Greenhouse Gas Control Technologies, Volume II J. Gale and Y. Kaya (Eds.) 2003 Elsevier Science Ltd.

. All rights reserved.

1451

ENZYMATIC SYNTHESIS OF PYRUVIC ACID AND L-LACTIC ACID FROM CARBON DIOXIDE
Masaya Miyazaki, Hiroyuki Nakamura, and Hideaki Maeda Micro-space Chemistry Laboratory, AIST Kyushu, National Institute of Advanced Industrial Science and Technology (AIST), 807-1 Shuku, Tosu, Saga 841-0052, Japan

ABSTRACT A new enzymatic synthesis of pyruvic acid and L-lactic acid from acetaldehyde and carbon dioxide has been developed. The reaction of pyruvic acid synthesis utilizes reverse reaction of pyruvate decarboxylase in sodium bicarbonate buffer. The reaction proceeded at high pH, and gave better yield in higher concentration of buffer. The maximum yield was obtained in 500mM sodium bicarbonate buffer at pHI 1. We also developed a one-pot, two-step enzymatic synthesis of L-lactic acid from acetaldehyde and carbon dioxide. The reaction employing reversal of pyruvate decarboxylase and hydrogenation of pyruvate by Llactic dehydrogenase. The maximum yield was obtained at pH 9.5 in 500mM sodium bicarbonate buffer.

INTRODUCTION

Because of recent global warming, the emission of carbon dioxide became serious problem. While carbon dioxide gas is not in itself a particularly strong greenhouse gas, the contribution to the atmosphere since the beginning of the industrial revolution has been very considerable. The increase in global carbon dioxide levels over the last hundred years or more have been reported. Furthermore, there is some indication of a rising global temperature. A number of meetings, particularly in Kyoto in December 1997, have recommended reducing the total emission of carbon dioxide. Several efforts to immobilize or utilize carbon dioxide have been achieved. Still, no conclusive method is available for this purpose. Recent interest in the problems of environmental pollution has forced the development of a green chemistry process for the chemical industry [1]. Several environmentally-friendly processes have been developed. Biocatalytic processes such as fermentation and enzyme reactions have attracted attention as environmentally safe chemical process [2,3]. Biodegradable polymers have also proved interesting for environmental safety [4,5]. Among the biodegradable plastics, poly(L-lactic acid) has been widely studied. Several methods have been developed to produce lactic acid from many sources [6]. However, most of them produce the racemic form and require relatively longer time and multiple steps to obtain pure L-lactic acid. These disadvantages are problematic, not only for large-scale production but also for environmental safety. Therefore, a simple method, that can produce optically pure L-lactic acid, is desired. We are interested in the development of novel enzymatic reactions and reactor systems characterised by environmentally safe chemical processes. Pyruvate decarboxylase (EC 4.1.1.1) is known as a catalyst for

1452 the decarboxylation reaction of pyruvic acid, to produce acetaldehyde, and has been utilized for C-C bond formation such as chiral Gt-hydroxy ketones, which are versatile building blocks for organic and pharmaceutical chemistry [7, 8]. The reverse reaction of this enzyme is also of interest as a catalytic procedure for carboxylation. Several studies have been performed which imitate these enzyme reactions using CO2 as the reactant. A previous study using a-lactylthiamin showed that production of pyruvate was achieved at higher pH (>10) [9]. However, these reactions require organic solvents and severe conditions. In the present study, we demonstrated the usefulness of the reverse reaction of pyruvate decarboxylase in the production of pyruvic acid from acetaldehyde and carbon dioxide. We also developed a one-reactor, twostep enzymatic synthetic procedure for L-lactic acid from acetaldehyde, by combining this reverse reaction of pyruvate decarboxylase and reduction of pyruvic acid by L-lactic dehydrogenase to produce L-lactic acid.

