WO2012087153A1

WO2012087153A1 - Enrichment of marine oils with omega-3 polyunsaturated fatty acids by lipase-catalysed hydrolysis

Info

Publication number: WO2012087153A1
Application number: PCT/NO2011/000354
Authority: WO
Inventors: Kahveci DERYA; Xu XUEBING
Original assignee: Marine Bioproducts AS
Current assignee: Marine Bioproducts AS
Priority date: 2010-12-23
Filing date: 2011-12-22
Publication date: 2012-06-28
Anticipated expiration: 2013-06-23

Abstract

The present invention relates to an enzymatic method for enrichment of the omega-3 PUFA content in the acylglycerol fraction of marine oils suitable for large scale application as well as to marine oils enriched with omega-3 polyunsaturated fatty acids and their use.

Description

ENRICHMENT OF MARINE OILS WITH OMEGA-3 POLYUNSATURATED FATTY

ACIDS BY LIPASE-CATALYSED HYDROLYSIS

FIELD OF THE INVENTION

The present invention relates to a method for enrichment of omega-3 polyunsaturated fatty acids comprised in acylglycerols of an oil by lipase-catalysed hydrolysis as well as to a processed marine oil and the use of the processed marine oil.

BACKGROUND TO THE INVENTION

Marine oils are generally recognized as being beneficial and/or preventive to human and animal health due their high content of polyunsaturated fatty acids (hereinafter "PUFA"), especially the omega-3 fatty acids eicosapentaenoic acid (20:5 n-3; EPA) and docosahexaenoic acid (22:6 n-3; DHA). Positive effects of omega-3 fatty acids have been reported for numerous conditions such as cardiovascular diseases, atherosclerosis, several types of cancer, dyslipidemia, hypertension, diabetes, obesity, inflammatory diseases, neurological/neuropsychiatric disorders, asthma and rheumatoid arthritis.

Moreover, current knowledge clearly states that EPA and DHA have different roles in their overall effect on human and animal health. DHA, being a main component of nervous tissue, is vital for growth and development of infants and developing animals in general, especially for their vision and cognitive functions. Beside other effects feeding of DHA has been shown to result in a significant reduction in blood pressure and heart rate. The same was not shown for EPA. EPA, on the other hand, being quantitatively the main omega-3 PUFA compound, has anti-inflammatory effects due to its regulatory role in gene expression. A recent meta-analysis revealed that significant improvement of mood in patients with depression was obtained

exclusively by EPA.

Thus, not only the content of omega-3 fatty acids in oils is of importance for its later value and application but also the quantitative composition of the comprised omega- 3 fatty acids as well as their ratio i.e. EPA/DHA. Oils comprising omega-3 PUFA are frequently extracted from wild caught marine resources such as from fish and krill. However, marine oils can also be obtained from farmed marine organisms such as farmed fish. In the recent years, the use of plant oils of terrestrial origin substantially not comprising any omega-3 fatty acids has increased in feed for aquaculture production, especially for Salmon ids such as the Atlantic salmon. Since salmon is dependent on receiving these fatty acids with their diet, fish fed a low amount of omega-3 fatty acids will also store a reduced content of these fatty acids in their lipids.

Oils obtained from these fish will consequently have a lower omega-3 fatty acid content, which reduces the value of these oils for applications such as for human and animal consumption, pharmaceutical compositions, functional feed and health products. This is one major concern of the fish farming industry when making efficiently use of those oils from farmed fish. Application of oils obtained from farmed animals are preferred on the other hand due to their in general low degree of contamination with environmental pollutants and their stable quality compared to oils obtained from wild catches.

Efforts have been made to improve the content of omega-3 polyunsaturated fatty acids (PUFA) to overcome the disadvantages mentioned above. The available methods for concentrating/enrichment of PUFA include adsorption chromatography, fractional or molecular distillation, enzymatic splitting, low-temperature crystallization, supercritical fluid extraction and urea complexation. Only few of these methods are suitable for large-scale processes and each of these techniques is recognised for its own advantages and draw backs. Today, there is still a challenge to develop cost- effective methods for the production of PUFA-concentrates in order to meet the growing demand for omega-3 fatty acids and to make use of low concentrated raw materials as from farmed fish. Enzymatic processes for concentrating omega-3 fatty acids were previously recognised as particularly advantageous when handling PUFA, since these fatty acids are highly sensitive to oxidation. Enzymatic methods allow the application of mild reaction conditions, meaning lower temperature and pressure, which is important when dealing with omega-3 fatty acids. Low temperature also improves the feasibility of the process. Furthermore, enzymatic processes are considered as more environmentally friendly compared to chemical ones.

Along with alcoholysis, lipase-catalysed hydrolysis is one of the most widely used enzymatic reaction, for the purpose of improvement of the omega-3 concentration of fish oil. By the lipase catalysed hydrolysis the amount of omega-3 fatty acids in the acylglycerols is increased (enriched) in relation to the total amount of fatty acids comprised in the acylglycerols.

The key property of the lipase-catalysed process lies in the fatty acid (FA) selectivity of lipases, since most of them discriminate against PUFA and preferably hydrolyse saturated and monounsaturated fatty acids (SFA and MUFA) bound in the

acylglycerols. Saturated and monounsaturated fatty acids are thus released as free fatty acids during the lipase-catalysed hydrolysis. Thereby the amount of PUFA fatty acids increases in the acyl glycerol fraction in respect to the total amount of fatty acids present in the acyl glycerol fraction. The enzymatic process is commonly followed by a separation process such as membrane filtration or molecular complexation whereby the hydrolysed free fatty acids are removed and an oil is obtained which has an acylglycerols fraction with an increased/enriched content of omega-3 fatty acids.

Lipase-hydrolysed concentrating of oils are described in the prior art. However, the efficiency of the disclosed processes and thus their applicability in industrial scale is often not satisfactory. The selectivity of the lipases decreases typically during the process since the availability of the SFA and MUFA eventually decreases, leading to a loss of PUFA e.g. of EPA. Carvalho et al. (Enzymatic hydrolysis of Salmon Oil by native lipases: Optimization of process parameters, J. Braz. Chem. Soc, 2009, Vol 20, No. 1 , page 117-124) discloses a lipase catalysed enrichment of DHA and EPA in salmon oil with a reaction time of 24 hours using the microbial lipases Asperigillus niger, Rhizopus javanicus and Penicillium solitium, while WO 2009/040676 discloses a method for increasing the concentration of PUFA in oils by use of the lipase Thermomyces lanuginosus. Yadwad et al. (Application of lipase to concentrate the docosa- hexaenoic acis (DHA) fraction of fish oil. Biotechnology and bioingeneering. 1991. Void 38, page 956-959) describes the use a lipase from Rhizopus niveus for increasing the amount of DHA in fish oils.

Some lipases can also discriminate between EPA and DHA, allowing to produce omega-3 PUFA concentrates with dominance of DHA or EPA. The lipase from Candida rugosa (also referred to as Candida cylindracea) has been recognized as a lipase which can discriminate between different omega-3 fatty acids. Thus, Candida rugosa is one of the enzymes which has been used in concentrating of

polyunsaturated fatty acids in triglycerides. Both, Okada and Morissey (Production of n-3 polyunsaturated fatty acid concentrate from sardine oil by lipase catalysed hydrolysis. Food Chemistry, 2007, Vol. 103, side 141 1 -1419) and Mbatia et al.

(Enzymatic enrichment of omega-3 polyunsaturated fatty acids in Nile perch (Lates niloticus) viscera oil. Eur. J. Lipid Sci. Technolo. 2010, Vol 1 12, page 997-984) disclose a method for concentrating n-3 polyunsaturated fatty acids in marine oil by lipase catalyses hydrolysis with different microbial lipases also including Candida rugosa. JP07051075 suggests two hydrolysis steps applying two different lipases having different characteristics. In the first step a lipase of Candida rugosa is used to concentrate the amount of DHA in the triglyceride fraction. In a second hydrolysis step the obtained glyceride mixture is hydrolysed by a lipase from a Penicillium- derived lipase, that does not hydrolyse triglycerides to obtain a glycerides fraction with high DHA amounts. Even though a large amount of work has been conducted in 80s and 90s of the last century, a simple, selective and efficient system to concentrate omega-3 fatty acids in fish oil with low content of PUFA is still in high pursue, especially with all the push to enhance the commercial omega-3 fatty acid supply. Thus, there is a need for more efficient processes in order to obtain the highest possible PUFA content in the fish product.

Moreover, known methods used to concentrate oils comprising omega-3 mainly focussed on an increase of the total omega-3 PUFA content. Methods using lipases in the concentration process do generally not discriminate between different omega- 3 fatty acids such as EPA and DHA.

At present simple, mild and at the same time, highly selective as wells as cost- and time-efficient methods applicable in industrial scale are thus lacking for this purpose.

