik, R., Hübner, G. and König, S. (2002), Pyruvate decarboxylase from Kluyveromyces lactis. European Journal of Biochemistry, 269: 3256–3263. doi: 10.1046/j.1432-1033.2002.03006.x
Note: a web site is available at http://www.biochemtech.uni-halle.de/
- Issue published online: 4 AUG 2009
- Article first published online: 4 AUG 2009
- (Received 20 March 2002, accepted 17 May 2002)
- thiamine diphosphate;
- microscopic rate constants;
Pyruvate decarboxylase (EC 184.108.40.206) was isolated and purified from the yeast Kluyveromyces lactis. The properties of this enzyme relating to the native oligomeric state, the subunit size, the nucleotide sequence of the coding gene(s), the catalytic activity, and protein fluorescence as well as circular dichroism are very similar to those of the well characterized pyruvate decarboxylase species from yeast. Remarkable differences were found in the substrate activation behaviour of the two pyruvate decarboxylases using three independent methods: steady-state kinetics, stopped-flow measurements, and kinetic dilution experiments. The dependence of the observed activation rate constant on the substrate concentration of pyruvate decarboxylase from K. lactis showed a minimum at a pyruvate concentration of 1.5 mm. According to the mechanism of substrate activation suggested this local minimum occurs due to the big ratio of the dissociation constants for the binding of the first (regulatory) and the second (catalytic) substrate molecule. The microscopic rate constants of the substrate activation could be determined by a refined fit procedure. The influence of the artificial activator pyruvamide on the activation of the enzyme was studied.
pyruvate decarboxylase (2-oxo acid carboxy lyase, EC 220.127.116.11)
pyruvate decarboxylase from brewer's yeast
pyruvate decarboxylase from Kluyveromyces lactis
pyruvate decarboxylase from Pisum sativum
pyruvate decarboxylase from Saccharomyces cerevisiae
pyruvate decarboxylase from Zymomonas mobilis
The cytosolic pyruvate decarboxylase (PDC) is a key enzyme at the branching point of alcoholic fermentation and respiration in yeast, some bacteria and plant seeds. It catalyses the nonoxidative decarboxylation of pyruvate to acetaldehyde and carbon dioxide. Thiamine diphosphate (TDP) and Mg2+ are both required as cofactors in this reaction. In yeast and bacteria, the catalytically active enzyme is composed of four subunits. PDC species from plant seeds are able to form higher oligomeric states [1–3]. In the genome of the yeast Kluyveromyces lactis one gene was found to code for PDC . Its nucleotide sequence has 85% identity to PDC1 from Saccharomyces cerevisiae. All amino acids that are likely to be involved in the regulation and catalysis of PDC are conserved. The phenomenon of substrate activation has been described for all PDC species investigated so far, except for the enzyme from Zymomonas mobilis. In 1978, Hübner et al.  described the kinetics of substrate activation of ByPDC (PDC from brewer's yeast) showing a hyperbolic dependence of the activation rate constant on the substrate concentration. Furthermore, it was demonstrated that the artificial substrate surrogate pyruvamide is able to activate PDC. The following minimal model of the catalytic mechanism was derived :
A substrate molecule binds rapidly to a regulatory site of the inactive enzyme Ei and triggers an isomerization towards an active enzyme conformation SEa. In a subsequent step the active conformation state binds a second substrate molecule and catalyses its decarboxylation to yield acetaldehyde (AA). The isomerization step proceeds slowly compared to the substrate binding. Ka is the dissociation constant of the substrate binding to the regulatory site preceding the isomerization. The isomerization constant Kiso is equal to the ratio k–iso/kiso. The Km value for substrate conversion is defined as:
A comparison of the crystal structures of native and pyruvamide-activated ByPDC clearly demonstrated that this isomerization is realized by a rearrangement of the dimers within the tetramer. This 30° rotation resulted in a disorder-order transition of two loop regions and thus in closing two of four active sites of the enzyme [7,8].
Here, PDC from K. lactis was characterized. The substrate activation behaviour of this enzyme displayed a complex dependence of the activation rate constant on the substrate concentration. The dissociation constants of the substrate at the regulatory and the catalytic site were determined and compared to other PDC species.
Materials and methods
All reagents used for enzyme purification and activity measurements were of analytical grade and purchased from Merck, Serva, and Sigma–Aldrich. Columns and media were from Amersham Pharmacia Biotech.
