The Polar Lipidome of Cultured Emiliania huxleyi: A Source of Bioactive Lipids with Relevance for Biotechnological Applications

Polar lipids from microalgae have aroused greater interest as a natural source of omega-3 (n-3) polyunsaturated fatty acids (PUFA), an alternative to fish, but also as bioactive compounds with multiple applications. The present study aims to characterize the polar lipid profile of cultured microalga Emiliania huxleyi using hydrophilic interaction liquid chromatography coupled with high-resolution mass spectrometry (HILIC–MS) and fatty acids (FA) analysis by gas chromatography (GC–MS). The lipidome of E. huxleyi revealed the presence of distinct n-3 PUFA (40% of total FA), namely docosahexaenoic acid (22:6n-3) and stearidonic acid (18:4n-3), which give this microalga an increased commercial value as a source of n-3 PUFA present in the form of polar lipids. A total of 134 species of polar lipids were identified and some of these species, particularly glycolipids, have already been reported for their bioactive properties. Among betaine lipids, the diacylglyceryl carboxyhydroxymethylcholine (DGCC) class is the least reported in microalgae. For the first time, monomethylphosphatidylethanolamine (MMPE) has been found in the lipidome of E. huxleyi. Overall, this study highlights the potential of E. huxleyi as a sustainable source of high-value polar lipids that can be exploited for different applications, namely human and animal nutrition, cosmetics, and pharmaceuticals.


Introduction
Microalgae have attracted great interest in various biotechnological applications, including human and animal nutrition, cosmetics, and pharmaceuticals [1], representing a sustainable feedstock to supply new products and materials, as well as various high-value compounds. Furthermore, different strains of microalgae can grow rapidly in well-controlled production systems (e.g., photobioreactors) [2], allowing scaling to high volumes for large-scale industrial production of microalgal biomass.

Biomass Production
The strain of Emiliania huxleyi RCC1250 used was obtained from Roscoff Culture Collection (RCC; strain: AC453; origin: Alboran sea, Western Mediterranean). The starter inoculum was grown in 50 mL Erlenmeyer flasks in a climatic chamber (20 • C with 12:12 h light:dark cycle, using an irradiance of 50 µmol m −2 s −1 ) and the culture medium recommended by RCC, K/2 medium modified by Ian Probert [31]. Thereafter, the concentrated inoculum was scaled up to 500 mL and 5 L round flasks supplemented with NaHCO 3 (0.087 g L −1 ) and modified algae medium [32] to a final nitrate concentration of 0.4 mM. All cultures were grown at the standard room temperature of the laboratory (22 ± 2 • C) under natural light without aeration, for about 15 days. All flasks were agitated manually, two times a day, to keep the cells in suspension. Cultures were monitored by optical density (750 nm) and microscopic observations. Samples were harvested during the late exponential growth phase by centrifugation (1735× g for 15 min) and freeze-dried (LyoQuest Telstar, Terrassa, Spain) for the proximal composition and lipidome analysis.

Moisture and Ash Determination
The residual moisture of freeze-dried biomass was determined by drying samples (50 mg × 3 replicates from a bulk sample) at 105 • C for 15 h in ceramic crucibles. After cooling to room temperature on a desiccator, the crucibles were then weighed to determine moisture by gravimetry. For the ash determination [33], the same biomass samples (previously dried at 105 • C) were incinerated in a muffle furnace at 575 • C for 6 h, cooled to room temperature, and weighed. In complement to this common method used in a muffle, the ash content was also determined by thermogravimetry (TG), using a protocol optimized for microalgae as follows: 5-10 mg of biomass submitted under an air atmosphere to a heating rate of 20 • C min −1 from room temperature to 600 • C (held for 30 min) [34].

Total Sugar Content Determination
Emiliania huxleyi biomass (2 mg × 3 replicates) was subjected to a prehydrolysis with 0.2 mL of 72% H 2 SO 4 (w/w) for 3 h at room temperature, followed by 2.5 h hydrolysis with 1 M H 2 SO 4 at 100 • C. Total sugar content was then determined by the phenol-sulfuric acid method [35]. For that, 1 mL of concentrated sulfuric acid was added to 80 µL of biomass hydrolysate in an ice bath, followed by addition of 150 µL of phenol solution (5%, w/v). Blanks were prepared without the addition of the phenol solution. After vigorous manual shaking, the tubes were placed in a bath at 100 • C for 10 min and then cooled in an ice bath. A calibration curve was prepared by performing the same procedure with glucose standards (0-0.6 mg mL −1 ). After new shaking, the absorbance of samples and standards was measured at 490 nm on a microplate UV-Vis spectrophotometer.

