A validated, fast method for quantification of sterols and gut

Jun 19, 2018 - There has been an increasing interest during recent years in the role of the gut microbiome on health and disease. Therefore, metabolit...
1 downloads 0 Views 789KB Size
Subscriber access provided by University of South Dakota

Article

A validated, fast method for quantification of sterols and gut microbiome derived 5#/#-stanols in human faeces by isotope dilution LC-high resolution MS Hans-Frieder Schoett, Sabrina Krautbauer, Marcus Hoering, Gerhard Liebisch, and Silke Matysik Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b01278 • Publication Date (Web): 19 Jun 2018 Downloaded from http://pubs.acs.org on June 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A Validated, Fast Method for Quantification of Sterols and Gut Microbiome Derived 5α/β-Stanols in Human Faeces by Isotope Dilution LC-High Resolution MS Hans-Frieder Schött, Sabrina Krautbauer, Marcus Höring, Gerhard Liebisch*, Silke Matysik* Institute for Clinical Chemistry and Laboratory Medicine, University Hospital Regensburg, Franz-Josef-Strauss-Allee 11, 93053 Regensburg, Germany ABSTRACT: There has been an increasing interest during recent years in the role of the gut microbiome on health and disease. Therefore, metabolites in human faeces related to microbial activity are attractive surrogate marker to track changes of microbiota induced by diet or disease. Such markers include 5α/β-stanols as microbiome-derived metabolites of sterols. Currently, reliable, robust, and fast methods to quantify faecal sterols and their related metabolites are missing. We developed a liquid chromatography-high resolution mass spectrometry (LC-MS/HRMS) method for the quantification of sterols and their 5α/β-stanols in human faecal samples. Faecal sterols were extracted and derivatized to N,N-dimethylglycine esters. The method includes cholesterol, coprostanol, cholestanol and sitosterol, 5α/β-sitostanol, campesterol and 5α/β-campestanol. Application of a biphenyl column permits separation of isomeric 5α- and 5β-stanols. Sterols are detected in parallel reaction monitoring (PRM) mode and stanols in full scan mode. HRMS allows differentiation of isobaric β-stanols and the [M+2]-isotope peak of the coeluting sterol. Performance characteristics meet the criteria recommended by FDA and EMA guidelines. Analysis of faecal samples from healthy volunteers revealed high inter-individual variability of sterol and stanol fractions. Interestingly, cholesterol and sitosterol showed similar fractions of mainly 5β-stanols. In contrast, campesterol is substantially converted to 5α-campestanol and might be a poorer substrate for bacterial metabolism. Robust and fast quantification of faecal sterols and their related stanols by LC-MS/HRMS offers great potential to find novel microbiome-related biomarker in large scale studies.

Recent findings suggest that the gut microbiome plays an important role in health and disease such as obesity 1,2, regulation of the immune system 3,4, incidence of cancer 5,6, and homeostasis of blood lipids 7,8. Food intervention 9-11, and antibiotic treatment 12,13 are recent attempts to modulate the gut microbiome and thereby influence human health. Human faeces contain, depending on the diet, different amounts of faecal sterols (FS) i.e. cholesterol, sitosterol and campesterol which are the most intensively studied faecal lipids. Beside these sterols also 5α and 5β hydrogenated species namely cholestanol and coprostanol, as well as 5α/β-sito- and campestanols derived from plant sterols are the main sterol metabolites found in human faeces 14-17 (see Suppl. Table S1 and Figure S1A). However, the role of faecal sterols in the development of colon cancer and atherosclerosis by reduction of the sterol absorption is largely unknown. The analysis of faecal stanols together with their sterol precursor might improve our understanding of the (patho)-physiological role of these metabolites. Moreover, as surrogate markers they could be applied to monitor changes of microbiota and their role in health and disease. Until today mainly gas chromatographic methods for the determination of faecal sterols are described with the drawback of long analysis run times. Additionally, most of the described methods do not cover 5α/β-stanols and focus on the quantifica-

tion of cholesterol and its metabolites only (see Suppl. Table 1) 18-24. Here we present a fast, validated liquid chromatography-high resolution tandem mass spectrometry (LCMS/HRMS) method for the quantification of cholesterol and plant sterols and their corresponding 5α/β-stanols in human faeces.

