Edited by: Joseph Boudrant, CNRS, France
Reviewed by: Ana M. R. B. Xavier, Universidade de Aveiro, Portugal; Alicia Estevez Garcia, Institut de Recerca i Technologia Agroalimentàries, Spain; Yann Guiavarc’H, University of Lorraine, France
Specialty section: This article was submitted to Process and Industrial Biotechnology, a section of the journal Frontiers in Bioengineering and Biotechnology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Enzymatic glycerolysis of Echium oil (
In recent years, partial glycerides, such as diacylglycerols (DAG) or monoacylglycerols (MAGs), are achieving increasing popularity from different perspectives. On the one hand, the most popular function of these lipids is related to their amphiphilic nature and surface-active properties, being well known as emulsifier ingredients in the food, pharmaceutical, and cosmetic industries. On the other hand, partial glycerides, such as self-emulsifying lipid delivery systems, are attractive lipids in the formulation of potential vehicles of drugs and bioactive compounds of poor solubilization, in order to favor their bioaccessibility at intestinal level or to protect against their degradation under gastrointestinal (GI) conditions, and ultimately to reach the most efficient bioactivity of compounds (Martin et al.,
Oil represents one of the most important excipients in emulsifying formulations because it can solubilize marked amounts of the lipophilic drug, facilitate emulsification, and increase transport and absorption of lipophilic drug via GI tract. Molecular nature of the triglyceride plays an important role, and surprisingly, very few lipid-based formulations have reached the pharmaceutical market place (Gupta et al.,
One of the tools available for the production of mono- and diacylglycerol mixtures is solvent-free glycerolysis, which is a simple process and easily scalable. In this procedure, there is no production of free fatty acids (FFAs) that otherwise, should be removed at the end of the process. No organic solvents are utilized which avoids time-consuming, costly solvent recovery and increases the acceptability of the final product for food applications and for the development of emulsifying systems (Kristensen et al.,
Several glycerolysis systems utilizing or not organic solvents, with immobilized or non-immobilized enzymes, and in microemulsion or other media have been reported. A common strategy called solid-phase reaction consists of carrying out the glycerolysis reaction below the critical temperature (
Alternatively, incorporation of surfactants may improve homogeneity, increase the interfacial area, provide higher reaction rates, and enhance the efficiency of the biocatalyst for the system oil/glycerol/enzyme. This is especially relevant for lipases which are characterized as enzymes that act at the interface. The characteristic of a surfactant is the formation of micellar systems that (1) increases enzymatic stability at the process conditions, (2) reduces mass transfer limitations, and (3) improves reaction conversion due to the increased solubility among substrates (Valério et al.,
It is well known that glycerol has a negative impact on lipase activity and stability by being adsorbed onto the support of the immobilized lipases reducing the diffusion of the hydrophobic substrate to the active site of the lipase (Fureby et al.,
The ratio of 1,3- vs. 1,2-diacylglycerols is nutritionally important, and it has also implications in the product composition of glycerolysis reactions. It has been reported that non-specific lipases can remove fatty acid residues from sn-2 position of the triacylglycerol molecules forming 1,3-diacylglycerols. We also believe that other undesirable secondary reactions besides glycerolysis can take place to produce 1,3-diacylglycerols (Scheme
The addition of organic solvents to the reaction mixture is one strategy to improve the system homogeneity and stability as well as to reduce viscosity and mass transfer limitations. However, even in glycerolysis reactions, organic solvent-assisted phase split behavior has been described (Damstrup et al.,
In summary, considering the vast information found in the literature on glycerolysis reactions (Kaewthong et al.,
Echium oil was acquired to Harke Nutrition (Mülheim and der Ruhr, Germany). The fatty acid profile of the oil was comprised of palmitic acid (7%), stearic acid (4%), oleic acid (15%), linoleic acid (15%), gamma-linolenic acid (12%), alpha-linolenic acid (33%), and stearidonic acid (14%). The balance of the oil (1%) consisted of other minor fatty acids. Monoolein (99%), diolein (99%), triolein (99%), oleic acid (C18:1) (99%), and limonene were purchased from Sigma-Aldrich (MO, USA). All solvents utilized were HPLC grade acquired to Lab Scan (Giliwice, Poland). Novozym® 435 and Lipozyme® RM IM were a gift from Novozymes A/S (Bagsvaerd, Denmark). Lipase PLG (from
Ten grams of reaction mixture containing approximately 9 g of Echium oil and 1 g of glycerol were incubated in an orbital shaker (IKA KS 4000, Staufen, Germany) at 40°C and 200 rpm for 48 h. Echium oil and glycerol were preincubated for 10 min prior enzyme [10% (w/w) of the total reaction mixture] was incorporated to start the reaction.