Neutral Sugars and Uronic Acids Analysis
Emiliania huxleyi biomass (2 mg × 3 replicates) was subjected to the same prehydrolysis and hydrolysis conditions described for phenol-sulfuric acid method. For determination of uronic acids (UA) content by m-phenylphenol colorimetric method [36], a volume of 500 µL of biomass hydrolysate was recovered after 1 h of hydrolysis at 100 • C with 1 M H 2 SO 4 and diluted in 1 mL of distilled water. d-Galacturonic acid solutions (0-80 µg mL −1 ) were used to construct a calibration curve. To each tube containing 500 µL of sample/standard, 3 mL of 50 mM sodium borate prepared in concentrated sulfuric acid was added. After vigorously shaking, the tubes were placed in a bath at 100 • C for 10 min, followed by an ice bath. Then, 100 µL of MFF (m-phenylphenol 0.15% (w/v) in 0.5% (w/v) NaOH) was added to two of the three tubes of each sample and standard (replicates), followed by shaking and incubation in the dark for 30 min. The absorbance of each tube was measured at 520 nm after homogenization.
For neutral sugars analysis, after hydrolysis at 100 • C for a total of 2.5 h, 200 µL of internal standard 2-desoxyglucose (0.1 mg mL −1 ) was added. Alditol acetates were prepared from 1 mL of each hydrolysate. After neutralization with 200 µL of NH 3 25%, alditol derivatives were obtained by reduction with 100 µL of NaBH 4 (15% (w/v) in 3 M NH 3 ) at 30 • C for 1 h. After cooling in an ice bath, the excess of borohydride was destroyed by the addition of glacial acetic acid (2 × 50 µL). The alditol derivatives (only 300 µL) were then acetylated with 3 mL of acetic anhydride in the presence of 450 µL of 1-methylimidazol at 30 • C for 30 min. To decompose the excess of acetic anhydride, distilled water (3 mL) was added with tubes in an ice bath. Alditol acetates were then extracted by adding 2.5 mL of dichloromethane. After vigorous manual shaking and centrifugation for 1 min at 1400× g, the aqueous phase was removed. An additional volume of dichloromethane (2.5 mL) and water (3 mL) was added, and the aqueous layer was removed using the same procedure. The organic phase was washed two times by the addition of distilled water (2 × 3 mL) and then evaporated to dryness. The dried material was dissolved in anhydrous acetone (2 × 1 mL), followed by the evaporation of acetone to dryness. Gas chromatography with a flame ionization detector (GC-FID) was used to identify alditol acetates [37]. The gas chromatograph (Perkin-Elmer Clarus 400) was equipped with a DB-225 capillary column (Agilent J&W GC columns, USA) with 30 m of length, 0.25 mm of internal diameter and 0.15 µm of film thickness. The injector temperature was 220 • C and the detector temperature was 230 • C. The oven was programmed for an initial temperature of 200 • C for 1 min, raised at 40 • C min −1 to 220 • C (held for 7 min), then raised at 20 • C min −1 to 230 • C (held for 1 min). Hydrogen was used as carrier gas at a 1.7 mL min −1 flow rate.

Nitrogen Determination and Protein Estimation
Nitrogen content of freeze-dried samples (2 mg × 3 replicates) was obtained by elemental analysis on a Leco Truspec-Micro CHNS 630-200-200 elemental analyzer at combustion furnace temperature 1075 • C and afterburner temperature 850 • C. Nitrogen was detected using thermal conductivity. The protein content was estimated from the nitrogen determination using 4.78 as nitrogen-protein conversion factor [38].

