The Kinetic Mechanism of Formate Dehydrogenase from Bakery Yeast

Doklady Biochemistry and Biophysics, Vol. 382, 2002, pp. 26–30. Translated from Doklady Akademii Nauk, Vol. 382, No. 3, 2002, pp. 401–405. Original Russian Text Copyright © 2002 by Serov, Popova, Tishkov.

BIOCHEMISTRY, BIOPHYSICS, AND MOLECULAR BIOLOGY

The Kinetic Mechanism of Formate Dehydrogenase from Bakery Yeast A. E. Serov, A. S. Popova, and V. I. Tishkov Presented by Academician A.A. Bogdanov September 24, 2001 Received October 25, 2001

NAD+-dependent formate dehydrogenase (FDG, EC 1.2.1.2) catalyzes oxidation of formate to carbon dioxide coupled with reduction of NAD+ to NADH: çëéé– + NAD+ → ëé2 + NADH. A characteristic feature of this reaction is the absence of proton-release stages that exist in all reactions catalyzed by other dehydrogenases. In addition, formate is structurally the simplest of dehydrogenase substrates. This allows us to regard this reaction as a model that can be used in the studies of the mechanism of hydride ion transfer from the substrate to the C-4 atom of the nicotine amide ring in the dehydrogenase active site. NAD+-dependent FDGs comprised of two identical subunits and containing no metal ions or prosthetic groups in the active site are widely spread in nature. Currently, data banks contain complete sequences of more than twenty FDGs from bacteria, yeast, Phycomycetes, and higher plants. Some of these enzymes have been obtained in a homogenous state. However, the mechanism of FDG action has been studied in detail only in the case of the bacterial enzyme from Pseudomonas sp. 101 [1–3]. The comparison of the amino acid sequences of FDGs from different sources is shown in Fig. 1. Because of the high homology of these enzymes within the specified groups (more than 85–90%) and for convenience, we do not present here all the sequences. We show only two sequences from each group: bacteria Pseudomonas sp. 101 (PseFDG) [4] and Moraxela sp. C-2 (MorFDG, EMBL accession O08375), higher plants potato (MisFDG, EMBL accession Z21493) and barley (BarFDG, EMBL accession D88272), imperfect yeast Candida methylica (CmeFDG, EMBL accession X81129) and Pichia angusta (formerly called Hansenula polymorpha, HanFDG, EMBL accession A94993), Phycomycetes Neurospora crassa (NeuFDG, EMBL accession L13964) and Aspergilus nidulans (AspFDG, EMBL accession Z11612), and bakery yeast Saccharomyces cerevisiae (SceFDG, EMBL accession Z75296). The comparison

Faculty of Chemistry, Moscow State University, Vorob’evy gory, Moscow, 119899 Russia

of the amino acid sequences of FDGs from different sources reveals a high extent of homology between the enzymes of this type. Even evolutionally distant enzymes such as FDGs from bacteria and higher plants share more than 50% of homology. The extent of homology in the active site region is even higher (90– 95%). The feature that makes bacterial FDGs significantly different from the other FDGs is a longer N-terminal sequence. With respect to the other part of the sequence, all FDGs can be divided into two groups. Group I comprises FDGs from bacteria and higher plants, and group II, FDGs from bakery yeast and Phycomycetes. Figure 1 shows that the enzymes of group I differ from those of group II in the alignment gaps in the catalytic domain region (two residues in the helix α4 and five residues in the β-turn βG) and the coenzymebinding domain (one residue between the structural elements αB and βB). FDG from bakery yeast S. cerevisiae is an intermediate variant, although this enzyme is closer to FDGs of group II. In the case of the α4 helix, alignment revealed a gap similar to that characteristic of group I FDG. In the other cases, gaps in aligned sequences are absent, and the bakery yeast FDG is homologous to group II FDGs. The bakery yeast FDG also contains an additional loop comprised of 13 residues between the αC and the portion of the βB β-sheet in the coenzyme-binding region. Most importantly, the bakery yeast FDG is characterized by several significant differences in the region of residues comprising the active site of the enzyme. Earlier, it was shown that, in the FDG from the bacterium Pseudomonas sp. 101, Gln313 and His332 play a crucial role in FDG-mediated catalysis [2]. As is seen from Fig. 1, Gln313 is located between two Pro residues, and one more conserved Pro residue is located before His332. The results of X-ray analysis of the bacterial FDG holoenzyme indicate that these Pro residues ensure rigid spatial fixation of Gln313 and His332 [5]. Figure 1 shows that, in the case of the bakery yeast FDG, Pro312 and Pro331 are substituted with lysine and valine residues, respectively (the numeration is given according to the bacterial FDG sequence). In addition, the S. cerevisiae FDG sequence contains other amino acid residues in the positions that are absolutely conserved in the FDGs of both

