High-Throughput Quantitative Lipidomics Analysis of Nonesterified

Subscriber access provided by UNIVERSITY OF LEEDS

Article

A High-throughput Quantitative Lipidomics Analysis of Non-esterified Fatty Acids in Human Plasma Nicolas Christinat, Delphine Morin-Rivron, and Mojgan Masoodi J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00198 • Publication Date (Web): 17 May 2016 Downloaded from http://pubs.acs.org on May 19, 2016

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 free 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 accessible to all readers and 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.

Journal of Proteome Research 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 25

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

Journal of Proteome Research

A High-throughput Quantitative Lipidomics Analysis of Non-esterified Fatty Acids in Human Plasma Nicolas Christinat, Delphine Morin-Rivron, and Mojgan Masoodi* Lipid Biology, Department of Gastro-Intestinal Health & Microbiome, Nestlé Institute of Health Sciences, EPFL Innovation Park, Bâtiment H, 1015 Lausanne, Switzerland

ABSTRACT We present a high-throughput, non-targeted lipidomics approach using liquid chromatography coupled to high resolution mass spectrometry for quantitative analysis of non-esterified fatty acids. We applied this method to screen a wide range of fatty acids from medium chain to very long chain (8 to 24 carbon atoms) in human plasma samples. The method enables us to chromatographically separate branched-chain species from their straight-chain isomers as well as separate biologically important ω-3 and ω-6 polyunsaturated fatty acids. We used 51 fatty acid species to demonstrate the quantitative capability of this method with quantification limits in the nanomolar range; however, this method is not limited only to these fatty acid species. High-throughput sample preparation was developed and carried out on a robotic platform that allows extraction of 96 samples simultaneously within 3 hours. This highthroughput platform was used to assess the influence of different types of human plasma collection and preparation on the non-esterified fatty acid profile of healthy donors. Use of the 1 ACS Paragon Plus Environment


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 2 of 25

anticoagulants EDTA and Heparin has been compared to simple clotting, and only limited changes have been detected in most non-esterified fatty acid concentrations.

Keywords: Lipidomics, liquid chromatography, mass spectrometry, non-esterified fatty acids, branched-chain fatty acids, human plasma.

INTRODUCTION Fatty acids, as primary building blocks of complex lipids, are important components of living organisms and have multiple biological functions. For instance, they are major components of cell membranes and play important roles in energy storage and gene regulation. Under their free form, known as non-esterified fatty acids (NEFA), they have been linked to multiple human diseases, primarily metabolic disorders such as diabetes1, obesity2, or insulin resistance3, but also to brain disease4. In this context, measurements of NEFA plasma levels and profiles are valuable tools for assessing the alteration of fatty acid metabolism, and these applications have received considerable interest in the past years5,6. Understanding the profile of non-esterified fatty acids is very important in any biological system in order to elucidate their biological functions that rise from their signaling properties. Different methods such as thin-layer chromatography (TLC) and electrophoresis have been used for quantifying fatty acids in various matrices7. However, gas chromatography (GC) coupled to flame ionization detection (FID) or mass spectrometry (MS) has traditionally been the method of choice for detection and quantification of both total and NEFA fractions of circulating fatty acids8,9.

GC-MS provides

good

sensitivity together with

high

chromatographic resolution, allowing for the separation and quantification of positional isomers such as ω-3 and ω-6 isomers of polyunsaturated fatty acids (PUFA)10. This methodology however requires derivatization of fatty acids in order to achieve the desired 2 ACS Paragon Plus Environment

