Preparation of Chiral Triacylglycerols, sn-poo and sn-oop, via Lipase-mediated Acidolysis Reaction

Transcription

1 Journal of Oleo Science Copyright 2018 by Japan Oil Chemists Society doi : /jos.ess17149 Preparation of Chiral Triacylglycerols, sn-poo and sn-oop, via Lipase-mediated Acidolysis Reaction Yukihiro Yamamoto 1*, Hiroki Yoshida 2, Toshiharu Nagai 3 and Setsuko Hara 2 1 Faculty of Life and Environmental Sciences, Prefectural University of Hiroshima, Nanatsuka-cho 562, Shobara, Hiroshima, JAPAN 2 Faculty of Science and Technology, Seikei University, Kichijoji-kitamachi 3-3-1, Musashino, Tokyo, JAPAN 3 Tsukishima Foods Industry Co., Ltd. Higashi Kasai , Edogawa-Ku, Tokyo, JAPAN Abstract: It is well known that lipases are useful tools for preparing various structured triacylglycerols (TAGs). However, the lipase-mediated preparation of chiral TAGs has never been reported. This study aimed to prepare chiral TAGs (viz., 1-palmitoyl-2,3-dioleoyl-sn-glycerol (sn-poo) or 1,2-dioleoyl-3- palmitoyl-sn-glycerol (sn-oop)) via lipase mediated acidolysis, using triolein (TO) and palmitic acid (P) as substrates. Three commercially available lipases (viz., Lipozyme RM-IM, Lipozyme TL-IM, and Lipase OF ) were used. Lipozyme RM-IM resulted in an increase 1P-2O (sn-poo + sn-oop + 1,3-dioleoyl-2- palmitoyl-sn-glycerol) content with reaction time, which plateaued at 2~24 h (max. yield 47.1% at 4 h). The highest sn-poo/sn-oop ratio of ca. 9 was obtained at 0.25 h, and the rate got close to 1 with reaction time (sn-poo/sn-oop = 1.3 at 24 h). Lipozyme TL-IM resulted in a lower 1P-2O synthesis rate than Lipozyme RM-IM, where its highest sn-poo/sn-oop ratio of ca. 2 was obtained at 0.25 h and did not vary much further with reaction time. In the case of Lipase OF, its reaction rate for 1P-2O synthesis was lower than that of the other two lipases, and the highest sn-poo/sn-oop ratio of ca. 1.4 was obtained at 0.5 h, reaching closer to 1 with a longer reaction time. Reaction solvents (viz., hexane, acetone, and benzene) also affected the 1P-2O preparation, where the highest 1P-2O content was obtained with the solvent-free system. Furthermore, the solvent-free system showed a higher reaction rate for 1P-2O synthesis than did the hexane system, with no effect on chiral specificity of the lipase for the TAG molecules. These results suggested that among three types of commercial lipase, Lipozyme RM-IM is the most useful for the preparation of chiral TAGs by acidolysis reaction. Key words: chiral triacylglycerol, lipase, acidolysis reaction 1 Introduction The chemical and physical properties of triacylglycerols TAGs are affected not only by their fatty acid compositions but also by the positions of those fatty acids within the bound TAG molecul 1 3. There are a few reported studies concerning the physical properties of chiral TAGs 4, 5. For example, blends of 1,2-decanoyl-3-palmitoryl-sn-glycerol and 1-palymitoyl-2,3-decanoyl-sn-glycerol formed a metastable β polymorphic conglomerate and a stable β polymorphic racemic compound 4. Furthermor, the most thermodynamically stable form of enantiopure TAGs was β polymorph 4. In 2013, Mizobe et al. reported that enantiopure TAGs had a higher crystallization temperature and higher melting point than those of racemic compounds of 1-oleolyl-2,3-dipalmitoyl-sn-glycerol and 1,2-dipalmitolyl- 3-oleoyl-sn-glycerol 5. These findings are very important for foods containing oils and fats, because the crystallization behavior of oils and fats gives effects food quality significantly. Although basic studies are required in this field, the high cost of enantiopure TAGs has hindered research efforts in this regard. Lipases are very useful tools for modifying or preparing various structured TAGs because they are able to catalyze complicated reactions under mild