EXPERIMENTAL SECTION Chemicals and reagents Sodium hydroxide (p.a.), isooctane (2,2,4-trimethylpentane) >99%, cholesterol (cholest-5-en-3β-ol), cholesterol-d5 (D5cholest-5-en-3β-ol), campesterol (24(R)-methylcholest-5-en3β-ol), N,N-dimethylglycine (DMG), N-(3dimethylaminopropyl)-N′-ethylcarbodiimide (EDC), 5αsitostanol ((24(R)-ethyl-5α-cholestan-3β-ol), cholestanol (5αcholestan-3β-ol), and cholestanol-d7 (D7-5α-cholestan-3β-ol) were purchased from Sigma Aldrich Chemie GmbH (Steinheim, Germany). 2-Propanol HPLC grade was purchased from Karl Roth GmbH (Karlsruhe, Germany). Methanol LiChrosolv gradient grade for liquid chromatography, N,NDimethylpyridin-4-amine (DMAP), ammonium acetate and formic acid were obtained from Merck KGaA (Darmstadt, Germany). Campestanol-d7 (D7-24(R)-methyl-5α-cholestan3β-ol) and 5α-campestanol (24(R)-methyl-5α-cholestan-3β-ol)

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

were purchased from Toronto Research Chemicals (Toronto, ON, Canada). Sitostanol-d5 (D5-5α-stigmastan-3β-ol) 95%D was obtained by Medical Isotopes Inc. (Pelham, NH, USA). Sitosterol-d5 (D5-24(R)-ethylcholest-5-en-3β-ol), campesterold5 (D5-24(R)-methylcholest-5-en-3β-ol) were purchased from Sugaris (Münster, Germany). 5β-sitostanol (24(R)-ethyl-5βcholestan-3β-ol) was ordered from Chiron AS (Trondheim, Norway). Sitosterol (24(R)-ethyl-cholest-5-en-3β-ol) was purchased from Avanti Polar Lipids (Alabaster, Alabama, USA). Coprostanol-d5 (D5-5β-cholestan-3β-ol) was obtained from CDN Isotopes (Point-Claire, Quebec, Canada). All chemicals were of high purity grade for analysis. Purified water was produced by Millipore Milli Q UF-Plus water purification system (Molsheim, France). Stock solutions All sterol/stanol stock solutions (1.0 mg/mL) were prepared in methanol. The internal standard (ISTD) working solution contained sitosterol-d5 (8.8 µg/mL), campesterol-d5 (8.8 µg/mL), sitostanol-d5 (3.5 µg/mL), campestanol-d7 (0.4 µg/mL), cholesterol-d5 (8.8 µg/mL), cholestanol-d7 (1.6 µg/mL), coprostanol-d5 (8.8 µg/mL) solved in methanol. Samples For method development faeces samples were obtained from 22 healthy volunteers. Samples were collected in 50 mL flatbottom polypropylene tubes, stored immediately at -20°C and transported to the laboratory on ice. Until further processing samples were stored at -80°C. Sample amounts varied between 5-40 g wet weight and were collected from a single defaecation. The volunteers were requested to take samples from different faeces locations in order to get a more representative sample. Preparation of faecal homogenates Raw faeces homogenate was prepared by using up to 2.0 g of faeces, adding 2.5 mL 70%-isopropanol and homogenizing in a gentleMACS™ Dissociator (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). This homogenate was diluted with 2.5 mL 70%-isopropanol and again homogenized. Between preparation steps samples were kept on ice. The dry weight (dw) of the raw faeces homogenate was determined by overnight drying of 1.0 mL of this mixture. For further analysis the raw faeces homogenate was diluted to a final concentration of 2.0 mg dw/mL (diluted faeces homogenate DFH). DFH were stored at -80°C until further processing. Sufficient homogenization was evaluated by repeated determination of dry weight from 1 ml of raw faeces homogenate. The determined dry weight of these portions showed coefficients of variation 97 % for all compounds. Aliquots of the charcoal-stripped DFH and pooled DFH, respectively were supplemented with a combined standard solution to obtain six calibrators in appropriate concentration ranges. The concentration range of the calibrators was estimated based on previous studies and literature values 17,22,25.

