- Email: [email protected]
Stepwise ethanolysis of tuna oil using immobilized Candida antarctica lipase
JOURNALOF BIOSCIENCEAND BIOENGINEERING Vol. 88, No. 6, 622-626. 1999
Stepwise Ethanolysis of Tuna Oil Using Immobilized Candida antarctica Lipase YOM1 WATANABE, YUJI SHIMADA, * AK10 SUGIHARA, AND YOSHIO TOMINAGA Osaka Municipal Technical Research Institute, I-6-50 Morinomiya, Joto-ku, Osaka 5364553, Japan Received 14 July 1999/Accepted 6 September 1999
Ethanolysis of fish oil under mild conditions has been strongly desired for preparing the starting materials for the purification of ethyl docosahexaenoate. Thus, we attempted ethanolysis of tuna oil using immobilized Candida antarctica lipase. The immobilized lipase was inactivated in the presence of 2/3 molar equivalent of ethanol against the total fatty acids in tuna oil. To avoid such inactivation, the first step of ethanolysis was conducted at 4WC in a mixture of tuna oil and l/3 molar equivalent of ethanol using 4% immobilized lipase. After a 10-h reaction, ethanol was consumed and 33% of tuna oil was converted to its corresponding ethyl esters (E-FAs). The reactant is named Gly/E-FA33. The lipase was not inactivated in the presence of 213 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33. These findings and the consideration of several factors affecting ethanolysis of tuna oil led to the development of the two- and three-step ethanolyses. The two-step reaction was performed as follows: the first step was carried out at 40°C for 12 h in a mixture of tuna oil and l/3 molar equivalent of ethanol with 4% immobilized lipase; the second step was performed for 36 h (total reaction period, 48 h) after adding 2/3 molar equivalent of ethanol. On the other hand, the three-step reaction was conducted as follows: the first step was conducted under the same conditions as those in the two-step ethanolysis; in the second and third steps, l/3 molar equivalent of ethanol was added after 12 and 24 h, respectively; and in the third step, the mixture was shaken for 24 h (total, 48 h). Both types of ethanolyses achieved the conversion of 95% or more of tuna oil to its corresponding E-FAs. To investigate the lipase stability, the two- and three-step ethanolyses were repeated by transferring the enzyme to a fresh substrate mixture of the first step after finishing one cycle of reaction. The two- and three-step reactions maintained over 95% of the conversion for 70 d and over 100 d, respectively. [Key words: ethanolysis,
Candida antarctica lipase, tuna oil, ethyl esters, immobilized
Ethyl eicosapentaenoate (E-EPA) has been purified from fatty acid ethyl esters (E-FAs) originating from sardine oil, by a combination of rectification and urea adduct fractionation (1). Purified E-EPA has been used in the treatment of arteriosclerosis obliterans and hyperlipemia since 1991 in Japan (2). Ethyl docosahexaenoate (E-DHA) has physiological activities similar or superior to those of E-EPA (3-9), and has been desired as a pharmaceutical substance. We recently succeeded in enriching E-DHA in the unreacted E-FA fraction by alcoholysis of E-FAs originating from tuna oil with lauryl alcohol using Rhizopus delemar or Rhizomucor miehei lipase which acts on docosahexaenoic acid (DHA) very weakly (10, 11). In this selective alcoholysis, the E-FAs were prepared from tuna oil by a chemical process with an alkaline catalyst. However, because heating under alkaline conditions often results in the isomerization of DHA, enzymatic reaction under mild conditions is preferable. Recently, Breivik et al. (12) reported that immobilized Candida antarctica lipase efficiently catalyzed ethanolysis of glycerides containing DHA and eicosapentaenoic acid. However, we found that the lipase, although immobilized, could not be repeatedly used in their system. In order to achieve the industrial ethanolysis of fish oil with lipase, repeated use of the enzyme is essential because of its high cost. Thus, we attempted the construction of a new enzymatic system that can be industrially available for the continuous production of E-FAs from tuna oil. In this work, we demonstrate that the decrease in the * Corresponding
author.
enzyme]
extent of ethanolysis is attributed to the inactivation of immobilized Candida lipase by ethanol, and that the inactivation can be eliminated by stepwise addition of ethanol. It is also shown that the conversion of tuna oil to its corresponding E-FAs can be maintained at over 95% for 100d under the reaction conditions determined on the basis of several factors affecting the reaction. MATERIALS
AND METHODS
Materials Tuna oil was a gift from Maruha Corp. (Tokyo). The molarity of the oil was calculated from its saponification value of 186. Immobilized C. antarctica lipase (Novozym 435) was purchased from Novo Nordisk (Bagsvaerd, Denmark). Trycaprylin (purity, 98%) and ethanol (99.5%) were of reagent grade. Ethanolysis A reaction mixture was composed of tuna oil, ethanol, and immobilized Candida lipase. The reaction was performed at 40°C in a 20- or 50-ml screwcapped vessel with shaking at 130 oscillations/min. The continual batch reaction was conducted by transferring the immobilized lipase to a fresh substrate mixture after one cycle of reaction. The extent of ethanolysis (the conversion of tuna oil) was expressed as the content of EFAs in the total reaction mixture. Preparation of a mixture of glycerides and 33% E-FAs Ethanolysis was carried out at 40°C for 24 h in a mixture of 71.37 g of tuna oil, 3.63 g of ethanol (l/3 molar equivalent against the total fatty acids in the oil), and 4% immobilized Candida lipase. The reaction was performed in a lOO-ml screw-capped vessel. Ethanolysis was repeated 10 times by transferring the immobilized en-
VOL.