1607-6729/02/0102-0026$27.00 © 2002 MAIK “Nauka /Interperiodica”

THE KINETIC MECHANISM OF FORMATE DEHYDROGENASE

27

Fig. 1. Comparison of amino acid sequences of formate dehydrogenases from different sources: the bacteria Pseudomonas sp. 101 (PseFDG) and Moraxela sp. C-2 (MorFDG), the higher plants potato (MisFDG) and barley (BarFDG), the imperfect yeast Candida methylica (CmeFDG) and Pichia angusta (formerly called Hansenula polymorpha, HanFDG), the Phycomycetes Neurospora crassa (NeuFDG) and Aspergilus nidulans (AspFDG), and the bakery yeast Saccharomyces cerevisiae (SceFDG). The numeration of amino acid residues and positions of the secondary structure elements are given for FDG from Pseudomonas sp. 101 [5]. X denotes the residues contained in the enzyme active site. The portion of the sequence at the N-termini of potato FDG shown in italic is a signal peptide.

groups (Fig. 1). Taking into account the specific structure of the active site and the entire primary structure of the bakery yeast FDG compared to FDGs from other sources, we began a systematic study of the mechanism of action of this enzyme. The gene encoding the S. cerevisiae FDG was cloned and expressed in E. coli in a soluble active form (as much as 45% of the total proDOKLADY BIOCHEMISTRY AND BIOPHYSICS

Vol. 382

tein). In this work, we studied the kinetic mechanism of the enzyme and the effect of ionic strength on the enzyme affinity for NAD+. Earlier, the kinetic mechanism of the FDG-catalyzed reaction was determined for the enzymes from the bacterium Pseudomonas sp. 101 [6–8]; imperfect yeast C. boidinii [9, 10], C. methylica [11], and 2002

SEROV et al. (a)

[Formate]/Vm, ef 0.16

1 2

0.14

3 4 5

0.12

3e – 3 2e – 3

(b)

8e – 6 7e – 6

2

+

[NAD+]/v 4e – 3

1

0.10

6e – 6

0.08

5e – 6

0.06

1e – 3

4e – 6

0.04 0 –2e – 4

2e – 4

4e – 4

–1e – 3

6e – 4 [NAD+], å

0.02

[Formate]KMNAD ef/Vm,ef

28

3e – 6 0.01 0.02 0.03 0.04 0.05 [Formate], å

0

Fig. 2. (a) A Heinz–Wulf plot of the initial reaction rate on NAD+ concentration vs. fixed concentrations of formate: (1) 4.5 mM, (2) 9.0 mM, (3) 18.0 mM, (4) 22.5 mM, and (5) 45 mM. (b) Secondary plots of the tangents of the slopes and the intercepts from the primary plots vs. formate concentration (0.1 M phosphate buffer, pH 7.0, 37°C).

P. angusta (H. polymorpha) [12]; and bean Phaseolus aureus [13]. It has been shown that the bacterial enzyme is characterized by random substrate binding and product release [6, 7] with a rapid attaining the equilibrium between the free enzyme and double enzyme–substrate complexes (FDG–NAD+) and FDG– formate) [8]. In the case of the yeast FDG, the Bi–Bi ordered mechanism, with NAD+ binding first, is realized. The bean FDG catalyzes reactions according to a special variety of the ordered kinetic mechanism (Theorell–Chance mechanism). This mechanism is characterized by a very high reaction rate during the triple enzyme–substrate complex formation, catalytic stage, and ëé2 release from the triple complex (FDG– NADH–CO2). To determine the kinetic mechanism of the bakery yeast FDG, we first studied the dependence of the initial reaction rate on the NAD+ concentration in the presence of several fixed formate concentrations and vise versa. The obtained dependencies expressed as a Heinz–Wulf plot ([S]/v–[S]) are a series of straight lines crossing in the same point both at varying NAD+ concentrations (Fig. 2a) and at varying formate concentrations (not shown). Secondary plots of the tangents of the straight line slopes and the ordinate vs. concentration of the second (fixed) substrate were also linear for both formate (Fig. 2b) and NAD+ (not shown). The data obtained indicate that a triple complex (FDG–NAD+– formate) is formed in the course of enzymatic reaction and that the dependence of the reaction rate on the concentration of substrates is described by equation [13] V m[N][F] -, v = ------------------------------------------------------------------------------------N E N F Kd KM + KM[F] + KM[N] + [N][F]

