Krill Sterol and Lipid Class Fatty Acid Data

Brief description

Fatty acid analysis is a powerful tool in food web research for estimating dietary sources in marine predators. However, the utility of fatty acids as dietary indicators from whole lipid samples, rather than from separate lipid classes, has been questioned. Samples are often collected at a single time point, precluding seasonal dietary comparisons. We investigated variations in the fatty acid composition of structural (phospholipids) and storage lipids (triacylglycerols) of Antarctic krill (Euphausia superba) using fisheries samples obtained over one year. Seasonal variation was observed in fatty acid biomarkers within triacylglycerol and phospholipid fractions of krill. Fatty acids in krill triacylglycerols (thought to better represent recent diet), reflected omnivorous feeding with highest percentages of flagellate biomarkers (18:4n-3) in summer, and diatom biomarkers (16:1n-7c) in autumn, winter and spring. Carnivory biomarkers (∑ 20:1 + 22:1 and 18:1n-9c/18:1n-7c) in krill were greater in autumn. Phospholipid fatty acids were less variable and higher in 20:5n-3 and 22:6n-3, which are essential components of cell membranes. Sterol composition did not yield detailed dietary information, but percentages of the major krill sterol, cholesterol, were significantly higher in winter and spring compared with summer and autumn. Unexpectedly, 18:4n-3 and copepod markers ∑ 20:1 + 22:1 were not strongly associated with the triacylglycerol fraction during some seasons. Krill may mobilise 18:4n-3 to phospholipids for conversion to long chain polyunsaturated fatty acids, which would have implications for its role as a dietary biomarker. For the first time, we demonstrate the dynamic seasonal relationship between specific biomarkers and krill lipid classes.

Lineage

Maintenance and Update Frequency: notPlanned

Statement: 3.3.1. Krill sample collection
Krill sample collection is described in detail in Ericson et al. (2018a). Briefly, krill were caught on board the FV Saga Sea (Aker Biomarine) during their 2016 fishing season (December 2015 – September 2016), from three different locations; the West Antarctic Peninsula (WAP), South Orkney Islands (SOI) and South Georgia (SG) (Figure 3.1). Twenty krill day-1 were randomly sampled from the catch by a fisheries observer (there was no selection by size or maturity stage). These samples were transported to Hobart, Tasmania on dry ice and stored at –80ºC.

3.3.2. Initial lipid extraction and fatty acid analyses
Three male and three female krill were collected from the fisheries samples at 2-week intervals, for lipid extraction. Krill were individually extracted in separation funnels using a modified method of Bligh and Dyer (1959), with a solvent mixture of methanol (MeOH)/dichloromethane (CH2Cl2)/water (H2O) at 20:10:7, by vol. To separate phases, 10 mL CH2Cl2 and 10 mL saline MilliQ H2O were added the following day. The lower lipid layer was drained into a round bottomed flask and solvent was removed using a rotary evaporator, to concentrate the total lipid extract (TLE). The TLE was stored at –20 ºC in a pre-weighed glass vial with added solvent (CH2Cl2), to ensure that oxidation of the sample did not occur.
These total lipid extracts (TLE) obtained from Ericson et al. (2018a) were used for the present study. Krill with a large range of selected biomarker percentages in the TLE were included. Only males were included to eliminate gender as a potential confounding variable (Clarke 1980; Mayzaud et al. 2000). Males that had < 25% TAG (as % of total lipids) were excluded, as low TAG percentages during the reproductive season may make them less suitable for dietary analysis (Stübing & Hagen 2003; Virtue et al. 1996). Total lipid extracts from 12 males from the 2016 catch were selected from summer, autumn and combined winter/early spring (total N = 36). Summer krill were sampled from the WAP and the SOI in January and February 2016, autumn krill were sampled from the WAP between March - May 2016, and winter/early spring (referred to as ‘winter/spring’ hereafter) krill were sampled from SG between June – September 2016. Samples selected for this study are shown in Appendix IV, along with their total lipid (mg g -1 DM; dry mass), TAG and PL percentage data (% of total lipid) obtained from their TLE. Detailed analyses for the full suite of krill collected in 2016 can be found in Hellessey et al. (2018).