88,
ETHANOLYSIS
1999
zyme to a fresh substrate mixture. Each reaction mixture was combined, and the lower layer (glycerol) was removed. The amount of the resulting mixture was 723 g, and the E-FA content was 32.9 wt%. The mixture is named Gly/E-FA33. Analysis The content of E-FAs in the reaction mixture was quantified on a Hewlett-Packard 5890 gas chromatograph (Avondale, PA, USA) equipped with a DB-5 capillary column (0.25 mm x 10 m; J&W Scientific, Folsom, CA, USA), using tricaprylin as an internal standard. The analysis was performed according to the conditions described elsewhere (13). RESULTS Inactivation of Cundida lipase by ethanol Ethanolysis of tuna oil was conducted at 40°C for 24 h in a 10-g mixture containing various amounts of ethanol and 4% immobilized Candida lipase (Fig. 1). The reaction was repeated four times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol. When ethanolysis of tuna oil was performed with less than 2/3 molar equivalent of ethanol against the total fatty acids in the oil, the first ethanolysis proceeded efficiently. However, when the reaction was repeated in the mixture containing more than 2/3 molar equivalent of ethanol, the conversion was decreased. The result suggests the inactivation of the lipase in the presence of more than 2/3 molar equivalent of ethanol. At least an equal molar equivalent of ethanol against the total fatty acids in tuna oil is required for the complete conversion of the oil to its corresponding E-FAs. Because the lipase was inactivated in the presence of more than 2/3 molar equivalent of ethanol, we attempted the stepwise addition of ethanol. The first step was performed in a mixture of tuna oil and l/3 molar equivalent of ethanol, as described in Materials and Methods, and a mixture of glycerides and 32.9% E-FAs (Gly/EFA33) was obtain led. To investigate the effect of the 100 b‘;; 80 .; 60 E P c 40 0” 20 0
I l/6
OF TUNA OIL WITH LIPASE
amount of ethanol in the second step of ethanolysis, the reaction was conducted at 40°C for 24 h in a 10-g mixture of Gly/E-FA33 and various amounts of ethanol using 4% immobilized Candida lipase (Fig. 2). The reactions were repeated three times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol, and the fourth-cycle reaction was carried out in a mixture of Gly/E-FA33 and 2/3 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33. A molar equivalent of ethanol of less than 2/3 did not decrease the conversion of glycerides to its corresponding E-FAs. However, the conversion decreased when the glycerides were allowed to react with more than the equal molar amount of ethanol. When the fourth-cycle reaction was performed by transferring the enzyme to a mixture containing 2/3 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33, the decreased conversion in the reaction with one molar equivalent of ethanol was completely recovered, but the conversion in the reaction with 5/3 or 7/3 molar equivalent of ethanol was not. These results indicate that the decrease in the conversion is due to the irreversible inactivation of some of the lipase by ethanol. Other factors affecting ethanolysis of tuna oil Ethanolysis was carried out at 40°C in a mixture of 9.52 g of tuna oil, 0.48 g of ethanol (l/3 molar equivalent against the total fatty acids in the oil), and 1 to 10% immobilized Cundida lipase (Fig. 3). The reaction rate depended on the amount of enzyme, and 1, 4, and 10% lipase converted more than 30% of the oil to its corresponding E-FAs after 24, 10, and 5 h, respectively (the consumption of ethanol was over 90%). When tuna oil was shaken at 30°C with l/3 molar equivalent of ethanol using 4% immobilized lipase, some of the component(s) was crystallized in the early stage of 100
0 II3
i l/3
213
2l3 Ethanol/Fatty
i 313
513
Ethanol/Fattyacids(mol/mol)
FIG. 1. Effect of amount of ethanol on ethanolysis of tuna oil with Cundidu lipase. The reaction was repeated four times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol. The amount of ethanol was shown as molar equivalents against the total fatty acids in tuna oil. The conversion was expressed as the amount of ethanol consumed for ester conversion of the oil (when the ethanol amount was less than one molar equivalent), and as the ratio of the amount of fatty acid ethyl esters (E-FAs) to the total amounts of ethyl esters and unreacted glycerides (more than one molar equivalent). Symbols: 0, the first cycle of reaction; B, the second cycle; M, the third; I, the fourth.
623
313
513
acids (mol/mol)
713
I
FIG. 2. Effect of amount of ethanol on ethanolysis of glycerides with immobilized Candida lipase in the presence of E-FAs originating from tuna oil. A mixture of glycerides and E-FAs (Gly/E-FA33; 32.9% E-FAs) was used as a substrate. The reaction was repeated three times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol, and the fourth cycle of reaction was performed in a mixture of Gly/E-FA33 and 2/3 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33. The amount of ethanol was shown as molar equivalents against the total fatty acids in Gly/E-FA33. The conversion ratio was expressed as the amount of ethanol consumed (when amount of ethanol was l/3 molar equivalent), and as the ratio of the amount of E-FAs to the total amounts of E-FAs and unreacted glycerides (more than 213 molar equivalent). Symbols: 0, the first cycle of reaction; B, the second cycle; m, the third; I, the fourth.
624
WATANABE
ET AL.
J. BIOSCI. BIOENG..
E
: 60 8 jj 40 8 E 20 5 0 0
2
4 6 8 Amount of enzyme (%)
10
reaction and the viscosity of the reaction mixture increased. Thus, the effect of temperature on the ethanolysis of tuna oil was investigated in a range of 35 to 60°C (Fig. 4). The ester conversion at 5 h increased with increasing temperature, but the conversion at 15 h reached 33% even at 35°C. Ethanolysis of tuna oil by stepwise addition of ethanol We attempted to carry out two types of reactions (twoand three-step reactions) by considering the factors affecting ethanolysis of tuna oil. The two-step reaction was performed as follows: the first step was carried out in a mixture of tuna oil and l/3 molar equivalent of ethanol for 24 h, and the second step was started by adding 2/3 molar equivalent of ethanol to the reaction mixture. The three-step reaction was conducted as follows: the first step was performed under the same conditions as those of the two-step reaction, and in the second and third steps, l/3 molar equivalent of ethanol was added every 24 h. The reactions were carried out at 40°C using 4% immobilized lipase. Typical time courses are shown in Fig. 5. In the first step of the two- and three-step reactions,
40 30 20 10
5 0 60 Temperature
10
20
30
40
50
60
70
Reaction time (h)
FIG. 3. Effect of amount of enzyme on ethanolysis of tuna oil. Tuna oil was shaken at 40°C with l/3 molar equivalent of ethanol against the total fatty acids in tuna oil and 1 to 10% immobilized Cczndidu lipase. Symbols: 0, conversion of tuna oil to the corresponding E-FAs after a 5-h reaction; l , conversion of tuna oil to E-FAs after a 10-h reaction; q , conversion of tuna oil to E-FAs after a 24-h reaction.