Page 3 of 25

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


separation and sensitivity11. Although this procedure, in particular for preparation of methyl ester of fatty acids, has been optimized over the years, it remains laborious and timeconsuming12. In recent years, liquid chromatography (LC) coupled with mass spectrometry has gained popularity as an alternative to GC13. Although derivatization protocols for LC-MS/MS detection have been widely used14, one of its biggest advantages is the possibility to directly analyze non-derivatized fatty acids. This greatly simplifies sample preparation and broadens the field of application while preserving excellent sensitivity and resolution15. LC-MS(/MS) is particularly well suited to the quantification of NEFA and has been successfully applied to fatty acid profiling in various matrices such as human plasma16,17, tissue18, cell culture19 or shellfish20. Most of the published methods use a C18 stationary phase for separation coupled with mass spectrometry. Sub-classes of NEFA are usually targeted, depending on their chain length or degree of saturation. Recently, methods for quantitative analysis of a broader range of NEFA have been published. Koletzko and coworkers presented a high-throughput method for the quantification of a broader range of NEFA (30-40 species) and its applications to bio-fluid analysis21. High-resolution mass spectrometry19,22 or less conventional stationary phases21 have also allowed quantifying up to 30-40 NEFA species. Although this method provides valuable insight into biological samples, it lacks the resolving power to distinguish between positional isomers, thus losing valuable biological information. In order to fill this gap, we developed a method for the quantification of a large panel of NEFA in human plasma, ranging from medium-chain to very long-chain fatty acids. Special care was given to the separation of positional isomers, and most PUFA isomers as well as branched-chain fatty acids (BCFA) have been successfully separated. Untargeted analysis was performed using a LTQ Orbitrap mass spectrometer. High-resolution mass spectrometry,

3 ACS Paragon Plus Environment


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

which has already been successfully used to study fatty acids metabolism

Page 4 of 25

19,22

, offers a great

advantage compared to triple quadrupole instruments, namely the ability to quantify known species while simultaneously screening for unknowns. During the development of our method, we used this technical advantage to expand our NEFA panel, adding biologically important but often neglected BCFA23-25 to classically quantified saturated and unsaturated fatty acids. To the best of our knowledge, this is the first method for the simultaneous quantification of these three types of NEFA in human plasma. EXPERIMENTAL SECTION Chemicals and Reagents Acetonitrile and isopropanol (LC-MS grade) were purchased from VWR Internationals (Leuven, Belgium) and Merck (Darmstadt, Germany) respectively. Water was purified inhouse using a Milli-Q Advantage A10 system from Merck Millipore (Billerica, MA, USA). Ammonium acetate (≥99.99% purity) was supplied by Sigma-Aldrich (St-Louis, MO, USA). External calibration was performed using a GLC-566 standard (Nu-Check Prep Inc., Elysian, MN, USA), containing 38 fatty acid species as well as individual chemicals of 18 other fatty acid species. These fatty acids were purchased from Larodan Fine Chemicals AB (Malmoe, Sweden), Nu-Check Prep Inc. (Elysian, MN, USA) and Sigma-Aldrich (St-Louis, MO, USA). Uniformly labelled octanoic acid (d15, 98 %), decanoic acid (d19, 98 %), dodecanoic acid (d23, 98%), hexadecanoic acid (d31, 98%) and octadecanoic acid (d35, 98%) were purchased from Sigma-Aldrich (St-Louis, MO, USA). 10-pentadecenoic acid and 10,13-nonadecadienoic acid were purchased from Nu-Check Prep. Inc. (Elysian, MN, USA) and used as internal standards. Fatty acid Nomenclature


Page 5 of 25

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


Throughout the text, fatty acids are named according to the number of carbon atom and number of double bonds in their chain, using the scheme “C number of carbon/number of double bond”. The position of the first double bond is indicated by “ω-x” where x is the first double bond carbon atom from the methyl (the “ω”) end of the chain. For instance, stearic acid is represented as C18:0 and arachidonic acid as C20:4 ω-6. The position of the methyl group of branched-chain fatty acids is indicated by the suffix iso or anteiso. Preparation of Internal and External standards Individual 100 µM stock solutions of each internal standard (octanoic-d15 acid, decanoic-d19 acid, dodecanoic-d23 acid, 10-pentadecenoic acid, hexadecanoic-d31 acid, octadecanoic-d35 acid, and 10,13-nonadecadienoic acid) were prepared from the corresponding commercially available powder and stored at -20°C. On the day of analysis, 55 µl of each stock solution was placed in a 50 ml volumetric flask and the volume was completed with isopropanol to yield sufficient precipitation reagent for preparing 96 samples. External standards were prepared as follows: A total of 114.05 mg of GLC-566 standard was dissolved in 10 mL of methanol. A volume of 1 mL of this solution was further diluted with 9 mL of methanol to obtain a concentration of 30-250 µM of each fatty acid. In parallel a second mix containing 18 additional fatty acids in methanol was prepared from individual stock solutions. An HTS PAL liquid handler (CTC Analytics, Zwingen, Switzerland) was then used to mix the stock solutions in order to prepare calibration standards. Different volumes of the two stock solutions and internal standards solution (10 µM in methanol) were mixed, and the volume of each standard was manually adjusted to 1 mL with water/acetonitrile (50/50) to yield a series of 10 calibration standards with approximate concentrations of 0.005−70 µM, covering the range of physiological concentrations in human plasma. MS Instrumentation and Method Settings 5 ACS Paragon Plus Environment


