LWT - Food Science and Technology 126 (2020) 109284
`
`Contents lists available at ScienceDirect
`LWT - Food Science and Technology
`
`journal homepage: www.elsevier.com/locate/lwt
`
`Separation and concentration of ω-3 PUFA-rich phospholipids by hydration
`of krill oil☆
`Casey Showman, Kimberly Barnes, Jacek Jaczynski∗∗, Kristen E. Matak∗
`Division of Animal and Nutritional Sciences, West Virginia University, PO Box 6108, Morgantown, WV, 26505-6108, USA
`
`T
`
`A R T I C L E I N F O
`
`A B S T R A C T
`
`Keywords:
`Krill oil
`Omega-3 fatty acids
`Phospholipids
`Water degumming
`Nutraceuticals
`
`Krill oil (KO) is unique in contrast to fish oil because the ω-3 PUFAs, particularly eicosapentaenoic (EPA) and
`docosahexaenoic (DHA), are almost exclusively esterified on phospholipids (PL) rather than triglycerides (TAG)
`which improves bioavailability and water solubility. Therefore, the primary aim of this study was to use the
`principles of water degumming to separate PL from TAG in KO. Water was mixed with KO in varying ratios,
`centrifuged for separation and freeze-dried. Separation into gum and oil fractions occurred in the 75:25 and
`50:50 KO:H2O samples. TLC-densitometry revealed 67.6 ± 1.97% and 49.73 ± 3.90% PL concentrations
`(p < 0.05) in the lipid recovered from the 75:25 and 50:50 KO:H2O sample, respectively. The 75:25 KO:H2O
`gum fraction did not contain detectable triglycerides (TAG); however, the 50:50 KO:H2O gum fraction had
`23.88 ± 6.36% (p < 0.05). The fatty acid profile of the 75:25 KO:H2O gum fractions presented with more EPA
`and DHA than the oil fraction (p < 0.05). Phosphatidylcholine (PC) with EPA and DHA fatty acids attached
`concurrently at sn-1 and sn-2 (i.e., PC-EPA/EPA, PC-EPA/DHA, and PC-DHA/DHA) and was present in the gum
`fractions of both samples. This approach offers a simple separation technique to separate health-beneficial PL
`from KO.
`
`1. Introduction
`Cardiovascular disease (CVD) is the leading cause of death in the
`United States (Bunea, El Farrah, & Deutsch, 2004). Omega-3 poly-
`unsaturated fatty acids (ω-3 PUFAs) have demonstrated health benefits
`that may protect against it (Bunea et al., 2004). Specific ω-3 PUFAs of
`importance include eicosapentaenoic acid (EPA, 20:5 ω-3) and doc-
`osahexaenoic acid (DHA, 22:6 ω-3). Typical sources of EPA and DHA
`are fatty fish; however, krill also contains significant amounts of these
`PUFAs.
`Antarctic krill (Euphausia superba) are small crustaceans with one of
`the largest biomasses of any multicellular animal species (Nicol, James,
`& Pitcher, 1987). The lipid content, on a dry matter basis, ranges from
`12 to 50% based on maturity, season and location of harvest
`(Kolakowska, 1991; Saether, Ellingsen, & Mohr, 1986). Krill are a un-
`ique species due to the composition of lipid in their bodies. In contrast
`to fish oil, the ω-3 PUFAs, particularly EPA and DHA, in krill oil are
`almost exclusively esterified on phospholipids (PL) rather than trigly-
`cerides (TAG) (Kutzner et al., 2016; Tou, Jaczynski, & Chen, 2007;
`
`Winther, Hoem, Berge, & Reubsaet, 2010). Interestingly, some of these
`ω-3 PUFAs in KO appear uniquely esterified on a single PL molecule
`concurrently occupying sn-1 and sn-2 positions.
`PLs are most often found within cell membranes. Their structure
`consists of a polar, hydrophilic phosphate head and two hydrophobic
`fatty acid (FA) chains. The polarity and hydrophobic interactions allow
`for the formation of lipid bilayers (Gilbert, 2000). The level of satura-
`tion of the FA chain varies. Unsaturated FA allow the bilayer to be more
`fluid in colder temperatures. This is due to cis-double bonds imposing
`stress on the FA chain, which results in a non-linear shape of the FA.
`These molecules increase their stearic hindrance, which inhibits the
`chains from lining up closely (Mathews, van Holde, & Ahern, 2000).
`More saturated FA chains would cause the lipid bilayer to be more rigid
`in the colder temperatures of the krill's natural environment.