Yeast strain and media
The K. lactis strain JA-6 was a gift from I. Eberhardt (Laboratorium of Molecular Cell Biology, Biological Department, Katholic University Leuven, Belgium; present address, Department of Molecular Biology, Gent University, Belgium). The yeast complete medium contained 1% (w/v) yeast extract, 2% (w/v) peptone, 5% (w/v) glucose, 0.1 mm thiamine and 0.1 mm magnesium sulfate. The yeast cultures (1 L) were inoculated with precultures (10% of the volume) and grew for at least 24 h at 29 °C at a shaker frequency of 110 r.p.m.
Purification of PDC from K. lactis
The purification procedure was developed on the basis of methods established for other PDC from other species [2,9–11]. All steps were carried out at 4 °C. Frozen K. lactis cells (60 g) were thawed in 200 mL of 100 mm sodium phosphate pH 6.1, 5 mm dithiothreitol, 0.1 mm TDP, 0.1 mm magnesium sulfate, 5% (v/v) glycerol, 10 µm phenylmethylsulfonyl fluoride and disrupted using glass beads (0.3–0.5 mm diameter) in a beat beater homogeniser (6 times for 30 s separated by 5 min cooling periods). The glass beads were washed three times with 50 mL buffer. After centrifugation (14 500 g for 30 min), 0.75% (w/v) streptomycin sulfate was added to the supernatant under continuous stirring at 4 °C for 30 min. The solution was centrifuged at 14 500 g for 30 min; the precipitate was discarded and 27% (w/v) ammonium sulfate was added to the supernatant. After centrifugation at 4 °C, 10% (w/v) ammonium sulfate was added to the supernatant. The solution was stirred for 30 min and centrifuged again. The pellet was resuspended in 20 mL of 100 mm Mes/NaOH pH 6.0, 150 mm ammonium sulfate. Insoluble protein was removed by centrifugation at 20 000 g for 15 min. The supernatant was loaded on a Sephacryl S200 H column (5.0 × 100 cm, flow rate 1 mL·min−1) equilibrated and eluted with the same buffer. Fractions containing KlPDC (PDC from K. lactis) activity were precipitated by 50% (w/v) ammonium sulfate. The pellet was resuspended in 10 mL of 20 mm Bistris pH 6.8. The solution was desalted on a Superdex G50 column (1.6 × 30 cm, flow rate 1 mL·min−1) and applied to an anion exchange column Resource Q (6 mL, flow rate 1 mL·min−1). The protein was eluted using an increasing ammonium sulfate gradient (0–0.2 m, 100 mL). Fractions containing KlPDC were stabilized by 1 m ammonium sulfate and loaded on a hydrophobic interaction column Resource Phe (1 mL, flow rate 1 mL·min−1.). The protein was eluted using a stepwise decreasing ammonium sulfate gradient. KlPDC was detected at ammonium sulfate concentrations below 400 mm. The fractions containing highly purified KlPDC were precipitated by 50% (w/v) ammonium sulfate and stored at −20 °C.
Enzyme assay and protein determination
Catalytic activity was measured in 0.1 m Mes/NaOH pH 6.0, 5 mm TDP, 5 mm magnesium sulfate at 340 nm and 30 °C (Uvikon 941, Kontron Instruments) using the established coupled optical test  with alcohol dehydrogenase from yeast (alcohol dehydrogenase, Sigma, 45 U·mL−1) and NADH at enzyme concentrations of 5–12.5 µg·mL−1. At a substrate concentration above 40 mm the catalytic activity was measured at 366 nm. In order to ensure that the lag phase in product formation was not due to insufficient activity of the auxiliary enzyme alcohol dehydrogenase, its concentration was varied between 4 and 45 U·mL−1; no effect on the duration of the lag phase was observed. The PDC activity in the reaction mixture was not larger than 0.4 U·mL−1 and the activity of alcohol dehydrogenase was determined for the reverse reaction that is about 20 times higher in the direction from the aldehyde to the alcohol. Consequently, the auxiliary enzyme should not limit the substrate activation of KlPDC.
Substrate activation measurements were performed on a stopped-flow spectrophotometer (SX.18MV, Applied Photophysics) using the same buffer mentioned above. One syringe contained the substrate pyruvate and the auxiliary enzyme alcohol dehydrogenase (Sigma, 1 mg·mL−1, respectively, 900 U·mL−1) and NADH, the other KlPDC (50–100 µg·mL−1), 10 mm TDP, and 10 mm magnesium sulfate. The solutions were mixed in a ratio of 1 : 1.