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 25

A LTQ Orbitrap Elite mass spectrometer (ThermoFisher Scientific, Bremen, Germany) was operated in negative ionization mode and detection was performed over the mass range m/z 110-380 with a resolving power of 60’000 (at m/z = 400). The mass spectrometer was interfaced to the UPLC system using a HESI probe. The spray voltage was set to -3kV. The heater and capillary temperatures were set to 300°C and 350°C respectively. Sheath gas and auxiliary gas flow rate were set to 35 and 10 AU respectively. The instrument was calibrated every four days according to manufacturer specifications. Chromatographic Conditions A I-Class UPLC system (Waters Corporation, Milford, MA, USA) combining a binary pump, a FTN autosampler and a column oven was used. Chromatographic separation was performed on a Waters ACQUITY UPLC CSH C18 Column, (130 Å, 1.7 µm, 150 x 2.1 mm) using a binary gradient with solvent A being a 2:3 mixture of water/acetonitrile with 10 mM ammonium acetate and solvent B being a 1:1 mixture of acetonitrile/isopropanol with 10 mM ammonium acetate. After 2 minutes of equilibration with 90% eluent A, eluent B was linearly increased to 46% over 12 minutes and subsequently to 100% over 3.5 minutes. The equipment was then rinsed for 3 minutes at 100% B and allowed to re-equilibrate in starting conditions for 2.5 minutes. The eluent flow rate was set to 450 µl/min. Column oven temperature was set to 55°C and the autosampler injection volume to 1 µl. Collection and Handling of Plasma Samples Pooled human plasma (2K EDTA) used to assess repeatability of the measurement was purchased from Biopredic International (Rennes, France) and was collected from blood donors in blood centers. All blood donors were asked to sign a consent form, and personnel of the Nestlé Institute of Health Sciences did not have access to any direct identifiers or key codes that could link biological material to its donor. Use of human plasma for this project


Page 7 of 25

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


was approved by an independent ethical committee and the material was handled according to the Swiss Human Research Act. Plasma EDTA, plasma heparin, and clotted blood were obtained from the Harmonized Micronutrient clinical trial (NCT01823744), participants were healthy and between 9 to 13 years old. Each participant signed the Statement of Informed Assent and parents of each participant signed informed consent. These samples were used to evaluate the effect of plasma preparation on NEFA profile. Sample Preparation All sample pipetting steps were performed using a Star robotic unit (Hamilton Bonaduz AG, Bonaduz, Switzerland) except mentioned otherwise. 50 µl of plasma was placed in a 96, 2000 µL, white border deep-well plate (Eppendorf AG, Hamburg, Germany) and 450 µl of IS solution in isopropanol – see above for preparation – was added in order to precipitate proteins. The plate was shaken for 30 minutes at 700 rpm in a Thermomixer Comfort C (Eppendorf AG, Hamburg, Germany) and then placed in an Eppendorf 5810R centrifuge (Eppendorf AG, Hamburg, Germany) for 10 minutes at 1500 rpm. During all operations, the samples were maintained at 4°C. 150 µl of supernatant was then pipetted and placed in a skirted twin.tec PCR Plate 96 (Eppendorf AG, Hamburg, Germany). After solvent evaporation in a Concentrator Plus speedvac (Eppendorf AG, Hamburg, Germany), the extract was reconstituted in 75 µl acetonitrile-water (1:1) and shaken for 5 minutes at 1000 rpm in the Thermomixer maintained at 4°C. The plate was sealed with an aluminium foil and directly placed in the autosampler for analysis. Data Analysis Quantitative analyses were performed with Xcalibur software 2.2 SP1 (ThermoFisher Scientific, Bremen, Germany). Chromatograms were extracted using a mass tolerance of 5