As a consequence of the problem addressed above, the aim of the present invention is to provide a simple, mild, efficient and at the same time highly selective method for the improvement of fish oils by lipase- catalysed hydrolysis, which can be used in an industrial large scale in order to obtain omega-3 PUFA at higher yield and purity at lower cost than the existing methods. Furthermore, the invention also aims at providing a method for the concentration of omega-3 PUFA in oils wherein the content in EPA and DHA is balanced in the final product.

SUMMARY OF THE INVENTION

According to a first aspect, the present invention concerns a method for enrichment of the amount of omega-3 polyunsaturated fatty acids comprised in acylglycerols of an oil by lipase-catalysed hydrolysis, wherein the oil is mixed with an aqueous solution in a water to oil ratio of 2:1 to 5:1

(v/w),

a lipase, which selectively hydrolyses saturated and monounsaturated fatty acids bound to said acylglycerols and discriminates in its hydrolytic activity against omega-3 polyunsaturated fatty acids bound to said acylglycerols, is added to the oil- water mixture in a concentration of 3-5.5 % by weight based on the amount of oil.

Preferably, the free fatty acid are separated from the acylglycerol fraction after the hydrolysis, and the lipase-catalysed hydrolysis of the separated acylglycerol fraction is repeated. More preferably, the same enzyme is used in the first and second hydrolysis step of the repeated hydrolysis.

It is further preferred that the lipase is a microbial lipase, preferably from a species selected from the group consisting of Pseudomonas, Candida, Rhizopus, and Rhizomucor.

More preferred, the lipase is selected from the group consisting of Candida rugosa, Burkholderia cepacia, and Rhizopus oryzae.

In a preferred embodiment of the present invention, the incubation temperature in the lipase-catalysed hydrolysis is in the range of about 30-50°C, more preferably of about 40 to 50 °C, and most preferably of about 45 °C.

It is also preferred that the incubation time in the lipase-catalysed hydrolysis is between 2 hours and 8 hours, more preferably between 2 and 4 hours.

It is also preferred that said omega-3 polyunsaturated fatty acids are chosen from DHA, EPA and/or DPA.

Its is preferred that said lipase discriminates in its activity between EPA and DHA. In a preferred method the hydrolysis conditions are as follows

- the water-to-oil ratio is in the range of 2:1 to 5:1 (v/w),

- the incubation temperature is in the range of about 30-50°C, and

- the incubation time is between 2 hours and 4 hours. More preferred the lipase is added in a concentration of about 3 % by weight based on the amount of oil, and the hydrolysis conditions are as follows

- the incubation temperature is about 45°C,

- the water-to-oil ratio is about 3:1 (v/w),

- the incubation time is between 2 and 4 hours. In another preferred embodiment the lipase is added in a concentration of about 3-4 % by weight based on the amount of oil, and the hydrolysis conditions are as follows

- the incubation temperature is about 30°C,

- the water-to-oil ratio is in the range of about 2.5-5:1 (v/w),

- the incubation time is about 4 hours. It is also preferred that said oil to be enriched with omega-3 polyunsaturated fatty acids which are comprised in the acylglycerols is a marine oil obtained from a fish, a crustacean, a bacterium, a macroalgae and/or a microalgae.

Preferably, the oil is obtained from one or several fish species selected from the group consisting of Salmonids, Gadoids, Clupeids, Engraulidae, Scromboids, and Elasmobranchs, which can be wild caught or farmed fish.

Is further preferred that said oil is from a farmed marine animal comprising a reduced amount of omega-3 polyunsaturated fatty acids compared to an oil obtained from the same species in the wild.

Preferably, the oil has a polyunsaturated fatty acid content of no more than 16 % mol, preferably of no more than 14 % mol, at the start of the enrichment process with polyunsaturated omega-3 fatty acids. In a preferred embodiment the acylglycerol fraction comprising the enriched omega-3 polyunsaturated fatty acids is separated from the free fatty acids after the enzymatic hydrolysis, preferably by use of short path distillation.

More preferred the separation comprises the following steps:

- separation of the oil and water phase by centrifugation, and

- separating of the oil phase comprising the concentrated omega-3 polyunsaturated fatty acids from the free fatty acids by short path distillation.

Particular preferred is that the oil phase is fed to a short path distillation unit at a rate of 2 mLJmin under 10³ mbar of vacuum, wherein the feeding tank is set at 35°C, the condenser at 45°C, and the evaporator at 145°C.

In another aspect, the present invention relates to processed marine oil wherein the amount of polyunsaturated fatty acids in the acylglycerols of said oil has been enriched by a method according to any of the preceding paragraphs.

In yet another aspect the present invention relates to a processed marine oil wherein the content of omega-3 polyunsaturated fatty acids in the acylglycerols has been enriched characterised in that at least 33 mol % of the fatty acids comprised in the acylglycerols are omega-3 polyunsaturated fatty acids and the EPA/DHA ratio is 0.4 or more.

Preferably, at least 55 mol % of the fatty acids comprised in the acylglycerols are omega-3 polyunsaturated fatty acids and the EPA/DHA ratio is 0.3 or more.

Preferably, the oil is a fish oil, preferably from a farmed salmonid, more preferred from a farmed salmon.

It is further preferred that the oil has an omega-3-PUFA content of less than 18 % before enrichment of the acylglycerols with omega-3 PUFA and has an omega-3- PUFA content of more than 36 % after the enrichment, preferably of more than 48 %. It is also preferred that the oil has a DHA content of less than 10 % before enrichment of the acylglycerols with omega-3 PUFA and has a DHA content of more than 20 % after the enrichment, preferably of more than 30 %.

Furthermore, it is preferred that the oil has an EPA content of less than 7 % before enrichment of the acylglycerols with omega-3 PUFA and has an EPA content of more than 8 %, preferably of more than 10 % after the enrichment.

It is preferred that the oil is from a pelagic fish, preferably form a species chosen from the family consisting of Clupeidae and Engraulidae, more preferably chosen from the species herring (Clupea harengus) and anchoveta (Engraulis ringens) . Preferably, the oil has an omega-3-PUFA content of less than 14 % before enrichment of the acylglycerols with omega-3 PUFA and has an omega-3-PUFA content of more than 22 % after the enrichment.

Preferably, the oil has a DHA content of less than 8 % before enrichment of the acylglycerols with omega-3 PUFA and has DHA content of more than 13 % after the enrichment.

Preferably, the oil has an EPA content of less than 6 %before enrichment of the acylglycerols with omega-3 PUFA and has an EPA content of more than 8 %.

Preferably, the DHA content comprised in the acylglycerols is increased by a factor of at least 1.9, preferably of at least 2.5 in a lipase-catalysed omega-3 concentration process, having one hydrolysis step and by a factor of at least 3.9 in a repeated hydrolysis.

Preferably, the EPA content comprised in the acylglycerols is increased by a factor of at least 1.5 in a lipase-catalysed omega-3 concentration process, having one hydrolysis step and by a factor of at least 2 when the hydrolysis is repeated. Preferably, the content of polyunsaturated fatty acids comprised in the acylglycerols is increased by a factor of at least 1.7, preferably of at least 2.4 in a lipase-catalysed concentration process, having one hydrolysis step and by a factor of at least 3 in a repeated hydrolysis. Another aspect of the present invention relates to the use of a processed marine oil according to any of preceding paragraphs as an ingredient for a feed, a functional feed, a health product, a cosmetic composition, or a pharmaceutical composition.

Preferred embodiments are also defined in the dependent claims.

It will be appreciated that features of the invention described in the foregoing can be combined in any combination without departing from the scope of the invention.

DESCRIPTION OF THE INVENTION

Embodiments of the invention will now be described, by the way of examples with reference to the following diagrams, wherein

Figure 1 shows the time courses of hydrolysis reactions catalyzed by different lipases from Penicillium camembertii (PC), Rhizomucor javanicus (RJ), Rhizopus niveus (RN), Rhizopus delemar (RD), Burkholderia cepacia (BC), Rhizopus oryzae (RO), Candida rugosa (CR), and Rhizomucor miehei (RM).

Figure 2 shows changes in total omega-3 PUFA content related to the hydrolysis degree (HD, %) during hydrolysis catalyzed by lipases from Burkholderia cepacia (BC), Rhizopus oryzae (RO), and Candida rugosa (CR).

Figure 3 shows relationship between hydrolysis degree (HD, %) and hydrolysis resistant value (HRV, %) for (A) EPA, (B) DHA, and (C) OA using lipases from Burkholderia cepacia (BC), Rhizopus oryzae (RO), and Candida rugosa (CR). Figure 4 shows the main effect of (A) temperature on total omega-3; (B) reaction time on EPA/DHA; (C) enzyme load (%) on OA/total omega-3 in the lipase-catalysed hydrolysis of salmon oil catalysed by a lipase from Candida rugosa. DHA,

docosahexaenoic acid; EPA, eicosapentaenoic acid; OA, oleic acid.

Figure 5 shows the response surface plot demonstrating the effect of temperature and reaction time on the total omega-3 response at a fixed enzyme load at 3% while the ratio between oil and water phases was equal to 1 (lipase from Candida rugosa).