Page 2 of 12

Preparation of DMG derivatives 100 µL ISTD mix, 200 µL DFH (2 mg dw/mL), 200 µL of an aqueous 5M NaOH solution and 500 µL of 70%-isopropanol were combined in a 15.0 mL tube and sealed with a screw cap. Sterol esters were hydrolysed at 60°C for 60 min in a water bath under constant agitation. After alkaline hydrolysis samples were neutralized by adding 1.0 mL of 1M hydrochloric acid and extracted with 3.0 mL of isooctane. 1200 µL of the upper isooctane layer were pipetted into an auto sampler vial, and evaporated to dryness in a vacuum concentrator. The residue was dissolved in a mixture of 60 µL DMG (0.5 M) and DMAP (2 M) in chloroform and 60 µL EDC (1 M) in chloroform at 45°C for 60 min. The derivatization reaction was stopped by adding 500 µL of methanol. Excess solvent was evaporated, and the residue was dissolved in 300 µL of methanol. The sample was centrifuged, and 100 µL were pipetted into a micro insert for analysis. Liquid chromatography high resolution mass spectrometry analysis of faecal sterols and stanols The analysis of sterols and stanols was performed using an LC-MS/HRMS system consisting of an UltiMate 3000 RS column oven, an UltiMate 3000 XRS quaternary UHPLC, and an UltiMate 3000 isocratic pump (Thermo Fisher Scientific, Waltham, MA USA) coupled to hybrid quadrupole-orbitrap mass spectrometer QExactive (Thermo Fisher Scientific, Bremen, Germany). The system was equipped with a heated electrospray ionization source. The ion source was operated in positive ion-mode using the following settings: Ion spray 3500V, sheath gas 58, aux gas 16, sweep gas 3 and aux gas heater temperature of 463°C. Capillary temperature was set to 281°C and the S-lens RF level to 55. Data analysis was performed with TraceFinder 3.1 Clinical (Thermo Fisher Scientific Waltham, MA USA) which extracts target ions and generates calibration equations. Quantification of faecal sterols by LC-MS/HRMS The separation was achieved on a Kinetex™ 2.6µm Biphenyl, 50x2.1mm column (Phenomenex, Aschaffenburg, Germany) at 40°C. The injection volume was 5 µL. Mobile phase A consisted of methanol/water 5/95 (v/v), mobile phase B was methanol/acetonitrile 10/90 (v/v), both containing 2mM ammonium acetate. The gradient elution started at 72% B with a flow rate of 500 µL/min. Solvent B is raised to 84.5% B until 3.5 min followed by a final increase to 100% B in 0.1 min. The flow rate was increased to 800 µL/min at 3.6 min and kept 0.5 min for column cleaning. After 4.1 min solvent B was set back to 72% and hold for 0.5 min. Method run time was 4.6 min. Isocratic pump was set to 0.2 mL/min methanol. Data were acquired in parallel reaction monitoring (PRM) mode with the following settings: Resolution 17,500; AGC target: 1e6, maximum IT 50 ms, MSX count 2, isolation window 0.8 m/z, and mass extraction window of ±5ppm. PRM monitors full product ion spectra of selected precursor ions, i.e. in our setting analyte and its internal standard were multiplexed and analyzed together (multiplexing of 2). To reduce the contamination of the mass spectrometer, the column flow was directed into the detector from 2.0-4.0 min by a divert valve. Collision energy for fragmentation was set to 15 eV.

ACS Paragon Plus Environment

Page 3 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Quantification of faecal stanols by LC-HRMS The same LC and MS conditions were applied as described above for sterol analysis except the following changes: The solvent B is raised to 88% B until 4.5 min, followed by final increase to 100% B in 0.1 min. The flow rate was increased to 800 µL/min at 4.6min followed by column cleaning at 800 µL/min kept for 0.6 min. After 5.2 min, solvent B was set back to 72% and hold for 0.4 min. The method run time was 5.7 min. Data were acquired in full MS mode with the following settings: Resolution 140,000 (at m/z 200), AGC target 1e6, maximum IT 150ms, a full scan range from m/z 469 to 511, and mass extraction window of ±3ppm. The sterols and 5α/βstanols were quantified using the ratio to the corresponding ISTD. Target masses used for quantification are listed in Table 1. Method Validation Method validation was performed according to the EMA and FDA guidelines on bioanalytical method validation 26,27 (for details see Data Supplement).

RESULTS AND DISCUSSION Method development Aim of the current study was to develop a comprehensive, robust and fast LC-MS/HRMS method to quantify FS in human samples. Sterols and stanols were derivatized to N,Ndimethylglycine (DMG) esters to permit efficient ionization 28. For optimized separation of stereoisomeric compounds, different stationary phases with high stereochemical selectivity were tested, i.e. biphenyl- and pentafluorophenyl (PFP) columns. With the PFP column the separation of different sterols i.e. cholesterol, sitosterol and campesterol was achieved although their corresponding epimeric 5α/β-stanols have not been separated (Suppl. Figure S2). In contrast, the biphenyl stationary phase permits chromatographic separation of isomeric 5α/βstanols which cannot be differentiated by their mass spectra (Figure 1).