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 8 of 25

ppm and signals were integrated with the ISIS algorithm. Calibration curves were linearly fitted with a weighting factor of 1/x. Screening experiments were performed using Thermo Scientific LipidSearch software (Mitsui Knowledge Industry, Tokyo, Japan). All data postprocessing was performed with an R-based package developed in-house. RESULTS AND DISCUSSION Sample Preparation. We used 57 non-esterified fatty acids (Table S1) to optimize sample preparation. The standards were spiked in an aqueous solution of bovine serum albumin, which served here as a plasma surrogate. During this optimization phase, we tried to keep the procedure as simple as possible in order to easily automate it on a robotic unit later. In this respect, protein precipitation is a simple sample-preparation technique, which has been successfully used for analysis of non-esterified fatty acids by LC-MS/MS 21,36. We tested protein precipitation with different organic solvents such as acetonitrile, methanol, and isopropanol as well as different sample to organic ratios. When used in an approximate sample-organic ratio of 1:10, all three solvents gave fatty acid recoveries in the 70-120% range. We chose isopropanol because this solvent has the lowest volatility and highest viscosity of the three, making it better suited for automatic pipetting by a robotic unit. Eventually we opted for a 1:9 sample-isopropanol ratio with a sample load of 50 µl. The rest of the preparation involved classical steps of mixing, centrifugation, evaporation, and reconstitution. During the latter step, the extracts were slightly concentrated and the overall dilution factor for the sample preparation was 5. The next step involved automation of pipetting steps on a Hamilton Star robotic unit in order to decrease inter-sample variation and increase throughput, as this unit typically uses a 96wells plate format. Another advantage of the robotic unit is that it can operate at 4°C, which limits sample degradation, e.g. the release of fatty acids from intact lipids and fatty acid


Page 9 of 25

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


oxidation. Once fully optimized, the method allowed us to extract a complete 96-well plate in about 3 hours. LC/MS Method Development GLC-566 reference mixture containing 38 NEFA was used as a reference to develop and optimize our LC-MS method. Various chromatography columns and solvent systems were tested but no combination of these allowed for separation of all fatty acid isomers. Very good separation

was

obtained

when

a

15

cm

long

CSH

C18

column

and

a

water/acetonitrile/isopropanol gradient containing ammonium acetate as an additive were used. Within the 17.5 minutes of the gradient most isomers could be baseline separated. Only ω-3/ω-6 isomers of C20:3 could not be separated. C18:1 ω-7 and ω-9 peaks showed a slight overlap but the separation latter proved to be sufficient for simultaneously quantifying both isomers. NEFA were detected as deprotonated molecules in full scan MS mode. We chose MS detection rather than MS/MS because non-esterified fatty acids do not usually show very rich fragmentation spectra and often non-specific MRM transitions are used for quantification, thus lowering the specificity of the analysis. In our method, the high resolution, high massaccuracy of the measurement provides the necessary specificity, with a 5 ppm mass tolerance being sufficient to separate analytes from contaminants and matrix molecules. A good illustration of the importance of the mass tolerance window is given by tetracosapentaenoic acid. When C24:5 chromatogram is extracted with a 5 ppm mass tolerance (Figure 2, trace A), virtually no background noise can be detected. On the other hand, a 50 ppm window (Figure 2, trace B) does not allow to filter out the contaminant signal at m/z 357.2652. Although in the latter case signals corresponding to ω-3 and ω -6 isomers can still be detected, the quantification limits of these two NEFA will be negatively affected by the higher background noise. The 5 ppm mass tolerance window also offers the advantage to avoid potential isotopic 9 ACS Paragon Plus Environment