Figure 6 shows the response surface plot demonstrating the effect of enzyme load and water-to-oil ratio on the EPA/DHA response at a fixed temperature of 30°C and reaction time of 4 h (lipase from Candida rugosa).

Figure 7 shows the response surface plot demonstrating the effect of enzyme load and water-to-oil ratio on the OA/total omega-3 response at a fixed temperature of 30°C and time of 4 h (lipase from Candida rugosa).

Figure 8 shows the time course of the change in total omega 3 PUFA content during hydrolysis of different fish oils (lipase from Candida rugosa).

Figure 9 shows the lipid class compositions of product and residue after applying the process for large scale hydrolysis followed by short path distillation of the hydrolysed product. MG = monoacylglycerols, DG = diacylglycerols, TG = triacylglycerols, FFA = free fatty acids (lipase from Candida rugosa).

Figure 10 shows a comparison of the fatty acid profiles of substrate, residue and distillate fractions after enzymatic hydrolysis and short path distillation (lipase from Candida rugosa).

Figure 11 shows the changes in fatty acid content of the acylglycerol (·, o) and FFA (a,□) fractions obtained by single step (filled symbols) and repeated (non-filled symbols) hydrolysis throughout the reactions (lipase from Candida rugosa).

EXPERIMENTAL SECTION Several experiments were carried out with the aim to develop an optimized process applicable in large scale for concentrating (enrichment) of omega-3 PUFA in the acylglycerol fraction of an oil comprising omega-3 polyunsaturated fatty acids by lipase-catalyzed hydrolysis. Lipases used in the experiments were all lipase which hydrolyze saturated and monounsaturated fatty acids selectively, and discriminate against omega-3 polyunsaturated fatty acids. Some selected lipases, which are representative for this group of lipases, were tested in experiment 1 A and B for their individual hydrolytic potential in the up-concentrating process.

The lipase from Candida rugosa being the one with the best hydrolysis result in experiment 1 was chosen as an example candidate from this group for the optimization of the process conditions for this group of lipases with special focus on large scale applications (experiments 2-5). The hydrolyzed fatty acids are released as free fatty acids, which eventually can be separated from the acylglycerol fraction comprising PUFA by different fractionating/-separation methods.

The following experiments and processes describe particular embodiments of the process and process optimization according to the invention:

Experiment 1 : concentrating of PUFA content in marine oils by lipase-catalvsed hydrolysis

Upgrading of farmed salmon fish oil obtained from by-products was carried out by lipase-catalyzed hydrolysis to increase omega-3 polyunsaturated fatty acids (PUFA) content. The lipases tested were lipases which hydrolyze saturated and monounsaturated fatty acids selectively, and discriminate against omega-3 polyunsaturated fatty acids.

Material and methods:

Salmon oil from by-products of farmed salmon was produced according to the enzymatic process by Sorensen et al. (2004) involving hydrolysis of by-products by a protease enzyme. Non-immobilized lipases from Penicillium camembertii (PC; 244.9 U/g), Rhizomucorjavanicus (RJ; 751.4 U/g), Rhizopus niveus (RN; 589.5 U/g),

Rhizopus delemar (RD; 1092.5 U/g), Burkholderia cepacia (BC; previously known as Pseudomonas cepacia; 705.1 U/g), and Rhizopus oryzae (RO; 914.7 U/g) were from Amano (Virginia, VA, USA) while Candida rugosa lipase (CR; previously known as Candida cylindracea; 1489.2 U/g) was purchased from Fluka (Buchs, Switzerland). Immobilized lipase from Rhizomucor miehei (RM; 282 U/g) was provided by

Novozymes A/S (Bagsvaerd, Denmark). Fatty acid methyl ester standard was purchased from Nu-Chek-Prep (Elysian, MN, USA). All other reagents and solvents used were from Sigma-Aldrich Co. (St. Louis, MO, USA) and of chromatographic grade.

Determination of enzyme activities Enzyme activities were determined by the Japanese Industrial Standard method

(Japanese Industrial Standards Committee. Japanese Industrial Standard, JIS K601. 1988) substituting salmon oil for olive oil. Free fatty acids (FFA) released by the hydrolysis reaction (30 min) were titrated against 0.5 N NaOH in the presence of 1% methanolic phenolphthalein solution. One unit of enzyme activity (U) was defined as the amount of enzyme that liberated 1 pmol of FA per min at 37°C.

Screening of lipases for lipase-catalyzed hydrolysis of salmon oil

Three grams of salmon oil and 6 ml_ of 150 mM phosphate buffer (pH 7.0) were placed in a 15 mL vial and sonicated for 30 min. The substrate mixture was then transferred to a sealed jacketed glass reactor of 25 mL volume and heated to 37°C under 700 rpm stirring. After the substrates reached the reaction temperature, lipase (40 U/g oil) was added to initiate the reaction. Samples were taken periodically for analysis. 20 pL sample was dissolved in 800 pL chloroform : methanol (2:1 , v/v) and analyzed by thin layer chromatography coupled with a flame ionization detector (TLC-FID) to quantify the lipid classes (triacylglycerols, TG; diacylglycerols, DG; monoacylglycerols, MG; FFA, Free fatty acids). 100 pL of sample was mixed with 1 mL of ethanol and 1 mL of distilled water. Acylglycerol fraction was extracted twice by 2 mL hexane after saponification of FFA by adding 0.5 mL of 0.5 M ethanolic KOH. The ethanolic water phase was acidified by 0.3 mL of 2 M HCI, and FFA fraction was extracted by hexane similarly. TLC-FID analysis confirmed that all FFA was reserved in FFA fraction while TG, DG and MG were combined in the acylglycerol fraction. Fractions were methylated according to the AOCS method Ce- 1 b 89 [15] and analyzed by gas chromatography (GC).

Lipid class analysis by TLC-FID Samples were analyzed by thin layer chromatography coupled with a flame ionization detector (latroscan MK-6 s, Bechenheim, Germany). Aliquots of 20 μΙ_ were dissolved in 0.8 ml_ of chloroform/methanol mixture (2:1 , v/v), and 1 μΙ_ of diluted sample was spotted onto silica-coated Chromarod quartz rods by a semiautomatic sample spotter. Samples were developed with the developing system of n-hexane, diethyl ether, and acetic acid (45:25:1 , v/v/v). The rods were dried for 2 min at 120°C prior to analysis. The area percentages of TG, DG, MG, and FFA were used for the calculation of product yields. Hydrolysis degree is defined as (100-TG), %. All analyses were done in duplicate. The adopted values are the means at the 95% confidence limit.

Fatty acids methyl ester analysis by GC

The FA compositions of fractions were investigated by gas chromatography (Thermo Trace GC Ultra, USA) equipped with an autosampler, a flame ionization detector and a Omegawax 250 fused silica capillary column (30 m x 0.25 mm x 0.25 μητι film thickness; Supelco, Bellefonte, PA, USA). Helium was used as the carrier gas with a flow rate of 1 mL/min. A temperature program was set as follows: increasing from 170°C to 215°C at a rate of 1 °C/min, held at 215°C for 15 min. The injector and detector temperatures were set at 250°C and 270°C, respectively. FAs were identified by comparing their retention times with standard mixtures and expressed as wt% after correction for detector response factors.

Results:

A. Hydrolysis efficiency

The time course of the hydrolysis for different lipases is shown in figure 1. The hydrolysis reaction reached equilibrium for all the lipases tested after 12 to 24 h. After the first 30 minutes, the HD (hydrolysis degree) values obtained were 1 1.97%, 19.68%, and 9.42% for the PC, RJ, and RN lipases respectively. The HD was 35.33%, 47.24%, 47.33%, and 54.97%, respectively, by the lipases from CR, RM, BC, and RO after the first 30 min.

There was no relationship observed between the positional specificity and hydrolysis efficiency of the lipases. For instance, HD value obtained by CR lipase (91.89%), which has no positional specificity, was equal to that obtained by RD lipase

(91.72%), which hydrolyzes sn-1 ,3 positions specifically. Besides, no differences between immobilized and free lipases were observed in terms of hydrolysis efficiency. RM, the only immobilized lipase used in the study, had a moderate HD level.

B. Evaluation of the lipases based on omega-3 PUFA content in acylqlvcerols and fatty acid selectivity of lipases Three of the lipases from experiment 1 A having the highest degree of hydrolysis degree (originating from CR, BC, and RO) were additionally investigated for their efficiency to increase the omega-3 PUFA content in acylglycerols comprised in marine oil. Table 1 shows the FA content in the acylglycerol fraction of salmon oil after 24 h of hydrolysis catalyzed by the selected lipases. RO and BC lipases resulted in a moderate improvement in omega-3 PUFA content by increasing the value from 13.77% to 17.88% and 14.52%, respectively. CR lipase, had the highest increase in omega-3 PUFA content. The value was increased by approximately 50% (20.54%) after 2 h and reached to 27.81 % at the end of the 24 h period. CR lipase is previously recognized as an enzyme from this group which has a high hydrolysis efficiency and discrimination against EPA and DHA. This lipase had no positional specificity, hydrolyzing FAs at all sn- positions randomly even at HD levels as low as 20%. On the other hand, it had a strict acyl chain specificity, resulting in selective hydrolysis of SFA and MUFA at a much higher rate compared to omega-3 PUFA. Table 1. FA content in the acylglycerol fraction before the start of the hydrolysis (0 h) and after 24 h of hydrolysis catalyzed by selected lipases (RO, BC and CR).