Figure 1) Extracted ion chromatograms of sterols and stanols of a human faeces sample: Figure 1 A-F) shows the chromatograms of faecal sterols (blue), stanols (green) and deuterated internal standards (red) on a biphenyl column. A) cholesterol B) campesterol; C) sitosterol; D) coprostanol and cholestanol; E) 5α- and 5β-campestanol; F) 5α- and 5β-sitostanol.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Page 4 of 12

Table 1: Analytical characteristics of the LC-MS/HRMS and LC-HRMS method. Sterols were quantified by PRM analysis. Stanols were quantified by full scan HRMS. Compound

formula (DMG-derivative)

tR [min]

mass transition

calibration range [nmol/mg dw]

LOD [nmol/mg dw]

LOQ [nmol/mg dw]

Cholesterol

C31H5302N

2.29

472.4 → 369.3508

y = (0.565 ± 0.031)*x + (-0.299 ± 0.258)

0.193 - 137.68

0.029

0.39

β-Sitosterol

C33H5702N

2.75

500.4 → 397.3831

y = (1.247 ± 0.10)*x + (-0.025 ± 0.131)

0.039 - 32.28

0.039

0.16

Campesterol

C32H5502N

2.51

486.4 → 383.3670

y = (1.340 ± 0.153)*x + (-0.003 ± 0.153)

0.019 - 9.06

0.019

0.22

Cholesterol-d5

C31H44D502N

2.28

477.4 → 374.3842

Sitosterol-d5

C33H52D502N

2.74

505.5 → 402.4144

Campesterol-d5

C32H50D502N

2.49

491.5 → 388.3984

Compound

formula (DMG-derivative)

tR [min]

target mass

calibration range [nmol/mg dw]

LOD [nmol/mg dw]

LOQ [nmol/mg dw]

Coprostanol

C31H5502N

2.34

474.4306

y = (0.516 ± 0.023)*x + (-0.070 ± 0.096)

0.090 - 138.25

0.090

0.301

Cholestanol

C31H5502N

2.54

474.4306

y = (1.734 ± 0.110)*x + (-0.011 ± 0.019)

0.008 - 3.78

0.003

0.009

5β-sitostanol

C33H5902N

2.78

502.4619

y = (0.349 ± 0.039)*x + (-0.019 ± 0.047)

0.033 - 44.49

0.033

0.111

5α-sitostanol

C33H5902N

3.00

502.4619

y = (2.594 ± 0.108)*x + (-0.012 ± 0.060)

0.008 - 7.09

0.008

0.026

5β-campestanol

C32H5702N

2.56

488.4462

y = (3.672 ± 0.053)*x + (-0.026 ± 0.021)

0.009 - 21.36

0.009

0.032

5α-campestanol

C32H5702N

2.78

488.4462

y = (3.672 ± 0.053)*x + (-0.026 ± 0.021)

0.009 - 21.36

0.015

0.049

Coprostanol-d5

C31H50D502N

2.33

479.4580

Cholestanol-d7

C31H48D702N

2.53

481.4745

5α-sitostanol-d5

C33H54D502N

2.99

507.4932

5α-campestanol-d7

C32H50D702N

2.77

495.4901

Calibration curve

Calibration curve

a) calibration curve (n=5)

ACS Paragon Plus Environment

a)

a)

Page 5 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Table 2: Intraday and interday assay imprecision of sterols and 5α/β-stanols in faeces intra day imprecision Mean (n=6)

Cholesterol

Coprostanol

Cholestanol

Sitosterol

5β-sitostanol

5α-sitostanol

Campesterol

5β-campestanol

5α-campestanol

sample 1

[nmol/mg dw] 3.24

sample 2

coefficient of variation

inter day imprecision Mean (n=4)