`PLs have several functions in the pharmaceutical and food produc-
`tion industries due to their amphiphilic character. PLs are commonly
`called lecithin. One of the most common uses of lecithin in the food
`industry is for stabilizing oil/water emulsions (Ceci, Constenla, &
`Crapiste, 2008). In the pharmaceutical industry, PLs serve as drug
`
`☆ This work was supported by the USDA Hatch program (WVA00722).
`∗ Corresponding author. Division of Animal and Nutritional Sciences, West Virginia University, PO Box 6108, Morgantown, WV, 26505-6108, USA.
`∗∗ Corresponding author. Division of Animal and Nutritional Sciences, West Virginia University, PO Box 6108, Morgantown, WV, 26505-6108, USA
`E-mail addresses: clshowman@mix.wvu.edu (C. Showman), kim.barnes@mail.wvu.edu (K. Barnes), jacek.jaczynski@mail.wvu.edu (J. Jaczynski),
`kristen.matak@mail.wvu.edu (K.E. Matak).
`https://doi.org/10.1016/j.lwt.2020.109284
`Received 19 November 2019; Received in revised form 9 March 2020; Accepted 13 March 2020
`Available online 15 March 2020
`0023-6438/ © 2020 Elsevier Ltd. All rights reserved.
`
`RIMFROST EXHIBIT 1174 Page 0001
`
`

`

`C. Showman, et al.
`
`delivery systems because of their unique biocompatibility and hydro-
`philic properties (Li et al., 2015). Dietary PLs have a high bioavail-
`ability and can have a protective effect on diseases such as coronary
`heart disease (Küllenberg, Taylor, Schneider, & Massing, 2012).
`There are several species of PL present in krill oils which includes
`phosphatidylcholine (PC), phosphatidylinositol
`(PI), phosphatidy-
`lethanolamine (PE), phosphatidylserine (PS), and phosphatidic acid
`(PA). The PL species are separated based on the chemical structure
`attached to the phosphate head. This structure also makes them either
`hydratable or nonhydratable. Nonhydratable PL are commonly sepa-
`rated from oil and neutral lipids through processes such as acid de-
`gumming, whereas hydratable PLs are separated with the addition of
`water (Dijkstra & Segers, 2007). The predominant PL species in krill is
`PC and interestingly, Castro-Gomez, Holgado, Rodriguez-Alacala,
`Montero, and Fontecha (2015) reported that 6.5% of the total PC in KO
`contained EPA/DHA and EPA/EPA concurrently on the sn-1 and sn-2
`positions of one PC molecule. PC is fully hydratable due to its chemical
`structure, swelling capacity, and ordered structure formation (Diez,
`1995; Larsson, 1994). The head group on a PC molecule carries a large
`dipole moment due to the negative charge on the phosphate and po-
`sitive charge on the trimethylamine group causing PC to strongly favor
`water (Dijkstra, 2018).
`Water degumming is an industrial process to separate PLs in crude
`oil and produce commercial lecithin. In the degumming process, the
`amount of water added is typically equal to that of the amount of gum
`to be separated (Braae, 1976). The process includes mixing water with
`crude oil to hydrate the polar lipids causing the oil and gum to separate.
`The resulting products are a refined oil and gum that consists of TAGs
`and PLs, respectively. Crude KO is typically extracted from krill with
`conventional organic and inorganic solvents, supercritical fluids such as
`CO2, as well as solvent-free methods (Bruheim, Tilseth, & Mancinelli,
`2018; Gigliotti, Davenport, Beamer, Tou, & Jaczynski, 2011; Katevas &
`Guerra, 2018; Sampalis & Harland, 2016; Soerensen & Jensen, 2014).
`However, these methods yield crude KO whose composition is variable
`and comprises variety of lipid classes without targeting purification/
`concentration of health-beneficial krill PLs; while some of
`these
`methods are inherently costly.
`To our best knowledge, degumming KO has not been tested for the
`separation of ω-3 PUFA-rich PLs. Therefore, the primary aim of this
`study was to use the principles of water degumming to separate phos-
`pholipids from triglycerides in krill oil by hydration. It is hypothesized
`that water degumming will be an adequate, simple method to separate
`PL-esterified ω-3 PUFAs. The secondary objective of this study was to
`verify the presence and attempt quantification of unique krill PC whose
`sn-1 and sn-2 are concurrently occupied by various permutations of
`EPA/DHA (i.e., PC-EPA/EPA, PC-EPA/DHA, and PC-DHA/DHA).
`2. Materials and methods
`2.1. Sample
`Virgin krill oil (KO) from Antarctic krill (Euphausia superba) was
`purchased from Jedwards International, Inc. (Braintree, MA, U.S.A.).
`Upon arrival, KO was stored in an amber glass bottle, flushed with ni-
`trogen, and stored at 2–5 °C throughout the experiment.