';; b L t; a, 2 s t % E
0
(“C)
FIG. 4. Effect of temperature on ethanolysis of tuna oil. Tuna oil was shaken at temperature in the range of 35 to 60°C with l/3 molar equivalent of ethanol against the total fatty acids in tuna oil and 4% immobilized Candida lipase. Symbols: @, conversion of tuna oil to E-FAs after a 5-h reaction; I, conversion of tuna oil to E-FAs after a 15-h reaction.
FIG. 5. Typical time courses of the two- and three-step ethanolyses of tuna oil using immobilized Cundidu lipase. The first reaction was performed at 40°C in a mixture of 28.55 g of tuna oil, 1.45 g of ethanol (l/3 molar equivalent against the fatty acids in the oil), and 1.2 g of immobilized lipase. In the two-step reaction ( ~3), 2.90 g of ethanol (213 molar equivalent) was added after 24 h, as indicated by a downward arrow. In the three-step reaction (0), 1.45 g of ethanol (l/3 molar equivalent) was added once every 24 h a total two times as indicated by upward arrows. The single-step ethanolysis (A) was carried out in a mixture of 26.03 g of tuna oil, 3.97 g of ethanol (one molar equivalent), and 1.2 g of immobilized lipase.
the conversion of tuna oil to E-FAs reached 33% at lOh, showing that the ethanol added was completely consumed. After the 24-h reaction, 2/3 and l/3 molar equivalents of ethanol were added to the reaction mixtures in the two- and three-step reactions, respectively. The addition of 2/3 molar equivalent of ethanol converted the oil to its corresponding E-FAs faster than the addition of l/3 molar equivalent of ethanol. The ester conversion reached 96% after 24 h (48 h in total) of adding 2/3 molar equivalent of ethanol. Because the addition of the second l/3 molar equivalent of ethanol converted 66% of the oil to E-FAs after 16 h (40 h in total), the third l/3 molar equivalent of ethanol was added again after 48 h. As a result, 96% of the oil was converted to E-FAs after 16 h (64 h in total). These results show that the two- and three-step reactions are effective for the conversion of tuna oil to its corresponding E-FAs. The single-step ethanolysis was additionally employed at 40°C in a mixture of tuna oil, one molar equivalent of ethanol, and 4% immobilized lipase, for comparison with the above two reactions (Fig. 5). The conversion rate in the single-step reaction was faster than those in the two- and three-step reactions, and the conversion reached 94% after 24 h. The single-step reaction was, therefore, more effective than the two- and three-step reactions, but the lipase was drastically inactivated by repeating the reaction (Fig. 1). If a lipase stable in equimolar ethanol is found, the enzyme will become a good catalyst for the efficient ethanolysis of tuna oil. Continual ethanolyses of tuna oil by two- and threestep additions of ethanol One cycle of the two-step reaction was performed as follows: the first step was carried out at 40°C for 12 h in a 10-g mixture composed of tuna oil, l/3 molar equivalent of ethanol, and 4% immobilized Candida lipase; the second step was conducted for 36 h after adding 2/3 molar equivalent of ethanol (total reaction period, 48 h). Meanwhile, the three-step reaction was performed as follows: the first step was carried out under the same conditions as those in the two-step reaction; the second and third l/3 molar equiva-
ETHANOLYSIS
VOL. 88, 1999
0
10
20
30
40
50
60
Cycle number FIG. 6. Continual ethanolyses of tuna oil by the two- and threestep reactions using immobilized Candida lipase. The two- and threestep reactions were conducted as described in the text. Symbols: 0, two-step reaction; 0, three-step reaction.
lents of ethanol were added after 12 and 24 h, respectively, and the third step was continued for 24 h (total, 48 h). The two- and three-step reactions were repeated by transferring the enzyme to a fresh substrate mixture of the first step. Figure 6 shows the lipase stability in the continual two- and three-step ethanolyses. The conversion in the two-step ethanolysis did not decrease up to 37 cycles, and then decreased rapidly. However, the three-step reaction maintained over 95% of the conversion for 54 cycles (108 d). To investigate the activity of lipase used for 54 cycles in the three-step reaction, the time course of the ethanolysis in the 55th cycle was investigated and compared with that of the ethanolysis in the first cycle (Fig. 7). The difference in the reaction rates showed that the enzyme activity decreased to 32% of the original activity. It was also found that some ethanol remained in the reaction mixture at the end of the first step (12 h). From these results, the marked decrease in the enzyme activity in the two-step reaction could be explained as follows: (i) the lipase was gradually inactivated after a long period of use, (ii) some ethanol remained in the reaction mixture at the end of the first step, (iii) when 213 molar equivalent of ethanol was added in the second step, the
4oI
0 0
5
10
15
20
Time(h) FIG. 7. Time courses of the first and 55th ethanolyses of tuna oil. The reaction was carried out at 40°C in a 10-g mixture consisting of tuna oil, l/3 molar equivalent of ethanol against the total fatty acids in the oil, and 400 mg of immobilized Candida lipase. The ethanolyses were conducted using fresh enzyme and the enzyme from the threestep reaction in Fig. 6 after 54 cycles. Symbols: 0, ethanolysis with fresh enzyme; 0, ethanolysis with enzyme after 54 cycles.