Lipid Extraction
Lipids were extracted using an adapted Folch method [23]. Microalgal biomass (20 mg × 3 replicates) was mixed with 2 mL of CH 2 Cl 2 :CH 3 OH (2:1) in a glass PYREX tube and homogenized by vortexing 2 min. After incubation at 30 • C for 30 min, the mixture was centrifuged (Selecta JP Mixtasel, Abrera, Barcelona, Spain) for 10 min at 626× g and the organic phase was collected in a new glass tube. The biomass residue was re-extracted twice with 2 mL of CH 2 Cl 2 :CH 3 OH (2:1) until a colorless pellet was obtained. The combined organic phases were dried under N 2 stream.
To eliminate nonlipid contaminants, extracts were redissolved in 2 mL CH 2 Cl 2 and 1 mL CH 3 OH. After vortexing for 1 min, 0.75 mL of Milli-Q water was added. The mixture was then vortexed for 1 min to allow mass transfer from the polar to nonpolar phase, followed by phase separation by centrifugation at 626× g for 10 min. The organic phase was collected in a new glass tube and the aqueous phase was re-extracted with 2 mL of CH 2 Cl 2 . The combined organic phases were then transferred to preweighed amber vials, dried under a stream of N 2 , weighed, and stored at −20 • C. Lipid content was estimated as a dry weight percentage.

Glycolipids and Phospholipids Quantification
Glycolipid quantification was achieved by the orcinol colorimetric method [39,40]. Briefly, 2 mL of orcinol solution (0.2% in 70% H 2 SO 4 ) was added to 200 µg of each dried lipid extract (n = 3). After incubation at 80 • C for 20 min, samples were cooled to room temperature and absorbance was read at 505 nm. The sugar amount in the lipid extract was determined from a calibration curve prepared by performing the same reaction on known amounts of glucose (up to 40 µg, from an aqueous solution containing 5 mg mL −1 of sugar). The content of glycolipids was calculated by multiplying the amount of sugar by 2.8 [40,41].
Phospholipids were quantified by a molybdovanadate method [42], as routinely performed in the authors' laboratory [25,43]. For that, 125 µL of 70% perchloric acid was added to 200 µg of each dried lipid extract (n = 3), previously transferred to a glass tube washed with 5% nitric acid. Samples were then incubated at 180 • C for 1 h on a heating block. After cooling to room temperature, 825 µL of Milli-Q water, 125 µL of ammonium molybdate (2.5 g 100 mL −1 in Milli-Q water), and 125 µL of ascorbic acid (0.1 g mL −1 in Milli-Q water) were added to each sample, with vortex mixing between each addition. Samples were incubated in a water bath at 100 • C for 10 min, and then immediately cooled down in a cold-water bath. Phosphate standards from 0.1 to 2 µg of phosphorus (P) were prepared from a sodium dihydrogen phosphate dihydrate (NaH 2 PO 4 ·2H 2 O, 100 µg mL −1 of P), using the same experimental procedure as samples without the heating block step. The absorbance was measured at 797 nm. For each lipid extract, the amount of total phospholipid was calculated by multiplying the amount of P (determined by linear regression) by 25. In both methods, the absorbance of standards and samples was measured on a microplate UV-Vis spectrophotometer (Multiskan GO, Thermo Scientific, Hudson, NH, USA).

Fatty Acid Analysis by Gas Chromatography-Mass Spectrometry (GC-MS)
Fatty acid methyl esters (FAMEs) were prepared by base-catalyzed transmethylation using 70 µg of each dried lipid extract (n = 3), followed by addition of 1 mL of internal standard (1.08 µg mL −1 of methyl nonadecanoate in n-hexane) and 200 µL of a methanolic solution of potassium hydroxide (2 M). After 2 min vortexing, 2 mL of an aqueous solution of sodium chloride (10 mg mL −1 ) was added and the sample was centrifuged for 5 min at 626× g [44]. The upper organic phase (600 µL) was collected and completely dried under a nitrogen stream. FAMEs were then redissolved in 30 µL of n-hexane and 3 µL was used for GC-MS analysis on a GC system (Agilent Technologies 6890 N Network, Santa Clara, CA, USA) equipped with a DB-FFAP column (30 m of length, 0.32 mm of internal diameter, and 0.25 µm of film thickness) (J&W Scientific, Folsom, CA, USA). The GC equipment was connected to an Agilent 5973 Network Mass Selective Detector. An electron impact mode was used at 70 eV, m/z range 50-550 and 1 s cycle in a full scan mode acquisition.
The GC oven was programmed from an initial temperature of 80 • C for 3 min, followed by successive linear increases at 25 • C min −1 to 160 • C, at 2 • C min −1 to 210 • C, and 30 • C min −1 to 250 • C, standing at 250 • C for 10 min. The injector temperature was 220 • C, detector temperature was 280 • C, and helium (carrier gas) was used at a flow rate of 1.4 mL min −1 . FA identification was based on retention times and MS spectra of FA standards (Supelco 37 Component FAME Mix, Sigma-Aldrich, Saint Louis, MI, USA), complemented with the analysis of MS spectra from Wiley 275 library and "The Lipid Web" [40]. FA quantification was performed using calibration curves obtained from FAME standards under the same instrumental conditions.