MATERIALS AND METHOD

General
Pyruvate decarboxylase (from brewer's yeast) and L-lactic dehydrogenase (from rabbit heart) were obtained from SIGMA (St. Louis, MO, U. S.A.). Thiamine pyrophosphate and NADH were purchased from Wako Chemical Ind. (Osaka, Japan). Sodium bicarbonate buffers were prepared immediately before experiment. All experiments were performed in thermostated bioshaker (Titech Co., Tokyo Japan) with vortex shaking. The HPLC analysis was performed using Waters Alliance 2596 system equipped with Wakosil C22 analytical column (d~4.0 x hl 50 mm, Wako Chemical Ind.). IH-NMR spectra was recorded by Bruker DSX300 spectrophotometer.

Synthesis of Pyruvic Acid

A typical run was performed as follows. To a solution of acetaldehyde (100 ~tM) in sodium bicarbonate buffer (1 ml) in a 1.5 ml microcentrifuge tube, pyruvate decarboxylase (1 unit) and thiamin pyrophosphate (the final concentration was 10 ~tM) were added at 4 C. The reaction mixture was warmed to 25 C quickly, and then shaken on a vortex mixer at room temperature. After 1 h, the reaction mixture was chilled on ice, and then subjected to RP-HPLC analysis immediately. The amount of pyruvic acid was calculated from the peak area of RP-HPLC analysis calibrated with commercially available pyruvic acid as standards. The yield was estimated based on acetaldehyde.

Synthesis of L-Lactic Acid

The reaction was initiated by adding pyruvate decarboxylase (1 unit) and L-lactic dehydrogenase (1 unit) to a solution of acetaldehyde (0.1 laM), thiamine pyrophosphate (0.1 ktM), NADH (0.2 ~tM) in various concentrations and pHs of NaHCO3/Na2CO3 buffer at room temperature. The reaction was performed 1 h, and then the reaction was terminated by adding 1M HCI and analyzed by RP-HPLC. The amount of each compound was estimated by peak area calibrated by commercially available standards. The absolute configuration of lactic acid was confirmed by the optical rotation, and the yield was calculated based on acetaldehyde. For confirmation, the yielded lactic acid was subjected to the analysis. The IH-NMR spectra and optical rotation value were identical to the commercially available L-lactic acid.

RESULTS

Synthesis of Pyruvic Acid

First, we tried the reverse reaction of pyruvate decarboxylase (Scheme 1). Previous organic synthesis showed that hydrolysis of lactylthiamin requires a higher pH [9]. Thus, we chose a sodium bicarbonate buffer system, because not only is this buffer suitable at higher pHs, but it also can be used as the source of carbon dioxide.

1453
Pyruvate decarboxylase > Thiamine

O Scheme 1: Synthetic process of pyruvic acid from carbon dioxide

H3C-,,~H O

CO2

O HaC~o

We firstly evaluated the effect of pH on the reaction. The reaction was performed using acetaldehyde (100 ktM), thiamin (0.1 ~tM), and pyruvate decarboxylase (1 unit) in 0.1 M NaHCO3/Na2CO3 buffer at various pHs (pH 8.5-11.5). The result is shown in Figure 1A. Higher pHs gave a better yield of pyruvic acid. The maximum yield was obtained at pH 11 (61%). The present result agreed well with a previous observation obtained from the hydrolysis of ct-lactylthiamin. In that case, the best yield from hydrolysis was obtained at pH 12. In our case, the best yield was obtained at pH 11, but the yields decreased at much higher pH. Although the hydrolysis proceeds at a maximum rate at pH 12, the enzyme might not be stable over pH 11. Therefore, the maximum yield was obtained at pH 11. Thus, we decided to perform further experiments at pHil.
100

BO

..........

: ..........

~. . . . . . . . . . .

~ ..........

,~

.o

40

o
8

..........i......................................................i...........i.........
i ; i i
B.5 9 9.5 10 pH 10.5 11 11.5 12

Figure l" Effect of pH on the reverse reaction of reverse reaction of pyruvate decarboxylase. The
experiment was performed as described in Material and Method section.

...

60

~"

40

,-- '

.....................................

~ ........................................

20

.... "...................................................................................

,
1 o

.....

| 1 oo

. . . . . . . .
1 ooo

Concentration/raM

Figure 2: Effect of ionic strength of bicarbonate buffer on the reverse reaction of reverse reaction of pyruvate decarboxylase. The experiment was performed as described in Materials and Method section.