OF TUNA OIL WITH LIPASE
625
amount of ethanol was actually more than 2/3 molar equivalent, (iv) a higher concentration of ethanol inactivated the lipase, thus, decreasing conversion. On the other hand, because the amount of ethanol added in the three-step reaction was l/3 molar equivalent, the amount of ethanol in the reaction mixture did not exceed 2/3 molar equivalent. Thus, the lipase was not inactivated by ethanol. These results also suggest that the lipase can be used for a longer period in the two-step reaction by adding the second 2/3 molar equivalent of ethanol after the complete consumption of the first l/3 molar equivalent of ethanol. DISCUSSION We have demonstrated that more than 95% of tuna oil is efficiently converted to its corresponding E-FAs by the two- and three-step ethanolyses with stepwise addition of ethanol. Because the conversion rate in the second step of the two-step ethanolysis was faster than that in the second step of the three-step reaction (Fig. 5), the period for one cycle of reaction could be shortened by employing the two-step reaction. Indeed, in the first cycle of reaction, the two-step reaction required 34 h to obtain more than 95% conversion (the first step, 10 h; the second step, 24 h), while the three-step reaction required 42 h (the first step, 10 h; the second step, 16 h; the third step 16 h). However, ethanol remaining in the first-step reaction mixture was responsible for the inactivation of lipase in the two-step reaction, although the ethanol did not inactivate lipase in the three-step reaction (Fig. 6). The inactivation of lipase in the two-step reaction can be avoided by monitoring the amount of ethanol in the reaction mixture. It is, therefore, inferred that the lipase can be used in the two-step reaction for the same period as that in the three-step reaction. All things considered, we conclude that the two-step reaction with monitoring of the amount of ethanol is more suitable for the production of E-FAs from tuna oil than the three-step reaction. It is natural that a short reaction time is desirable because of the lability of polyunsaturated fatty acids. We are now studying continuous E-FAs production from tuna oil using a fixed bed bioreactor to further shorten the reaction period. In the early stage of reaction, ethanol is completely dissolved in tuna oil and the concentration is high. As ethanolysis proceeds, glycerol is generated and the reaction mixture separates into two layers: Gly/E-FAs and glycerol layers. The ethanol concentration in the GIy/EFAs layer decreases because of the movement of some of the ethanol to the glycerol layer. In general, a low concentration of the substrate affects the equilibrium and the velocity of the reaction. Therefore, if glycerol can be removed from the reaction mixture, the efficiency of ethanolysis will be further enhanced. Immobilized Candida lipase was inactivated by l/2 molar equivalent of methanol (13). We considered that the excess amount of methanol remained as droplets dispersed in the oil, which inactivated the lipase. If this is correct, methanol that is completely dissolved in the substrate mixture should not inactivate the lipase. In the present study, 2/3 molar equivalent of ethanol inactivated the lipase in the reaction mixture of tuna oil and ethanol (Fig. 1). However, the lipase was negligibly inactivated by one molar equivalent of ethanol in the mixture of Gly/E-FA33 and ethanol (Fig. 2). The solubili-
626
WATANABE
ET AL
ties of ethanol in tuna oil and Gly/E-FA33 were l/2 and one molar equivalent against the total fatty acids, respectively. These results strongly suggest that the above hypothesis is correct. ACKNOWLEDGMENTS We thank Professor Hideki Fukuda, Graduate School of Science and Technology, Kobe University, and Dr. Hideo Noda, Kansai Chemical Engineering Co. Ltd., for valuable discussions. This work was supported by the New Energy and Industrial Technology Development Organization. REFERENCES 1. Noda, H., Noda, Y., Hata, K., and Fujita, T.: Purification of ethyl eicosapentaenoate by distillation under high vacuum. Kagaku Kogaku, 55, 623-625 (1991). (in Japanese) 2. Hara, K.: Pharmaceutical application of eicosapentaenoic acid. Yushi, 46, 91-99 (1993). (in Japanese) 3. Grundt, H., Nilsen, D. W. T., Hetland, @., Aarsland, T., Baksaas, I., and Grande, T.: Improvement of serum lipids and blood pressure during intervention with n-3 fatty acids was not associated with changes in insulin levels in subjects with combined hyperlipidaemia. J. Intern. Med., 237, 249-259 (1995). 4. Kromhout, D., Bosschietor, E. B., and Coulander, C. D. L.: The inverse relation between fish consumption and 20-year mortality from coronary heart disease. New Engl. J. Med., 312, 1205-1209 (1985). 5. Pilipson, B. E., Rothrock, D. W., Connor, W. E., Harris, W. S., and Blingworth, D. R.: Reduction of plasma lipids, lipoproteins and apoproteins by dietary fish oils in patients with hypertriglyceridemia. New Engl. J. Med., 312, 1210-1216
J. B~oscr. BIOENC (1985). 6. Crawford, M. A.: The role of essential fatty acids in neural development: implications for perinatal nutrition. Am. J. Clin. Nun., 57(suppl), 703%710s (1993). 7. Carlson, S. E., Werkman, S. H., Rhodes, P. G., and Tolley, E. A.: Visual-acuity development in healthy preterm infants: effect of marine-oil supplementation. Am. J. Clin. Nutr., 58, 35-42 (1993). 8. Lanting, C. I., Fidler, V., Huisman, M., Touwen, B.C. L., and Boersma, E. R.: Neurological differences between g-yearold children fed breast-milk or formula-milk as babies. Lancet, 344, 1319-1322 (1994). 9. Nishikawa, M., Kimura, S., and Akaike, N.: Facilitatory effect of docosahexaenoic acid on fV-methyl-o-aspartate response in pyramidal neurones of rat cerebral cortex. J. Physiol., 475, 8393 (1994). 10. Shimada, Y., Sugihara, A., Yodono, S., Nagao, T., Maruyama, K., Nakano, H., Komemushi, S., and Tominaga, Y.: Enrichment of ethyl docosahexaenoate by selective alcoholysis with immobilized Rhizopus delemar lipase. J. Ferment. Bioeng., 84, 138-143 (1997). 11. Shimada, Y., Mamyama, K., Sugihara, A., Baba, T., Komemushi, S., Moriyama, S., and Tominaga, Y.: Purification of ethyl docosahexaenoate by selective alcoholysis of fatty acid ethyl esters with immobilized Rhizomucor miehei lipase. J. Am. Oil Chem. Sot., 75, 1565-1571 (1998). 12. Breivik, H., Haraldsson, G. G., and Kristinsson, B.: Preparation of highly purified concentrates of eicosapentaenoic acids and docosahexaenoic acid. J. Am. Oil Chem. Sot., 74, 14251429 (1997). 13. Shimada, Y., Watanabe, Y., Samukawa, T., Sugihara, A., Noda, H., Fukuda, H., and Tominaga, Y.: Conversion of vegetable oil to biodiesel using immobilized Candida antarctica lipase. J. Am. Oil Chem. Sot., 76, 789-793 (1999).