condition and in simple steps. There are many publications related to the lipasemediated preparation of structured TAGs 6 8. Recently, we reported a preparation of a human milk fat substitute from lard using Novozym435 a lipase from Candida antarctica and Lipozyme RM-IM a lipase form Rhizomucor miehei 9. Moreover, lipases have attracted particular attention in the pharmaceutical field because they often recognize the chirality of organic compounds 10, 11. For * Correspondence to: Yukihiro Yamamoto, Faculty of Life and Environmental Sciences, Prefectural University of Hiroshima, Nanatsuka-cho 562, Shobara, Hiroshima, JAPAN yyamamoto@pu-hiroshima.ac.jp Accepted September 6, 2017 (received for review June 30, 2017) Journal of Oleo Science ISSN print / ISSN online

2 Y. Yamamoto, H. Yoshida, T. Nagai et al. example, lipase from Arthrobacter sp. was used for the resolution of four arylazetideones derivatives precursors of an anticancer reagent in buffer solution ph 7.0 with some co-solvents, and resulted in enantiomeric excess of both alcohols and acetates of 99 and conversions close to A racemic mixture of ibuprofen an anti-inflammatory reagent was resolved through esterification with 1-propanol in the presence of the lipase from Candida rugosa immobilized onto magnetic beads 13, resulting in an enantiomeric excess of S -ester of 83. It is known that lingual lipase in infant rat preferably hydrolyzes ester bond at sn-3 position of milk fat TAGs to release medium-chain fatty acid as a source of energy 14. Indeed, medium-chain fatty acid composition at the sn-3 position of milk fat TAGs from cow, goat and donkey contains makes up of total fatty acids 15. Besides this, in the hydrolysis reaction using trioctanoin and triolein TO as a prochiral substrate, the lipase from Mucor miehei showed highest preference for sn-1 position, whereas the lipase cutinase from Fusarium solani showed the highest preference for sn-3 position, among 25 lipases tested by Ewa et al. 16. However, studies on lipasemediated preparation of chiral TAGs have never been reported. In this study, 1-palmitoyl-2,3-dioleoyl-sn-glycerol sn-poo and 1,2-dioleoyl-3-palmioyl-sn-glycerol sn-oop were prepared via lipase mediated acidolysis reaction using TO and palmitic acid P as the substrates Fig. 1 because P and oleic acid C18:1, O are abundant fatty acids in natural products and food materials, such as palm oil, lard, and butter. Thereafter, the distribution of 1P-2O sn-poo sn-oop 1,3-dioleoyl-2-palmitoyl-sn-glycerol sn-opo and sn-poo/sn-oop ratio in the synthesized TAGs were analyzed, using high performance liquid chromatography HPLC equipped with a chiral column 17, to evaluate the effects of lipase variants and reaction mediums on chiral TAG synthesis. This study aimed to reveal whether lipases have chiral selectivity for TAG synthesis in acidolysis reaction. 2 Experiments 2.1 Materials TO of 99.9 purity was gifted from Tsukishima Foods Industry Co., Ltd. Tokyo, Japan. TO with 73.1 purity and P were purchased from Tokyo Chemical Industry Co., Ltd. Tokyo, Japan. Lipozyme RM-IM Rhizomucor miehei; 150 U/g; regiospecificity: at the sn-1 and sn-3 positions; optimal temperature: 60 and Lipozyme TL-IM Thermomyces lanuginosus; 75 U/g; regiospecificity: at the sn-1, and sn-3 position; optimal temperature: 60 were purchased from Novo Nordisk A/S Bagsvӕrd, Denmark. Lipase OF Candida cylindrasea; 360,000 U/g; regiospecificity: random; optimal temperature: 37 was obtained from Meito Sangyo Co., Ltd. Aichi, Japan. These three enzymes are food compatible. Other reagents and solvents used were of analytical grade. 