Phytosterols in human serum as measured using a liquid chromatography tandem mass spectrometry

Phytosterols are lipophilic compounds found in plants with structural similarity to mammalian cholesterol. They cannot be endogenously produced by mammals and therefore always originate from diet. There has been increased interest in dietary phytosterols over the last few decades due to their association with a variety of beneficial health effects including low-density lipoprotein cholesterol lowering, anti-inflammatory and anti-cancerous effects. They are proposed as potential moderators for diseases associated with the central nervous system where cholesterol homeostasis is found to be imperative (multiple sclerosis, dementia, etc.) due to their ability to reach the brain. Here we utilised an enzyme-assisted derivatisation for sterol analysis (EADSA) in combination with a liquid chromatography tandem mass spectrometry (LC-MS n ) to characterise phytosterol content in human serum. As little as 100 fg of plant sterol was injected on a reversed phase LC column. The method allows semi-quantitative measurements of phytosterols and their derivatives simultaneously with measurement of cholesterol metabolites. The identification of phytosterols in human serum was based on comparison of their LC retention times and MS 2 , MS 3 spectra with a library of authentic standards. Free campesterol serum concentration was in the range from 0.30-4.10 µ g/mL, β -sitosterol 0.16-3.37 µ g/mL and fucosterol was at lowest concentration range from 0.05-0.38 µ g/mL in ten individuals. This analytical methodology could be applied to the analysis of other biological fluids and tissues.


Introduction
Plant sterols and plant stanols also referred as phytosterols, are a class of lipophilic compounds found in plants with high structural similarity to cholesterol, a well-known molecule endogenously found in humans, particularly notorious for its relation to cardiovascular disease (CVD) [1][2][3].Phytosterols are components of plant membranes, which, like cholesterol, regulate membrane fluidity and permeability [4].Since phytosterols cannot be endogenously produced by mammals they originate from dietary uptake, food such as in vegetables, fruits, grains, cereals, vegetable oils and margarines [2,5,6].More than 250 steroids have been described in plants [7][8][9].The most frequently occurring plant sterols are β-sitosterols, sitostanol, campesterol and stigmasterol in food and human body [10,11].Cholesterol averages around 50 mg/kg of total lipid in plants, whereas in mammals it can be as high as 5 g/kg (or more) [12].The Food and Drug Administration (FDA) and the European Food Safety Authority (EFSA) recognise phytosterols as "safe" and have authorised health claims regarding risk reduction of coronary heart disease with a daily dietary intake of at least 2 g/day [13,14].This has led to increased availability and consumption of phytosterol fortified foods and food supplements.The daily intake in a typical western diet averages to about 300 mg for phytosterols and 20 mg for phytostanols [15][16][17].However, their absorption efficacies are reported to be much lower >2 % and >0.02 % for phytosterols and phytostanols respectively.In contrast, the absorption efficiency for cholesterol is estimated at 50-60 % [3,18,19].
Sterols are chemically composed of a steroid core consisting of perhydrophenathrene (A, B and C ring) fused with a cyclopentane (D ring), a hydroxyl group on C3 of the A-ring, a methyl groups attached at C18 and C19, and variable C17-side chain (R) attached to the D-ring (Fig. 1a,  b).The difference between distinct sterols lies in the addition of an extra methyl, ethyl or hydroxyl group to the side chain or/and the steroid core.R groups of phytosterols typically contain nine or ten carbons, as opposed to eight in cholesterol [4].In contrast to plant sterols, plant stanols contain a saturated core.(Fig. 1d).Phytosterols can occur free forms and in four conjugated forms in which the hydroxyl group at C3 is esterified with a fatty acid (FA) or hydroxycinnamic acid (HA) and glycosylated to a hexose (GH) or acyl hexose (AGH) (Fig. 1c).
Phytosterols and cholesterol can be oxidised to oxyphytosterols and oxysterols respectively (Fig. 1d).Oxyphytosterols are present in low levels in food but are tentatively endogenously produced in humans following intestinal absorption by similar biochemical pathways as oxysterols [4,13,14,20].Both cholesterol and phytosterols are prone to autoxidation under conditions such as heat and light during food processing, or by reactive oxygen species in tissues [21,22].Oxyphytosterols in food such as 7-oxo, 7-hydroxy-, and 5, 6-epoxy-phystosterols are autoxidised on the sterol-ring (Fig. 1d), while side-chain oxidation is mediated by specific enzymes [3,14].For example, 24-hydroxycholesterol and 27-hydroxycholesterol are hydroxylated by cytochrome P450 oxidase CYP46A1 and CYP27A1, respectively.Plat J reported an average serum cholesterol concentration in the general population around 5 mmol/L, whereas concentrations of plant sterols by 400 times lower and stanols by 100,000 times lower than cholesterol [14].