3.3.3. Separation of lipid classes via column chromatography
Aliquots were taken from the TLE and analysed via column chromatography, to investigate the fatty acid composition within the lipid classes of krill. For all chosen samples (N = 36), one gram of activated silica was added to a glass column and washed through using chloroform (CHCl3) to pack the column. Ten milligrams of total lipid extract were added to the packed column. Triacylglycerols were eluted with 10 ml CHCl3, followed by elution of glycolipids with 20 ml acetone (C3H6O), and elution of phospholipids with 20 ml methanol (MeOH), to produce extracts for triacylglycerol, phospholipid and glycolipid fractions (total lipid class fractions; TLCF). All TLCF were reduced via rotary evaporation and added to 1.5 ml glass vials with Teflon caps. Accurate lipid class separation was confirmed by running 1 µl aliquots of all lipid class fractions through an Iatroscan TLC-FID analyser (see Hellessey et al. 2018 for detailed methods) following column chromatography. Once accurate separation was verified, TAG and PL lipid fractions were used to prepare fatty acid methyl esters (FAME) for fatty acid analysis.
A subsample of each TLCF was transferred to a glass test tube with a Teflon® screw-cap and 3 mL of methylating solution (MeOH /CH2Cl2/hydrochloric acid (HCl), 10:1:1, by vol) was added. Each test tube was then heated at 90 – 100 ºC for 75 min, then cooled for 5 min before addition of 1mL H2O and 1.8 mL hexane (C6H14)/CH2Cl2 solution (4:1, by vol) to extract the FAME. Samples were centrifuged for 5 min, and the upper layer (FAME) was transferred to a vial. An additional 1.8 mL of C6H14/CH2Cl2 solution was added to the test tube, and the sample was centrifuged again, before adding the top layer of FAME to the vial. This process was carried out three times in total, to ensure that all of the FAME had been extracted and added to the vial (samples in the vial were blown down with nitrogen (N2) gas in between transfers).
To prepare samples for gas chromatography (GC-FID), 1.5 mL of internal injection standard (23:0 FAME) was added to each vial. Samples were analysed via GC-FID using an Agilent Technologies 7890A System equipped with a non-polar Equity®-1 fused silica capillary column (15 m length x 0.1 mm internal diameter x 0.1 μm film thickness). Samples (0.2 μl) were injected in spitless mode at an oven temperature of 120 ºC with helium the carrier gas. The oven temperature was raised at a rate of 10 ºC min-1 up to 270 ºC, then a rate of 5 ºC min-1 up to 310 ºC. Quantification of fatty acid peaks (expressed as a % of the total fatty acid area) was conducted using Agilent Technologies ChemStation software. Initial identification was based on comparison of retention times with known (Nu Check Prep) and fully characterized laboratory (tuna oil) standards.
Gas chromatography-mass spectrometry (GC-MS) was carried out using a Thermo Scientific 1310 GC-MS coupled with a TSQ triple quadruple, to further confirm component identification. Selected samples were injected using a Tripleplus RSH auto sampler using a non-polar HP-5 Ultra 2 bonded-phase column (50 m length x 0.32 mm internal diameter x 0.17 μm film thickness). The HP-5 column was a similar polarity to the column used for GC-FID analyses. The initial oven temperature (45 ºC) was held for 1 min, then rose at a rate of 30 ºC min-1 to 140 ºC, then at a rate of 3 ºC min-1 to 310 ºC, and held for 12 min. Helium was the carrier gas. Operating conditions of the GC-MS were as follows: electron impact energy 70 eV; emission current 250 μamp; transfer line 310 ºC; source temperature 240 ºC; scan rate 0.8 scans sec-1; mass range 40 – 650 Da. Mass spectra were acquired and processed with the software Thermo Scientific XcaliburTM. Nu Check Prep and tuna oil standards were also used for assistance in identification of peaks.


3.3.4. Sterol analysis
An additional 300 µl aliquot was taken from each of the TAG fractions for saponification. Each aliquot was transferred into a glass test tube fitted with a Teflon lined screw cap, blown down under N2 gas and treated with 2 mL of saponifying solution (5% potassium hydroxide (KOH) in MeOH/MilliQ H2O, 80:20, by vol), then heated at 60 °C for 3 h. Samples were cooled and 1 ml of MilliQ H2O and 1.8 ml of C6H14 : CH2Cl2 solution was added to extract the total non-saponifiable neutral lipids (TSN). Samples were then centrifuged for 5 min and the upper layer containing TSN was transferred to a vial, and another 1.8 ml of C6H14/CH2Cl2 was added to the test tube. This process was carried out three times and samples were blown down each time using N2 gas.
Samples of TSN lipids obtained above were silylated by treatment with N2 gas and addition of 50 µl N,O-bis (Trimethylsilyl) trifluoroacetamide, then heated overnight at 60 °C. Prior to analysis, samples were blown down using N2 gas and 1000 µl of internal injection standard (23:0 FAME) was added to each vial. Samples were blown down again under N2 gas and transferred to glass inserts with 200 µl CH2Cl2. Samples were then run through a GC-FID and GC-MS as described above, to obtain sterol composition and content, and to confirm component identifications.

3.3.5. Statistical analyses
Principal components analysis (PCA) of fatty acid data was carried out in PRIMER 6 using Pearson correlation, due to large differences in fatty acid variances. Data was transformed (log x+1) prior to PCA analysis. Fatty acid biomarker data was analysed in RStudio (version 1.1.453) using one-way ANOVA with either season or lipid class as a factor, or two-way ANOVA with season and lipid class as factors, and a season*lipid class interaction. Tukey comparisons were used to investigate significant differences between levels of season. Sterol data was also analysed in RStudio, using one way ANOVA with season as a factor. Data for all analyses was log or square root transformed when it did not meet assumptions of normality or homogeneity of variances. A Welch’s test was used for sterol data that had heterogeneous variances and data transformation did not normalise the data.