All dilution experiments were carried out in the same buffer system mentioned above and started at a substrate concentration of 2 mm. After 65 s the reaction mixture was diluted manually by adding buffer containing the cofactors (0.1 m Mes/NaOH, pH 6.0, 5 mm TDP, 5 mm magnesium sulfate, 45 U·mL−1 alcohol dehydrogenase and 0.3–1.0 mm NADH) in a ratio of 1 : 2, 1 : 3 and 1 : 5. The enzyme concentration was between 22.5 and 30 µg·mL−1 after dilution.
The protein concentration in the crude extract was determined according to Bradford  with bovine serum albumin as standard protein. In all other cases the protein concentration was calculated from the UV spectra at 280 nm using the molar extinction coefficient of 61 950 m−1·cm−1 for the KlPDC subunit, derived from the amino-acid sequence using the software package of expasy.
SDS/PAGE (10% acrylamide) was carried out according to the method of Laemmli . Gels were stained with Coomassie Brillant Blue G 250.
Determination of the molecular mass
This was performed on a Fractogel EMD BioSEC(S) column (1.6 × 60 cm, Merck kgaA, flow rate 1 mL·min−1) in 0.1 mm Mes/NaOH pH 6.0.
Homogeneous KlPDC was desalted on a HiTrap column (5 mL) using 10 mm ammonium acetate pH 6.4. Enzyme solution (0.7 mg·mL−1) was mixed with an equal volume of 90 mm 3,5-dimethoxy 4-hydroxycinnamic acid. The molecular mass was detected on a REFLEX (Bruker-Franzen Analytik) time of flight (TOF) mass spectrometer with matrix-assisted laser desorption ionization (MALDI) using a nitrogen laser at 337 nm and an acceleration voltage of 30 kV. Bovine serum albumin (Merck) served as a protein standard.
Theoretical background of substrate activation
For the studies on the mechanism of substrate activation Eqn (1) can be applied to analyse the corresponding progress curves. A0 is the initial absorbance at zero time, ΔSS, the steady-state velocity of absorbance change, Δ0, the corresponding initial velocity at zero time, and kobs, the observed first-order rate constant of the substrate activation.
The zero time slopes were found to be very small in the absence of the activator pyruvamide (Δ0/ΔSS = 0.018). Inclusion of Δ0, however, in Eqn (1) enables the application of this equation to progress curves recorded in the presence of pyruvamide. The dependence of the observed rate constant on the substrate concentration given in Eqn (2) could be derived from Scheme 1.
The dissociation constants are defined in Scheme 1. The observed rate constant of substrate activation kobs is composed of two functions. The first one, including Km, describes a property of the catalytic centre of the enzyme, the second, hyperbolic one, including Ka, a property of the regulatory substrate binding site. From Eqn (2), extrema of kobs can be expected at a substrate concentration of
In the case of Ka/Km<Kiso<Km/Ka, a maximum will occur and in the case of Km/Ka<Kiso<Ka/Km, a minimum will occur, otherwise [S]ext will be negative.
From Scheme 1 the following steady-state rate equation can be derived:
where: Vmax is the maximum velocity, A = Ka · Kiso · Km, the overall dissociation constant of the complex SEaS; Ka · Kiso is the overall dissociation constant of the complex SEa, and B = Km · (Kiso + 1). This function follows a sigmoid shape characterized by an apparent Hill coefficient between one and two. In the case of √A ≫ B, the Hill coefficient is almost two, in the case of √A ≪ B, the Hill coefficient is close to one. Therefore, the larger the ratio A/B, the more pronounced the sigmoidicity of the dependence of the steady-state rate on the substrate concentration appears. In terms of the activation process, this means that the weaker the binding of the activating substrate molecule to the regulatory centre and/or the more unfavoured is the activated state against the inactive one, the larger the deviation from the Michaelis–Menten behaviour.
From Eqn (4) the value S0.5, defining the substrate concentration at which the velocity is Vmax/2, can be obtained as
These equations are valid only for enzymes without significant initial activity (Δ0/ΔSS close to zero). This is not the case for PsPDC (Δ0/ΔSS = 0.25, from reference ). However, we have involved this enzyme species in our analysis for sake of comparison.