`2.2. Hydration fractionation
`KO was fractionated using deionized distilled water (ddH2O).
`Various ratios of KO and ddH2O were used. Hydration trials included
`100% KO, 75:25 KO:H2O, 50:50 KO:H2O, and 25:75 KO:H2O wt/wt.
`Krill oil and ddH2O were weighed to a final weight of 50 g into glass
`beakers, flooded with nitrogen, and covered with foil. Each ratio was
`stirred at 200 rpm for 24 h at room temperature (~20 °C). After 24 h,
`each sample was placed into a centrifuge tube. Water soluble and in-
`soluble fractions were separated by centrifuging at 5000 g for 5 min at
`
`LWT - Food Science and Technology 126 (2020) 109284
`
`22 °C. Centrifugation generally resulted in two visual fractions, gum
`and oil, that were manually decanted. Special attention was paid to
`prevent fraction contamination during decanting. Separated gum and
`oil fractions were freeze dried (VirTis, Model #35L) to remove H2O
`followed by storage at −80 °C. Each of the four ratios were completed
`in triplicates. Weights were recorded after stirring, separation of each
`fraction (i.e., gum and oil after centrifugation and decanting), and
`freeze-drying.
`2.3. Thin layer chromatography and densitometry
`Thin Layer Chromatography (TLC) was applied to separate lipid
`classes using a method described by Gigliotti et al. (2011). Fractions
`separated as described above were diluted to 5 mg/mL in a 1:1
`chloroform: methanol solvent (vol:vol). Twenty μL of the mixture was
`pipetted onto a silica TLC plate (Merck TLC Silicagel 60 plates with
`60 Å pore size, Darmstadt, Germany). For the mobile phase, the TLC
`plate was developed in a hexane: diethyl ether: acetic acid (80:20:1.5;
`vol:vol:vol) for approximately 1 h until the solution reached 2.5 cm
`from the top of the plate. After drying for 5 min, TLC plates were
`sprayed with 50% sulfuric acid in water solution (1:1 vol:vol) and
`placed in a 120 °C drying oven for 45 min to develop spots indicating
`lipid classes. Standards for mono-, di-, and triglycerides, phospholipids,
`cholesterol (Supelco, Bellefonte, PA), and free fatty acids (oleic acid, Nu
`Chek Prep, Elysian, MN) were also plated and run for identification of
`each lipid class. TLC was applied to all fractions from each of the four
`KO:H2O ratio tested.
`For densitometry analysis, images of the TLC plates were captured
`with Gel Doc XR + gel imaging system (Bio-Rad Gel Doc XR+ #
`170–8170 and ChemiDoc XRS + Imaging Software Version 6,
`California, United States) using transluminating white light and pixel
`density was measured. Densitometry values are expressed as percent
`lipid class (mean ± SD) of the total bands detected within the TLC
`lane.
`2.4. Fatty acid profile analysis
`2.4.1. Extraction
`The Bligh and Dyer (1959) method using chloroform-methanol
`(C:M) mixture (2:1 v/v) was followed to extract total lipids from each
`fraction (i.e., gum and oil). A 0.05 g sample was used from each fraction
`for analysis. Trizma/EDTA buffer (50 mM, pH = 7.4) and a chloroform:
`methanol: glacial acetic acid (C:M:A; 400:200:3 mL) solvent were
`added. Nonadecanoic acid (19:0) was added (125 μL) as an internal
`standard for quantification of fatty acids. Samples were incubated at
`room temperature for 10 min, then centrifuged at 900 g for 10 min at
`10 °C. The separated lower layer was filtered through a pre-rinsed
`Whatman 1-PS filter (Whatman 2200090, Buckinghamshire HP7, 9NA,
`UK). The upper layer was run through extraction methods using a 4:1
`chloroform: methanol solvent, centrifuged to separate, and again the
`lower layer was filtered through the same filter. Filtrates were flushed
`with nitrogen gas at 60 °C to dry.
`2.4.2. Methylation
`The extracted lipid was transmethylated according to Fritshe and
`Johnston (1990). Methylation occurred by adding 4 mL of 4% H2SO4, in
`anhydrous methanol, to the lipid and incubating in a water bath at
`90 °C for 60 min. Deionized distilled water was added to stop the re-
`action after incubation period. Chloroform was added to extract fatty
`acid methyl esters (FAMEs), and mixture was centrifuged at 900 g for
`10 min at 10 °C. The separated lower layer was filtered through an-
`hydrous Na2SO4. The collected layer was dried with nitrogen gas at
`60 °C, diluted with isooctane, and stored at −20 °C until analyzed.