2.2 Acidolysis reaction TO 0.50 g and P was added to 100 ml eggplant flask at the P/TO molar ratio of 1.4 and then dissolved with 10 ml of hexane. Thereafter, enzymes were added 10 wt for Lipozyme RM-IM and Lipozyme TL-IM, and 50 wt for Lipase OF, vs total substrate weight. The ratio of substrates and the amount of enzymes used in this study were based on preliminary experiments. The highest 1P-2O yield was obtained when the reaction was conducted with a P/ PO ratio of 1.4 data not shown, although the influence of the substrate ratio on enzyme enantioselectivity was not investigated. The reactions were carried out at 650 rpm of agitation under optimum temperature of each enzyme and terminated by filtering enzymes. The filtrates were then evaporated to obtain the reaction mixture. 2.3 Separation of TAGs from reaction mixture Prior to conduct chiral recycle HPLC-mass spectrometry MS /MS analysis Section , the TAGs were separated from other lipid classes of lipids, such as free fatty acids and slight amount of diaclyglycerols DAGs and monoaclyglycerols MAGs. Separation of TAGs 0.50 g was conducted by Florisil column 10 mm i.d. 100 mm length chromatography, using 10 ml of hexane/diethyl ether 90 : 10, v/v to elute fraction 1 13 and 50 ml of chloroform to elute the final fraction 14. The purity of the TAG fractions was checked by thin layer chromatography-frame ionization detector TLC-FID analysis Section Fig. 1 Reaction scheme of lipase-mediated acidolysis between TO and P. This scheme shows 1P-2O generation. 208

3 Preparation of Chiral Triacylglycerols 2.4 Analysis of the TAGs Lipid composition The lipid compositions of the reaction mixture or Florisil column eluates were determined by TLC-FID analysis. The Iatroscan MK-6s and silica gel rod S-IV devices LSI Medience Corp., Tokyo, Japan were used, and a mixture of benzene/chloroform/acetic acid 35:15:1, v/v/v was used as the development solvent TAG molecular species composition: calculation of the 1P-2O yields The composition of TAG molecular species in the reaction mixtures were analyzed using HPLC system equipped with the reversed-phase column Crest Pak C18S mm; JASCO Corp., Tokyo, Japan, a refractive index detector RID-10A, Shimadzu Corp., Kyoto, Japan, and an integrator Chromatopac C-R8A, Shimadzu Corp.. The mobile phase was a mixture of acetone/acetonitrile 1 : 1, v/ v, and the flow rate was 0.8 ml/min. TAG molecular species were identified using the equivalent carbon number and theoretical carbon number of TAGs, as described in the literature 18, 19. The area ratios of chromatograms were defined as the yields of each TAG generated by the acidolysis reaction TAG positional isomer compositions in 1P-2O: reversed-phase recycle HPLC-MS/MS analysis Prior to chiral HPLC analysis of 1P-2O, asymmetric 1P-2O sn-oop sn-poo should be analyzed and fractionated by reversed-phase recycle HPLC on polymeric ODS column. The TAG fraction separated from reaction mixture Section 2. 3 was applied to this HPLC system. The recycle HPLC system was composed of a recycle pump PU712R; GL Sciences Inc., Tokyo, Japan, an autosampler GL-7420; GL Sciences Inc., a column oven CO705C; GL Sciences Inc., a UV-visible detector UV702, GL Sciences Inc., and two automatic valves VALVE UNIT 401; FLOM Co., Ltd., Tokyo, Japan, which passed the analytes through the same columns repeatedly, thus reducing the extra column volume of the autosampler during a recycle mode. An APCI-MS device Quattro micro API; Waters Corporation, Milford, MA was used as the detector in combination with the recycle HPLC system. Tandemjointed polymeric ODS columns Inertsil ODS-P, 4.6 mm i. d. 250 mm, 5 μm; GL Sciences Inc. were employed, and the column temperatures was set at 10. Acetonitrile/2- propanol/hexane 3:2:1, v/v/v was used as the mobile phase at a flow rate of 1.0 ml/min. The injection volume was 10 μl, and the wavelength of the UV-visible detector was 205 nm. The APCI-MS parameters were as follows: polarity, APCI positive; corona current, 3.0 μa; source temperature, 120 ; desolvation temperature, 450 ; cone gas flow, 50 L/h; desolvation gas flow, 200 L/h; cone voltage 40, V; collision energy, 25eV; and data acquisition mode, selected reaction monitoring SRM mode m/z POO NH 4 PO. sn-opo and rac-oop sn-oop/sn-poo 1/1 were dissolved in acetone, individually adjusted to 1,000 μg/ml, and used as standard solutions. A 10 μl aliquot of the standard solution was injected and the peaks of the TAG positional isomers ie., sn-opo/ sn-oop sn- POO, were sufficiently resolved by four passages through the columns. Thereafter, they were forced into the mass spectrometer through the recycle valve. TAG fraction separated from reaction mixture Section 2. 3 was also analyzed in the same way, where the ratio of the TAG positional isomers sn-opo/ sn-oop sn-poo was obtained through comparison with the peak area ratio of the standards. The asymmetric 1P-2O was fractionated at the same time Chiral TAG compositions in asymmetric 1P-2O: chiral recycle HPLC-MS/MS analysis The TAG fraction separated from reaction mixture Section was applied to the chiral recycle HPLC-MS/ MS system. The same HPLC-MS/MS system used in Section was used for this analysis except for HPLC conditions, which were as follows: a column: CHIRALCEL OD-3R and guard cartridge 3 μm, 4.6 mm 150 mm; and 3 μm, 4.0 mm 10 mm; Daicel Corp., Osaka, Japan ; column temperature, 25 C; flow rate, 0.5 ml/min; and the mobile phase, methanol. The SRM transitions for asymmetric 1P-2O were m/z POO H OO and POO H PO, respectively. In this analysis, sn-oop and sn-poo were detected in this order at 7th recycles, with the retention time of 179 and 183 min, respectively. sn-oop and sn-poo were defined from retention times of authentic compounds. The peak areas of sn-oop and sn-poo were used as their composition, because they are a pair of enantiomers with the same physical properties. 3 Results and Discussion 3.1 Separation of TAGs from the reaction mixture Table 1 shows the yield and purity of each TAG fraction obtained Florisil column chromatography of the reaction mixture. The TAGs started eluting from fraction 2, and none were obtained from fraction 7. Lipids obtained from fraction 14 contained unknown TAGs and other lipids, such as fatty acids, DAGs, and MAGs. Finally, using 0.50 g of reaction mixture, g 49.0 of TAGs were obtained from fraction Preparation of 1P-2O and chiral analysis with Lipozyme RM-IM, Lipozyme TL-IM, and Lipase OF Lipozyme RM-IM Figure 2 shows the compositions of TAGs obtained by Florisil column chromatography of the reaction mixture resulting from Lipozyme RM-IM catalysis. In order to prepare chiral TAGs sn-poo or sn-oop in good yield, the 209

4 Y. Yamamoto, H. Yoshida, T. Nagai et al. Table 1 Yield and TAG purity of each fraction obtained florisil column chromatography of 0.50 g of reaction mixture. Frction No Yield (g) TAG (%) Others (%) 64.1 Elution solvents: hexane-diethyl ether (90 : 10, v/v) for fraction 1~13 and chloroform for fraction 14. Ten milliliter of solvent for fraction 1~13 and 50 ml for fraction 14 was used. Fig. 2 Content of 1P-2O A and its chiral TAG compositions B generated by Lipozyme RM-IM -mediated acidolysis between TO and P. Reaction conditions: substrate molar ratio P/TO of 1.4, enzyme dose of 10 wt, hexane volume of 10 ml, agitation speed of 650 rpm, and reaction temperature of 60. yield of 1P-2O is important. Therefore, we first checked the content of 1P-2O in TAGs obtained by Florisil column chromatography of the reaction mixture. The 1P-2O content increased with reaction time and reached a plateau at 2 h Fig. 2 A, and the maximum content was obtained at 4 h With increasing 1P-2O production, the content of 2P-1O likewise increased, indicating acidolysis reaction was proceed properly. However, a longer reaction time 8 h and 24 h unexpectedly caused a non-specific