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 25

cross-talk between fatty acids having the same number of carbon atoms but different degrees of saturation, should these species not be chromatographically separated. Finally, the m/z range was set in order to cover all investigated NEFA and the resolving power (instrument setting 60,000 at m/z 400) permitted resolution of all species in the mass spectra at sufficient scan speed for reliable quantification. Once our method was optimized for the reference mixture, we investigated whether other non-esterified fatty could be detected. We thus performed an initial screen on human plasma in order to identify potential new targets. Plasma samples were extracted as described in the experimental section. For data processing, we used LipidSearch software. This software program possesses a build-in database of short-chain to very long-chain fatty acids, and after being provided with basic experiment parameters such as formed adducts, type of instrument, and mass tolerance, it can extract chromatograms and perform peak picking. The screening experiment allowed us to find multiple distinct chromatographic peaks that had not been previously identified. We found signals corresponding to masses of the long-chain and very long-chain polyunsaturated fatty acids C16:2, C18:4, C20:3, C20:4, C22:5, C24:4, C24:5 (2 signals), and C24:6. Although these signals usually showed low intensities (~3-4e3), the background noise was also very limited, making them good candidates to include in our quantification method. In order to confirm their identity, we compared their retention time and isotopic pattern with those of commercially available standards. Using this methodology we were able to identify stearidonic acid (C18:4 ω-3), mead acid (C20:3 ω-9), 4,7,10,13,16docosapentaenoic acid (C22:5 ω-6), 9,12,15,18,21-tetracosapentaenoic acid (C24:5 ω3),9,12,15,18-tetracosatetraenoic acid (24:4 ω-6), 6,9,12,15,18-tetracosapentaenoic acid (C24:5 ω-6), and nisinic acid (C24:6 ω-3) (Figure 1). While we were unable to confirm the identity of C16:2 and C20:4 due to lack of commercially available standards, 7,10hexadecadienoic acid (C16:2 ω-6) and 8,11,14,17-eicosatetraenoic acid (C20:4 ω-3) were


Page 11 of 25

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


predicted based on accurate mass and relative RT to our reference standards, which is reasonable as both of them have already been detected/reported in human plasma or serum26,27. In addition to polyunsaturated fatty acids, we also identified monounsaturated and saturated fatty acids. Most monounsaturated fatty acid traces showed multiple peaks. Focusing on the most intense signals, we were able to successfully identify 10-heptadecenoic acid (C17:1 ω7). Investigations to identify signals from C18:1 and C20:1 are still ongoing. Although the observed signals were tested for a few predicted candidates such as petroselenic acid (C18:1 ω-12), 8-eicosenoic acid (C20:1 ω-12), and 5-eicosenoic acid (C20:1 ω-15) standards, none of them matched observed retention times. Further investigation is required to elucidate the structure of unknown signals. Our next candidates will be trans isomers of unsaturated fatty acids as they are known to be detectable in plasma by classical methods28. More surprisingly, we found additional chromatographic peaks matching the exact mass of many saturated fatty acids but eluting earlier than the reference standards. C15:0 and C17:0 traces each showed a total of three well-separated peaks, while C19:0, C20:0, and C23:0 traces each showed two. Comparison with commercially available standards showed that these peaks belong to branched-chain fatty acids. Indeed, our method was able to separate 12methyl-tetradecanoic acid (C15:0-anteiso), 13-methyl-tetradecanoic acid (C15:0-iso), 14methyl-hexadecanoic acid (C17:0-anteiso), 15-methyl-hexadecanoic acid (C17:0-iso), phytanic acid (3,7,11,15, tetramethyl hexadecanoic acid) and its alpha-oxidation product pristanic acid (2,6,10,14 tetramethyl pentadecanoic acid) by their straight-chain isomers (Figure 3). While pristanic and phytanic acid have already been detected and quantified in human plasma29,30, the presence of branched C15:0 and C17:0 was unexpected. These fatty acids have been detected in human samples such as vernix and meconium of healthy term newborns31, and are known to be present in bacteria and food although so far, no reports have



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 12 of 25

established their presence in human plasma. BCFAs are bioactive lipids and also essential membrane components of many bacterial species. However, their role in human nutrition and health has not been fully explored. A study in neonatal rats showed that BCFA altered the gastrointestinal microbial ecology towards organisms that use BCFA in their membranes and reduced the incidence of necrotising enterocolitis32. Furthermore, BCFA induce apoptosis in human breast cancer cells, and inhibit tumour growth in cultured cells and in vivo

33,34

. The

estimated mean intake of BCFA in the American diet reaches 500 mg/d, delivered primarily from ruminant food products35. These data all point to the potential and important nutritional benefits that BCFA may have for the development and maintenance of microbiota, and possibly other functions. Method Validation The entire method-sample preparation and LC-MS analysis- was validated according to the guidelines of the Food and Drug Administration (FDA). Specifically, we assessed the method’s extraction recovery, reproducibility, accuracy, precision, and dynamic range (linearity, limit of detection & quantification).. In order to assess dynamic range and lower limit of quantification (LLOQ) of the method, we performed 10-point calibrations for every NEFA, using triplicates for each calibration level. Linearity of the obtained calibration curves was assessed in the range of 0.005-100 µM and a coefficient of determination (R2) > 0.99 calculated over a minimum of 6 calibration points was considered acceptable. LLOQ was defined as the lowest concentration that could be robustly quantified (coefficient of variation (CV) ≤ 20% in 3 different samples), and limit of detection (LOD) was calculated as the minimum concentration resulting in a signal-to-noise ratio of 3. Using these criteria, 53 fatty acids species were successfully validated (Table S2).