coefficient of variation

7.0%

[nmol/mg dw] 3.30

5.69

2.2%

5.42

3.8%

sample 3

2.76

4.6%

2.66

6.8%

sample 4

2.89

1.6%

2.82

5.6%

sample 1

20.78

1.1%

21.38

6.2%

sample 2

28.78

1.8%

29.52

5.6%

sample 3

37.14

1.6%

38.52

4.4%

sample 4

9.33

1.9%

9.71

5.1%

sample 1

0.56

3.0%

0.56

3.5%

sample 2

0.80

4.6%

0.77

3.3%

sample 3

1.04

1.8%

1.03

2.8%

sample 4

0.42

0.9%

0.45

6.5%

sample 1

1.93

4.0%

1.93

6.8%

sample 2

1.89

4.0%

1.70

4.9%

sample 3

1.36

0.8%

1.25

8.4%

sample 4

2.31

1.7%

2.30

3.6%

3.5%

sample 1

11.67

3.0%

11.96

8.9%

sample 2

10.54

1.1%

10.42

10.4%

sample 3

18.12

3.3%

18.01

11.8%

sample 4

6.36

1.6%

6.63

13.2%

sample 1

0.73

3.6%

0.75

1.8%

sample 2

0.44

1.4%

0.44

1.8%

sample 3

0.78

1.9%

0.78

1.0%

sample 4

0.45

1.7%

0.49

8.0%

sample 1

0.54

4.7%

0.49

12.9%

sample 2

0.73

4.7%

0.68

2.0%

sample 3

0.54

2.3%

0.49

10.4%

sample 4

1.01

3.0%

0.91

9.9%

sample 1

0.29

1.8%

0.34

15.1%

sample 2

0.38

2.4%

0.45

16.0%

sample 3

0.76

4.3%

0.93

21.4%

sample 4

0.30

3.2%

0.38

21.1%

sample 1

0.50

2.4%

0.50

4.6%

sample 2

0.25

2.1%

0.24

2.5%

sample 3

0.55

1.3%

0.55

0.4%

sample 4

0.36

2.0%

0.36

3.0%

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 12

Table 3: Apparent recovery data of sterols and 5α/β-stanols in human faeces (Concentrations determined in 5 replicates) Spiked concentration [nmol/mg dw]

Compound

Concentration ± standard deviation [nmol/mg dw] 8.63 ± 0.50 11.43 ± 0.41 22.09 ± 0.64 73.71 ± 2.52

Apparent recovery [%]