Figure 2 shows the changes in total omega-3 PUFA content related to HD during hydrolysis catalyzed by lipases CR, BC and RO. The changes in omega-3 PUFA content were significantly different although the HD values were similar for all the lipases at the end of 24 h. CR lipase resulted in constant increase in total omega-3 PUFA content with increasing HD above 40%. The other lipases showed a slight decrease in total omega-3 PUFA above 50% of HD.

The hydrolysis resistant value (HRV) was calculated according to Tanaka et al.

(1993) as

HRV (%) = {(100 x GC_a - B x GC_b) ÷ (100 x GC_a)} x 100 where GC_a is the content of each FA in the original oil and GCb is the content of each FA in the FFA fraction of the product, both of which were measured by GC (wt %), and B is the ratio of FFA in the hydrolyzed oil, measured by TLC-FID (vol %). A high HRV indicates a high resistance of the FA of interest to hydrolysis.

HRV values for EPA, DHA, and OA were calculated for the hydrolysis reactions catalyzed by the tested lipases. Figures 3A, 3B, and 3C show the relationship between HRV and HD for EPA, DHA, and OA, respectively.

All of the lipases used had a high resistance towards hydrolysis of EPA at the initial stages of the reactions (Figure 3A). Starting from HD of approximately 40%, HRV decreased steadily until the end of the 24 h period. The HRV calculated for EPA was the highest for CR lipase until HD of 60% and decreased with increased hydrolysis. A similar trend was observed for the resistance of DHA against hydrolysis (Figure 3B). The HRV pattern for OA (Figure 3C) was significantly different compared to those for EPA and DHA. CR lipase had the lowest value throughout the reaction. All lipases seemed to have a high discrimination against hydrolysis of EPA and DHA, although differences in the degree of discrimination were observed between the lipases as well as in respect to the hydrolysis time used.

Considering HRV in relation to HD for EPA, DHA and OA, the concentration of omega-3 PUFA in salmon oil depends on one hand on the discrimination of the lipase against EPA and DHA, as well as the selective hydrolysis of monounsaturated fatty acids such as for OA, the target FA. Apart of these individual characteristics of each lipase, the efficiency of the lipases depends on the other hand on the chosen reaction conditions as well as the combination of reaction conditions which is considered to be important for the efficiency of the lipase catalysed concentrating process of oils.

Experiment 2: Optimized process of concentrating omeqa-3 PUFA in a fish oil Material and methods:

Salmon oil from by-products of farmed salmon was produced according to the patented enzymatic process by Sorensen et al. (2004; EP 1 575 374 B1 ) involving hydrolysis of by-products by a protease enzyme. The major FA found in the substrate was oleic acid (OA) with a share of 35.51 %, while the omega-3 PUFA content was as follows: 4.8% EPA, 2.04% docosapentaenoic acid (DPA), 6.93% DHA. Lipase from Candida rugosa (64000 U/g) was donated by Meito Sangyo Co., Ltd. (Tokyo, Japan). Fatty acid methyl ester standard was purchased from Nu-Chek- Prep (Elysian, MN, USA). All other reagents and solvents used were from Sigma- Aldrich Co. (St. Louis, MO, USA) and of chromatographic grade.

Experimental design

Central composite design (CCD) was used for the optimization studies. CCD is a 2^k factorial design with star and center points. Star distance was chosen to be 1 in order to provide a face-centered CCD, since the ranges selected for the factors cover the area of interest, and thus, it was not necessary to have extreme levels beyond these ranges.

The following conditions of temperature, incubation time, enzyme concentration and water-to-oil ration were tested in the reaction:

- temperature range: 30-50°C

- incubation time 4-8 h

- enzyme load 3 -8 % (wt), i.e. weight based on oil amount

- water-oil-ratio (v/w): 1 :1 - 5:1.

The oil amount was kept constant at 3 g. Twenty eight experimental settings consisting of 4 centre points were generated by Modde 8.0.2 (Umetrics, Umea, Sweden). Reaction conditions and observed responses are given in Table 2.

Table 2. Central composite design and responses for the lipase-catalyzed hydrolysis of salmon oil. Te, temperature; t, time; En, enzyme load; W_r, water-to-oil ratio: OA, oleic acid.

Factors Responses (wt%)

Exp. OA/Total no Te (°C) t (h) En (wt %) W_r (w/w) Tot. omega-3 EPA/DHA omega-3

1 30 4 3 ¹ 28.11 0.5 8.92

2 50 4 3 ¹ 32.27 0.47 7.26

3 30 8 3 ¹ 32.58 0.45 8.2

4 50 8 3 ¹ 33.31 0.39 6.05

5 30 4 8 ¹ 20 0.55 7.65

6 50 4 8 ¹ 31.65 0.46 8.02

7 30 8 8 ¹ 32.18 0.44 7.2

8 50 8 8 ¹ 32.56 0.43 7.31

9 30 4 3 5 21.41 0.63 11.06

10 50 4 3 5 30.52 0.47 7.93

11 30 8 3 5 28.03 0.58 11.46

12 50 8 3 5 33 0.42 7.93

13 30 4 8 5 20.02 0.53 8.29

14 50 4 8 5 33.3 0.46 7.54

15 30 8 8 5 27.99 0.39 6.45

16 50 8 8 5 35.47 0.31 5.63

17 30 6 5.5 3 24.98 0.54 7.33

18 50 6 5.5 3 34.88 0.36 6.1

19 40 4 5.5 3 32.41 0.45 7.19

20 40 8 5.5 3 33.45 0.44 7.3

21 40 6 3 3 32.86 0.48 7.86

22 40 6 8 3 31.96 0.44 7.03

23 40 6 5.5 1 31.08 0.4 5.85

24 40 6 5.5 5 32.44 0.44 7.2

25 40 6 5.5 3 33.75 0.41 7.9

26 40 6 5.5 3 32.62 0.48 7.94

27 40 6 5.5 3 32.6 0.48 7.36

28 40 6 5.5 3 32.42 0.46 7.34 Fatty acids methyl ester analysis bv gas chromatography

100 μΙ_ of sample were mixed with 1 mL of ethanol and 1 mL of distilled water. The acylglycerol fraction, consisting of triacylglycerols (TG), diacylglycerols (DG) and monoacylglycerols (MG), was extracted twice by 2 mL hexane after saponification of free fatty acids (FFA) by adding 0.5 mL of 0.5 M ethanolic KOH. The ethanolic water phase was acidified by 0.3 mL of 2 M HCI, and FFA fraction was extracted by hexane similarly. Fractions were methylated according to the AOCS method Ce-1 b (2007). The FA compositions of fractions were investigated by gas chromatography (GC) (Thermo Trace GC Ultra, USA) equipped with an autosampler, a flame ionization detector and a Omegawax 250 fused silica capillary column (30 m x 0.25 mm x 0.25 μιη film thickness; Supelco, Bellefonte, PA, USA). Helium was used as the carrier gas with a flow rate of 1 mL/min. A temperature program was set as follows: increasing from 170°C to 215°C at a rate of 1 °C/min, held at 215°C for 15 min. The injector and detector temperatures were set at 250°C and 270°C, respectively. FAs were identified by comparing their retention times with standard mixtures and expressed as wt% after correction for detector response factors.

Statistical analysis

The data were analyzed by means of RSM using Modde 8.0.2. Dependent variables were chosen to be total content of EPA, DPA and DHA in the acylglycerol fraction (total omega-3), the ratio of EPA to DHA in the acylglycerol fraction (EPA/DHA), and the ratio of OA to the sum of EPA, DPA and DHA in the FFA fraction (OA/total omega-3). The responses were first fitted to factors by multiple regressions, and then the models generated were used to evaluate the effects of various factors. The first- and second-order coefficients were generated by regression analysis (Table 3). The accuracy of the models was evaluated by coefficient of determination (R²), absolute average deviation (AAD) and a test for lack of fit from analysis of variances

(ANOVA). AAD was calculated as follows:

Yi.obs and yi,pr are the observed and predicted responses, respectively, and p is the number of experimental run. R² must be close to 1.0 and the AAD between the predicted and observed data must be as small as possible.

Quadratic polynomial regression models were assumed for predicting the responses. The model proposed for each response was

Y = βο fitjXtXj

where Y is the response (total omega-3; EPA/DHA; OA/total omega-3), /¾ is the intercept, β, is the first-order model coefficient, β„ is the quadratic coefficient for the variable, j8,y is the interaction coefficient for the interaction of variables /^' and j, X, and Xj are the independent variables (Te, t, En, W_r). Five additional reactions were performed at the optimized levels to verify the models by chi-square (χ²) test.