Page 13 of 25

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


The recovery of the extraction method was determined at two concentration levels for the 53 NEFA that were successfully quantified. Recovery was determined using fatty acid standards in albumin solutions (45 g/L). 5 µL aliquots of standard solution (high and low concentrations) were spiked in 45 µL of albumin solutions (50 g/L). The mixture was shaken for 15 minutes at 4°C prior and samples were extracted as described in the experimental section. 51 out of the 53 NEFA showed acceptable recovery ranging from 80% to 119%. For two of them, C19:0 and C22:0 the recovery was extremely variable (40 to 218%). We thus concluded that these two NEFA could not be reliably extracted, – presumably due to their low solubility in isopropanol – and they were excluded from the method for quantitative analysis. Post-preparation stability was assessed by comparing calculated concentrations obtained during three consecutive days after sample workup. The extracts were stored at 4°C (autosampler conditions) and analyzed as triplicates. We found accuracies between 80.0% and 120.0% for all NEFA species. These results prove post-preparation stability for at least 3 days stored at 4°C. The reproducibility of the analysis was assessed by performing repeated sample preparations and analysis of 9 aliquots of human plasma samples during five consecutive days. Briefly, human plasma samples were aliquoted prior to analysis and kept at -80 °C. On the day of analysis, the aliquots were defrosted for 2 h at 4°C. After rapid mixing, sample preparation was performed as described in the experimental section. Out of the 51 NEFA, 47 were detected and 38 were present in sufficient amount to be accurately quantified. For all amounts above LLOQ, intra-assay variations (calculated as RSD on calculated amounts) was within 14.6%. Inter-assay variation was within 14.8%, proving the repeatability of the analysis. Detailed results are shown in Table 2. Human Plasma Screening.



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 14 of 25

In addition to assessing the repeatability of the assay, the measurement of human plasma also offered us values for comparison with other published results. The LipidMaps consortium performed an extensive analysis of various lipid classes in human plasma36. In particular, they performed triplicate measurements of 31 NEFA species by GC-MS. Compared to the LipidMaps group, we found slightly higher concentrations of NEFA in our plasma samples. Plausible explanations for these differences are the use of different analytical techniques (LCMS vs GC-MS), as well as differences in the number of plasma donors (6 vs 100 for LipidMaps plasma) . These hypotheses are further supported by the fact that our results are much closer to those reported by Koletzko and colleagues in their LC-MS/MS measurement of NEFA in plasma of 8 overnight-fasted subjects21. Although comparison with this method is complicated by its inability to separate positional isomers, measurements are usually within ±50% of each other. Low-level NEFA such as stearidonic acid show broader variation but measurements of abundant palmitic, oleic, and linoleic acids are in good agreement. Table S3 shows a comparison of NEFA levels measured in the different studies. Our investigation of plasma sample preparation gave us the opportunity to compare NEFA levels in an adult and child population. Again, the results from the reproducibility experiment were compared to those of a pooled sample of all 30 individual plasma EDTA samples (table S3). Measured levels of saturated fatty acids were equal or higher in the adult group. For instance, palmitic acid had equivalent levels for adults and children, while stearic acid was 2.5-fold higher in adult population. Compared to other saturated fatty acids, octanoic acid concentration was remarkably different in the two populations: while we measured about 1 µM in children samples, the corresponding value for adults was about ten times higher. Similar results were observed for monounsaturated and branched-chain chain fatty acids. Plasma concentrations in adults were usually within a 1 to 2.5 times higher than those measured in children. Only C11.1 and C12:1 were found outside this range.


Page 15 of 25

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


PUFA showed a completely opposite pattern as their concentrations are generally equivalent or higher in children’s plasma. For instance, DHA concentration was about 3 µM in both groups. At the other end of the spectrum, linoleic acid concentration is two times higher in children. It is important to point out that this is just an approximate estimate based on pooled plasma samples, and must be further investigated in a well-designed clinical cohort.