Cholesterol

unspiked Spike low Spike medium Spike high

2.73 13.67 68.36

Coprostanol

unspiked Spike low Spike medium Spike high

2.42 12.11 60.56

15.18 17.34 26.92 72.17

± ± ± ±

0.48 0.25 0.56 3.60

89.2 97.0 94.1

Cholestanol

unspiked Spike low Spike medium Spike high

0.09 0.45 2.24

0.53 0.62 0.99 2.66

± ± ± ±

0.01 0.01 0.02 0.13

97.5 102.1 95.2

Sitosterol

unspiked Spike low Spike medium Spike high

0.77 3.85 19.24

3.85 4.58 7.57 21.77

± ± ± ±

0.03 0.07 0.18 0.82

94.8 96.7 93.2

5β-sitostanol

unspiked Spike low Spike medium Spike high

3.59 17.96 89.82

8.03 11.25 25.01 95.39

± ± ± ±

0.29 0.31 0.44 7.64

89.6 94.5 97.3

5α-sitostanol

unspiked Spike low Spike medium Spike high

0.16 0.81 4.07

0.54 0.70 1.30 4.28

± ± ± ±

0.02 0.02 0.03 0.22

94.0 93.5 92.0

Campesterol

unspiked Spike low Spike medium Spike high

0.11 0.57 2.86

1.45 1.56 1.98 3.97

± ± ± ±

0.11 0.05 0.09 0.10

95.0 92.6 88.1

5α-campestanol

unspiked Spike low Spike medium Spike high

0.08 0.39 1.96

0.38 0.47 0.79 2.27

± ± ± ±

0.01 0.00 0.02 0.10

111.3 103.9 96.7

Table 4: Concentrations of sterols and 5α/β-stanols in human faeces Compound

mean ± standard deviation

median

min

max

[nmol/mg dry weight] *

Cholesterol Coprostanol

20.24 20.41

± ±

20.05 15.97

11.03 16.83

2.62 0.06

65.66 64.49

Cholestanol

0.74

±

0.38

0.62

0.23

1.77

Sitosterol

6.01

±

4.69

4.58

1.16

16.65

5β-sitostanol

7.82

±

6.95

5.88

0.02

23.60

5α-sitostanol

0.61

±

0.27

0.60

0.10

1.23

Campesterol

2.22

±

1.69

1.74

0.50

5.66

5β-campestanol 5α-campestanol

1.17 0.51

± ±

2.90 0.22

0.55 0.48

0.00 0.07

13.70 1.01

*values of 22 volunteers

ACS Paragon Plus Environment

102.5 98.4 95.2

Characteristics of the method Validation confirmed the selectivity of the method. The specificity of the method was investigated using qualifier ions in six different faecal samples. The ion ratios quantifier/qualifier correspond to those of authentic standards with a maximum deviation of ±15% for all compounds. Calibration lines were linear over a wide range (Table 1) and verified by lack-of-fit testing. LOD and LOQ for stanols were assessed by signal-tonoise ratio and for sterols by functional testing because in PRM analysis almost no noise was present. The limits of detection and quantification were determined in the range of 0.003-0.09 nmol/mg dw and 0.026-0.301 nmol/mg dw, respectively. Imprecision of the method was tested in four real samples and showed coefficients of variation below 15% except for 5β-campestanol which reached CVs up to 22% (Table 2). The apparent recovery at low, medium and high levels was between 88% and 111% (Table 3). Internal standard corrected matrix factor (isMF) was determined in seven different samples at high and low spike concentration with a maximum variation of 16% CV. Derivatized samples were stable at least for 10 days at a temperate of 4-8 °C. Carryover of analytes was not detected. Faecal sterol and stanol concentration range in human faeces The FS concentrations were determined for 22 healthy volunteers (Table 4). In all samples both sterols and their related stanols were detected except in the sample of an infant where neither plant stanols nor coprostanol were detected. Highest concentrations were detected for cholesterol and its related 5βstanol coprostanol followed by sitosterol and campesterol and their related stanols. Interestingly, cholesterol and sitosterol were mainly converted to their 5β-stanols while campesterol

was mainly converted to 5α-campestanol (Figure 2). 100

80

60

40

20

ol ta n pe s

am 5a -C

Si to s

ta no l

ta no l 5a -

ho le s

st an ol

C

ol

pe

st an am 5ß -C

Si to 5ß -

C

op ro

st

st er

an ol

ol

ol pe

te r am C

Si to s

te r

ol

0

ho le s

Based on the following considerations the method was split into two separate runs, an LC-MS/HRMS run for sterols and a second LC-HRMS run for stanols: i) Due to their double bond at C-atom 5 DMG-derivatives of sterols form an abundant product ion of the steroid backbone representing a fragment with a high specificity for sterols. A schematic illustration of the fragmentation is shown in Suppl. Figure S1B. ii) Stanols do not contain that double bond and their DMG-derivatives generate primarily a DMG-derived fragment at m/z 104.0711 which has low specificity for stanols. iii) Because of the fast chromatography all compounds elute in a narrow retention time window resulting in an insufficient number of scans per chromatographic signal in MS/MS experiments. iv) Full scan HRMS enables the separation of the stanol species from the isobaric interference of the second isotope peak of the coeluting sterol (Suppl. Figure S3). Coprostanol, cholestanol, 5αsitostanol, 5α-campestanol were quantified with the corresponding deuterated standard. 5β-Sitostanol and 5βcampestanol are quantified with the deuterated 5α-stanols since no deuterated 5β-stanols are commercially available. 5βcampestanol was quantified with the calibration curve of 5αcampestanol since a 5β-campestanol standard is not available. Target compounds, internal standards and retention times are listed in Table 1.

% of total sterol and derived stanols

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

C

Page 7 of 12

Figure 2) Faecal sterol fractions in healthy volunteers: Displayed are the fractions of cholesterol, sitosterol, campesterols and their corresponding 5α/β-stanols as % of their sum (sterol + 5α + 5β-stanols). Each symbol represents an individual subject.

CONCLUSION Here we describe the first LC-MS/HRMS method for the quantification of cholesterol, plant sterols and their corresponding 5α/β-stanols derived from microbiota mediated metabolism. Chromatographic separation of stereoisomeric 5α/β-stanols was achieved on a biphenyl column. Finally, two runs were conducted to achieve both determination of faecal sterols and stanols. PRM mode was used for analysis of sterols and full scan mode was used for stanols. High mass resolution in full scan MS is required to resolve isobaric interferences from the M+2-isotope of the sterol to the corresponding 5βstanol due to their partial co-elution. Compared to existing GC methods an overall run time of 11 minutes and the analysis of both sterols and corresponding 5α/β-stanols are clear advantages of the presented LC-MS/HRMS method. Except for 5β-sitostanol and 5β-campestanol, all compounds were quantified using stable isotope labelled internal standards resulting in a high precision of the method. For 5β-sitostanol and 5βcampestanol we selected the corresponding deuterated 5αstanol as appropriate internal standard. The deuterated sterols sitosterol-d5 and campesterol-d5 might also be used for calculation with similar results. However, the correct choice of the internal standard is a delicate subject in exact quantification as pointed out recently by our group 29. The LOD and LOQ are sufficient for the determination of FS also in young children whose developing microbiome generates only a minor stanol fraction 30. In summary, the validation of the method confirms the excellent performance and robustness of the developed method and underlines its suitability for broad study cohorts and clinical screening attempts. In a first application we determined FS concentrations in 22 volunteers. The concentrations reported here are in concordance with results described in other studies (Suppl. Table S1). Our data support previous findings that the microbiome mediates the conversion of sterols to stanols. Furthermore, the