Cholesterol
Plant sterols, RS = "steroid core", R= "side-chain" Over 20 % of free cholesterol in humans is found in the brain, where steady cholesterol homeostasis is essential for proper functioning of neurons [23].Cholesterol biosynthesis and metabolism is tightly regulated in the central nervous system (CNS) [24,25].Pathophysiological studies have linked disturbed metabolism of cholesterol to neurological and neurodegenerative diseases such as multiple sclerosis, Alzheimer's disease, and dementia [1,2,10,14].Phytosterols have been proposed to have therapeutic effects in the pathogenesis of neurodegenerative diseases, potentially by modulating cholesterol homeostasis in CNS [26,27].As phytosterols have been supplemented in functional food products, this leads to increase dietary exposure of both phytosterols and their oxidation products [3,12,28].Recent studies show that phytosterols and oxyphytosterols can traverse the blood brain barrier (BBB) and accumulate in the brain [13,20].Based on this finding, researchers have started to investigate the physiological role of phytosterols and their oxidation products [13,20].
Many methodologies were exploited to analyse phytosterols in biological samples, from the traditional methods using thin-layer chromatography (TLC), to liquid chromatography (LC) [11,13,29,30].Nowadays, phytosterols and their oxidation products analyses are dominated by GC-MS using selected-ion monitoring (SIM) or LC-MS/MS method combined with multiple reaction monitoring (MRM) [31,32].In summary, the protocol usually starts with the addition of ethylenediamine tetraacetic acid (EDTA) and butylated hydroxytoluene (BHT) to minimise autoxidation of sterols, following by alkaline hydrolysis to unesterified sterols.For measuring only free sterols, hydrolysis is omitted.This follows by solid-phase extraction (SPE) using either a reversed-or normal-phase SPE cartridges.Sterols are then derivatised to their trimethylsilyl ester (TMS) derivatives to enhance volatility prior to GC-MS analysis [29,33,34].The LC-MS/MS analyses are current popular choice [35].To enhance sensitivity of measurements and phytosterol solubility, different derivatisation agents were used to transform sterols to their picolinyl ester-, [36] nicotinyl esters-, [37] and Girard P (GP) hydrazones [29,38,39].
Here, we utilised an enzyme-assisted derivatisation for sterol analysis (EADSA) technology for measurement of non-esterified free phytosterols in human serum.EADSA is a chemical approach, where 3βhydroxy-5-ene or 3β-hydroxy-5α-hydrogen sterols convert to 3-oxo-4ene or 3-oxo sterols using cholesterol oxidase from Streptomyces sp.The sterols possessing an oxo group are then derivatised with the commercially available Girard P reagent (Fig. 2).The derivatisation reaction with the Girard P hydrazine is carried out directly on the products of the cholesterol oxidase reaction, without any further purification or extraction.The incorporation of the GP hydrazone group into the 3-oxo sterol structure increases the mass of the sterol by 134 Da.The resulting GP hydrazones are then separated from excess GP hydrazine using a recycling protocol on a reversed phase SPE cartridge.The benefits of EASDA are: -(a) sterols solubility increase, and (b) enhancement in an electrospray (ES) signal by 1000 in comparison to underivatised sterols versions [40,41].Most importantly, sterol GP-hydrazones generate informative MS 3