1454 Next, we examined the effects of concentration of bicarbonate buffer on the reaction (Figure 1B). Higher ionic strength of the bicarbonate buffer strongly influenced the yield, as expected. The maximum yield of the reaction was 81% at 500 mM NaHCO3/Na2CO3 buffer. This yield was sufficient to use as an organic process, and much higher than that obtained by the reaction in DMF under 20 atm of CO2 [9]. Not only does the latter reaction require multiple steps, but the use of the organic solvent DMF is problematic for environmental safety reasons. The enzymatic reaction does not require any organic solvent and gave a better yield. It has been reported that the thiamin itself could catalyze a reaction analogous to that of the enzyme, but preparation of the intermediate c~-lactylthiamin from acetaldehyde was unsuccessful. Thus, it is difficult to reverse the reaction without an enzyme, and pyruvate decarboxylase is the best catalyst for carboxylation of acetaldehyde.

Synthesis of L-lactic acid Pyruvic acid can be easily hydrogenated asymmetrically by L-lactic dehydrogenase in the biological system. Therefore, it is possible to produce L-lactic acid from acetaldehyde and carbon dioxide by combining the two enzyme reactions, reverse reaction of pyruvate decarboxylase and L-lactic dehydrogenase (Scheme 2). In the present study, we designed and evaluated a one-pot, two-step enzymatic synthetic procedure for Llactic acid from acetaldehyde as an initial experiment.

H3C',,~H
CO2 0

Pyruvate decarboxylase "H3C O

O OH

L-Lactic dehydrogenase ~

O H3C OH OH

Scheme 2: Synthetic process of L-lactic acid from carbon dioxide First, the effect of pH (7 to 10.5) on the reaction was examined, because the reverse reaction of pyruvate decarboxylase proceeds at higher pH, the yields should become better than that of neutral pH. The results were summarized in Figure 3. In the present study, the yield of lactic acid was maximized at pH 9.5. Much higher pH gave lower yield of L-lactic "acid and gave relatively better recovery of pyruvic acid. This result can be explained as follows. The first step proceeds at higher pH, whereas the second step, hydrogenation by L-lactic dehydrogenase, gave lower yield at much higher pH, yielded decrease of production of L-lactic acid and therefore the recovery ofpyruvic acid was increased (Figure 3). 100 . . . . . . . . . . . . . . . . . .

80

.........

: ...................

-. . . . . . . . . . . . . . . . . . . .

; ...................

~ ..................

~.'.

60

................................................

"-"
.~A"

'

: ...................................... i
,~lk - -,i[" "

.,A--

40 .............................................--:.~.....................................
!

20

.........

i .......

.......

' .

O.

10

pH

Figure 3: Effect of pH on the reaction yields of the two-step reaction. Each plot show the yields of pyruvic acid (O), L-lactic acid (11), and sum up of both yields (A). The experiments were performed as described in Material and Method section.

1455 We also examined the effect of concentration of bicarbonate buffer at pH 9.5, because the reverse reaction of pyruvate decarboxylase prefers higher concentration of bicarbonate buffer, s shown in Figure 4, higher concentration of bicarbonate gave better yield, as expected. The maximum yield was obtained at 500 mM bicarbonate buffer (51%). The reverse reaction of pyruvate decarboxylase yielded about 43% of pyruvic acid production in 500 mM bicarbonate buffer at pH 9.5. However, the combined yield of lactic acid and pyruvic acid exceeds 65%, meaning the yield of the reverse reaction of pyruvate decarboxylase was improved than that by alone. This might result from change of equilibrium condition of the reaction, namely pyruvic acid consumption by the second step might promote better yield of the first step. Further studies are required to solve the details of this mechanism.
B{'] . . . . . . . . . .

70

.........................................................................

-;-,t ..........

60

. . . . . :. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

".......... ,,~: ......................... z /

~o
"u '~ 40

..... i ........................................

~ ..................................