Hydrophilic Interaction Liquid Chromatography-Mass Spectrometry (HILIC-MS)
Lipid extracts were analyzed by hydrophilic interaction liquid chromatography (HILIC) on a high-performance LC (HPLC) system (Ultimate 3000 Dionex, Thermo Fisher Scientific, Bremen, Germany) with an autosampler coupled online to a Q-Exactive ® mass spectrometer with Orbitrap ® technology, as previously used for other algae [28][29][30]. Mobile phase A consisted of 25% water, 50% acetonitrile, and 25% methanol (per volume), with 2.5 mM ammonium acetate; mobile phase B consisted of 60% acetonitrile and 40% methanol, with the same amount of ammonium acetate in mobile phase A. Initially, 10% of mobile phase A was held isocratically for 2 min, followed by a linear increase to 90% of mobile phase A within 13 min and a maintenance period of 2 min, returning to the initial conditions in 8 min and maintained for 10 min. A volume of 10 µL of each sample, from a mixture containing 10 µL of lipid extract in CH 2 Cl 2 (1 µg µL −1 ), 4 µL of a mixture of phospholipid standards (0.02 µg of dMPC, dMPE, NPSM, and LPC (19:0); 0.08 µg of dPPI, dMPA, and tMCL; 0.012 µg of dMPG; 0.04 µg of Cer (d18:1/17:0) and dMPS) and 86 µL of starting eluent, was introduced into the Ascentis Si column HPLC Pore column (10 cm × 1 mm, 3 µm, Sigma-Aldrich, St. Louis, MI, USA) with a flow rate of 50 µL min −1 at 35 • C. The mass spectrometer with Orbitrap ® technology was operated in simultaneous positive (electrospray voltage 3.0 kV) and negative (electrospray voltage −2.7 kV) modes with high resolution with 70,000 and AGC target of 1 × 10 6 , the capillary temperature was 250 • C, and the sheath gas flow was 15 U. In MS/MS experiments, a resolution of 17,500 and AGC target of 1 × 10 5 was used and the cycles consisted in one full scan mass spectrum and 10 data-dependent MS/MS scans were repeated continuously throughout the experiments with the dynamic exclusion of 60 s and intensity threshold of 2 × 10 4 . Normalized collision energy™ (CE) ranged between 20, 25, and 30 eV. Data acquisition was performed using the Xcalibur data system (V3.3, Thermo Fisher Scientific, Bremen, Germany).
LC-MS data processing was performed using MZmine 2 (version 2.32). Raw data was filtered for a RT from 0 to 30 min with m/z range of 400-1600 and a relative m/z tolerance of 5.0 ppm, a noise level of 1 × 10 4 and typical retention time (RT) tolerance of 2 min were used. Chromatogram building was applied by means of the ADAP chromatogram builder module. The isotopic peaks grouper algorithm was applied with a maximum charge of 1 and the representative isotope was set to "lowest m/z". Join aligner algorithm was set with weight for m/z 100 and weight for RT 10. Gap filling algorithm was also used with intensity tolerance of 100% and RT correction. The identification of the chromatographic peaks of interest was performed using in-house database (Supplementary Tables S1 and S2) created based on the information available on the LIPID MAPS. Duplicate peak filter algorithm with filter mode set to "new average" was applied. The peak identification was validated based on the typical RT of the respective lipid class and mass accuracy observed in LC-MS spectra (<5 ppm), as well as LC-MS/MS spectra interpretation that allowed to confirm the polar head group identity and the fatty acyl chains of the most molecular species.