Human serum samples
The blood samples were collected from six healthy volunteers of age 23-30, three females and three males, with BMI 18-22 during public engagement events (including young scientist days) organised by the Center for Adolescent Rheumatology Versus Arthritis, Division of Medicine, University College London (UCL).Two blood samples were collected from patients with Juvenile Onset Systemic Lupus Erythematosus (JSLE) attended the adolescent and young adult lupus clinics at University College London Hospital, females ages 20 and 26 with BMI 20.Two samples were from patients with alternating hemiplegia of childhood (AHC), female and male age 33, BMI 22. Informed written consent or parental consent/participant assent was acquired from both patients and healthy volunteers as age-appropriate under the ethical approval reference: REC11/LO/0330 and in accordance with the Declaration of Helsinki.All information was stored as pseudoanonymised data.Blood serum was used for mass spectrometry analyses.The blood serum samples were collected in serum gel S/9 mon-ovette® (Sarstedt).Serum was separated from blood cells by centrifugation at 3500 rpm for 10 min at 18 • C (Heraues multifuge 4KR centrifuge, Osterode, Germany).To avoid autoxidation 10 µL of a methanolic butylated hydroxytoluene (BHT 25 mg/mL) solution was added to 1 mL serum.

Recycling solid phase extraction
Recycling SPE was carried out to remove excess GP reagent as previously published [21,44].A 200 mg Sep-Pak Vac cartridge was washed with 6 mL 100 % methanol, followed by 6 mL 10 % methanol, and then conditioned with 4 mL 70 % methanol.The GP reaction mixture (3.25 mL in 70 % methanol) after oxidation and GP derivatisation procedures was directly applied on the C 18 SPE cartridge, followed by 1 mL of 70 % methanol (this is a wash of the reaction vessel), and 1 mL of 35 % methanol.The effluent was collected into a glass beaker.The combined effluent (now 5.25 mL) was diluted with 4 mL of water.The resulting mixture (now 9 mL in 35 % methanol) was again applied to the C 18 SPE column, followed by a wash with 1 mL of 17 % methanol.The effluent was collected into the glass beaker.To the combined effluent, 9 mL of water was added to give 19 mL of about 17.5 % methanol.This was again applied to the C 18 SPE column and the effluent was discarded.At this point, most GP-hydrazones were retained on the C18 cartridge.The cartridge was then washed with 6 mL 10 % methanol in water to remove excess of GP hydrazine.Our validation experiments confirm the required volumes of elution solvents required to elute fully derivatised phytosterols (Appendix A).We found that cholesterol-and phytosterol-GP-hydrazones eluted with three 1-mL portions of 100 % methanol and collected in 1.5 mL-microcentrifuge tubes (SPE-2-Fr1, 2, 3), followed by four 1-mL portions of 100 % ethanol (SPE-2-Fr-4, 5, 6, 7) and another application of two 1-mL portions of 100 % DCM (SPE-2-Fr-8, 9) from the SPE-2 C 18 cartridge.