Stepwise Ethanolysis of Tuna Oil Using Immobilized Candida antarctica Lipase YOM1 WATANABE, YUJI SHIMADA, * AK10 SUGIHARA, AND YOSHIO TOMINAGA Osaka Municipal Technical Research Institute, I-6-50 Morinomiya, Joto-ku, Osaka 5364553, Japan Received 14 July 1999/Accepted 6 September 1999
Ethanolysis of fish oil under mild conditions has been strongly desired for preparing the starting materials for the purification of ethyl docosahexaenoate. Thus, we attempted ethanolysis of tuna oil using immobilized Candida antarctica lipase. The immobilized lipase was inactivated in the presence of 2/3 molar equivalent of ethanol against the total fatty acids in tuna oil. To avoid such inactivation, the first step of ethanolysis was conducted at 4WC in a mixture of tuna oil and l/3 molar equivalent of ethanol using 4% immobilized lipase. After a 10-h reaction, ethanol was consumed and 33% of tuna oil was converted to its corresponding ethyl esters (E-FAs). The reactant is named Gly/E-FA33. The lipase was not inactivated in the presence of 213 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33. These findings and the consideration of several factors affecting ethanolysis of tuna oil led to the development of the two- and three-step ethanolyses. The two-step reaction was performed as follows: the first step was carried out at 40°C for 12 h in a mixture of tuna oil and l/3 molar equivalent of ethanol with 4% immobilized lipase; the second step was performed for 36 h (total reaction period, 48 h) after adding 2/3 molar equivalent of ethanol. On the other hand, the three-step reaction was conducted as follows: the first step was conducted under the same conditions as those in the two-step ethanolysis; in the second and third steps, l/3 molar equivalent of ethanol was added after 12 and 24 h, respectively; and in the third step, the mixture was shaken for 24 h (total, 48 h). Both types of ethanolyses achieved the conversion of 95% or more of tuna oil to its corresponding E-FAs. To investigate the lipase stability, the two- and three-step ethanolyses were repeated by transferring the enzyme to a fresh substrate mixture of the first step after finishing one cycle of reaction. The two- and three-step reactions maintained over 95% of the conversion for 70 d and over 100 d, respectively. [Key words: ethanolysis,
Candida antarctica lipase, tuna oil, ethyl esters, immobilized
Ethyl eicosapentaenoate (E-EPA) has been purified from fatty acid ethyl esters (E-FAs) originating from sardine oil, by a combination of rectification and urea adduct fractionation (1). Purified E-EPA has been used in the treatment of arteriosclerosis obliterans and hyperlipemia since 1991 in Japan (2). Ethyl docosahexaenoate (E-DHA) has physiological activities similar or superior to those of E-EPA (3-9), and has been desired as a pharmaceutical substance. We recently succeeded in enriching E-DHA in the unreacted E-FA fraction by alcoholysis of E-FAs originating from tuna oil with lauryl alcohol using Rhizopus delemar or Rhizomucor miehei lipase which acts on docosahexaenoic acid (DHA) very weakly (10, 11). In this selective alcoholysis, the E-FAs were prepared from tuna oil by a chemical process with an alkaline catalyst. However, because heating under alkaline conditions often results in the isomerization of DHA, enzymatic reaction under mild conditions is preferable. Recently, Breivik et al. (12) reported that immobilized Candida antarctica lipase efficiently catalyzed ethanolysis of glycerides containing DHA and eicosapentaenoic acid. However, we found that the lipase, although immobilized, could not be repeatedly used in their system. In order to achieve the industrial ethanolysis of fish oil with lipase, repeated use of the enzyme is essential because of its high cost. Thus, we attempted the construction of a new enzymatic system that can be industrially available for the continuous production of E-FAs from tuna oil. In this work, we demonstrate that the decrease in the * Corresponding
author.
enzyme]
extent of ethanolysis is attributed to the inactivation of immobilized Candida lipase by ethanol, and that the inactivation can be eliminated by stepwise addition of ethanol. It is also shown that the conversion of tuna oil to its corresponding E-FAs can be maintained at over 95% for 100d under the reaction conditions determined on the basis of several factors affecting the reaction. MATERIALS
AND METHODS
Materials Tuna oil was a gift from Maruha Corp. (Tokyo). The molarity of the oil was calculated from its saponification value of 186. Immobilized C. antarctica lipase (Novozym 435) was purchased from Novo Nordisk (Bagsvaerd, Denmark). Trycaprylin (purity, 98%) and ethanol (99.5%) were of reagent grade. Ethanolysis A reaction mixture was composed of tuna oil, ethanol, and immobilized Candida lipase. The reaction was performed at 40°C in a 20- or 50-ml screwcapped vessel with shaking at 130 oscillations/min. The continual batch reaction was conducted by transferring the immobilized lipase to a fresh substrate mixture after one cycle of reaction. The extent of ethanolysis (the conversion of tuna oil) was expressed as the content of EFAs in the total reaction mixture. Preparation of a mixture of glycerides and 33% E-FAs Ethanolysis was carried out at 40°C for 24 h in a mixture of 71.37 g of tuna oil, 3.63 g of ethanol (l/3 molar equivalent against the total fatty acids in the oil), and 4% immobilized Candida lipase. The reaction was performed in a lOO-ml screw-capped vessel. Ethanolysis was repeated 10 times by transferring the immobilized en-
VOL.