`2.4.3. Fatty acid profile and quantification
`Fatty acid methyl esters were analyzed using a Varian CP-3800 Gas
`2 RIMFROST EXHIBIT 1174 Page 0002
`
`

`

`C. Showman, et al.
`
`Instruments; Walnut Creek,
`Chromatograph (Varian Analytical
`California, U.S.A) equipped with a flame ionization detector (FID;
`Varian Inc., Walnut Creek, California, U.S.A.). A silica capillary column
`(100 m length, 0.25 mm diameter) was used to separate the FAMEs. A
`method of 140 °C held for 5 min followed by a temperature ramp of
`4 °C/min to 220 °C held for 15 min was adopted; totaling to 85 min for
`each FAME separation. Temperatures were held at 270 °C and 300 °C
`for the injector and detector respectively. Identification of FAMEs was
`based on retention times compared to FAME 37 standard (SupelcoTM
`quantitative standard FAME 37; Sigma-Aldrich, St. Louis, Missouri,
`U.S.A.). The Star GC workstation version 6 software (Varian Inc.,
`Walnut Creek, California, U.S.A.) was used to determine the peak area
`and relative amounts of each fatty acid in the samples. Fatty acids are
`expressed as mg FA/g sample based on quantified amounts (mg/g).
`
`2.5. Ultra high-performance liquid chromatography mass spectrometry
`2.5.1. Sample preparation
`Quantification of phosphatidylcholine (PC) in the gum fractions was
`accomplished by reverse phase ultra high-performance liquid chroma-
`tography mass spectrometry (UHPLC-MS). The methodology is similar
`to that developed by Winther et al. (2011). Briefly, each sample and a
`stock solution of PC-free matrix were dissolved in 60:40 (vol:vol) di-
`chloromethane:methanol mixture to a concentration of 1 mg/mL.
`Phosphatidylcholine
`22:6/22:6
`standard
`(Avanti Polar
`Lipids,
`Alabaster, AL) was used as an external calibrator. The standard was
`mixed into PC free matrix solution in a set of dilutions with con-
`centrations of 1, 5, 10, 50, and 100 μg/mL. A calibration curve of
`chromatographic peak area to standard concentration was established
`from set of dilutions to quantify krill oil samples.
`
`2.5.2. HPLC setup and conditions
`Chromatographic separation was performed in reverse phase on a
`Thermo Accela UHPLC with an Agilent Eclipse XDB-C18 column with
`particle diameter of 5 μm. Column dimensions were 2.1 × 100 mm.
`There were two mobile phases used to elucidate the PC. Phase A con-
`sisted of 90% water with 10% formic acid while phase B consisted of
`60% methanol, 40% acetonitrile, and 0.1% formic acid. A linear gra-
`dient of the two phases, similar to that of Winther et al. (2011), was
`used for separation of PC species. Briefly, for 4 min after injection the
`system was kept constant at 65% mobile phase B and then increased to
`100% mobile phase B for 22 min. The mobile phase flow rate was set to
`0.26 mL/min.
`
`2.5.3. Mass spectrometry
`Detection of PC species was obtained using a Thermo Q-Exactive
`with method setup, data acquisition, and quantification done by soft-
`ware Xcalibur 3.0 (Thermo Scientific, USA). The mass spectrometer ran
`at a voltage of 4.00 kV, sheath gas flow rate of 30 (arbitrary units),
`auxiliary gas flow rate of 5, and sweep gas of 5. Capillary temperature
`was set to 200 °C. Mass spectrometry resolution was set at 70,000 (m/z
`200) for full mass scan and 17,500 for tandem mass. Mass scan range
`was set to 100–1000 m/z in profile mode. Parallel reaction monitoring
`analysis was used. Both the precursor ions and the fragment ions were
`selected to monitor the transition using the ratio of respective peak
`areas to quantify each PC species. Masses for each PC species, de-
`termined by previous studies (Castro-Gomez et al., 2015), were used to
`identify the molecules by mass spectrometry.
`The precursor m/z window was set as ± 1. A full scan of precursor
`ions was carried out. Then the selected precursor ions were subjected to
`high-energy collision dissociation (HCD), followed by the selection of
`specific product ions in the Orbitrap Mass Analyzer (Thermo Scientific,
`USA).
`
`LWT - Food Science and Technology 126 (2020) 109284
`
`2.6. Statistical analysis
`Hydration fractionation trials were performed in triplicates (n = 3).
`At least duplicate TLC with densitometry and fatty acid profile mea-
`surements were performed for each triplicate fractionation trial.
`UHPLC-MS was conducted once for each triplicate. SAS JMP Pro ver-
`sion 13 (SAS Institute Inc., North Carolina, USA) was used for all ana-
`lyses. One-way analysis of variance (ANOVA) with post-hoc analysis
`using Tukey-Kramer's test (p < 0.05) was used to determine statistical
`differences.