Reaction conditions and responses, both predictive and observed, are given in Table 4. Large scale production was performed by using 200 g of oil, resembling the same reaction conditions as reaction 1 in Table 4.

Table 3. Regression coefficients and probability of predicted quadratic polynomial models for different responses. Te, temperature; t, time; En, enzyme load; W_r, water- to-oil ratio: OA, oleic acid.

Variables Total omega-3 EPA/DHA OA/Total omega-3

Coefficient P-value Coefficient P-value Coefficient P-value

Intercept (βο) -24.15 <0.01 0.88 <0.01 16.07 <0.01

Linear (β,)

Te 2.54 <0.01 -0.01 <0.01 -0.07 <0.01 t 2.39 <0.01 -0.02 <0.01 -1.13 0.02

En -2.46 0.29 -0.03 0.03 -1.7 <0.01 w_r -2.38 0.09 0.11 0.38 1.94 0.01

Quadratic (fin)

Te² -0.03 0.01 0.0001 0.62 -0.0008 0.84

0.09 0.68 0.001 0.78 0.11 0.24

En* -0.02 0.88 0.003 0.36 0.1 0.1

-0.2 0.41 -0.005 0.42 -0.07 0.48

Interaction ( ¾^■)

Te x t -0.08 <0.01 0.0001 0.79 -0.004 0.61

Te x En 0.03 0.04 0.0004 0.29 0.02 <0.01

Te x W_r 0.05 <0.01 -0.0009 0.07 -0.01 0.06 t x En 0.1 0.17 -0.002 0.19 -0.04 0.18 t x W_r 0.01 0.91 -0.002 0.42 -0.004 0.91

En x W_r 0.17 0.04 -0.006 <0.01 -0.13 <0.01

Table 4. Verification of the models by chi-square (χ²) test. Te, temperature; t, time; En, enzyme load; W_r, water-to-oil ratio; OA, oleic acid; Obs, observed; Pred., predicted.

b: Calculated according to the model equation obtained from the regression analysis c x„2² =∑ Yobs, observed value; y_pr, predicted value. Cutoff point was

y_Pr

9.488 at a = 0.05 and degree of freedom = 4.

Results and Discussion

Model fitting

The main objective of the study was to concentrate/enrich the omega-3 PUFA content in the acylglycerol fraction by releasing the rest of the FAs through lipase- catalyzed hydrolysis. Responses obtained from the experimental design are given in Table 2. R² and AAD values of the model generated based on total omega-3 response were 0.94 and 2.76, respectively. ANOVA (data not shown) also revealed that the probability for the regression of the model was significant (P<0.001 ), confirming that the model was statistically good, and it had no lack of fit (P>0.05). According to the regression coefficients (Table 3), the most significant linear effect at the significance level of 95% was that of temperature. Moreover, the quadratic effect of temperature as well as the interaction of temperature with the other three factors had significant effects on the response. Figure 4A shows the main effect of temperature on the total omega-3 content. As suggested by the positive linear effect and negative quadratic effect of temperature on the response, there exists an optimum value for the factor, which is located in the range of 40-50°C. Temperature plays an important role in enzymatic reactions, mainly by determining the reaction rate. Although increased temperature enhances the hydrolytic rate, it can also lead to thermal deactivation of lipase. Moreover, increased temperature results in oxidation of PUFA to a higher extent, which could also be an explanation of the lower rate of increase in the total omega-3 PUFA content at elevated temperatures.

In order to have a balanced content of EPA and DHA in the product, the reaction conditions were optimized for EPA/DHA response as well. Responses obtained from the experimental design are given in Table 2. R² and AAD values of the model were 0.86 and 4.57, respectively. ANOVA (data not shown) also revealed that the probability for the regression of the model were significant (P<0.001 ), meaning that the model was statistically good, and it had no lack of fit at 95% level of significance. Regression analysis (Table 3) showed that all of the factors except water-to-oil ratio had significant linear effects on the EPA/DHA response (P<0.05). None of the quadratic effects were statistically significant, while the interaction between enzyme load and water-to-oil ratio had a significant effect on the response. Figure 4B shows the main effect of time on EPA/DHA response. Short reaction time was not sufficient to have a reasonable hydrolysis degree. After 4 h of reaction (reaction 9 in Table 2), 82.3% of the initial TAG was unhydrolyzed (data not shown), which explains why EPA/DHA value (0.63) was closer to the original value, and in fact, was the highest value obtained. It was suggested that hydrolysis reaction catalyzed by Candida rugosa lipase took place in two steps: TG molecules without DHA were hydrolyzed at the earlier stages of the reaction. As the reaction progressed, DHA containing-TG molecules are hydrolyzed as well which results in faster clearance of EPA compared to DHA. Temperature and time had similar effects on the response: The higher the level was, the lower the EPA/DHA value was, which was a result of the faster hydrolysis of EPA compared to DHA, confirmed by the higher increase of the EPA level in the FFA fraction compared to DHA (data not shown).

Oleic acid was the dominant FA in the salmon oil composition, with 35.51 % share, due to the feed given in the salmon farm. Thus, OA was chosen as the target FA to be removed from the oil. Monitoring the ratio between OA and total omega-3 PUFA released to the FFA fraction by hydrolysis enabled to compare the selectivity of the lipase towards these FAs. Responses obtained from the experimental design are given in Table 2. R² and AAD values of the model were 0.9 and 4.73, respectively. ANOVA (data not shown) revealed that the probability for the regression of the model were significant (P<0.001 ), confirming that the model was statistically good, and it had no lack of fit at 95% level of significance. According to the regression coefficients (Table 3), all of the factors had significant linear effects on the OA/total omega-3 response at a significance level of 95%. None of the quadratic effects were statistically significant, while the interaction between temperature and enzyme load as well as that between enzyme load and water-to-oil ratio had significant effects on the response. Figure 4C shows the main effect of enzyme load on OA/total omega-3 level. Since the linear effect of enzyme load was negative, while the quadratic effect of it was positive, there exists an optimum level for the factor, which lies in the range of 3-5.5%. The availability of the enzyme increases with increased load of it, and thus, the selectivity of the hydrolytic reaction decreases, resulting in hydrolysis of omega-3 PUFA as well as OA. However, further increase of the enzyme amount did not have additional effect on the response, since the reaction medium was saturated with enzyme.

Analysis of response surface plots

The effects of interactions between investigated factors were visualized by varying two factors within the experimental range, while holding the other two at constant values so as to maximize the responses. The most significant interaction effect on total omega-3 was that between temperature and time. Figure 5 is the response surface plot indicating the effect of temperature and time on the total omega-3 response. Enzyme load and water-to-oil ratio are fixed at their lowest levels, because increasing both of these factors resulted in a decreased response. Total omega-3 content was doubled at the lowest level of the factors, and increased with increasing reaction time throughout the range investigated. Temperature had a positive effect on the response as well; however, the negative quadratic effect of temperature became significant at levels higher than 40°C, which resulted in slightly decreased total omega-3 content. This is believed to be a result of the decreased lipase activity at elevated temperatures. The omega-3 content of the substrate could be increased above 32.5% with a temperature of 40-50°C by a 4-h reaction while enzyme load was 3% and the weight ratio between oil and water phases was equal to 1.

Figure 6 shows the response surface plot indicating the effect of enzyme load and water-to-oil ratio on the EPA/DHA level while the other two factors are fixed at their lowest levels, since increasing both of these factors resulted in decreased response. None of the combinations in the design ranges resulted in a similar EPA/DHA value with that of salmon oil, which was 0.7. Hydrolysis degree increases with increased water-to-oil ratio, due to the available interphase between the substrates. EPA is more prone to hydrolysis compared to DHA, which was depicted by decreased EPA/DHA level at water-to-oil ratios higher than 3. However, since the linear effect of enzyme load is much more significant than that of water-to-oil ratio, EPA/DHA response was highly dependent on the amount of enzyme. Increased enzyme load enhanced the reaction rate, resulting in faster hydrolysis of EPA, and thus, decreased the EPA/DHA value. EPA/DHA level could be maintained above 0.55 with 2-5 water-to-oil ratio and 3-6% enzyme load by a 4-h reaction at 30°C.

The response surface plot representing the interaction effect of enzyme load and water-to-oil ratio on OA/total omega-3, while temperature and time are fixed at their lowest levels, is given in Figure 7. The highest level for the response was obtained when the enzyme load was 3%, and decreased as the amount of enzyme increased. This result is due to the high availability of enzyme in the medium, and thus, decreased selectivity of the hydrolytic reaction, resulting in hydrolysis of omega-3 PUFA as well as OA. Furthermore, the FA selectivity of the Candida rugosa lipase was believed to be interfered by increased FFA content in the medium at elevated content of water, as a result of the increased hydrolytic rate as well as the

simultaneous esterification of the released FFA, which was OA in the present case. These phenomena would also lead to decreased OA/total omega-3 level when the water in the medium was more than 3-fold of oil. A reasonably selective process, with an OA/total omega-3 level in the FFA fraction above 9, could be obtained by a 4-h reaction catalyzed by 3-4% lipase with 2.5-5 fold of water based on oil at 30°C.