Results and discussion
Purification of PDC from K. lactis
In contrast to brewer's yeast, K. lactis is a Crabtree-negative yeast exhibiting repressed alcoholic fermentation under oxygen saturation. According to Breunig et al. , K. lactis expresses PDC at high glucose and low oxygen concentration. Considerably high amounts of KlPDC were obtained under these growth conditions. A summary of the purification procedure is illustrated in Table 1. Twenty-five micrograms of purified PDC with a catalytic activity of about 40 U·mg−1 were obtained from 60 g yeast cells from 4 L of cell culture. The yield and catalytic activity are comparable to those of PDC from other organisms [1–3,9,10,18–22]. Only one type of subunit was detectable in the SDS/PAGE (Fig. 1) in contrast to ByPDC [9,23] and PDC species from plant seeds that exhibit two types of subunits with slightly different sizes. The molecular mass of 61.5 ± 0.2 kDa for the subunit was determined by mass spectrometry (Fig. 1) and corresponds to the size derived from SDS/PAGE and the value calculated from the amino-acid sequence deduced from the nucleotide sequence of the KlPDC gene (61821 Da). Size-exclusion chromatography of the pure enzyme revealed a single peak with a molecular mass of about 200 kDa pointing to a tetrameric structure of the native enzyme as the typical native state of PDC. The N-terminus of the subunit is blocked.
|Step||Total activity (U)||Total protein (mg)||Specific activity (U·mg−1)||Yield of activity (%)|
|Crude extract (broken cells)||4500||5400||0.8||100|
|Streptomycin sulfate precipitation||4300||4600||0.9||96|
|Ammonium sulfate precipitation||2400||1900||1.3||53|
The circular dichroism and fluorescence spectra of the apoenzyme and holoenzyme of KlPDC are very similar to those of ByPDC and ScPDC (data not shown; [11,24,25]).
Dependence of the catalytic activity of KlPDC on the substrate concentration
Pyruvate decarboxylase from K. lactis shows a sigmoid dependence of the catalytic activity on the substrate concentration (Fig. 2) like other PDC species from plant seeds and yeasts [2,10,22,26–30]. The only known exception is PDC from Zymomonas mobilis that displays a hyperbolic dependence . However, the sigmoidicity of the plot of velocity vs. pyruvate concentration of KlPDC, expressed by the ratio of the parameters A and B in Eqn (4)(Table 2), is significantly more pronounced at substrate concentrations below 1.5 mm than in the case of other PDCs. At high substrate concentrations (above 100 mm), a weak substrate inhibition was detected and a Ki value of 1.2 m was estimated. An S0.5 value of 1.85 mm was calculated according to Eqn (5) at pH 6.0. The S0.5 value increased continuously with increasing pH (data not shown) as for other PDC species [2,22].
|A (mm)2||2.95 ± 0.21||0.68||0.24||–|
|B (mm)||0.25 ± 0.10||0.54||0.73||–|
|A/B (mm)||11.80 ± 5.56||1.26||0.32||–|
|S 0.5 (mm)||1.85 ± 0.15||1.10||1.00|
|K a (mm)||207.00 ± 12.40||9.30||3.31||–|
|k iso (s−1)||3.03 ± 0.09||0.65||1.15||–|
|k −iso(s−1)||0.18 ± 0.11||0.10||0.12||–|
|K iso||0.06 ± 0.04||0.16||0.11||–|
|K a·Kiso (mm)||12.42 ± 8.67||1.49||0.36|
|K m (mm)||0.24 ± 0.11||0.47||0.66||0.4|
|k cat[monomer] (s−1)||40 ± 4.0||60||60||180|
|k cat/Km (s−1·mm−1)||167 ± 93||129||90||450|
|K i (m)||1.2 ± 0.06||0.2–0.3||–||0.58|
Characterization of the substrate activation behaviour of KlPDC
Substrate activation was studied by the stopped-flow technique. A distinct lag phase in the product formation dependent on the substrate concentration was observed under all conditions used (Fig. 3). Progress curves were analysed using the combined zero- and first-order function shown in Eqn (1). The initial catalytic activity (at zero time) was determined by calculating the ratio between the initial and steady-state velocity of absorbance change (Δ0/ΔSS). As illustrated in Fig. 3, KlPDC is potentially inactive in the absence of the substrate, as found for ByPDC . However, the main difference of KIPDC to all other substrate activated PDC species analyzed so far is manifested in the plot of kobs vs. the substrate concentration (Fig. 4A). Whereas in all other cases this dependence was found to be hyperbolic [5,16], a complex function corresponding to Eqn (2) with a minimum at 1.5 mm pyruvate was obtained for KlPDC. According to the values for ByPDC given in Table 2, the calculated curve for the dependence of kobs on the substrate concentration displays a weak minimum at 0.4 mM pyruvate too. However, because of the high errors of the measurements at these low