Extraction of phytosterols from human serum
One hundred µL of serum was added drop-wisely into a 2-mL Eppendorf tube containing 1 mL of absolute ethanol and 0.5 µL of [ 2 H 7 ] cholesterol (2 µg/µL in propan-2-ol).The sample was sonicated for 5 min.Then, 330 µL of water was added to the tube and ultrasonicated for a further 5 min, and the sample was then centrifuged at 14,000 xg at 4 • C for 30 min.The resulting sample contained 70 % ethanol.A 200-mg Sep-Pak C 18 cartridge was rinsed with 4 mL of ethanol and then conditioned with 6 mL of 70 % ethanol.We adapted a previously published protocol [42] where authors validated the method with a solution of cholesterol and 24(R/S)- [26,26,26,27,27,27-2 H 6 ] hydroxycholesterol in 70 % ethanol, they found that cholesterol was retained on the column even after a 5.5-mL column wash of 70 % ethanol, whereas 24(R/S)- [26,26,26,27,27,27-2 H 6 ] hydroxycholesterol elutes in the flow-through and column wash.They also mentioned that after a further column wash with 4 mL of 70 % ethanol, cholesterol was eluted from the column in 2 mL of absolute ethanol.They further applied an additional 2 mL of absolute ethanol on the column to elute more hydrophobic sterols.Therefore, in this work, the flow-through (1.43 mL) and a column wash of 9.5 mL of 70 % ethanol were collected in a 15 mL round-bottom flask to elute more polar phytosterols from the column (Figure S1).As some phytosterols are more lipophilic than cholesterol, we eluted them in 8 mL ethanol (Figure S1, fraction B).Just mention here we applied sequentially 1 mL of ethanol eight times on the cartridge.The cartridge was further stripped with 2 mL DCM to elute even more lipophilic phytosterols (Figure S1, fraction C).All three fractions were combined into the same round-bottom flask.Finally, the solvent was dried using a rotary evaporator.Then the samples were reconstituted with 100 µL propan-2-ol and thoroughly vortexed, before they were subjected to cholesterol oxidation, GP-derivatisation and SPE-2 on a C18 column, as described above.

Direct infusion single-stage TOF-MS
A Premier XE Q-TOF-MS connected to a 2777 C autosampler (Waters, UK) was utilised to screen SPE-2 fractions for each sample.The ESI was operated in positive mode.The capillary voltage was 2.5 kV and sample cone voltage was 50 V.The desolvation and source temperatures were 150 • C. The desolvation gas flow was 450 L/h and the cone gas flow was 100 L/h.The mass range was m/z 100-800 and the scan rate was 1 s − 1 .Samples (10 µL) were directly infused in 50 % of mobile phase A (MeOH, Propan-2-ol, Formic acid 50:50:0.1,v/v/v) and 50 % mobile phase B (MeOH, Formic acid 100:0.1,v/v/v) with a flow of 0.2 mL/min.The analysis time was 2 minutes.Then, one fraction for each the O/GP derivatised phytosterol with the highest [M] + signal was selected for a LC-MS n analysis.

Direct infusion ES multi-stage fragmentation mass spectrometry
A Thermo Finnigan LTQ linear ion trap mass spectrometer (Thermo Fisher Scientific, UK) was operated with the following settings: -spray voltage 1.00-1.20 kV, capillary temperature 200 0 C, no sheath or auxiliary gas used.The mass range m/z 50 -700 was scanned, and centroid data was collected.MS, MS 2 and MS 3 spectra were recorded.MS 2 experiment was performed on a precursor ion.MS 2 spectra were dominated by [M-79] + and [M-107] + fragment ions.MS 3 scans were performed on fragment-ions resulting from a neutral loss of 79 Da or 107 Da in the MS 2 .For acquisition of both MS 2 and MS 3 spectra, the collision energy setting was 35 % with the isolation width at 1.00.MS, MS 2 and MS 3 scans consisted of three averaged micro scans each with a maximum injection time of 200 ms.

Capillary-LC-ES-MS n
LC-MS n analyses were performed using an Accela HPLC system interfaced to the LTQ MS.The LC system is comprised of an autosampler, degasser and pump system.The injection volume was 10 µL.Chromatographic separation was achieved on a Hypersil Gold C 18 column (1.9 µm particles, 100 mm×21 mm, Fisher Scientific).Mobile phase A was composed of 33.3 % methanol, 16.7 % acetonitrile, with 0.1 % formic acid and mobile phase B was 60 % methanol, 40 % acetonitrile, 0.1 % formic acid.Initially B was at 50 % and was raised to 70 % over 3 min, then was raised to 99 % B over the next 17 min and stayed at 99 % B for 3 min, before returning to 50 % B in 6 s and re-equilibrating for a further 5 min 54 s, giving a total analysis time 26 min.The flow rate was 180 µL/min.The eluent was directed to the ESI source of the LTQ mass spectrometer.The ESI was operated in positive mode with a capillary temperature of 280 • C. The spray and capillary voltages were set to 4.5 kV and 33 V respectively.The sheath, auxiliary and sweep gas flow rates were 40, 10 and 0 respectively.The ion trap analyser set-up for seven scan events during one LC-MS analysis.In event 1 full scan of m/z 80-600, then follows by other 6 events set either for MS 2 and/or MS 3