Biomass Composition
The mean moisture content of E. huxleyi biomass was 5.5 ± 0.3%, while the ash content obtained by the traditional method at 575 • C in a muffle was 50.8 ± 0.7% ( Figure 1). The ash content of E. huxleyi was also determined using TG at a temperature of 600 • C and a percentage of 48.0% was obtained.

Betaine Lipids
The betaine lipids identified in this study included diacylglyceryl-N,N,N-trimethyl-homoserine (DGTS), monoacylglyceryl-N,N,N-trimethyl-homoserine (MGTS), and diacylglyceryl carboxyhydroxymethylcholine (DGCC). Among these classes of betaine lipids, DGCC is the least reported in microalgae, in which its occurrence has been associated with the phylum Haptophyta [47,48]. Additionally, lipid species of unknown structure, previously reported in E. huxleyi and classified based on exact mass, predicted formula and MS/MS as betaine-like lipids (BLL), due to lack of phosphate and the presence of reduced nitrogen in the headgroup [17], were identified in this study.
Betaine lipids were observed in positive LC-MS spectra as [M + H] + ions (Table 4)       In this study, for the first time, a species of MMPE was identified in E. huxleyi. A previous study reported the presence of dimethylphosphatidylethanolamines (DMPE) in E. huxleyi [18], but not of MMPE. The MMPE identified was MMPE (30:1), which is an isomer of PE (31:1). These isomeric species were discriminated because they eluted in different retention times (Supplementary Figure S12a    In negative mode, sGSL were observed as These were similar to the fragmentation pathways previously observed in ion-trap MS/MS data [19].

Ceramides
In this study, ceramides were identified in positive LC-MS spectra as [M + H] + ions. The most abundant ceramides were assigned as Cer(d38:2) at m/z 592.5672 and Cer(d40:1) at m/z 622.6143 (Supplementary Figure S16). The species of this class were confirmed by LC-MS/MS; for example, in the spectrum of Cer(d40:2) it is possible to identify the sphingodienine ion at m/z 262.2525 (Supplementary Figure S19).