reaction to produce tripalmitin TP. Given that the yield of TAGs from the reaction was 49.0 Section 3.1, the maximum amount of 1P-2O from 0.50 g of starting TO under this reaction condition was calculated as 0.12 g g 47.1 /100. Lipozyme RM-IM showed distinctive specificity for chiral recognition, as revealed by 1P-2O compositions of the chiral TAGs produced at each reaction time point monitored Fig. 2 B. At the reaction time of 0.25 h, the compositions of sn-poo and sn-oop were 89.7 and 10.3, respectively. The sn-poo/sn-oop ratio at that time was ca /10.3, indicating that the rate of sn-poo generation was higher than that of sn-oop. In other words, Lipozyme RM-IM showed preference for the sn-1 position of the TAG molecule in acidolysis reaction with TO and P. The sn-poo/sn-oop ratio reached closer to 1 with a longer reaction time sn-poo/sn-oop 1.3 at 24 h. Although the yield was very low e.g., the content of asymmetric 1P-2O was only 13.3 in 0.25 h of reaction, high purity of sn-poo was obtained in the short reaction time. Here, the absolute amount of sn-poo from 0.50 g of starting TO was calculated as 0.03 g g 13.3 / / 100. A small amount of sn-opo was generated at 24 h. It is possible that acyl migration from sn-poo or sn-oop to sn-opo had occurred to form the thermodynamically stable conformation symmetric conformation in the long reaction time at Lipozyme TL-IM Figure 3 shows the compositions of TAGs obtained by Florisil column chromatography of reaction mixture generated by Lipozyme TL-IM -mediated acidolysis between TO and P. The 1P-2O content increased with reaction time to reach a final maximum yield of 44.9 Fig. 3 A, which was almost the same as the yield with Lipozyme RM-IM Fig. 2 A. However, at 0.25 h, 0.5 h, and 1 h, the 1P-2O contents were 13.3, 21.8, and 37.7, respectively, in the reaction with Lipozyme RM-IM whereas they were 6.1, 19.0, and 24.9, respectively, when Lipozyme 210

5 Preparation of Chiral Triacylglycerols Fig. 3 Content of 1P-2O A and its chiral TAG compositions B generated by Lipozyme TL-IM -mediated acidolysis between TO and P. Reaction conditions: substrate molar ratio P/TO of 1.4, enzyme dose of 10 wt, hexane volume of 10 ml, agitation speed of 650 rpm, and reaction temperature of 60. Fig. 4 Content of 1P-2O A and its chiral TAG compositions B generated by Lipase OF -mediated acidolysis between TO and P. Reaction conditions: Reaction conditions: substrate molar ratio P/TO of 1.4, enzyme dose of 50 wt, hexane volume of 10 ml, agitation speed of 650 rpm, and reaction temperature of 37. TL-IM was used for the catalysis. These results showed that the rate of 1P-2O generation by Lipozyme TL-IM was slightly lower than that of Lipozyme RM-IM. The ratio of sn-poo/sn-oop in 1P-2O at the reaction time of 0.25 h was ca /34.0 Fig. 3 B, but it decreased slightly with longer reaction time, reacting ca. 1.5 at 24 h. This result was completely different from that obtained with Lipozyme RM-IM Fig. 2 B. It was suggested that chiral recognition was differed from the two lipases even having the same regiospecificity at sn-1, 3 position of TAG molecule. Even though Lipozyme TL-IM preferably recognized sn-1 positon of TAG molecule, its rate of sn-poo generation was lower than that by Lipozyme RM-IM Lipase OF Figure 4 shows the compositions of TAGs obtained by Florisil column chromatography of the reaction mixture generated by Lipase OF -mediated acidolysis between TO and P. In this case, the content of 1P-2O increased gradually with reaction time, and finally reached a maximum yield of 40.0 at 72 h Fig. 4 A. At the reaction time of 24 h, the 1P-2O content was 26.6, whereas that generated with Lipozyme RM-IM and Lipozyme TL-IM were 46.0 and 44.9, respectively Figs. 2 A and 3 A. This indicates that Lipase OF has a lower reaction rate for 1P-2O generation than Lipozyme RM-IM and Lipozyme TL-IM. 211