Optimization of the reaction conditions and large scale processing

A χ² test using five additional experimental sets chosen from the given ranges of reaction conditions was performed to examine the adequacy of the models established (Table 4). The χ² test for total omega-3, EPA/DHA, and OA/total omega- 3 responses indicated that there were no significant (P<0.05) differences between the observed and predicted values since the χ² values were much smaller than the cutoff point (9.488) at a = 0.05 and degree of freedom = 4. Thus, the models were proven to be predictive of the responses.

Large scale hydrolysis was performed by selecting the reaction conditions so as all of the responses investigated were predicted to be maximized (Table 4). Analysis of the product confirmed that the models were still predictive when the reaction was scaled up by ~67-fold. The final product contained 33.32% total omega-3 PUFA content with an EPA/DHA level of 0.49, and the hydrolysis was reasonably selective as depicted by the high (7.96) level of OA/total omega-3 in the FFA fraction.

Conclusions

Candida rugosa lipase, as an example representing lipases which selectively hydrolyse saturated and monounsaturated fatty acids bound to said acylglycerols and discriminate in its hydrolytic activity against omega-3 polyunsaturated fatty acids, significantly increased the content of omega-3 PUFA in the substrate.

Considering total content of omega-3 PUFA, as well as the ratio between them (EPA/DHA) in the product and the ratio of OA to total omega-3 PUFA released by hydrolysis, most optimum conditions were selected as follows: 45°C of temperature, 4 h of time, 3% of lipase (based on oil amount), 3.16-fold water (w/w, based on oil amount). Total omega-3 PUFA content of the product obtained at the given conditions was increased from 13.77% to 33.01%. EPA/DHA ratio was acceptable (0.48), while the ratio of OA to total omega-3 PUFA in the FFA fraction was significantly high (9.53). Application of the optimum conditions to a large scale production resulted in similar results as well, confirming the usability of the generated models for an industrial process according to the present invention. Thus, the applicant surprisingly was able to show that the following combination of process parameters are especially efficient in the omega-3 PUFA concentration process in an industrial scale:

1 ) Temperature

-incubation temperature between 30 and 50°C, preferably between 40 and 50°C, most preferred 45°C .

2) Lipase concentration:

- lipase added in a concentration of at least 3- 5.5 % by weight based on the amount of oil

3) Water-Oil ratio

- a water-oil ratio of 2-5:1.

4) Reaction time

- the reaction time should be at least 2 hours, preferably in the range of 2-8 hours.

The most optimal and preferred process for concentrating of omega-3 fatty acids in oils, especially in a large scale process applies the following conditions:

- the lipase is added in a concentration of at least 3 % by weight based on the amount of oil,

- the incubation temperature is about 45 °C,

- the water-to-oil ratio is about 3.16:1 (v/w),

- the incubation time is between 2 hours and 4 hours.

It is preferred that the lipase is from Pseudomonas, Candida, Rhizopus, Rhizomucor. Particularly preferred lipases are from Candida rugosa, Burkholderia cepacia and Rhizopus oryzae, however most preferred is the lipase from Candida rugosa.

The examined lipases are all belonging to the group of lipases which selectively hydrolyse saturated and monounsaturated fatty acids bound to said acylglycerols and discriminate in their hydrolytic activity against omega-3 polyunsaturated fatty acids. Although the optimization study has been based on the lipase from Candida rugosa as the most promising representative of the group, it is assumed that the optimized conditions will also be beneficial in lipase-catalyzed omega-3 fatty acid enrichment processes with the other lipases belonging to that group.

The process is especially suitable when used in an industrial applicable large-scale process due to its mild and at the same time cost and time-efficient features.

Experiment 3: Lipase-catalvzed hydrolysis of fish oil from different sources

Salmon oil obtained according to the enzymatic process as described by Sorensen et al. (2004) involving hydrolysis of by-products by a protease enzyme, pelagic oil obtained from herring (Clupea harengus) and two types of commercially available fish oils (from EPAX, Norway) were hydrolysed using Candida rugosa lipase according to the preferred conditions of experiment 2 (3% of lipase, incubation temperature of 45 °C, water-to-oil ratio of about 3.16:1 (v/w), and incubation time of 4 hours). Fatty acid compositions of substrates are given in Table 5.

Lipase-catalyzed hydrolysis

Three grams of oil and 9.5 mL of distilled water were placed in a sealed jacketed glass reactor of 25 mL volume and heated to 45°C under 300 rpm stirring. After the substrates reached the reaction temperature, lipase (3 wt% based on the oil amount) was added to initiate the reaction. Samples were taken periodically for analysis.

Lipid class analysis by TLC-FID

20 μί sample was dissolved in 800 pL chloroform : methanol (2:1 , v/v) and analyzed by thin layer chromatography coupled with a flame ionization detector (TLC-FID) to quantify the lipid classes (triacylglycerols, TG; diacylglycerols, DG; monoacyl- glycerols, MG; free fatty acids, FFA). Fatty acids methyl ester analysis by GC

100 μί of sample was mixed with 1 mL of ethanol and 1 mL of distilled water.

Acylglycerol fraction was extracted twice by 2 mL hexane after saponification of FFA by adding 0.5 mL of 0.5 M ethanolic KOH. The ethanolic water phase was acidified by 0.3 mL of 2 M HCI, and FFA fraction was extracted by hexane similarly. TLC-FID analysis confirmed that all FFA was reserved in FFA fraction while TG, DG and MG were combined in the acylglycerol fraction. Fractions were methylated according to the AOCS method Ce-1 b 89 and analyzed by gas chromatography (GC) after addition of internal standard (17:0).

Table 5. Fatty acid composition of fish oils before hydrolysis (given as mol%).

Fatty acid Salmon oil Pelagic oil EPAX 1050 EPAX 3000

14:0 3,09 6,42 0,27 7,02

16:0 1 1 ,46 1 1 ,77 2,09 17,70

16:1 4,41 4,44 0,72 9,70

18:0 2,85 1 ,25 1 ,32 3,90

18:1 n-9 31 ,40 12,58 2,96 1 1 ,15

18:1 n-7 3,13 1 ,61 0,83 3,74

18:2 10,29 1 ,37 0,42 1 ,42

18:3 3,58 0,91 0,29 0,89

18:4 0,99 2,01 0,78 2,82

20:0 0,34 0,10 0,05 0,01

20:1 4,17 14,65 0,99 0,96

20:2 0,76 0,23 0,19 0,20

20:3 n-6 0,21 0,09 0,26 0,38

20:4 n-6 0,52 0,36 0,58 1 ,57

20:3 n-3 0,33 0,13 0,14 0,10

20:4 n-3 1 ,21 0,56 0,70 0,98

20:5 5,46 5,44 10,18 20,83

22:1 4,38 26,94 5,96 0,48

22:5 2,79 0,73 12,41 2,77

22:6 8,1 1 7,23 58,49 12,94

24:1 0,52 1 ,19 0,36 0,46

∑SFA 17,74 19,54 3,73 28,62

∑MUFA 48,01 61 ,41 1 1 ,83 26,48

∑PUFA 34,25 19,05 84,44 44,89

Σω-3 PUFA 16,36 13,40 81 ,08 36,54 Results

Table 6. Fatty acid levels and fatty acid compositions of the acylglycerol fractions of hydrolysed salmon and pelagic oil products throughout the reaction.

The time courses of hydrolysis of fish oils are given in Figure 8. Detailed information on the reaction products is given in Tables 6 and 7. The highest increase in omega-3 PUFA content was obtained in salmon oil. The value increased from 16.36% to 38.71 %, which corresponds to a ~2.4-fold increase. Optimized reaction conditions resulted in a ~1.75-fold increase of omega-3 PUFA in the pelagic oil. The hydrolysis rate was observed to be lower for pelagic oil than for the salmon oil as indicated by the slower increase in the FFA level (Table 6). In the pelagic fish oil the omega-3 PUFA content increased by a factor of about 1.7, the DHA content by a factor of 1.9 and the EPA content by a factor of 1 ,5. Since EPAX 1050 already had a very high content of omega-3 PUFA, the enzyme was not able to catalyze hydrolysis at a significant level (Table 7). The omega-3 PUFA content of EPAX 3000 was increased from 36.54% to 52.25% at the optimized conditions (a ~1.43-fold increase). Similarly, high initial content of omega-3 PUFA resulted in lower hydrolysis, and thus, lower concentration (Table 7).

EPA/DHA ratio of the products

The EPA/DHA ratio in the salmon oil was 0.67 initially (Table 8). The ratio decreased to 0.44 at the end of the 4 h lasting reaction. The obtained EPA/DHA ratio in the present invention is considered satisfactory for the intended purpose and final product in spite of the short reaction time applied in the present invention. The only substrate with an initial EPA/DHA ratio higher than 1 was EPAX 3000, which did not decrease significantly after 4 h of reaction.