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

method discovers a high individual variance i.e. subjects with high (high metabolizer), medium, and low (low metabolizer) stanol fractions 14. In general, our data show that the conversion of cholesterol is comparable to that of sitosterol (Figure 2 and Figure S4; 18 out of 22 volunteers). However, the converted fraction of campesterol is lower than the fraction of cholesterol and sitosterol suggesting that campesterol might represent a poorer substrate for bacterial metabolism. Additionally, campesterol is, in contrast to cholesterol and sitosterol, substantially converted to 5α-campestanol. These findings indicate that a comprehensive FS determination used as surrogate markers can provide valuable information of the gut microbiome activity. However, since our number of subjects is low and the results might be influenced by characteristics of the study group the here described findings must be investigated in more detail in larger study cohorts. For reliable statistics much more data and representative sampling are necessary. Generally, harmonization of sterol determination in biological material is an international issue. First attempts were done for instance by the group of Mackay 31 In conclusion, the presented method provides a valuable tool to investigate faecal sterols and stanols as microbiome-related biomarkers and the relationship between gut microbiome and human health.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website.

Corresponding Author * Dr. Silke Matysik, Institute for Clinical Chemistry and Laboratory Medicine, University Hospital Regensburg, D-93042 Regensburg, Germany; Phone: +49-941-944-6281; Fax: +49-941944-6202; E-Mail: [email protected]

Author Contributions The manuscript was written through contributions of all authors. GL and SM contributed equally. All authors have given approval to the final version of the manuscript.

ACKNOWLEDGMENT This work was supported by the European Union's FP7 programme MyNewGut (grant agreement number 613979). We thank Simone Düchtel, Doreen Müller, Julia Schneider and Sebastian Roth for expert technical assistance.

REFERENCES

(1) Ridaura, V. K.; Faith, J. J.; Rey, F. E.; Cheng, J. Y.; Duncan, A. E.; Kau, A. L.; Griffin, N. W.; Lombard, V.; Henrissat, B.; Bain, J. R.; Muehlbauer, M. J.; Ilkayeva, O.; Semenkovich, C. F.; Funai, K.; Hayashi, D. K.; Lyle, B. J.; Martini, M. C.; Ursell, L. K.; Clemente, J. C.; Van Treuren, W.; Walters, W. A.; Knight, R.; Newgard, C. B.; Heath, A. C.; Gordon, J. I. Science 2013, 341, 1079-U1049.

Page 8 of 12

(2) Rodriguez, M. M.; Perez, D.; Chaves, F. J.; Esteve, E.; Marin-Garcia, P.; Xifra, G.; Vendrell, J.; Jove, M.; Pamplona, R.; Ricart, W.; PorteroOtin, M.; Chacon, M. R.; Real, J. M. F. Scientific Reports 2015, 5, 14. (3) Gomez, A.; Luckey, D.; Taneja, V. Clinical Immunology 2015, 159, 154-162. (4) Forbes, J. D.; Van Domselaar, G.; Bernstein, C. N. Frontiers in Microbiology 2016, 7, 18. (5) Louis, P.; Hold, G. L.; Flint, H. J. Nature Reviews Microbiology 2014, 12, 661-672. (6) Sivan, A.; Corrales, L.; Hubert, N.; Williams, J. B.; Aquino-Michaels, K.; Earley, Z. M.; Benyamin, F. W.; Lei, Y. M.; Jabri, B.; Alegre, M. L.; Chang, E. B.; Gajewski, T. F. Science 2015, 350, 1084-1089. (7) Fu, J. Y.; Bonder, M. J.; Cenit, M. C.; Tigchelaar, E. F.; Maatman, A.; Dekens, J. A. M.; Brandsma, E.; Marczynska, J.; Imhann, F.; Weersma, R. K.; Franke, L.; Poon, T. W.; Xavier, R. J.; Gevers, D.; Hofker, M. H.; Wijmenga, C.; Zhernakova, A. Circulation Research 2015, 117, 817-824. (8) Zhong, C. Y.; Sun, W. W.; Ma, Y. Y.; Zhu, H. L.; Yang, P.; Wei, H.; Zeng, B. H.; Zhang, Q.; Liu, Y.; Li, W. X.; Chen, Y. X.; Yu, L. Q.; Song, Z. Y. Scientific Reports 2015, 5, 12. (9) Walker, A. W.; Ince, J.; Duncan, S. H.; Webster, L. M.; Holtrop, G.; Ze, X. L.; Brown, D.; Stares, M. D.; Scott, P.; Bergerat, A.; Louis, P.; McIntosh, F.; Johnstone, A. M.; Lobley, G. E.; Parkhill, J.; Flint, H. J. Isme Journal 2011, 5, 220-230. (10) David, L. A.; Maurice, C. F.; Carmody, R. N.; Gootenberg, D. B.; Button, J. E.; Wolfe, B. E.; Ling, A. V.; Devlin, A. S.; Varma, Y.; Fischbach, M. A.; Biddinger, S. B.; Dutton, R. J.; Turnbaugh, P. J. Nature 2014, 505, 559-563. (11) Valentini, L.; Pinto, A.; BourdelMarchasson, I.; Ostan, R.; Brigidi, P.; Turroni, S.; Hrelia, S.; Hrelia, P.; Bereswill, S.; Fischer, A.; Leoncini, E.; Malaguti, M.; Blanc-Bisson, C.; Durrieu, J.; Spazzafumo, L.; Buccolini, F.; Pryen, F.; Donini, L. M.; Franceschi, C.; Lochs, H. Clinical Nutrition 2015, 34, 593-602. (12) Cho, I.; Yamanishi, S.; Cox, L.; Methe, B. A.; Zavadil, J.; Li, K.; Gao, Z.; Mahana, D.; Raju, K.; Teitler, I.; Li, H. L.; Alekseyenko, A. V.; Blaser, M. J. Nature 2012, 488, 621-+.