([M]
+ →[M-107] + →) spectrum, but with the fragment ions displaced in mass by 28 Da, corresponding to additional loss of CO (data not shown).These fragment ions are characteristic of the derivatised 3-oxo-4-ene structure in the absence of additional groups in the A and B rings.They are formed by cleavage in the B-ring and are described by two competing series of b-type fragment ions (Fig. 4).The *b ion-series corresponds to B-ring fragment ions which have formed via A library of fragmentation patterns MS 2 and MS 3 spectra for sterol GP hydrazones has been established and could be used by researchers for the structure elucidation of GP-tagged sterols [21,24,45,46,48].In summary, the *b and # b ion-series are indicative of the sterols possessing a 3-oxo-4-ene group before GP derivatisation and a 3β-ol-5-ene structure before treatment with cholesterol oxidase and GP reagent, with no additional substituents in the A and B rings.This pattern changes with the introduction of hydroxy-or oxo-groups in the B rings of sterols.
While the major cleavages in the sterol ring system occur in the Bring and give abundant fragment ions, minor but important fragment ions are generated by cleavages in the C-and D-rings, and in the C-17 side-chain giving the fragment ions of low abundance in MS 3 spectra.In the MS 3   S2).However, the peak at 367 in the spectrum of cholesterol (Figure S3c) corresponding to the doubly unsaturated cholestane carbonium ion, shifted to 395, and shifted to 381 in the MS 3 spectrum of the O/GP-derivatised campesterol (Figure S5c).The peak at m/z 424 is a loss of C 3 H 6 from the side-chain.The MS 3 spectrum of the GP-derivatised stigmasterol is characterised by peaks at   S6).There is also enhanced abundance of fragment ions at m/z 284 ( # e'-NH), 299 ( # e'), and 327 (*e') each formed because of cleavage of the C-17-C-20 bond.The presence of methyl group, rather than an ethyl group, attached to C-24 in brassicasterol leads to the same *e' and *h' fragment ions (Figure S7).
7-Oxo-β-sitosterol was derivatised by GP hydrazine at C-7 position even though it was oxidised by cholesterol oxidase enzyme during our sample preparation, giving the peak at m/z 562 (Supplemental Fig. 7a

Chromatographic separation of O/GP derivatised authentic standards phytosterols
Several capillary LC-MS methods were tested and summarised in Appendix A. Fig. 8 shows a chromatographic separation of O/GP authentic standards on the C 18 column using the finalised gradient of 26min.The O/GP lathosterol and cholesterol with 8-carbons saturated sidechain at C17 co-eluted at RT 7.38, cholestanol eluted at RT 7.39 min from the reversed phase column.The O/GP campesterol with 9-carbons sidechain at C17 eluted at RT 7.75 min and β-sitosterol with 10 carbons sidechain at C17 eluted at RT 8.28 min.The location of the double bond in the side chain also influences the retention on the reversed phase.For example, the O/GP desmosterol with 8-carbons sidechain at C17 eluted at RT 6.49 min, brassicasterol with 9-carbons in the sidechain eluted at RT 7.33 min, whereas fucosterol as well as stigmasterol have 10-carbons in the unsaturated sidechain at C17 eluted at RTs 7.42 min and 7.88 min, respectively.The O/GP cholesterol and lathosterol coelutedon the LC column using the 26-min gradient.However, the MS 3 spectrum of the O/GP lathosterol shows a characteristic fragment-ion at m/z 159, which is not present in the MS 3 spectrum of cholesterol (Figure S10c).The separation of the syn-and anti-conformers was also achieved for the O/GP campesterol, brassicasterol and β-sitosterol (Fig. 8).In general, the longer the sidechain the greater the retention time and the introduction of a double bond in the side chain reduces retention time.Also, the location of hydroxy-or keto-groups in the Bring and/or in the sidechain decrease the retention times.For example, the O/GP 7α,25-dihydroxycholesterol eluted at RT 4.51 and 5.08 min syn-and anti-conformers, 7α,27-dihydroxycholesterol at RT 4.82 and 5.49 min syn-and anti-conformers, 25-hydroxycholesterol at 6.03 min, 7-oxo-β-sitosterol eluted at 6.12 min, 27-hydroxycholesterol 6.33 min and 6.58 min syn-and anti-conformers from the C 18 column.The chromatographic separation was also achieved for the O/GP 7β-hydroxycholesterol RT 8.98 and 9.92 min syn and anti-conformers, 7αhydroxycholesterol RT 10.19 and 11.29 min syn-and anti-conformers using the 26-min gradient (Figure S19).The calibration curves and a limit of detection (LOD) and limit of quantification (LOQ) are presented in Appendix A.