`3. Results and discussion
`3.1. Fractionation of krill oil (KO) by hydration
`Fractionation into water soluble and insoluble fractions, or “gum”
`and “oil” fractions, respectively, occurred only in the 75:25 KO:H2O
`and 50:50 KO:H2O ratios. The 25:75 KO:H2O ratio formed an oil/water
`dispersion and since no fractionation occurred after centrifugation,
`further results are not reported. When a 50:50 KO:H2O ratio was used to
`separate lipids, the starting weight of krill oil was 25 g. After separa-
`tion, the weight of the oil fraction was 2.50 ± 1.36 g and the weight of
`the gum fraction was 20.53 ± 1.16 g. On the other hand, when a 75:25
`KO:H2O ratio was used the starting weight of the krill oil was 37.5 g.
`After separation, the weight of the oil fraction was 11.78 ± 1.42 g and
`the weight of the gum fraction was 19.83 ± 2.28 g. There was a minor
`but inevitable loss of sample due to the transfer from container to
`container in each step of the process because of the viscosity of KO.
`There were obvious visual differences in gum fractions obtained
`with the different KO:H2O ratios. These differences were not measured,
`but the 50:50 KO:H2O gum fraction had a similar appearance to the
`emulsion created by the 25:75 KO:H2O ratio, while the 75:25 KO:H2O
`gum fraction had a dark red and sludge-like appearance. This difference
`was likely due to the entrapment of oil in structures formed in the
`hydration of the 50:50 KO:H2O sample rather than the precipitation of
`polar lipids.
`Water is a polar solvent that is immiscible in oil; therefore, when
`water interacts with PLs they also become insoluble in the oil when
`hydrated. When there is an optimal ratio of KO to water the PLs form
`aggregates that are oil insoluble and can be separated from the oil (Lei,
`Ma, Kodali, Liang, & Davis, 2003). Lei et al. (2003) demonstrated a
`ternary phase diagram of soybean oil-water-soybean PC to assist in
`determination the appropriate amount of water to be added during
`degumming. They also showed that when an excess of water is added,
`above the PL saturation point, water droplets penetrate liposomes and
`disperse in the oil. This is likely why the 25:75 KO:H2O ratio had no
`separation, as well as the 50:50 KO:H2O ratio was approaching the
`saturation point of the PLs found in KO.
`3.2. Lipid class profiles
`Based on thin layer chromatography (TLC) and densitometry, water
`fractionation of PLs from neutral lipids, such as TAGs, was possible. TLC
`displayed how each major lipid class in KO fractionated after the ad-
`dition of water (Fig. 1). Densitometry gave the relative distribution of
`lipid classes within each fraction. Data are reported in Table 1 and are
`expressed as % lipid class ± standard deviation in that particular
`fraction. The 100% KO contained 19.78 ± 3.43% TAG and
`46.28 ± 0.32% PLs. The relative distribution of PLs is consistent with
`reported ranges between 46 and 64% (Kolakowska, 1991; Castro-
`Gomez et al., 2015, and Xie et al., 2017).
`Undetectable amounts of PLs remained in the 50:50 and 75:25
`KO:H2O oil fraction,
`likely because the PLs in KO are comprised
`of > 90% phosphatidylcholine (PC) which is fully able to absorb water
`(Table 1; Akanbi & Barrow, 2018; Castro-Gomez et al., 2015; Dijkstra &
`Segers, 2007). In addition, the 75:25 KO:H2O ratio gave rise to better
`3 RIMFROST EXHIBIT 1174 Page 0003
`
`

`

`C. Showman, et al.
`
`LWT - Food Science and Technology 126 (2020) 109284
`
`Fig. 1. Thin layer chromatography (TLC) displaying lipid classes for 100% krill oil (KO) and the 25:75 KO:H2O, 50:50 KO:H2O, and 75:25 KO:H2O and fractions
`separated by water degumming principles as well as standards used for identification. Along the x-axis, the lanes represent: MG DAG TAG = mono-, di-, triglyceride
`standard; FFA = free fatty acid (oleic acid) standard, cholesterol standard, PL = phospholipid standard, 100% KO, 25:75 KO:H2O (fractionation for oil and gum did
`not occur), 50:50 KO:H2O oil fraction, 50:50 KO:H2O gum fraction, 75:25 KO:H2O oil fraction, 75:25 KO:H2O gum fraction. Along the y-axis, lipid classes are
`abbreviated as TAG (triglyceride), FFA (free fatty acid), CHOL (cholesterol), DAG (diglycerides), and PL (phospholipids).