6 Y. Yamamoto, H. Yoshida, T. Nagai et al. In the Lipase OF -mediated reaction, the ratio of sn-poo/sn-oop in 1P-2O at the reaction time of 0.5 h was ca /39.7 Fig. 4 B, but it gradually reached closer to 1 with a longer reaction time. For example, the sn-poo/sn-oop ratio at 4 h was ca. 1.3, while that at 72 h was ca On the other hand, sn-opo was also generated at the first stage of the reaction, with a 4.6 yield even at 0.5 h Fig. 4 B. This can be explained by the fact that Lipase OF has no regiospecificity for TAGs and therefore possibly has no chiral specificity for the TAG molecule as well. 3.3 Effects of reaction solvents on 1 P-2O and chiral TAG preparation To examine the effects of reaction solvents on 1P-2O and chiral TAG preparation, the reaction was conducted with hexane being substituted by other solvents, such as acetone and benzene Fig. 5. A solvent-free system was also tested. For this experiment, Lipozyme TL-IM Figs. 5 and 6 and TO of low purity 73.1 Fig. 5 were used. The yields of 1P-2O were 41.1 by solvent-free, 40.1 by hexane, 25.5 by benzene, and 0 by acetone Fig. 5 A. Judging from the Snyder s polar parameter 19 value of 0.1, 2.7, and 5.1 for hexane, benzene, and acetone, respectively, these results indicate that polarity of reaction solvents affects the lipase activity for acidolysis reaction. Domingo et al. reported that, in lipase-catalyzed aza-michael addition reactions, the target compounds were successfully synthesized in a non-polar hexane medium, whereas aminolysis side reaction proceeded in polar solvents 20. In addition, water-miscible acetone might cause denaturation of the lipase by removing its essential water molecules. When 10 wt of enzyme was used, there were no difference in 1P-2O preparation between solvent-free and hexane systems Fig. 5 B. However, obvious differences appeared when 1 wt of enzyme was used. These results indicated that the reaction rate was higher in the solventfree system than in the hexane system. Several published reports have compared the reaction efficacies between solvent and solvent-free system For example, Sirirung et al. reported that, in the acidolysis reaction between 1,3-dicaprylin or 1,3-dilaurin and oleic acid, using lipase from Burkholderia cepacia Amano SP-D, the yields and reaction rates of TAG synthesis were considerably increased in solvent-free system under reduced pressure compare compared with the solvent hexane system 22. The reason for this, however, remains unclear. Nonetheless, a solvent-free system would mean a higher contact frequency between the enzyme and the substrate than that in the hexane system. Finally, to examine whether reaction solvents affect the chiral specificity of lipase for TAG molecule, the contents of asymmetric 1P-2O, sn-poo, and sn-oop in hexane and solvent-free system were compared Fig. 6. In hexane system, the sn-poo content increased with the reaction time, followed by sn-oop Fig. 6 A. At the final reaction time of 24 h, the 1P-2O yields of asymmetric 1P-2O, sn-poo, sn-oop, and sn-opo were 40.5, 24.2, 16.3, and 2.7, respectively. Under this reaction condi- Fig Effects of reaction solvents on 1P-2O preparation. Purity of TO used in this experiment was lower 73.1 than that of other experiments A Reactions were conducted using 10 ml of hexane, acetone, and benzene. Solventfree reaction was also conducted. Other reaction conditions; substrate molar ratio P/TO of 1.4, 10 wt of Lipozyme RM-IM, reaction time of 4 h, agitation speed of 650 rpm, and reaction temperature of 60. B Comparison of solvent-free or hexane system on yields of 1P-2O. Effects of enzyme Lipozyme TL-IM dose 10 wt and 1 wt was also compared. Other reaction conditions; substrate molar ratio of 1.4, agitation speed of 650 rpm, and reaction temperature of 60.