Identification of phytosterols in human serum
We adapted well validated protocol by Griffiths and colleagues [44,46] for the extraction of oxysterols but we modified this protocol to include phytosterols.Briefly, phytosterols were extracted from microliter quantities of human serum using reversed-phase SPE cartridge (SPE-1 protocol), we found experimentally to fully elute phytosterols require SPE-1-Fr-1 9.5 mL of 70 % ethanol (an oxysterol-rich fraction), SPE-1-Fr-2 6 mL 99.9 % ethanol and SPE-1-Fr-3 2 mL dichloromethane (this solvent was evaporated and this fraction was reconstituted in SPE-1-Fr-7).SPE-1 fractions were then combined as one sample (Supplemental Figure S1), followed by enzyme assisted derivatisation for sterol analysis (EADSA).EADSA consists of enzymatic conversion of 3β-hydroxy-5-ene-and 3β-hydroxy-5α-hydrogen-containing sterols to 3-oxo-4-ene and 3-oxo sterols followed by derivatisation with GP hydrazine to their corresponding GP hydrazones (Fig. 2).To remove the excess of GP hydrazine further purification and fractionation was achieved using a recycling SPE-2 protocol using a reversed phase SPE cartridge.As the O/GP-derivatised phytosterols are more hydrophobic than the O/GP oxysterols, cholenoic and cholestenoic acids, we determined experimentally to fully elute the O/GP derivatised cholesterol and phytosterols requires three 1-mL portions of 100 % methanol (SPE-2-Fr-1, 2, 3), four 1-mL portions of 99.9 % ethanol (SPE-2-Fr-4, 5, 6, 7) and two 1-mL portions of dichloromethane (SPE-2-Fr-8, 9)      3 (516.4→437.4→)shows three components eluted at (d) RT 6.45 min peak assigned as desmoterol and at RT 6.72 min possibly cholesta-5,x-dien-3β-ol (with double bond on the side chain) and 7/8-dehydrocholesterol was assigned to the chromatographic peak at RT 6.99 min, (f) The MS 3 (530→ 451→) spectrum at RT 7.31 min was brassicasterol, (g) the chromatographic peak at RT 7.75 min for the MS 3 (532→ 453→) was identified as campesterol.The identifications were based on comparison of RTs and MS 3 spectra with authentic standards.*b-series characteristic peaks at m/z 151, 163 and 177 possible a double bond in the sidechain at C17 which is built of nine carbons (Figure S23B), an authentic standard is required for a positive identification of this metabolite.The absence of *b-series fragments in MS 3 spectra were more likely to be sterols which have oxidation functionality on A or B ring, therefore the B-ring cleavage is less prominent.Fig. 10a shows RIC for the transition of MS 3 (546.4→467.4→)shows the chromatographic peak at RT 6.84 min corresponds to 3β-hydroxycholesta-5, 7-dien-26-oic acid as identified by comparing with MS 3 spectrum [46] and at RT 8.32 min, MS 3 of which was identical to the O/GP-tagged  + →[M-79] + →) fragmentation patterns for the 7-oxo derivatives differ from compounds with GP derivatisation at carbon position C-3 as shown for the 7-oxo-β-sitosterol (Fig. 6).The GP-derivatised 7-oxoβ-sitosterol was only measured in serum sample from a patient with AHC, the identification was based by RT and MS 3 spectrum of authentic standard.Fig. 11 shows the RIC for 548→469→ transition, the chromatographic peak at 3.31 min corresponds to 3β-hydroxycholest-(25 R)-5-en-26-oic acid.The chromatographic peak at 7.27 min possibly 3β,5α,6α-trihydroxy-sterol, as MS 3 spectrum shows a characteristic *b 2ion at m/z 177 indication that sterol with a 3β-hydroxy group and a planar A/B ring system and an unusually prominent fragment-ion observed in the MS 3 spectrum is at m/z 383 corresponding to [M-H 2 O-79-72] + (Figure S24A).The identity of analyte eluting from the C column at RT 6.48, 6.83 and 8.83 min were not possible due to the absence of authentic standards (Figure S24B,C).There were also several oxysterols, cholestenoic acid and unknown sterols measured in samples.The identification of some characterised and non-characterised sterols which followed a typical GP derivatised sterol fragmentation pattern ([M] + → [M-79] + →) are summarised in Table S6.For those sterols not in our current library as an authentic standard, they were identified by MS ([M] + → [M-79] + →) spectra comparison with published manuscripts using the EASDA technology combined with LC-MS n analysis [23,41,44, The chromatographic peaks at 1.37 and 1.75 min for the transition of MS 3 (564.5→485.5→) correspond to 7α-hydroxy-3-oxo-4-cholestenoic acid and 3β,7α-dihydroxycholest-(25 R)-5-en-26-oic acid (Figure S25A and B) as they matched to MS 3 from literature [21,41,47].
We analysed samples from patients with the autoimmune rheumatic disease juvenile-onset systemic lupus erythematosus (JSLE).Juvenileonset SLE (JSLE) is a severe inflammatory disease that can affect any part of the body, and JSLE patients are known to have altered lipid metabolism, resulting in increased risk of cardiovascular disease [52,53].In JSLE, these changes in lipid metabolism and potentially oxysterol metabolism are linked to a strong type 1 interferon (IFN) signature [54].Notably, Ch25h, a key rate limiting enzyme in metabolism of cholesterol into oxysterols, is an IFN-induced gene and changes in lipid profiles in patients with JSLE are associated with inflammation [55].It has also been demonstrated that the expression of receptors of GPR183, whose main ligand is the oxysterol 7α,25-dihydroxycholesterol, are altered in immune cells in adult-onset SLE.Taken together, this strongly suggests that oxysterol profiles are likely to alter in JSLE when compared to controls.We utilised the published 21-min [46] gradient and our 26-min for chromatographic separation of hydroxycholesterols and dihydroxycholesterols (Figure S26).We identified the O/GP 7α,25-dihydroxycholesterol, 7α,27-dihydroxycholesterol, 24S-, 25-and 27-hydroxycholesterols in patients with JSLE and healthy individuals.This preliminary analysis suggests that there is a potential reduction in 25-hydroxycholesterol in JSLE serum compared to healthy control serum with more limited differences in 7α,25-dihydroxycholesterol (Figure S27).However, more n-numbers are needed to confirm these preliminary results.