`
`separation of lipid classes with more concentrated PL (approximately
`68%) and undetectable amounts of TAG in the gum fraction; whereas
`the 50:50 KO:H2O gum fraction had a less concentrated PL content
`(approximately 50%) due to the presence of about 24% TAGs. The oil
`fractions of both 75:25 KO:H2O and 50:50 KO:H2O ratios contained
`mainly TAGs (> 80%).
`In addition, KO typically contains high and varying levels of free
`fatty acids (FFAs). Since krill are crustaceans, they also contain some
`levels of cholesterol (CHOL). Relatively high levels of FFAs were pre-
`sent
`in both 50:50 (14.51 ± 3.28%) and 75:25 KO:H2O
`(15.41 ± 1.52%) gum fractions. Unrefined oils typically contain some
`levels of FFAs. In this regard, unrefined KO usually contains much
`higher and more variable amounts of FFAs than oils from other sources.
`In addition to causing overall quality deterioration, FFAs are also
`considered impurities because they may promote oil hydrolytic de-
`gradation. FFAs may be removed by a neutralization step during oil
`refining. If neutralization is applied to KO, FFAs can be removed; and
`therefore, krill PL-esterified EPA/DHA can be further purified/con-
`centrated. If one assumes data presented in Table 1 for the 75:25
`
`KO:H2O gum fraction, removal of 15.41 ± 1.52% FFAs by neu-
`tralization could result in approximately 80% of PLs in that fraction
`when compared to 68% without neutralization. This increase in PLs
`would result is a proportional increase of EPA/DHA as well. Although
`the content of FFAs in unrefined KO is typically high as well as highly
`variable, neutralization step of KO would be beneficial in terms of
`higher content of PL-esterified EPA/DHA. Oil neutralization is rela-
`tively simple and a follow up study for neutralized KO is forthcoming.
`
`3.3. Fatty acid profile
`Major fatty acids (FAs) with content greater than 1 mg/g found in
`the 100% KO, 75:25 KO:H2O, and 50:50 KO:H2O ratios and their re-
`spective fractions (gum/oil) are listed in Table 2. These FAs include
`C14:0, C16:0, C16:1, C18:1, C18:3 (ALA), C20:2, C20:5 (EPA), and
`C22:6 (DHA). Results are presented as mg FA in g FAs ± standard
`deviation. There were significant differences in FA profiles for each oil
`and gum fraction (p < 0.05).
`For both the 50:50 KO:H2O and 75:25 KO:H2O samples, the gum
`
`Table 1
`Relative distribution of lipid classes in respective fractions (% lipid class/sample or fraction) of 100% krill oil (KO) and the 50:50 KO:H2O and 75:25 KO:H2O
`fractions separated by water degumming principles, determined by TLC-densitometry.
`Lipid Class
`CHOL
`FFA
`TAG
`19.78 ± 3.43B
`100% Krill Oil
`14.09 ± 0.20AB
`14.19 ± 4.84A
`50:50 KO:H2O
`82.38 ± 0.19A
`Oil
`2.69 ± 0.97D
`4.74 ± 1.28B
`11.87 ± 0.55C
`14.51 ± 3.28A
`23.88 ± 6.36B
`Gum
`75:25 KO:H2O
`80.67 ± 3.70A
`Oil
`2.97 ± 0.55D
`4.58 ± 0.29B
`15.54 ± 1.52A
`n.d.
`Gum
`16.84 ± 0.87A
`TAG-triglycerides, FFA-free fatty acids, CHOL-cholesterol, DAG-diglycerides, PL-phospholipids. n.d. Not detected.
`A,B,C Different letters mean significant differences within a column for a given class among fractions and ratios (p < 0.05).
`4 RIMFROST EXHIBIT 1174 Page 0004
`
`PL
`46.28 ± 0.32B
`n.d.
`49.73 ± 3.90B
`n.d.
`67.62 ± 1.97A
`
`DAG
`5.63 ± 0.89B
`10.19 ± 0.15A
`n.d.
`11.78 ± 2.57A
`n.d.
`
`

`

`C. Showman, et al.
`
`LWT - Food Science and Technology 126 (2020) 109284
`
`Table 2
`Profile of major fatty acid composition (mg fatty acid/g sample or fraction) of 100% krill oil (KO) and the 50:50 KO:H2O and 75:25 KO:H2O and fractions separated
`by water degumming principles.