88,
ETHANOLYSIS
1999
zyme to a fresh substrate mixture. Each reaction mixture was combined, and the lower layer (glycerol) was removed. The amount of the resulting mixture was 723 g, and the E-FA content was 32.9 wt%. The mixture is named Gly/E-FA33. Analysis The content of E-FAs in the reaction mixture was quantified on a Hewlett-Packard 5890 gas chromatograph (Avondale, PA, USA) equipped with a DB-5 capillary column (0.25 mm x 10 m; J&W Scientific, Folsom, CA, USA), using tricaprylin as an internal standard. The analysis was performed according to the conditions described elsewhere (13). RESULTS Inactivation of Cundida lipase by ethanol Ethanolysis of tuna oil was conducted at 40°C for 24 h in a 10-g mixture containing various amounts of ethanol and 4% immobilized Candida lipase (Fig. 1). The reaction was repeated four times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol. When ethanolysis of tuna oil was performed with less than 2/3 molar equivalent of ethanol against the total fatty acids in the oil, the first ethanolysis proceeded efficiently. However, when the reaction was repeated in the mixture containing more than 2/3 molar equivalent of ethanol, the conversion was decreased. The result suggests the inactivation of the lipase in the presence of more than 2/3 molar equivalent of ethanol. At least an equal molar equivalent of ethanol against the total fatty acids in tuna oil is required for the complete conversion of the oil to its corresponding E-FAs. Because the lipase was inactivated in the presence of more than 2/3 molar equivalent of ethanol, we attempted the stepwise addition of ethanol. The first step was performed in a mixture of tuna oil and l/3 molar equivalent of ethanol, as described in Materials and Methods, and a mixture of glycerides and 32.9% E-FAs (Gly/EFA33) was obtain led. To investigate the effect of the 100 b‘;; 80 .; 60 E P c 40 0” 20 0
I l/6
OF TUNA OIL WITH LIPASE
amount of ethanol in the second step of ethanolysis, the reaction was conducted at 40°C for 24 h in a 10-g mixture of Gly/E-FA33 and various amounts of ethanol using 4% immobilized Candida lipase (Fig. 2). The reactions were repeated three times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol, and the fourth-cycle reaction was carried out in a mixture of Gly/E-FA33 and 2/3 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33. A molar equivalent of ethanol of less than 2/3 did not decrease the conversion of glycerides to its corresponding E-FAs. However, the conversion decreased when the glycerides were allowed to react with more than the equal molar amount of ethanol. When the fourth-cycle reaction was performed by transferring the enzyme to a mixture containing 2/3 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33, the decreased conversion in the reaction with one molar equivalent of ethanol was completely recovered, but the conversion in the reaction with 5/3 or 7/3 molar equivalent of ethanol was not. These results indicate that the decrease in the conversion is due to the irreversible inactivation of some of the lipase by ethanol. Other factors affecting ethanolysis of tuna oil Ethanolysis was carried out at 40°C in a mixture of 9.52 g of tuna oil, 0.48 g of ethanol (l/3 molar equivalent against the total fatty acids in the oil), and 1 to 10% immobilized Cundida lipase (Fig. 3). The reaction rate depended on the amount of enzyme, and 1, 4, and 10% lipase converted more than 30% of the oil to its corresponding E-FAs after 24, 10, and 5 h, respectively (the consumption of ethanol was over 90%). When tuna oil was shaken at 30°C with l/3 molar equivalent of ethanol using 4% immobilized lipase, some of the component(s) was crystallized in the early stage of 100
0 II3
i l/3
213
2l3 Ethanol/Fatty
i 313
513
Ethanol/Fattyacids(mol/mol)
FIG. 1. Effect of amount of ethanol on ethanolysis of tuna oil with Cundidu lipase. The reaction was repeated four times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol. The amount of ethanol was shown as molar equivalents against the total fatty acids in tuna oil. The conversion was expressed as the amount of ethanol consumed for ester conversion of the oil (when the ethanol amount was less than one molar equivalent), and as the ratio of the amount of fatty acid ethyl esters (E-FAs) to the total amounts of ethyl esters and unreacted glycerides (more than one molar equivalent). Symbols: 0, the first cycle of reaction; B, the second cycle; M, the third; I, the fourth.
623
313
513
acids (mol/mol)
713
I
FIG. 2. Effect of amount of ethanol on ethanolysis of glycerides with immobilized Candida lipase in the presence of E-FAs originating from tuna oil. A mixture of glycerides and E-FAs (Gly/E-FA33; 32.9% E-FAs) was used as a substrate. The reaction was repeated three times by transferring the lipase to a fresh substrate mixture containing the indicated amount of ethanol, and the fourth cycle of reaction was performed in a mixture of Gly/E-FA33 and 2/3 molar equivalent of ethanol against the total fatty acids in Gly/E-FA33. The amount of ethanol was shown as molar equivalents against the total fatty acids in Gly/E-FA33. The conversion ratio was expressed as the amount of ethanol consumed (when amount of ethanol was l/3 molar equivalent), and as the ratio of the amount of E-FAs to the total amounts of E-FAs and unreacted glycerides (more than 213 molar equivalent). Symbols: 0, the first cycle of reaction; B, the second cycle; m, the third; I, the fourth.
624
WATANABE
ET AL.
J. BIOSCI. BIOENG..
E
: 60 8 jj 40 8 E 20 5 0 0
2
4 6 8 Amount of enzyme (%)
10
reaction and the viscosity of the reaction mixture increased. Thus, the effect of temperature on the ethanolysis of tuna oil was investigated in a range of 35 to 60°C (Fig. 4). The ester conversion at 5 h increased with increasing temperature, but the conversion at 15 h reached 33% even at 35°C. Ethanolysis of tuna oil by stepwise addition of ethanol We attempted to carry out two types of reactions (twoand three-step reactions) by considering the factors affecting ethanolysis of tuna oil. The two-step reaction was performed as follows: the first step was carried out in a mixture of tuna oil and l/3 molar equivalent of ethanol for 24 h, and the second step was started by adding 2/3 molar equivalent of ethanol to the reaction mixture. The three-step reaction was conducted as follows: the first step was performed under the same conditions as those of the two-step reaction, and in the second and third steps, l/3 molar equivalent of ethanol was added every 24 h. The reactions were carried out at 40°C using 4% immobilized lipase. Typical time courses are shown in Fig. 5. In the first step of the two- and three-step reactions,
40 30 20 10
5 0 60 Temperature
10
20
30
40
50
60
70
Reaction time (h)
FIG. 3. Effect of amount of enzyme on ethanolysis of tuna oil. Tuna oil was shaken at 40°C with l/3 molar equivalent of ethanol against the total fatty acids in tuna oil and 1 to 10% immobilized Cczndidu lipase. Symbols: 0, conversion of tuna oil to the corresponding E-FAs after a 5-h reaction; l , conversion of tuna oil to E-FAs after a 10-h reaction; q , conversion of tuna oil to E-FAs after a 24-h reaction.