ACS Paragon Plus Environment

Page 9 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

(13) Gibson, M. K.; Crofts, T. S.; Dantas, G. Current Opinion in Microbiology 2015, 27, 5156. (14) Wilkins, T. D.; Hackman, A. S. Cancer Research 1974, 34, 2250-2254. (15) Reddy, B. S.; Wynder, E. L. Cancer 1977, 39, 2533-2539. (16) Reddy, B.; Engle, A.; Katsifis, S.; Simi, B.; Bartram, H. P.; Perrino, P.; Mahan, C. Cancer Research 1989, 49, 4629-4635. (17) Weststrate, J. A.; Ayesh, R.; Bauer-Plank, C.; Drewitt, P. N. Food and Chemical Toxicology 1999, 37, 1063-1071. (18) Eneroth, P.; Hellstroem, K.; Ryhage, R. J Lipid Res 1964, 5, 245-262. (19) Korpela, J. T. Scandinavian Journal of Clinical & Laboratory Investigation 1982, 42, 529-534. (20) Arca, M.; Montali, A.; Ciocca, S.; Angelico, F.; Cantafora, A. J Lipid Res 1983, 24, 332-335. (21) Barker, G. M.; Radley, S.; Davis, A.; Setchell, K. D. R.; Oconnell, N.; Donovan, I. A.; Keighley, M. R. B.; Neoptolemos, J. P. International Journal of Colorectal Disease 1993, 8, 188-192. (22) Batta, A. K.; Salen, G.; Rapole, K. R.; Batta, M.; Batta, P.; Alberts, D.; Earnest, D. J.Lipid Res. 1999, 40, 1148-1154. (23) Batta, A. K.; Salen, G.; Batta, P.; Tint, G. S.; Alberts, D. S.; Earnest, D. L. Journal of Chromatography B-Analytical Technologies in

the Biomedical and Life Sciences 2002, 775, 153161. (24) Wu, J. M.; Hu, R. K.; Yue, J. Q.; Yang, Z. G.; Zhang, L. F. Journal of Chromatography A 2009, 1216, 1053-1058. (25) Keller, S.; Jahreis, G. J Chromatogr B Analyt Technol Biomed Life Sci 2004, 813, 199207. (26) In Guideline on Bioanalytical Method Validation. CHMP, European Medicines Agency, London., Use, C. f. M. P. f. H., Ed., 2014. (27) In Guidance for Industry: Bioanalytical Method Validation. FDA, US Department of Health and Human Services, Washington, DC., (FDA), U. F. a. D. A., Ed., 2001. (28) Pataj, Z.; Liebisch, G.; Schmitz, G.; Matysik, S. J Chromatogr A 2016, 1439, 82-88. (29) Krautbauer, S.; Buchler, C.; Liebisch, G. Anal Chem 2016, 88, 10957-10961. (30) Midtvedt, A. C.; Midtvedt, T. Journal of Pediatric Gastroenterology and Nutrition 1993, 17, 161-168. (31) Mackay, D. S.; Jones, P. J. H.; Myrie, S. B.; Plat, J.; Lutjohann, D. Journal of Chromatography B-Analytical Technologies in the Biomedical and Life Sciences 2014, 957, 116122.

Table of Contents (TOC) graphic:

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 12

10 ACS Paragon Plus Environment

Page 11 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Fig 1 254x190mm (96 x 96 DPI)

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Fig 2 254x190mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 12 of 12