Discussion
Emiliania huxleyi biomass was analyzed for the content (expressed as a percentage of freeze-dried sample weight) of moisture and ash, protein, sugar, and lipids. In terms of ash content, two methods were used: the traditional method at 575 • C in a muffle and TG at a temperature of 600 • C, which has been proposed as a more reasonable terminal temperature for the determining the ash content of microalgae [34]. A similar content was obtained for the two methods (48.0% by TG versus 50.8% by muffle furnace) with the advantages of the TG method, which requires only a few milligrams of biomass and is more automated than the muffle method. The high ash content of cultured E. huxleyi when compared with non-coccolithophore microalgae [53] can be due to the presence of calcium carbonate, as its decomposition is expected above 600 • C [54].
Protein content was determined as 15.3% with 4.78 as nitrogen-protein conversion factor. This factor was proposed for marine microalgae, as the traditional factor 6.25 could lead to an overestimation of the protein content in the biomass of microalgae [38].
Regarding sugar content, a similar total sugars content was obtained with different methods. Lipids content was determined as 20.3%, which is within the range of those previously reported for other haptophytes harvested at the late exponential growth phase (25.5 ± 2.4% for Pseudoisochrysis paradoxa and 24.3 ± 3.8% for Diacronema vlkianum) [48].
Together, the contents of glycolipids (84.9 ± 6.4 µg mg −1 extract) and phospholipids (19.1 ± 4.5 µg mg −1 extract) estimated by colorimetry represented approximately 10% of the total lipid extract. Indeed, other classes of polar lipids, in addition to glycolipids and phospholipids, were previously identified in E. huxleyi [15,18,22], namely betaine lipids which are not quantified by the colorimetric methods used. Polar lipids were previously estimated (by TLC) to represent approximately 50% of the total lipids extracted from eight isolates of E. huxleyi with chloroform:methanol (2:1, v/v), in logarithmic or stationary phases [13]. A similar content of polar lipids (also determined by TLC) was reported for other haptophytes (49.4% for P. paradoxa and 61.8% for D. vlkianum) [48]. On the other hand, the lipid extract can contain significant amounts of pigments (corroborated by the green colour of the extract), as well as sterols and long-chain alkenones, alkenoates, and alkenes [13]. Although neutral lipids are not the subject of the present study, these compounds may also be important from the perspective of potential applications. For example, alkenones from haptophytes have been reported as promising renewable phase change materials [55].
All classes of polar lipids identified in this study, except monomethylphosphatidylethanolamine (MMPE), have been reported in previous studies with other strains of E. huxleyi [15,18,22]. The lipid species identified in each class are not the same as those previously reported [18,22], which may be due to the strain and growth conditions of the microalgae, as well as to the different experimental conditions used for the extraction and analysis of lipids. Some fatty acids identified as components of the polar lipids were not seen in the fatty acid analysis by GC-MS, due to higher sensitivity of LC-MS compared to GC-MS.
In terms of bioactivity of sulfolipids, it should be noted that SQDG (16:0/16:0), also present in E. huxleyi, was previously described as a lipid with antimicrobial activity [68]. SQDG (20:5/16:0), also identified in E. huxleyi, was described with antimicrobial [69] and anti-inflammatory effects [70], which reinforces the potential use of this microalga as a source of bioactive lipids with potential health benefits. Neutral glycolipids from microalgae were also previously reported with bioactive properties, namely MGDG from Nannochloropsis sp. showed anti-inflammatory activity and MGDG from Chlorella vulgaris exhibited antitumor activity [8,71].
Betaine lipids have also been identified in several classes of microalgae [29,72] and some studies have been developed to identify their biological activities and possible applications. In particular, DGTS species isolated from the microalgae Nannochloropsis granulata were described as potential anti-inflammatory agents, exhibiting anti-inflammatory activity by inhibiting nitric oxide (NO) production in RAW264.7 macrophage cells with downregulation of inducible nitric oxide synthase expression [73]. Given the presence of betaine lipids in the biomass of E. huxleyi, it could be a future challenge to investigate its application as a source of these bioactive compounds From the point of view of relevance for biotechnological applications, phospholipids were also considered to be important bioactive compounds, namely anti-inflammatory PG were isolated from the red macroalga Palmaria palmata, exhibiting strong and dose-dependent NO inhibitory activity [70].
In respect of glycosphingolipids, sGSL were recently described as indicative of susceptibility to viral infection in E. huxleyi [17], and another class previously classified as host GSL (hGSL) due to its prevalence in uninfected E. huxleyi [14,19]. Viral GSL, previously described as potential biomarkers for viral infection [14], were not identified in this study with cultured E. huxleyi.
Previous studies have identified ceramides as highly bioactive compounds with significant effects on cell metabolism [74,75]. Wertz and his collaborators [76] investigated the uptake of several sphingoid bases by Escherichia coli and Staphylococcus aureus, and assessed subsequent ultrastructural damage, exploiting the potential for prophylactic or therapeutic purposes. Ceramides with antibacterial activity were also reported against pathogenic Neisseria [77]. Considering the complexity of the biological processes affected by this category of compounds [75], further work is still needed to establish the relevance of E. huxleyi sphingolipids for potential biotechnological applications.

Conclusions
In this study, the polar lipidome profile of the coccolithophore Emiliania huxleyi (strain AC453) was described in detail using HILIC-MS and GC-MS. E. huxleyi revealed the presence of distinct beneficial n-3 PUFA, representing about 40% of total FA. Their high amount underlines the importance of this microalga as a natural source of n-3 PUFA, in particular, 22:6n-3 (DHA) and 18:4n-3 (SDA), which are naturally esterified in various polar lipids. Among the classes of polar lipids identified in the biomass of E. huxleyi, glycerolipids, especially glycolipids, are the most recognized as bioactive compounds and linked to health benefits. Thus, the polar lipids of E. huxleyi show great potential for future biorefinery approaches, envisioning their application in the development of new products and materials, especially as ingredients in food and feed products, drugs and cosmetics.
Supplementary Materials: The following are available online at http://www.mdpi.com/2218-273X/10/10/1434/ s1, Supplementary Table S1. In-house database used for peak identification in MZmine (positive mode), Supplementary Table S2. In-house database used for peak identification in MZmine (negative mode), Figure S1. Representative examples of LC-MS chromatograms obtained by analysis of the polar lipids extracts from Emiliania huxleyi, acquired on (a) positive mode and (b) negative mode, Figure S2: Relative abundance of acidic and neutral glycolipids identified by LC-MS from Emiliania huxleyi, Figure S3