At different times, aliquots of 50 μL were taken from the reaction mixture, dissolved in 5 mL of methyl tert-butyl ether (MTBE), and filtered through a syringe filter of 0.45 μm of pore size to remove the biocatalyst since all biocatalysts tested had a particle size higher than 1 μm. Aliquots from these solutions were taken and evaporated under nitrogen to obtain a constant weight residue that was redissolved in MTBE up to a final concentration of 6 mg/mL. Approximately 0.2 μL of this final transparent solution was analyzed by gas chromatography.
Solvent-assisted glycerolysis was carried out utilizing the same amount of Echium oil and glycerol utilized in solvent-free reactions but by incorporating ca. 8 g of food grade limonene.
Approximately 20 g of reaction mixtures with identical ratio of Echium oil to glycerol to that described in glycerolysis reactions section, but containing 2% w/w of food grade monoolein, were prepared and homogenized by using an Emulsiflex C5 from Avestin Europe GmbH (Weinheimer, Mannheim, Germany). Five passes at approximately 500 bar were performed with each of the mixtures tested.
Separations were performed on a Hewlett-Packard 5890 series II gas chromatograph with on-column injection using a 7-m HP-5MS capillary column, 0.25 mm I.D. (Agilent Technologies, Santa Clara, CA, USA). An injector and detector temperatures of 40 and 340°C, respectively, were utilized. The temperature program was as follows: starting at 40°C and then heating to 250°C at 42°C min−1 with 15 min hold, followed by heating from 250 to 325°C at 15°C min−1 with 20 min hold. Calibration curves for MAGs, diacylglycerols, triacylglycerols, and FFAs were carried out with monoolein (99%), diolein (99%), triolein (99%), and oleic acid (C18:1) (99%), from Sigma-Aldrich (MO, USA). The peaks were computed using GC chemstation software.
Glycerolysis reactions were also carried out in a 1-L stainless steel reactor coupled to a paddle stirrer at 200 rpm (Kiloclave, Buchi Glass Uster, Switzerland) in the presence of Novozym 435 and PLG. Three consecutive trials reutilizing the same batch of each lipase were performed. The reaction mixture consisted of 500–600 g of Echium oil and ca. 10% of glycerol (based on the total weight of oil). After 10 min of preincubation at 40°C, 10% (w/w) of each lipase was incorporated to the reaction mixture. Both immobilized lipases were recovered at the end of the glycerolysis reaction by passing the product mixture and the biocatalyst through a filtering device with a 100-μm mesh to retain the enzyme inside the filter. This device was connected to the discharge valve of the reactor and the mixture was forced to pass through the filter with the help of nitrogen pressure (1–2 bar). The fluctuations on the amount of Echium oil utilized were based on the weight of the solid material recovered after filtering the product mixture. This solid material contained the recovered lipase impregnated of the product mixture, after each trial. As an example, more than 200 g of Novozym 435 impregnated in the product mixture from the first trial were recovered, and for this reason only, 500 g of Echium oil were utilized in the second trial with the mentioned impregnated Novozym 435. After 24 h, the reactions were finalized and the mixture was forced to pass through a 100-μm stainless steel filter with the help of 1–2 bar of nitrogen gas to separate the enzyme from the oily mixture.