Conclusion
Phytosterols have been supplemented in functional food products for their cholesterol-lowering ability, as well as other beneficial effects, leading to increased dietary exposure of both phytosterols and their oxidation products [3,13].As a result, the physiological effects of phytosterols and oxidation products are constantly researched.Recent studies have suggested both phytosterols and oxyphytosterols can transverse the blood-brain-barrier and accumulate in the brain [13,20].In this work, the EADSA method with a subsequent LC-MS n analysis was applied for the analysis of non-esterified phytosterols in human serum.This method uses an extraction of phytosterols from serum, then sterols were oxidised with cholesterol oxidase enzyme and derivatised with Girard P reagent and analysed using a LC-MS n .A library of authentic standards consisted of their retention time established on the C18 column and their corresponding MS, MS 2 and MS 3 spectra was created and utilised for the identification of phytosterols in human serum.The LC-MS n method was optimised for more hydrophobic sterols then cholesterol with a total LC run time of 26 min.The linear range was 0.2 pg/mL to 10 µg/mL for campesterol, β-sitosterol and brassicasterol.The limit of quantification was 100 fg of the O/GP-tagged for brassicasterol and desmosterol as inject on the LC column.Phytosterols were semi-quantified using deuterated cholesterol as the internal standard.We found free campesterol serum concentration was in the range from 0.30 to 4.10 µg/mL, β-sitosterol 0.16-3.37µg/mL, and fucosterol 0.05-0.38µg/mL in serum from 10 individuals (Table S5), and were in the same range as Lebcke J. and colleagues [56] reported free campesterol serum concentration in 49 individuals ranging from 0.55 to 4.73 mg/L and β-sitosterol from 0.32 to 2.29 mg/L those as determined using an APPI-LC-MS/MS methodology [56].Plant J reported campesterol concentration is generally higher in serum than β-sitosterol [14].Also, 7-oxoβ-sitosterol was identified in human serum based on its retention time and MS 3 spectra.This methodology opens possibilities to investigate the role of phytosterols in human health and disease and could apply for the identification of oxyphytosterols in biological samples.