7 Preparation of Chiral Triacylglycerols Fig. 6 Lipozyme TL-IM -mediated acidolysis between TO and P in hexane A and solvent-free B systems. Enzyme doses were 10 wt for hexane system and 1 wt for solvent-free system. Other reaction conditions: substrate molar ratio P/TO of 1.4, agitation speed of 650 rpm, and reaction temperature of 60. tion, the maximum sn-poo/sn-oop ratio was ca /16.3. Likewise, in the solvent-free system, the sn-poo content increased with reaction time, followed by sn-oop Fig. 6 B. At the final reaction time of 72 h, the 1P-2O yields of sn-poo sn-oop, sn-poo, sn-oop, and sn-opo were 39.0, 23.7, 15.3, and 1.0, respectively. Because the content of asymmetric 1P-2O 39.0 was almost the same as that in the hexane system 40.5, the same sn-poo/sn-oop ratio of ca /16.3 was obtained. From this result, it was suggested that reaction solvent, at least hexane, has no effect on chiral specificity of lipase for TAG molecule in acidolysis reaction with TO and P. The solvent-free system has an advantage in that it can be applied in the food industry, since it bears no harmful or restricted solvents. TAGs with high sn-poo/sn-oop ratio of 9 under the following reaction conditions: substrate molar ratio P/TO of 1.4, enzyme dose of 10 wt, hexane volume of 10 ml, agitation speed of 650 rpm, reaction temperature of 60, and reaction time of 0.25 h, although the 1P-2O yield was only In addition, the solvent-free system seemed to be more useful, giving higher reaction rate than the hexane system. Furthermore, the reaction solvent used had no effect on the chiral specificity of lipase for TAG molecules in this acidolysis reaction. Overall, these results suggest that Lipozyme RM-IM is the most useful lipase for preparing chiral TAGs by the acidolysis reaction with TO and P in a solvent-free system, especially in the food industry. Further studies are under the way to improve the 1P-2O yield while maintaining the highest sn-poo/sn-oop ratio of 9. 4 Conclusion In this study, the chiral TAGs, sn-poo and sn-oop, were prepared using three commercially available lipases Lipozyme RM-IM, Lipozyme TL-IM, and Lipase OF in an acidolysis reaction between TO and P. To the best of our knowledge, this is the first report concerning the enzymatic preparation of chiral TAGs. Lipozyme RM-IM and Lipozyme TL-IM, having regiospecificity for the sn-1 and sn-3 positions of TAG molecules, showed superior activity in 1P-2O preparation over that of Lipase OF, which has no regiospecificity to TAG molecule. Lipozyme RM-IM, in particular, produced chiral References 1 Chakra, W.; Claudio, C.; Peter, W.; Peter, F.; Benjamin, F.; Neeranat, T.; Patrick, P. Docosahexaenoic acid is more stable to oxidation when located at the sn-2 position of triacylglycerol compared to sn-1 3. J. Am. Oil Chem. Soc. 85, Yamamoto, Y.; Imori, Y.; Hara, S. Oxidation behavior of triacylglycerol containing conjugated linolenic acids in sn-1 3 or sn-2 position. J. Oleo Sci. 63, Dote, S.; Yamamoto, Y.; Hara, S. Effects of triacylglycerol molecular species on the oxidation behavior of 213