The molecular distiller pilot plant used in this project was acquired from POPE Scientific Inc. (Saukville, WI, USA). An external condenser and a cryogenic trap was installed immediately downstream of the still. The condensable, low-molecular-weight compounds are collected in the cryogenic trap upstream of vacuum system. Molecular distillation was carried out at 220°C and pressure of ca. 0.005 bar at a flow rate of ca. 250 mL/h. The methodology utilized with slight modifications has been utilized for the production of distilled MAGs (Fischer,
It is well known that the solubility of glycerol in the triglyceride at the reaction temperature is the determining factor for the yield of MAGs. Sonntag in 1982 reported a high yield of MAG in the glycerolysis reactions in different systems ranging from fine dispersion to superemulsion. Glycerol particle size of 10–0.05 μm and smaller, yields of 61–95% for MAGs. In all the cases examined, the reaction systems are still heterogeneous, although at superemulsion range, homogeneity is almost reached. An interesting conclusion is that besides high temperature and solubility, a high degree of contact may also be a feasible path for a high yield of MAGs. Therefore, a high shear liquid–liquid device besides to the solubility effects as a consequence of an increase of process temperature is expected to have a positive effect on glycerolysis reactions. Consequently, faster reactions at lower temperatures are feasible. Moreover, some of the undesirable taste and color due to the decomposition of triglycerides at high temperatures may be eliminated (Noureddini,
The previously mentioned three different glycerolysis reactions were carried out in the presence of Novozym 435. The results are shown in Figure
Similarly to the glycerolysis reaction in the presence of Novozym 435, three different glycerolysis reactions were also performed in the presence of lipase RM IM (Figure
Similarly to the two previous lipases, in the presence of PLG lipase, regardless of the reaction mixture assayed, TAG and glycerol conversions were ca. 60–70 and 50%, respectively (Figure
Regardless of the reaction mixture utilized, TAG conversion in the presence of
It should be also noted that the highest reaction rates were observed with this lipase. This result could be attributed to the absence of carrier in this commercial lipase that reduce particle size, increase the surface area, and reduce diffusion and mass transfer limitations. However, considering the high FFA content in the product mixture, glycerolysis in the presence of Lipase SL was not further investigated because it requires complicated purification steps to remove the released FFA from the acylglycerol mixture.
It can be concluded that high pressure homogenization or the addition of 2% of monoolein to the reaction mixtures did not offer any significant advantage over the glycerolysis reaction. On the contrary, in some cases, after homogenization, lower conversions were even observed that could be attributed to different reasons: glycerol particle size distribution, partial enzyme coating, slightly different degree of hydrolysis coupled to the glycerolysis reaction, enzyme carrier polarity, etc. Solvent-free glycerolysis takes place in a very heterogeneous reaction media, and it is affected by numerous factors that influence enzyme kinetics and reproducibility.
As it has been already previously emphasized, one of the most important factors in glycerolysis is the solubility and the surface contact between oil and glycerol. In this sense, in chemical glycerolysis, the catalyst form soaps, which promotes the reaction acting as emulsifiers. Similarly, the high temperature reduces mass transfer of the triglyceride to the glycerol phase, increases the mutual solubility of fat and glycerol phases, and conducts to a faster rate of reaction. However, decomposition of some fatty acids at high temperatures prevents utilization of temperatures above 250°C. Unfortunately, this strategy is inadequate for oils rich in polyunsaturated fatty acids because of the thermal instability caused by double bonds in the carbon chain (Noureddini,
Based on the triacylglycerol conversion, production of mono- and diacylglycerols, and hydrolysis level of the glycerolysis reactions, we chose Novozym 435 and PLG lipases without any pretreatment for further experiments.
One important issue in this type of reaction is the stirrer utilized in the process. In our opinion, mechanical stirrers should produce a much more homogeneous reaction mixture and hence, partially overcome some of the limitations pointed out in the present study. However, care should be taken with some of these mechanical stirrers, particularly magnetic stirrers, at lab scale. These devices frequently produce mechanical abrasion of the biocatalyst reducing its particle size, which change the catalyst properties, reduce reusability and half-life of the enzyme. Moreover, the results attained with magnetic stirrers are not always readily scalable. For these reasons, we have chosen an orbital shaker incubator despite the limitations and fluctuations described above.