substrate concentrations, the experimental verification of this minimum is difficult and it was not detected in previous studies [5,16]. On the basis of this phenomenon the kinetic studies of KlPDC allow a refined insight in the substrate activation behaviour of all pyruvate decarboxylases. The Ka and kiso values were derived from the plot of 1/kobs vs. 1/[S] at substrate concentration above 40 mm (Fig. 4B). The value kiso could not be determined directly at saturating substrate concentrations because of the high Ka value of 207 mm and the greater fitting errors at high pyruvate concentrations. The regulatory site of KlPDC shows a very low affinity for the primary binding of the substrate (Table 2) compared to other PDCs. The Ka value of KlPDC for pyruvate is twofold higher than that of ByPDC and sixfold higher than that of PsPDC. The low affinity of the substrate to the regulatory binding site is compensated by a fast isomerization (kiso) and a high affinity of the substrate to the catalytic centre. Km and k−iso were calculated by means of Eqns (6) and (7). The Km value is about three orders of magnitude smaller than the Ka value. Only this special ratio of all dissociation constants (Km/Ka<Kiso<Ka/Km) allowed the detection of a minimum in the plot of kobs vs. pyruvate concentration for the first time. The low Km value of KlPDC demonstrates a higher substrate affinity of the catalytic centre compared to those of ByPDC and PsPDC. The specificity constant kcat/Km is the highest found for activated PDC species so far and is about 40% of that of ZmPDC (Table 2), although the catalytic constant kcat of KlPDC is fourfold lower. All constants are summarized in Table 2. It was possible to generate a plot, which fits the data of the observed activation rate constants (kobs) using the calculated constants Ka, Km, kiso, and k–iso according to Eqn (2) (solid line in Fig. 4A). Moreover, a calculated plot according to Eqn (4) using the same constants from the substrate activation (Fig. 2, dashed line) is in coincidence with the fit of the experimental steady-state data of the KlPDC catalysis (Fig. 2, solid line).
The steady-state concentrations of each enzyme state are dependent on the substrate concentration and all enzyme states are in equilibrium as illustrated in Scheme 1. If this equilibrium is perturbed by rapidly lowering the substrate concentration a relaxation in the progress curve will be observed following a first-order reaction  as demonstrated in Fig. 5. The dilution process was carried out after a reaction time of 65 s when the enzyme was activated to 95% at 2 mm pyruvate (corresponding kobs = 0.05 s−1). The observed rate constants of the dilution process correspond to the observed rate constants of the substrate activation and show the same dependence on the substrate concentration (inset of Fig. 4A) as expected on the basis of the principle of microscopic reversibility.
Pyruvamide, a substrate surrogate of pyruvate, activates ByPDC without being converted . It was impossible to obtain a completely activated KlPDC, even at very high pyruvamide concentrations (above 400 mm, Fig. 6). This was to be expected because of the high Ka value of the substrate pyruvate. The estimated Ka value of pyruvamide binding is 90 mm (inset of Fig. 6A). In contrast to the results obtained with ByPDC, pyruvamide was found to be a mixed type inhibitor (competitive and noncompetitive) for KlPDC (data not shown). The S0.5 value increased and Vmax decreased with increasing pyruvamide concentration (Fig. 6B).
The quantitatively remarkable coincidence between the dependence of velocity vs. pyruvate concentration and the dependence of the activation rate constant kobs vs. pyruvate concentration in terms of the proposed model strongly points to the validity of the mechanism illustrated in Scheme 1. Moreover, it qualifies the model for the analysis of other enzyme variants with impaired substrate activation behaviour.
We thank Ines Eberhardt for providing the K. lactis strain JA6 (Laboratorium of Molecular Cell Biology, Biological Department, Katholic University Leuven, Belgium, Present address: Department of Molecular Biology, Gent University, Belgium). Klaus-Peter Rücknagel (Max-Planck Research Unit for Enzymology of protein folding) for N-terminal amino-acid sequencing and Angelika Schierhorn for the mass spectrometry measurements.
- 191980) Wheat germ pyruvate decarboxylase. Improved purification and properties in comparison with the yeast enzyme. Hoppe-S. Z. Physiol. Chem. 361, 1265.& (