(1)

where v and Vm are the effective and maximal reaction rates, respectively; [N] and [F] are the concentrations of N

F

NAD+ and formate, respectively; K M and K M are Michaelis constants for NAD+ and formate, respecN

tively; and K d is equilibrium dissociation constant of the double complex FDG–NAD+. The product of conN

F

stants K d K M (a term of Eq. (1)) characterizes the efficiency of the triple enzyme–substrate complex formation in the course of reaction. The table summarizes the values of constants determined experimentally by nonlinear regression (compare with the corresponding constants for bacterial, imperfect yeast, and bean FDGs). The dependence of the reaction rate on the substrate concentration obtained for FDG may correspond to several types of kinetic mechanisms. To discriminate between these mechanisms, we studied FDG inhibition by the reaction products and substrate analogues. We found that NAD+ is a competitive inhibitor of FDG with respect to NAD+ at both saturating and nonsaturating formate concentrations. In the case of varying formate concentrations, a noncompetitive inhibition at nonsaturating NAD+ concentrations and a complete absence of inhibition at saturating NAD+ concentrations were observed. Bicarbonate was a competitive inhibitor with respect to formate at any NAD+ concentration and incompetitive inhibitor with respect to NAD+ at nonsaturating formate concentrations; it did not inhibit FDG at high concentrations of the second substrate. The results similar to those obtained for bicarbonate were obtained for azide at different concentrations. Thus, our data indicate that the reactions catalyzed by the bakery

DOKLADY BIOCHEMISTRY AND BIOPHYSICS

Vol. 382

2002

THE KINETIC MECHANISM OF FORMATE DEHYDROGENASE

29

Kinetic parameters of the reaction catalyzed by NAD+-dependent formate dehydrogenases from different sources Source Parameter

S. cerevisiae (this study)

N

P. angusta Pseudomonas C. boidinii [9,10] C. methylica [11] (H. polymorpha) Ph. aureus [13] sp.101 [6–8] [12]

36 ± 5

64 ± 21

100

100 ± 30

370 ± 80

7.2 ± 1.0

5.5 ± 0.8

15 ± 6

16

13 ± 4

10 ± 1

1.6 ± 0.3

K D , mM

200 ± 40

220 ± 80

250

750 ± 300

590 ± 90

42 ± 8.0

Inhibition with NADH

(NAD+),

K M , µM F

K M , mM N

NADH

Ki

, mM

Inhibition with azide aside

Ki

, nM

Inhibition with bicarbonate Kinetic mechanism**

C* NC (formate)

(NAD+),

C C (formate)

(NAD+),

C NC (formate)

(NAD+),

C NC (formate)

C (NAD+), NC (formate)

12 ± 5

2.5 ± 0.3

C NC (formate)

19 ± 8

21 ± 7

IC (NAD+), C (formate)


200 ± 50

150 ± 60

150

80 ± 20

170 ± 40

ND

(NAD+),

(NAD+),

(NAD+),

(NAD+),

(NAD+),

NC (NAD+), NC (formate) Bi–Bi ordered (Theorell–Chance type)

NC NC (formate) Bi–Bi ordered

NC C (formate) Bi–Bi random steady-state

12 ± 5

(NAD+),

30 IC (NAD+), C (formate)




IC (NAD+), NC (formate)


ND***

* C, NC, and IC denote competitive, noncompetitive, and incompetitive inhibition types, respectively. ** The kinetic mechanism types are denoted according to Kleland [14]. *** ND, no data.

yeast FDG proceeds according to the Bi–Bi ordered mechanism, where NAD+ is the first substrate bound: N

F

CO2

NH .

E EN

EN

ENHCO2 ENH E

Similar results were obtained when determining the kinetic mechanisms for other yeast FDGs. However, with respect to the kinetic parameters, the bakery yeast FDG is much closer to the bacterial enzyme (table). At the second stage of our work, we studied the effect of ionic strength on NAD+ binding in the active site of FDG. The results of X-ray analysis of bacterial FDG indicate that electrostatic interactions between the negatively charged pyrophosphate moiety of NAD+ and positively charged Arg201 play a key role in the coenzyme binding in the active site of the enzyme [5]. The S. cerevisiae FDG sequence also contains an arginine residue at this position (Fig. 1). Taking into account the high extent of FDG homology, it can be assumed that, in this enzyme, the arginine residue will also be involved in NAD+ binding. It is known that the efficiency of electrostatic interactions strongly depends on ionic strength. Therefore, an increase in salt concentration in the solution should deteriorate NAD+ binding in the active site of FDG. Strange as it may seem, despite DOKLADY BIOCHEMISTRY AND BIOPHYSICS