`75:25 KO:H2O OIL
`50:50 KO:H2O GUM
`50:50 KO:H2O OIL
`100% KO
`7.54 ± 0.62b
`C14:0
`14.21 ± 0.86a
`5.95 ± 0.43bc
`14.57 ± 0.56a
`20.05 ± 0.37c
`21.32 ± 0.19ab
`20.50 ± 0.37bc
`21.14 ± 0.29abc
`C16:0
`6.78 ± 0.49a
`3.16 ± 0.14b
`6.98 ± 0.18a
`3.84 ± 0.14b
`C16:1 (C9)
`12.56 ± 0.61a
`7.43 ± 0.29b
`13.22 ± 0.19a
`8.42 ± 0.20b
`C18:1 (C9)
`5.17 ± 0.13a
`4.20 ± 0.01b
`5.14 ± 0.12a
`4.38 ± 0.09b
`C18:3 (C9,12,15)
`13.94 ± 0.30a
`7.65 ± 0.15c
`14.05 ± 0.19a
`8.91 ± 0.06b
`C20:2 (C11,14)
`14.12 ± 2.01c
`27.60 ± 0.73ab
`12.44 ± 0.32c
`24.48 ± 0.65b
`C20:5 (EPA)
`6.47 ± 0.90c
`16.77 ± 0.50b
`5.76 ± 0.06c
`14.94 ± 0.51b
`C22:6 (DHA)
`a,b,c Different letters mean significant differences within a row for a given FA among fractions and ratios (p < 0.05).
`
`75:25 KO:H2O GUM
`4.21 ± 1.22c
`21.67 ± 0.70a
`2.10 ± 0.54c
`5.82 ± 0.75c
`3.85 ± 0.15c
`5.88 ± 0.73d
`30.98 ± 2.34a
`19.82 ± 1.69a
`
`Fig. 2. Types of fatty acids found in 100% krill oil (KO) as well as oil and gum fractions of the 75:25 KO:H2O ratio trial. (A) Comparison of SFA (saturated fatty acids),
`UFA (unsaturated fatty acids), total ω3 (omega-3 fatty acids), and total ω6 (omega-6) fatty acids found in each sample. 100% KO is included as a reference for the FA
`comparison. Data are presented as mg/g FA in total FAs. (B) Comparison of SFA/UFA and ω6/ω3 ratios based on FA content of KO and separated 75:25 KO:H2O oil
`and gum fractions. a,b,c Different letters on the top of data bars indicate statistical differences (p < 0.05).
`
`fractions contained a greater concentration of EPA and DHA than the oil
`fractions (p < 0.05). Within all fractions and initial KO, the 75:25
`KO:H2O gum fraction had the greatest concentration of EPA and DHA
`(p < 0.05). The 50:50 KO:H2O ratio contained a lower concentration
`
`of EPA and DHA than the 75:25 KO:H2O sample, but only the DHA
`content was statistically different (p < 0.05). There were no differ-
`ences (p > 0.05) in FA profiles in the oil fractions tested. The con-
`centration of EPA and DHA in the fractions are consistent with others
`
`5 RIMFROST EXHIBIT 1174 Page 0005
`
`

`

`C. Showman, et al.
`
`LWT - Food Science and Technology 126 (2020) 109284
`
`Fig. 3. Representative mass spectrum of 75:25 KO:H2O ratio gum fraction. PC species of interest were identified by m/z. Each are labeled in figure. Relative
`intensities are shown on the y-axis with the mass on the x-axis.
`
`Table 3
`Identified phosphatidylcholine (PC) species containing EPA (20:5) and DHA
`(22:6) with the most abundant PC species by mass of molecule.
`PC species
`75:25 KO:H2O gum fraction
`Concentration (μg/mL)
`Mass (m/z)
`Relative Intensity
`54.77 ± 2.90
`780.56
`45.27 ± 1.07
`PC20:5/20:5
`48.98 ± 4.31
`828.56
`40.41 ± 0.86
`PC20:5/22:6
`15.30 ± 1.08
`852.56
`12.63 ± 0.08
`PC22:6/22:6
`121.12 ± 8.71
`878.57
`100
`PC16:0/20:5
`Mass of the PC species is expressed in Daltons. Data are expressed as mean ±
`standard deviation for PC species relative intensity and concentration.
`that reported the FA profiles of the individual lipid classes. Araujo, Zhu,
`Breivik, Hjelle, and Zeng (2014) separated neutral and polar lipids with
`different types of solvents and the fatty acid profiles, based on % nor-
`malized mole units, showed similar results. The FA composition of PL in
`the krill oil was 28.5% EPA and 13.5% DHA whereas there was 9.7%
`EPA and 3.8% DHA in the TAG (Araujo et al., 2014). Other studies
`reported much lower amounts of EPA and DHA bound to TAG with
`ranges of 1–4% for each fatty acid (Castro-Gomez et al., 2015; Gigliotti
`et al., 2011; Xie et al., 2017). The greater EPA and DHA content in the
`oil fraction may be due to diglycerides (DAGs) separated into this
`fraction (Table 1). Castro-Gomez et al. (2015) examined isolated frac-
`tions of KO and reported the DAG-rich fraction had 19.8% EPA and
`6.1% DHA. The content and distribution of these PUFAs in krill oil are
`what make the product interesting. A PL-rich fraction with a high
`content of ω-3 PUFAs can be a marketable food product.