';; b L t; a, 2 s t % E
0
(“C)
FIG. 4. Effect of temperature on ethanolysis of tuna oil. Tuna oil was shaken at temperature in the range of 35 to 60°C with l/3 molar equivalent of ethanol against the total fatty acids in tuna oil and 4% immobilized Candida lipase. Symbols: @, conversion of tuna oil to E-FAs after a 5-h reaction; I, conversion of tuna oil to E-FAs after a 15-h reaction.
FIG. 5. Typical time courses of the two- and three-step ethanolyses of tuna oil using immobilized Cundidu lipase. The first reaction was performed at 40°C in a mixture of 28.55 g of tuna oil, 1.45 g of ethanol (l/3 molar equivalent against the fatty acids in the oil), and 1.2 g of immobilized lipase. In the two-step reaction ( ~3), 2.90 g of ethanol (213 molar equivalent) was added after 24 h, as indicated by a downward arrow. In the three-step reaction (0), 1.45 g of ethanol (l/3 molar equivalent) was added once every 24 h a total two times as indicated by upward arrows. The single-step ethanolysis (A) was carried out in a mixture of 26.03 g of tuna oil, 3.97 g of ethanol (one molar equivalent), and 1.2 g of immobilized lipase.
the conversion of tuna oil to E-FAs reached 33% at lOh, showing that the ethanol added was completely consumed. After the 24-h reaction, 2/3 and l/3 molar equivalents of ethanol were added to the reaction mixtures in the two- and three-step reactions, respectively. The addition of 2/3 molar equivalent of ethanol converted the oil to its corresponding E-FAs faster than the addition of l/3 molar equivalent of ethanol. The ester conversion reached 96% after 24 h (48 h in total) of adding 2/3 molar equivalent of ethanol. Because the addition of the second l/3 molar equivalent of ethanol converted 66% of the oil to E-FAs after 16 h (40 h in total), the third l/3 molar equivalent of ethanol was added again after 48 h. As a result, 96% of the oil was converted to E-FAs after 16 h (64 h in total). These results show that the two- and three-step reactions are effective for the conversion of tuna oil to its corresponding E-FAs. The single-step ethanolysis was additionally employed at 40°C in a mixture of tuna oil, one molar equivalent of ethanol, and 4% immobilized lipase, for comparison with the above two reactions (Fig. 5). The conversion rate in the single-step reaction was faster than those in the two- and three-step reactions, and the conversion reached 94% after 24 h. The single-step reaction was, therefore, more effective than the two- and three-step reactions, but the lipase was drastically inactivated by repeating the reaction (Fig. 1). If a lipase stable in equimolar ethanol is found, the enzyme will become a good catalyst for the efficient ethanolysis of tuna oil. Continual ethanolyses of tuna oil by two- and threestep additions of ethanol One cycle of the two-step reaction was performed as follows: the first step was carried out at 40°C for 12 h in a 10-g mixture composed of tuna oil, l/3 molar equivalent of ethanol, and 4% immobilized Candida lipase; the second step was conducted for 36 h after adding 2/3 molar equivalent of ethanol (total reaction period, 48 h). Meanwhile, the three-step reaction was performed as follows: the first step was carried out under the same conditions as those in the two-step reaction; the second and third l/3 molar equiva-
ETHANOLYSIS
VOL. 88, 1999
0
10
20
30
40
50
60
Cycle number FIG. 6. Continual ethanolyses of tuna oil by the two- and threestep reactions using immobilized Candida lipase. The two- and threestep reactions were conducted as described in the text. Symbols: 0, two-step reaction; 0, three-step reaction.
lents of ethanol were added after 12 and 24 h, respectively, and the third step was continued for 24 h (total, 48 h). The two- and three-step reactions were repeated by transferring the enzyme to a fresh substrate mixture of the first step. Figure 6 shows the lipase stability in the continual two- and three-step ethanolyses. The conversion in the two-step ethanolysis did not decrease up to 37 cycles, and then decreased rapidly. However, the three-step reaction maintained over 95% of the conversion for 54 cycles (108 d). To investigate the activity of lipase used for 54 cycles in the three-step reaction, the time course of the ethanolysis in the 55th cycle was investigated and compared with that of the ethanolysis in the first cycle (Fig. 7). The difference in the reaction rates showed that the enzyme activity decreased to 32% of the original activity. It was also found that some ethanol remained in the reaction mixture at the end of the first step (12 h). From these results, the marked decrease in the enzyme activity in the two-step reaction could be explained as follows: (i) the lipase was gradually inactivated after a long period of use, (ii) some ethanol remained in the reaction mixture at the end of the first step, (iii) when 213 molar equivalent of ethanol was added in the second step, the
4oI
0 0
5
10
15
20
Time(h) FIG. 7. Time courses of the first and 55th ethanolyses of tuna oil. The reaction was carried out at 40°C in a 10-g mixture consisting of tuna oil, l/3 molar equivalent of ethanol against the total fatty acids in the oil, and 400 mg of immobilized Candida lipase. The ethanolyses were conducted using fresh enzyme and the enzyme from the threestep reaction in Fig. 6 after 54 cycles. Symbols: 0, ethanolysis with fresh enzyme; 0, ethanolysis with enzyme after 54 cycles.