Vol. 382

the more than 27-year history of FDG studies, such data are absent from the literature. Figure 3 shows the results of determination of Kå for NAD+ in the presence of different concentrations of formate. One can see that, at formate concentrations higher than 0.6 M, the Kå value increases. The dependence shown in Fig. 3 fits the equation with one exponent. Thus, we showed that, like FDGs from methylotrophic yeast, the bakery yeast FDG functions according to an ordered kinetic mechanism. However, with respect to its properties, it is closer to bacterial FDGs. In our laboratory, we obtained a mutant bakery yeast FDG that carried mutations returning the enzyme active site to the classical FDG configuration (K284P and V305P). The data on the kinetic mechanism of FDG will be used for correct interpretation of the changes in the kinetic properties of these mutant enzymes compared to wild-type FDG. In addition, we were the first to confirm the key role of electrostatic interactions in coenzyme binding in the active site of FDG. These data are also of great practical importance for mathematical simulation of the efficiency of enzymatic synthesis of chiral compounds with the use of dehydrogenases, where FDG is used for NADH regeneration [15]. Experimental. In this study, we used NAD+ and NADH obtained from Sigma (United States) of at least 2002

30

SEROV et al.

KMNAD+, µM 1200

REFERENCES 1. Tishkov, V.I. and Egorov, A.M., Biokhimiya, 1985, vol. 50, no. 7, pp. 1059–1066.

1000

2. Tishkov, V.I., Matorin, A.D., Rojkova, A.M., et al., FEBS Lett., 1996, vol. 390, no. 1, pp. 104–108.

800

3. Popov, V.O. and Lamzin, V.S., Biochem. J., 1994, vol. 301, no. 3, p. 625.

600

4. Tishkov, V.I., Galkin, A.G., and Egorov, A.M., Dokl. Akad. Nauk, 1991, vol. 317, no. 3, pp. 345–348.

400 200

5. Lamzin, V.S., Dauter, Z., Popov, V.O., et al., J. Mol. Biol., 1994, vol. 236, no. 3, pp. 759–785.

0

6. Popov, V.O., Rodionov, Yu.V., Egorov, A.M., and Berezin, I.V., Bioorg. Khim., 1978, vol. 4, no. 1, pp. 117–128.

0.5

1.0

1.5

2.0 2.5 [Formate], å

Fig. 3. Dependence of the effective Michaelis constant for NAD+ on formate concentration (0.1 M phosphate buffer, pH 7.0, 37°C).

99.5% purity and homogenous preparations of the bakery yeast recombinant FDG expressed in E. coli. The changes in the reaction rate in the presence of different concentrations of substrates and inhibitors were monitored spectrophotometrically at 340 nm by NADH production in 0.1 M phosphate buffer, pH 7.0, at 30°ë. The obtained kinetic dependencies were processed as described in [13].

7. Tishkov, V.I., Egorov, A.M., and Popov, V.O., Biokhimiya, 1983, vol. 48, no. 7, pp. 1172–1180. 8. Tishkov, V.I., Galkin, A.G., and Egorov, A.M., Biochimie, 1989, vol. 71, no. 4, pp. 551–557. 9. Kato, N., Sahm, H., and Wagner, F., Biochim. Biophys. Acta, 1979, vol. 566, no. 1, pp. 12–20. 10. Blanchard, J.S. and Cleland, W.W., Biochemistry, 1980, vol. 19, no. 19, pp. 4506–4513. 11. Zaks, A.M., Avilova, T.V., Egorova, O.A., et al., Biokhimiya, 1982, vol. 47, no. 4, pp. 546–551. 12. Mezentsev, A.V., Ustinnikova, T.B., Tikhonova, T.V., and Popov, V.O., Prikl. Biokhim. Mikrobiol., 1996, vol. 32, no. 6, pp. 589–595. 13. Peacock, D. and Boulter, D., Biochem. J., 1970, vol. 120, pp. 763–769.

ACKNOWLEDGMENTS

14. Cornish-Bowden, A., Principles of Enzyme Kinetics, London: Butterworths, 1975, Ch. 5.

This study was supported by the Russian Foundation for Basic Research (project no. 99-04-49156).

15. Hummel, H. and Kula, M.-R., Eur. J. Biochem., 1989, vol. 184, no. 1, pp. 1–13.

DOKLADY BIOCHEMISTRY AND BIOPHYSICS

Vol. 382

2002