`The 75:25 KO:H2O oil and gum fractions contained different lipid
`classes; and thus, different types of FAs (Fig. 2A). The FA profile of the
`gum fraction presented a greater concentration of ω-3 PUFAs
`(p < 0.05) than the oil fraction (Fig. 2A). The gum fraction was
`composed of approximately 55 mg/g ω-3 FAs while the oil fraction
`contained 26 mg/g when totaling all ω-3 FA present. The gum fraction
`also had approximately four times lower ω-6/ω-3 ratio when compared
`to the oil fraction (Fig. 2B). The ratios for oil and gum fractions are 0.65
`and 0.16, respectively.
`Fatty acid composition is important especially in relation to human
`health. Diets with a higher ω-6/ω-3 ratio are a promoter of cardio-
`vascular and inflammatory diseases (Simopoulos, 2002). Studies in-
`dicate that a lower level of ω-6 FAs in the blood decreases the occur-
`rence of cardiovascular disease (Lands, 2003). The recommended ratio
`of ω-6/ω-3 fatty acids is 5:1, but the Western diet is far below the
`optimal intake (Küllenberg et al., 2012). At the same time, the oil
`
`fraction had a greater saturated fatty acid/unsaturated fatty acid (SFA/
`UFA) ratio than the gum fractions of the 75:25 KO:H2O samples
`(Fig. 2B). The SFA/UFA ratios were 0.58 and 0.38 for the oil and gum
`fractions of the 75:25 KO:H2O, respectively.
`
`3.4. Identification and quantification of phosphatidylcholine (PC) species
`After hydration and fractionation of the PLs, ultra-high performance
`liquid chromatography mass spectrometry (UHPLC-MS) was used to
`identify and quantify the phosphatidylcholine (PC) species that con-
`currently contained EPA and DHA on sn-1 and sn-2 positions in the gum
`fraction of the 75:25 KO:H2O sample. Fig. 3 shows a representative
`mass spectrum in the range where the PC species of interest are iden-
`tified. Each PC molecule is identified with an arrow. Concentration (μg/
`mL) and relative intensity data for each PC species are shown in
`Table 3. Results indicate the presence of PC-20:5/20:5, PC-20:5/22:6,
`and PC-22:6/22:6 which are consistent with previous studies that found
`PC species in KO do have both EPA and DHA incorporated onto the
`same molecule (Castro-Gomez et al., 2015; Le Grandois et al., 2009;
`Winther et al., 2011). Typically, PLs contain a saturated and un-
`saturated FAs attached to sn-1 and sn-2, respectively. Having two highly
`unsaturated and long chain FAs such as EPA and DHA concurrently
`esterified on one PC molecule is unlikely because of stearic hindrance
`that these FAs exhibit (Berdanier, 2008). It was not determined if the
`EPA or DHA was on the sn-2 position; however, PC species with these ω-
`3 PUFAs on the sn-2 position can increase cardiovascular health as well
`as brain function (Parmentier, Mahmoud, Linder, & Fanni, 2007).
`Of the PLs of interest, PC20:5/20:5 was the most abundant, followed
`by PC20:5/22:6 in the gum fraction of the 75:25 KO:H2O sample
`(p < 0.05). Results are consistent with Winther et al. (2011) who
`reported similar order of abundance with only slightly lower relative
`intensities for each of the PC species of interest. A similar method of
`quantification was used in the previously mentioned study, but only
`relative intensity was reported by Winther et al. (2011). In addition to
`relative intensity, Table 3 also lists concentration of the PC species of
`interest.
`PLs have a different digestibility; and therefore, bioavailability
`when compared to TAGs. Due to the amphiphilic nature of PLs, they are
`not dependent on bile salts forming micelles for lipid digestion.
`Intestinal absorption of PLs is much higher than TAGs as well. Up to
`20% of dietary PLs can be absorbed passively in the intestine
`(Zierenberg & Grundy, 1982). Supplementation of PLs, especially ones
`containing EPA and DHA, will affect blood lipid profiles (Küllenberg
`et al., 2012). Incorporating these PLs esterified with EPA/DHA into
`6 RIMFROST EXHIBIT 1174 Page 0006
`
`

`

`C. Showman, et al.
`
`food or pharmaceutical products would enhance dietary intake of
`beneficial FAs. Lecithin, the name for commercial PLs,