In order to precipitate MAGs in the reaction mixture and shift the reaction equilibrium toward the formation of more products, glycerolysis reactions were carried out in two stages in the presence of Novozym 435 and PLG lipases. The first stage of 24 h at 40°C was followed by a second stage at 11°C. The results shown in Figure
The enzymatic glycerolysis carried out in limonene is shown in Figure
Three consecutive trials with Novozym 435 and with PLG reusing the same batch of enzyme were carried out. The results are shown in Table
Biocatalyst | Trial number | Time (h) | MAG% (w/w) | DAG% (w/w) | TAG% (w/w) |
---|---|---|---|---|---|
PLG | 1 | 24 | 22.1 | 29.5 | 47.7 |
PLG | 2 | 24 | 11.6 | 30.2 | 58.0 |
PLG | 3 | 24 | 4.1 | 25.8 | 66.7 |
Novozym 435 | 1 | 24 | 27.2 | 29.4 | 41.9 |
Novozym 435 | 2 | 24 | 27.5 | 30.7 | 41.0 |
Novozym 435 | 3 | 24 | 27.2 | 31.1 | 38.5 |
The first trial with Novozym 435 produced slightly higher level of MAGs compared with the results at lab scale, which indicates that mechanical stirring can have a positive effect in glycerolysis compared to orbital shaking at lab scale. On the contrary, worse MAG production than that attained at lab scale was observed in the first trial with PLG. As it was mentioned before, partial coating of the enzyme could be also responsible of these differences.
Surprisingly, the second trial reutilizing the same batch of PLG produced much worse triacylglycerol conversion and percentage of MAGs than those obtained in the first trial. Physical appearance of the enzyme after the second trial was very different from the first trial. Enzyme aggregates of several millimeters could be observed in the recovered lipase, which could be responsible of the worse results obtained. To overcome this problem, the immobilized enzyme was washed twice with ethanol 95% and dried by vacuum filtering to breakdown the formed aggregates and to recover the batch of enzyme in a state similar to that of the first trial. With this washed and dried lipase PLG, a third trial was carried out. Unfortunately, as soon as the enzyme got into contact with the reaction mixture, similar aggregates to those observed in the second trial were formed. The results of the third trial indicate even worse triacylglycerol conversion and MAG production than those reached in the second trial that could be attributed to a more severe coating and also to partial inactivation of the biocatalyst after the washing and drying treatment. On the contrary, similar triacylglycerol conversion and MAG production was attained in the presence of Novozym 435 in the three consecutive trials studied. For this reason, this biocatalyst was chosen as the most suitable for the solvent-free glycerolysis of Echium oil.
Approximately 1.5 kg of the product mixture attained in the pilot plant glycerolysis trials described before was utilized for fractionation via molecular distillation. The feed material and the composition of the two products obtained after fractionation are depicted in Table
Feed material | Residue | Distillate | |
---|---|---|---|
Weight (g) | 1543 | 1077 | 390 |
FFA (%) | 1.0 | 0.0 | 5.2 |
MAG (%) | 24.3 | 2.1 | 79.7 |
DAG (%) | 28.3 | 47.8 | 11.1 |
TAG (%) | 41.0 | 49.1 | 6.0 |
Solvent-free glycerolysis of Echium oil produced an acylglycerol mixture comprised of ca. 25% of MAGs, 30–40% of diacylglycerols and ca. 40% of triacylglycerols in the presence of PLG and Novozym 435 lipases, regardless the pretreatment performed on the mixture. Neither the high pressure homogenization, nor the addition of monoolein improved the results of the enzymatic glycerolysis. The reduction of the temperature transformed the mixture into a semisolid product reducing significantly the reaction rate. Slightly improvement of the MAG content was attained in the reactions carried out at 11°C. However, the long reaction times and the difficulties to recover the enzyme from the product mixture ruled out the utilization of this methodology. Lower conversion of the oil was obtained in the solvent assisted glycerolysis reaction. Scale-up of the solvent-free glycerolysis with PLG lipase revealed apparition of enzyme aggregates when the biocatalyst is reutilized even after ethanol washing of the enzyme to remove the product mixture from the biocatalyst. On the contrary, similar results of the reaction were obtained in the presence of Novozym 435 after three consecutive trials reutilizing the same batch of enzyme. Finally, molecular distillation produced two main fractions: a distillate is comprised of 80% of MAGs and a residue containing ca. 50% of both diacylglycerols and triacylglycerols. These products and fractions are intended to be used as bioactive acylglycerol mixture with emulsifying properties and also as precursors of other bioactive structured lipids.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This work was supported by the Ministerio de Economía y Competitividad (INNSAOLI, project number IPT-2011-1248-060000, subprograma INNPACTO) and the Comunidad de Madrid (ALIBIRD, project number S2013/ABI-2728).