Table 8. EPA/DHA ratios of the hydrolysis products throughout the reaction.

Past attempts to concentrate oils comprising omega-3 mainly focussed on an increase of the total omega-3 PUFA content and not on a balanced EPA and DHA content. These described efforts to concentrate omega-3 PUFA by lipases did generally not discriminate between EPA and DHA. Thus, the lipase-catalyzed hydrolysis often resulted in some loss of EPA, if not using selective lipase such as from the Pseudomonas family or the lipase from Candida rugosa of the present invention. The mechanism underlying the discrimination between EPA and DHA by the lipases of Pseudomonas and Candida rugosa depends on the structure of these FAs. The molecular conformation of c/^'s carbon-carbon double bonds in omega-3 PUFA, particularly EPA and DHA, causes steric hindrance and results in bending of the FA chains, which brings the terminal methyl groups very close to the ester bonds.

Because of this steric hindrance effect, enzymatic active sites cannot reach the ester-linkages of these FAs with their glycerol backbones, which results in protection of EPA and DHA from lipase-catalyzed hydrolysis. DHA, having one more double bond, is more resistant to hydrolysis compared to EPA. Consequently, the EPA/DHA ratio decreases during the hydrolysis compared to the starting material, as also described in literature.

The applicant was able to show that by using the process parameters of the present invention, loss of specific omega-3 fatty acids could be limited and a concentrated oil with a balanced EPA/DHA ratio could be obtained.

With the exception of the commercial EPAX 1050 oil which is an oil with a very high concentration of DHA, application of the optimised process of the present invention resulted in an clear increase of DHA (mol % fatty acids in the acylglycerol fraction) by a factor of about 1.5 to about 2.8 and of EPA by a factor in the range of about 1.3 to about 1.8. In all cases the enrichment with DHA was higher than with EPA. This was valid both for the different fish oils and the pelagic oil. Thus, the process according to the present invention was especially suitable and advantageous for the improvement of the DHA-content in the oil. Best specific enrichment results in terms of DHA and EPA were obtained for the salmon oil obtained from farmed fish, which were previously fed a diet with a reduced content of omega-3 PUFA and which therefore had a reduced content of omega-3 PUFA compared to fish fed a diet high in omega-3 fatty acids.

The concentrating process of the present invention resulted in an enrichment factor of about 2.4 in the salmon oil for omega-3 PUFA, about 1.8 for the pelagic oil and 1.4 for the commercial EPAX 3000 oil. This further supports the suitability of the process for the enrichment of marine oils, especially of the salmon oil and the pelagic oil with a reduced content of omega-3 PUFA. Due to the high efficiency of the process it is assumed that the process of the present invention may also be used for oils having an even lower content of omega-3 PUFA than in the present experiment such as for oils from farmed salmon fed very low amounts of omega-3 fatty acids. It is therefore preferred that also an oil of farmed fish having a very low content of PUFA can be used.

Experiment 4: Concentrating of Salmon oil using the process of large scale hydrolysis followed by short path distillation of the product

Material and Methods

Salmon oil from by-products of farmed salmon was produced according to the patented enzymatic process by Sorensen et al (2004) involving hydrolysis of byproducts by a protease enzyme. As in the previous experiments a lipase from

Candida rugosa lipase (64000 U/g) was used.

Lipase-catalvzed hydrolysis

Two hundred grams of salmon oil and 632 ml_ of distilled water (at 3.16 water-to-oil ratio) were placed in a sealed jacketed glass reactor of 1 L volume and heated to 45°C under 300 rpm stirring. After the substrates reached the reaction temperature, 6 g of lipase (3 wt% based on the oil amount) was added to initiate the reaction. Reaction took place for 3 h. Substrates and enzyme mixture was centrifuged twice at 10000 rpm for 15 min to separate oil from both water and lipase. Due to the losses during the process, centrifugation ended up with 169 g of product.

Short path distillation

100 g of product was fed to the short path distillation unit (KDL 5, UIC GmbH, Germany) at a rate of 2 mL/min under 10^~3 mbar of vacuum. The feeding tank, condenser and evaporator were set at 35°C, 45°C and 145°C, respectively. Both residue and distillate fractions were collected for lipid class and fatty acid

composition analyses.

Lipid class analysis by TLC-FID Lipid class analysis by TLC-FID was carried out as described in experiment 3.

Fatty acids methyl ester analysis by GC

Fatty acids methyl ester analysis by GC was carried out as described in experiment 2.

Results

Short path distillation yielded in 36.9 g of residue and 61.5 g of distillate, out of 100 g of feed. The loss during the purification process was only 1.6% by weight. Lipid class compositions of product and residue are illustrated in Figure 9. FFA level was reduced from 66.83% to 1.98% in the final product. Distillate was composed of 99.35% FFA and 0.65% MG (not shown).

Comparison of the fatty acid profiles of the salmon oil (substrate in the hydrolysis), residue and distillate fractions is given in Figure 10 and Table 9. Total PUFA, composed of EPA, DPA and DHA, was increased by 2.3-fold. Target fatty acid to be removed, oleic acid, was highly concentrated in the distillate fraction, but residue still had OA at the level of 18.69%. 63% of initial EPA and 91 % of initial DHA were recovered in the residue, while 20% of initial OA was still found in this fraction as well.

Table 9. Levels of main fatty acids in salmon oil, residue and distillate.

OA EPA DHA Total PUFA

Salmon oil 34,52 5,46 8,11 16,36

Residue 18,69 9,80 21 ,98 38,03

Distillate 41 ,77 3,32 0,92 5,20

Experiment 5. Repeated hydrolysis of salmon oil

A repeated hydrolysis (second hydrolysis) was performed on the residue obtained from short path distillation. The second hydrolysis was carried out with the same enzyme as in the first hydrolysis. Three grams of the residue obtained from Experiment 4 was subjected to a further hydrolysis under the same reaction conditions as in the first hydrolysis. The reaction was continued until the FFA level was approximately 40%. Results

Figure 11 compares the changes in FA content of the acylglycerols and FFA fractions obtained by single step and repeated hydrolysis throughout the reactions. A comparison of contents of major FAs in substrate and hydrolysis products are given in Table 10. The total omega-3 PUFA content was further concentrated to 50.58% by repeated hydrolysis. On the other hand, the loss of omega-3 PUFA to the FFA fraction was increased as well (Figure 1 1 ). Due to the reduced availability of SFA and MUFA, DHA was hydrolyzed to a certain degree in the second round of hydrolysis, which resulted in the increased omega-3 PUFA content in the FFA fraction obtained after repeated hydrolysis. The concentration achieved was, however, more than 3 times of the original level. Thus, the added value of the product was surprisingly further improved by the repeated hydrolysis approach even though using the same enzyme as in the first hydrolysis.

Table 10. FA content in the salmon oil before hydrolysis and in the acylglycerols after the first and the second hydrolysis (repeated hydrolysis).

Although the study using the optimized conditions in repeated hydrolysis has been based on the lipase from Candida rugosa as the most promising representative of the group, it is assumed that the repeated hydrolysis under the optimized conditions will also be beneficial in lipase-catalyzed omega-3 fatty acid enrichment processes with the other lipases belonging to that group.

Definitions of terms and abbreviations: DHA - docosahexanenoic acid (22:6 n-3)

DPA - docosapentaenoic acid (22:5 n-3)

DG - diacylglycerols

EPA - eicosapentaenoic acid (20:5 n-3)

FA - fatty acid

FFA - free fatty acids

MG - monoacylglycerols

MUFA - monounsaturated fatty acid

PUFA - polyunsaturated fatty acid

SFA - saturated fatty acid

TG - triacylglycerols

Total omega-3 PUFA - refers to the sum of EPA, DPA and DHA

v - volume

w- weight

Fish oil - by fish oil is meant an oil originating from fish comprising omega-3 polyunsaturated fatty acids such as EPA and/or DHA and/or DPA.

Pelagic oil - by pelagic oil is meant an oil obtained from a pelagic fish such as from herring (Clupea harengus) or anchoveta (Engraulis ringens), wherein the raw material (e.g. viscera after filleting, whole fish) is processed by a commonly known fish meal/ fish oil production method without enzymatic treatment.

Marine oil - by marine oil is meant a fish oil or oils produced from other animals or plants or oils produced by microorganisms such as bacteria or algae, which comprise polyunsaturated fatty acids such as EPA and/or DHA and/or DPA.

Marine/Fish oil concentrate or processed marine oil - by an oil concentrate or in the context of the present invention by a processed marine oil is meant an oil wherein the amount of omega-3 fatty acids in the acylglycerol fraction of the oil has been increased (enriched, up-graded, up-concentrated) in relation to the total amount of fatty acids comprised in said acylglycerol fraction by means of a

enrichment/concentrating process.