8 Y. Yamamoto, H. Yoshida, T. Nagai et al. oils containing α-linolenic acid. J. Oleo Sci. 65, R. Jhon, C.; Robert. W.L. Crystallization, polymorphism, and binary phase behavior of model enantiopure and racemic triacylglycerols. Cryst. Growth Des. 11, Mizobe. H.; Tanaka, T.; Hatakeyama, N.; Nagai, T.; Ichioka, K.; Hondoh, H.; Ueno, S.; Sato, K. Structures and binary mixing characteristics of enantiomers of 1-oleoyl-2,3-dipalmitoyl-sn-glycerol S-OPP and 1,2-dipalmitoyl-3-oleoyl-sn-glycerol R-PPO. J. Am. Oil Chem. Soc. 90, Utsugi, A.; Kanda, A.; Hara, S. Lipase specificity in the transacylation of triacylglycerin. J. Oleo Sci. 58, Kanda, A.; Namiki, F.; Hara, S. Enzymatic preparation of structured oils containing short-chain fatty acids. J. Oleo Sci. 59, Mounika, C.; Reddy. S.Y. Specialty fats enriched with behenic and medium chain fatty acids from palm stearin by lipase acidolysis. J. Am. Oil Chem. Soc. 89, Kotani. K.; Yamamoto, Y.; Hara. S. Enzymatic preparation of human milk fat substitutes and their oxidation stability. J. Oleo Sci. 64, Gotor-Fernandez, V.; Brieva, R.; Gotor, V. Lipases: Useful biocatalysts for the preparation of pharmaceuticals. J. Mol. Catal. B 40, Melo-Carvlho, A.C.L.; Sousa-Fonseca, T.; Mattos, M.C.; Oliveira, M.C.F.; Lemos, T.L.G.; Molinari, F.; Romano, D.; Serra, I. Recent advances in lipase-mediated preparation of pharmaceuticals and their intermediates. Int. J. Mol. Sci. 16, Anand, N.; Kapoor, M.; Ahmad, K.; Koul, S.; Parshad, R.; Manhas, K. S.; Sharma, L.R.; Qazi, G.N.; Tenja, S.C. Arthrobacter sp.: A lipase of choice for the kinetic resolution of racemic arylazetidinone precursors of taxanoid side chains. Tetrahedron 18, Marszaƚƚ, M.P.; Siódmiak, T. Immobilazation of Candida rugosa lipase onto magnetic beads for kinetic resolution of R, S -ibuprofen. Catal. Commun. 24, Staggers, J.E.; Fernando-Warnakulasuriya, G.J.; Wells, M.A. Studies on fat digestion, absorption, and transport in the suckling rat. II. Triacylglycerols: molecular species, stereospecific analysis, and specificity of hydrolysis by lingual lipase. J. Lipid Res. 22, The basic and applied science of fats, oils and other lipids. Japan oil chemists society, 2th ed. in Japanese. 16 Ewa, R.; Claire, C.; Francine, F.; Rovert, V. Stereoseletive hydrolysis of triglycerides by animal and microbial lipases. Chirality 5, Nagai, T.; Mizobe, H.; Otake, I.; Ichioka, K.; Kojima, K.; Matsumoto.; Gotoh, N.; Kurada, I.; Wada, S. Enantiomeric separation of asymmetric triacylglycerol by recycle high-performance liquid chromatography with chiral column. J. Chromatogr. A 1218, Plattner, R.D.; Spencer, G.G.; Kleiman, R. Triglyceride separation by reverse phase high performance liquid chromatography. J. Am. Oil Chem. Soc. 54, El-Hamdy, A.H.; Perkins, E.G. High performance reversed phase chromatography of natural triglyceride mixtures: critical pair separation. Am. Oil Chem. Soc. 58, Snyder, L.R. Classification of the solvent properties of common liquids. J. Chromatogr. A 92, José Domingo, R-R.; Jaime, E.; Agustín L-M.; Alain, M.; Edmundo, C. Thermodynamically controlled chemoselectivity in lipase-catalyzed aza-michael additions. J. Mol. Catal. B 112, Sirirung, W.; Aran, H-K.; Uwe, T.B. Lipase-catalyzed synthesis of structured triacylglycerol from 1,3-diacylglycerides. J. Am. Oil Chem. Soc. 81, Dejan, B.; Dusan, M.; Slavica, S-M.; Zorica, K. The Candida rugosa lipase catalyzed synthesis of amyl isobutyrate in organic solvent and solvent-free system: A kinetic study. J. Mol. Catal. B 38, Wei, W.; Feng, Y.; Zhan, Z.; Cao, X.; Feng, F. Synthesis of structured lipid 1,3-dioleoyl-2-palmitoylglycerol in both solvent and solvent-free system. LWT-Food Sci Technol, 60,