OF TUNA OIL WITH LIPASE
625
amount of ethanol was actually more than 2/3 molar equivalent, (iv) a higher concentration of ethanol inactivated the lipase, thus, decreasing conversion. On the other hand, because the amount of ethanol added in the three-step reaction was l/3 molar equivalent, the amount of ethanol in the reaction mixture did not exceed 2/3 molar equivalent. Thus, the lipase was not inactivated by ethanol. These results also suggest that the lipase can be used for a longer period in the two-step reaction by adding the second 2/3 molar equivalent of ethanol after the complete consumption of the first l/3 molar equivalent of ethanol. DISCUSSION We have demonstrated that more than 95% of tuna oil is efficiently converted to its corresponding E-FAs by the two- and three-step ethanolyses with stepwise addition of ethanol. Because the conversion rate in the second step of the two-step ethanolysis was faster than that in the second step of the three-step reaction (Fig. 5), the period for one cycle of reaction could be shortened by employing the two-step reaction. Indeed, in the first cycle of reaction, the two-step reaction required 34 h to obtain more than 95% conversion (the first step, 10 h; the second step, 24 h), while the three-step reaction required 42 h (the first step, 10 h; the second step, 16 h; the third step 16 h). However, ethanol remaining in the first-step reaction mixture was responsible for the inactivation of lipase in the two-step reaction, although the ethanol did not inactivate lipase in the three-step reaction (Fig. 6). The inactivation of lipase in the two-step reaction can be avoided by monitoring the amount of ethanol in the reaction mixture. It is, therefore, inferred that the lipase can be used in the two-step reaction for the same period as that in the three-step reaction. All things considered, we conclude that the two-step reaction with monitoring of the amount of ethanol is more suitable for the production of E-FAs from tuna oil than the three-step reaction. It is natural that a short reaction time is desirable because of the lability of polyunsaturated fatty acids. We are now studying continuous E-FAs production from tuna oil using a fixed bed bioreactor to further shorten the reaction period. In the early stage of reaction, ethanol is completely dissolved in tuna oil and the concentration is high. As ethanolysis proceeds, glycerol is generated and the reaction mixture separates into two layers: Gly/E-FAs and glycerol layers. The ethanol concentration in the GIy/EFAs layer decreases because of the movement of some of the ethanol to the glycerol layer. In general, a low concentration of the substrate affects the equilibrium and the velocity of the reaction. Therefore, if glycerol can be removed from the reaction mixture, the efficiency of ethanolysis will be further enhanced. Immobilized Candida lipase was inactivated by l/2 molar equivalent of methanol (13). We considered that the excess amount of methanol remained as droplets dispersed in the oil, which inactivated the lipase. If this is correct, methanol that is completely dissolved in the substrate mixture should not inactivate the lipase. In the present study, 2/3 molar equivalent of ethanol inactivated the lipase in the reaction mixture of tuna oil and ethanol (Fig. 1). However, the lipase was negligibly inactivated by one molar equivalent of ethanol in the mixture of Gly/E-FA33 and ethanol (Fig. 2). The solubili-
626
WATANABE
ET AL
ties of ethanol in tuna oil and Gly/E-FA33 were l/2 and one molar equivalent against the total fatty acids, respectively. These results strongly suggest that the above hypothesis is correct. ACKNOWLEDGMENTS We thank Professor Hideki Fukuda, Graduate School of Science and Technology, Kobe University, and Dr. Hideo Noda, Kansai Chemical Engineering Co. Ltd., for valuable discussions. This work was supported by the New Energy and Industrial Technology Development Organization. REFERENCES 1. Noda, H., Noda, Y., Hata, K., and Fujita, T.: Purification of ethyl eicosapentaenoate by distillation under high vacuum. Kagaku Kogaku, 55, 623-625 (1991). (in Japanese) 2. Hara, K.: Pharmaceutical application of eicosapentaenoic acid. Yushi, 46, 91-99 (1993). (in Japanese) 3. Grundt, H., Nilsen, D. W. T., Hetland, @., Aarsland, T., Baksaas, I., and Grande, T.: Improvement of serum lipids and blood pressure during intervention with n-3 fatty acids was not associated with changes in insulin levels in subjects with combined hyperlipidaemia. J. Intern. Med., 237, 249-259 (1995). 4. Kromhout, D., Bosschietor, E. B., and Coulander, C. D. L.: The inverse relation between fish consumption and 20-year mortality from coronary heart disease. New Engl. J. Med., 312, 1205-1209 (1985). 5. Pilipson, B. E., Rothrock, D. W., Connor, W. E., Harris, W. S., and Blingworth, D. R.: Reduction of plasma lipids, lipoproteins and apoproteins by dietary fish oils in patients with hypertriglyceridemia. New Engl. J. Med., 312, 1210-1216
J. B~oscr. BIOENC (1985). 6. Crawford, M. A.: The role of essential fatty acids in neural development: implications for perinatal nutrition. Am. J. Clin. Nun., 57(suppl), 703%710s (1993). 7. Carlson, S. E., Werkman, S. H., Rhodes, P. G., and Tolley, E. A.: Visual-acuity development in healthy preterm infants: effect of marine-oil supplementation. Am. J. Clin. Nutr., 58, 35-42 (1993). 8. Lanting, C. I., Fidler, V., Huisman, M., Touwen, B.C. L., and Boersma, E. R.: Neurological differences between g-yearold children fed breast-milk or formula-milk as babies. Lancet, 344, 1319-1322 (1994). 9. Nishikawa, M., Kimura, S., and Akaike, N.: Facilitatory effect of docosahexaenoic acid on fV-methyl-o-aspartate response in pyramidal neurones of rat cerebral cortex. J. Physiol., 475, 8393 (1994). 10. Shimada, Y., Sugihara, A., Yodono, S., Nagao, T., Maruyama, K., Nakano, H., Komemushi, S., and Tominaga, Y.: Enrichment of ethyl docosahexaenoate by selective alcoholysis with immobilized Rhizopus delemar lipase. J. Ferment. Bioeng., 84, 138-143 (1997). 11. Shimada, Y., Mamyama, K., Sugihara, A., Baba, T., Komemushi, S., Moriyama, S., and Tominaga, Y.: Purification of ethyl docosahexaenoate by selective alcoholysis of fatty acid ethyl esters with immobilized Rhizomucor miehei lipase. J. Am. Oil Chem. Sot., 75, 1565-1571 (1998). 12. Breivik, H., Haraldsson, G. G., and Kristinsson, B.: Preparation of highly purified concentrates of eicosapentaenoic acids and docosahexaenoic acid. J. Am. Oil Chem. Sot., 74, 14251429 (1997). 13. Shimada, Y., Watanabe, Y., Samukawa, T., Sugihara, A., Noda, H., Fukuda, H., and Tominaga, Y.: Conversion of vegetable oil to biodiesel using immobilized Candida antarctica lipase. J. Am. Oil Chem. Sot., 76, 789-793 (1999).