Enrichment of an oil with omega-3 PUFA /concentration of omega-3 PUFA in an oil - by enrichment of an oil with omega-3 PUFA or concentrating of omega-3 PUFA in an oil is meant that the amount of omega-3 PUFA or of a specific omega-3 fatty acid such as DHA or EPA is increased in the acylglycerol fraction of the oil in relation to the total amount of fatty acids comprised in said acylglycerol fraction by means of a enrichment/concentrating process.

Lipase which discriminates in its activity against certain fatty acids - by discriminating in its activity against certain fatty acids such as omega-3 fatty acids is meant that the lipase has a negative selectivity (or lower substrate affinity) towards these fatty acids and preferably hydrolyzes other fatty acids bound in acylglycerols e.g. those which are not omega-3 fatty acids such as saturated and monounsaturated fatty acids.

References

Sorensen S, Sandnes K, Hagen H, Pedersen K, 2004. Apparatus and method for hydrolysis of a protein containing raw material and application of the resulting hydrolysis products, (see also European Patent EP 1 575 374 B1 ).

Tanaka Y, Funada T, Hirano J, Hashizume R. Triglyceride specificity of Candida cylindracea lipase - effect of docosahexaenoic acid on resistance of triglyceride to lipase. J Am Oil Chem Soc 1993; 70: 1031-4.

Claims

PATENT CLAIMS

1. Method for enrichment of the amount of omega-3 polyunsaturated fatty acids comprised in acylglycerols of an oil by lipase-catalysed hydrolysis, wherein

the oil is mixed with an aqueous solution in a water to oil ratio of 2:1 to 5:1 (v/w),

a lipase, which selectively hydrolyses saturated and monounsaturated fatty acids bound to said acylglycerols and discriminates in its hydrolytic activity against omega-3 polyunsaturated fatty acids bound to said acylglycerols, is added to the oil- water mixture in a concentration of 3-5.5 % by weight based on the amount of oil. 2. Method according to claim 1 , wherein

the free fatty acid are separated from the acylglycerol fraction after the hydrolysis, and

the lipase-catalysed hydrolysis of the separated acylglycerol fraction is repeated. 3. Method according to claim 2, wherein the same enzyme is used in the first and second hydrolysis step of the repeated hydrolysis.

4. Method of according to any of the preceding claims, wherein the lipase is a microbial lipase, preferably from a species selected from the group consisting of Pseudomonas, Candida, Rhizopus, and Rhizomucor.

5. Method according to any of the preceding claims, wherein the lipase is selected from the group consisting of Candida rugosa, Burkholderia cepacia, and Rhizopus oryzae.

6. A method according to any of the preceding claims, wherein the incubation temperature in the lipase-catalysed hydrolysis is in the range of about 30-50°C, preferably of about 40 to 50 °C, and more preferably of about 45 °C.

7. A method according to any of the preceding claims, wherein the incubation time in the lipase-catalysed hydrolysis is between 2 hours and 8 hours, preferably between 2 and 4 hours.

8. A method according to any of the preceding claims wherein said omega-3 polyunsaturated fatty acids are chosen from DHA, EPA and/or DPA.

9. A method according to any of the preceding claims, wherein said lipase discriminates in its activity between EPA and DHA.

10. A method according to any of the preceding claims wherein the hydrolysis conditions are as follows

- the water-to-oil ratio is in the range of 2:1 to 5:1 (v/w),

- the incubation temperature is in the range of about 30-50°C, and

- the incubation time is between 2 hours and 4 hours. 1. Method according to any of the preceding claims, wherein

- the lipase is added in a concentration of about 3 % by weight based on the amount of oil, and the hydrolysis conditions are as follows

- the incubation temperature is about 45°C,

- the water-to-oil ratio is about 3:1 (v/w),

- the incubation time is between 2 and 4 hours.

12. Method according to any of the preceding claims, wherein

- the lipase is added in a concentration of about 3-4 % by weight based on the amount of oil, and the hydrolysis conditions are as follows

- the incubation temperature is about 30°C,

- the water-to-oil ratio is in the range of about 2.5-5:1 (v/w),

- the incubation time is about 4 hours. 3. Method according to any of the preceding claims wherein said oil to be enriched with omega-3 polyunsaturated fatty acids which are comprised in the acylglycerols is a marine oil obtained from a fish, a crustacean, a bacterium, a macroalgae and/or a microalgae.

14. Method according to any of the preceding claims wherein the oil is obtained from one or several fish species selected from the group consisting of Salmonids, Gadoids, Clupeids, Engraulidae, Scromboids, and Elasmobranchs, which can be wild caught or farmed fish.

15. Method according to any of the preceding claims wherein the oil is from a farmed marine animal comprising a reduced amount of omega-3 polyunsaturated fatty acids compared to an oil obtained from the same species in the wild. 16. Method according to any of the preceding claims wherein the oil has a polyunsaturated fatty acid content of no more than 16 % mol, preferably no more than 14 % mol, at the start of the enrichment process with polyunsaturated omega-3 fatty acids.

17. Method according to any of the claims 1-16 wherein the acylglycerol fraction comprising the enriched omega-3 polyunsaturated fatty acids is separated from the free fatty acids after the enzymatic hydrolysis, preferably by use of short path distillation.

18. Method according to claim 17, wherein the separation comprises the following steps:

- separation of the oil and water phase by centrifugation, and

19. Method according to claim 18, wherein the oil phase is fed to a short path distillation unit at a rate of 2 mL/min under 10^'3 mbar of vacuum, wherein the feeding tank is set at 35°C, the condenser at 45°C, and the evaporator at 145°C.

20. Processed marine oil wherein the amount of polyunsaturated fatty acids in the acylglycerols of said oil has been enriched by a method according to any of the claims 1-19.

21. Processed marine oil wherein the content of omega-3 polyunsaturated fatty acids in the acylglycerols has been enriched characterised in that at least 33 mol % of the fatty acids comprised in the acylglycerols are omega-3 polyunsaturated fatty acids and the EPA/DHA ratio is 0.4 or more.

22. Processed marine oil according to claim 2 , wherein at least 55 mol % of the fatty acids comprised in the acylglycerols are omega-3 polyunsaturated fatty acids and the EPA/DHA ratio is 0.3 or more.

23. Processed marine oil according to any of the claims 20-22, wherein the oil is a fish oil, preferably from a farmed salmonid, more preferably from a farmed salmon.

24. Processed marine oil according to any of the claims 20-23, wherein the oil has an omega-3-PUFA content of less than 18 % before enrichment of the acylglycerols with omega-3 PUFA and has an omega-3-PUFA content of more than 36 % after the enrichment, preferably of more than 48 %.

25. Processed marine oil according to any of the claims 20-24, wherein the oil has a DHA content of less than 10 % before enrichment of the acylglycerols with omega- 3 PUFA and has a DHA content of more than 20 %after the enrichment, preferably of more than 30 %.

26. Processed marine oil according to any of the claims 20-25, wherein the oil has an EPA content of less than 7 % before enrichment of the acylglycerols with omega- 3 PUFA and has an EPA content of more than 8 %, preferably of more than 10 % after the enrichment. 27. Processed marine oil according to any of the claims 20-22, wherein the oil is from a pelagic fish, preferably form a species chosen from the family consisting of Clupeidae and Engraulidae, more preferably chosen from the species herring (Clupea harengus) and anchoveta (Engraulis ringens) .

28. Processed marine oil according to any of the claims 20-23 or 27, wherein the oil has an omega-3-PUFA content of less than 14 % before enrichment of the acylglycerols with omega-3 PUFA and has an omega-3-PUFA content of more than 22 % after the enrichment.

29. Processed marine oil according to any of the claims 20-23 or 27-28, wherein the oil has a DHA content of less than 8 % before enrichment of the acylglycerols with omega-3 PUFA and has DHA content of more than 13 % after the enrichment. 30. Processed marine oil according to any of the claims 20-25, wherein the oil has an EPA content of less than 6 %before enrichment of the acylglycerols with omega-3 PUFA and has an EPA content of more than 8 %.

31. Processed marine oil according to any of the claims 20-30 wherein the DHA content comprised in the acylglycerols is increased by a factor of at least 1 ,9, preferably of at least 2.5 in a lipase-catalysed omega-3 concentration process, having one hydrolysis step and by a factor of at least 3.9 in a repeated hydrolysis.

32. Processed marine oil according to any of the claim 20-31 wherein the EPA content comprised in the acylglycerols is increased by a factor of at least 1.5 in a lipase-catalysed omega-3 concentration process, having one hydrolysis step and by a factor of at least 2 when the hydrolysis is repeated.

33. Processed marine oil according to any of the claims 20-32 wherein the content of polyunsaturated fatty acids comprised in the acylglycerols is increased by a factor of at least 1 ,7, preferably of at least 2.4 in a lipase-catalysed concentration process, having one hydrolysis step and by a factor of at least 3 in a repeated hydrolysis.

34. Use of a processed marine oil according to any of the claims 20-33 as an ingredient for a feed, a functional feed, a health product, a cosmetic composition, or a pharmaceutical composition.