:r30

~
i .A'-'--

._,

i
i

i'"

111

iiiiiiiiii
,.t
10

I'1

. . . . . . . .

|
1 O0

. . . . . . . .
1000

Concentration/raM

Figure 4" Effect of ionic strength on the reaction yields of the two-step reaction. Each plots show the yields ofpyruvic acid (0), L-lactic acid (m), and sum up of both yields (A). The experiments were performed as described in Material and Method section.

DISCUSSION Although the effect is much weaker than methane, carbon dioxide is considered to be a greenhouse gas, and therefore its immobilization is desired. The methods currently reported are mainly catalytic or electrochemical reactions, which require much energy [10-12]. One alternative method has been reported, which utilized carbonic anhydrase for the immobilization of carbon dioxide [ 13]. This biomimetic approach needs almost no energy for the reaction. However, the carbonic anhydrase just improves the solubility of carbon dioxide in aqueous media, and further treatment of dissolved gas is required. The biodegradable polymers have also been interested for environmental safety. Among the biodegradable plastics, the poly(L-lactic acid) has been widely studied. Several methods have been developed to produce the lactic acid from many sources. However, most of them produce racemic form and require relatively longer time and multiple steps to obtain pure L-lactic acid. These disadvantages are problematic not only for large-scale production but also for the environmental safety. Therefore, a simple method, which can produce optically pure L-lactic acid, is desired. Recent interest of poly(L-lactic acid) demands the effective method to produce the material, L-lactic acid. The fermentation method has been focused as effective solution, because it can produce relatively pure L-isomer [6]. By our method, the carbon dioxide in the aqueous phase can be condensed with acetaldehyde to produce the pyruvic acid and can easily be converted into lactic acid, which is a material ofbiodegradative plastic. This

1456 method utilizes an enzymatic reaction that proceeds in shorter times and gives the product in higher purity than that by fermentation. Although further studies are required to establish efficient pilot-scale production, these features are advantageous than the classical fermentation method for lactic acid production and the carbon dioxide will be able to be immobilized into the biological cycles. In conclusion, we have demonstrated the usefulness of the reverse reaction of pyruvate decarboxylase. This reaction might become a recommendable, environmentally safe carboxylation procedure for acetaldehyde. Also, we have developed a one-pot, two-step enzymatic synthesis method of L-lactic acid from acetaldehyde and carbon dioxide. This method might become a recommendable, environmentally safe procedure for Llactic acid production. Further studies are in progress in our laboratory.

ACKNOWLEDGMENT We thank Mitsukuni Shibue and Kazuya Ogino for technical assistance. REFERENCES Anastas, P. T. and Warner, J. C. (1998) Green Chemistry: Theory and Practice, Oxford. Koller, K. M. and Wong, C.-H. (200 l) Nature 409, 232. Schmid, A., Dordick, J. S., Hauer, B., Kiener, M., Wubbolts, M., and Witholt, B. (2001) Nature 409, 258. Amass, W., Amass, A., and Tighe, B. (1998) Polym. Int. 47, 89. Jacobsen, S., Degree, P. H., Fritz, H. G., Dubois, P. H., and Jerome, R. (1999) Polym. Eng. Sci. 39, 1311. Lunt, J. (1998) Polym. Degred. Stab. 59, 145. Sprenger, G. A. and Pohl, M. (1999) J. Mol. Catal. B: Enzym. 6, 145. Schorken, U. and Sprenger, G. A. (1998) Biochim. Biophys. Acta 1385, 229. Foppen, M.-A. E., de Lange, Y. M., van Rantwijk, F., Maat, L., and Kieboom, P. G. (1990) Recl. Tray. Chim. Pay-Bas 109, 359. Li, Y., Xu, G.-H., Liu, C.-J., Eliasson, B., and Xue, B.-Z. (2001) Energy & Fuels 15, 299. Aresta, M., Dibenedetto, A., and Tomassi, I. (2001) Energy & Fuels 15, 269. Liu, C.-J., Xu, G.-H., and Wang, T. (1999) Fuel Processing Technol. 58, 119. Bond, G. M., Stringer, J., Brandvold, D. K., Arzum Simsek, F., Medina, M.-G., and Egeland, G. (2001) Energy & Fuels 15, 309.

10. 11. 12. 13.