Fig. 1 .
Fig. 1.Chemical structures of plant sterols and their derivatives.(a) Basic structure of sterol with respective numbering.Most phytosterols (PS) have the same steroid core (A, B, C, D rings).The R group is a carbon side-chain.(b) The most abundant phytosterols:-campesterol, β-sitosterol, brasicasterol, fucosterol, stigmasterol and spinasterol.(c) The four main conjugates of PS.The hydroxyl group at C3 is esterified to a fatty acid (FA), where R1 is the carbon chain of the fatty acid or hydroxycinnamic acid (HA) or glycosylates to a hexose (HA) or acyl-hexose (AGH).The C3 hydroxyl group of HA is esterified to ferulic acid (here shown) or pcoumaric acid and AGH has a fatty acid that is esterified to the 6-OH of the hexose moiety.(d) Plant stanols, which contain a saturated core, also possible oxidation points on (plant) sterols are shown by asterisk and examples of diols, epoxy, and 7-oxo (plant) sterols.
scans.MS 2 transition was set on an expected derivatised phytosterols or predicted their metabolites/autoxidation products.The MS scans were performed on fragment-ions resulting from a neutral loss of 79 Da and 107 Da in the MS 2 , [M] + → [M-79] + →, and [M] + → [M-107] + →.A precursor-ion include list for and MS 3 [M] + → [M-79] + →, transitions for potential oxyphytosterols were set-up.MS 1 , MS 2 and MS scans contain three averaged microscans, each with a maximum ion fill

Fig. 4 .
Fig. 4. (a) MS 2 fragmentation, (b) MS 3 ([M] + → [M-79] + →) fragmentation of the O/GP-derivatised β-sitosterol.An asterisk preceding a fragment-ion describing letter e.g.*b 1-12,  indicates that the fragment-ion has lost the pyridine moiety from the derivatising group.A prime to the left of a fragment ion describing letter e.g.'*f, indicates that cleavage proceeds with the transfer of a hydrogen atom from the ion to the neutral fragment.A prime to the right of the fragment describing letter indicates that cleavage proceeds with hydrogen atom transfer to the fragment-ion e.g.*e'.The inset indicates fragmentation in the C-17 side chain of the GPderivatised sterols.Figures were taken from[21,41,44,47] with permission.

Fig. 9 .
Fig. 9. LC-MS 3 analysis using EADSA technology for serum sample from a healthy individual (a) RIC of MS 3 (544.4→465.4→)transition, (b) The MS 3 (544.4→465.4→)spectrum of the chromatographic peak at 7.42 min corresponding to fucosterol, (c) MS 3 (544.4→465.4→)spectrum recorded at 7.88 min corresponding to stigmasterol as identified based on comparison of RTs and MS 3 spectra with authentic standards.