Effect of human plasma preparation on NEFA profile. Using our high-throughput method we investigated whether different plasma preparation procedures had an impact on the NEFA profile. We compared blood collection into EDTA and heparin tubes to simple blood clotting by measuring NEFA concentrations of 90 samples coming from 30 healthy donors. Plasma EDTA and plasma heparin were prepared by collecting blood in the corresponding tube, followed by centrifugation at 2000g for 2 minutes and removal of the supernatant. Serum from clotted blood was prepared by leaving the withdrawn blood coagulates for 45 minutes at room temperature, followed by centrifugation at 2000g for 2 minutes and removal of the supernatant. The three types of samples were simultaneously prepared and then stored at -80°C until the day of analysis. NEFA analysis showed relatively similar profiles for the three types of samples. The level of only three species – caprylic acid, pelargonic acid, and arachidonic acid – was found to be different, suggesting that EDTA tubes are the best suited for plasma preparation for NEFA analysis (Figure 4). CONCLUSION In this publication we present a high-throughput non-targeted approach using LC-MS for quantitative analysis of NEFA in human plasma, as well as its validation parameters. The sample preparation procedure is fully automated, which represents a clear advantage in term 15 ACS Paragon Plus Environment


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 16 of 25

of throughput and reproducibility. Analysis by LC-MS allows for simultaneous quantification of multiple NEFA positional isomers while also performing a screen for unknown species. In plasma samples, about 40 NEFA were sufficiently abundant to be accurately quantified. In particular, branched-chain C15:0 and C17:0 could be identified and quantified. Although these fatty acids have important biological functions in other organisms, they have never been reported in human plasma until now. The method can potentially help identify other nonconventional NEFA, as multiple chromatographic signals were detected during plasma analysis but have not been yet unambiguously identified. We also performed a comparative analysis between different types of plasma preparation methods and between adult and child populations. Plasma EDTA appeared to be the most suited plasma preparation for NEFA analysis. Our measurements showed different NEFA profiles for adults and children. Adults tend to have higher saturated and monounsaturated concentrations and lower PUFA concentrations compared to children plasma.


Page 17 of 25

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


FIGURES

Figure 1: Reconstituted ion current chromatograms of ω-3 and ω-6 polyunsaturated fatty acids identified in a human plasma samples. All isomers are baseline separated except C20:3 ω-3/ω-6.

Figure 2: Chromatographic traces of C24:5 from human plasma extracted with a 5 ppm mass tolerance window (A) and a 50 ppm mass tolerance window (B). The spectrum on the top right corner shows the NEFA (m/z 357.2807) and contaminant (m/z 357.2652) signals.



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 18 of 25

Figure 3: Extracted chromatograms from human plasma extract showing the traces for C15:0 (A) and C17:0 (B). Three isomers (anteiso, iso, and straight chain) were separated.

Figure 4: Effect of anticoagulants on the blood plasma NEFA. Group-wise comparison of average NEFA concentration between ETDA (x-axis) and Heparin (y-axis in green) and clot blood (y-axis in red).


Page 19 of 25

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


ASSOCIATED CONTENT Supporting Information. This material is available free of charge via the Internet at http://pubs.acs.org.” Table S1: List of non-esterified fatty acids detected by LC-MS. Table S2: Method validation parameters. Table S3: Comparison between measured and reported NEFA levels in human plasma. AUTHOR INFORMATION Corresponding Author * Tel.: +41 21 6326156 Email address: [email protected] Funding Sources The authors declare no competing financial interest. ACKNOWLEDGMENT The authors would like to thank Dr Rinat Ran-Ressler for reviewing the article and Dr James Kaput and Dr Jacqueline Monteiro for providing the plasma samples. REFERENCES (1) Hirasawa, A.; Tsumaya, K.; Awaji, T.; Katsuma, S.; Adachi, T.; Yamada, M.; Sugimoto, Y.; Miyazaki, S.; Tsujimoto, G. Free fatty acids regulate gut incretin glucagon-like peptide-1 secretion through GPR120. Nat. Med. 2005, 11, 90-94. (2) Boden, G. Obesity and free fatty acids. Endocrinol. Metab. Clin. North Am. 2008, 37, 63546, viii-ix.



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 20 of 25

(3) Kahn, S. E.; Hull, R. L.; Utzschneider, K. M. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 2006, 444, 840-846. (4) Bazinet, R. P.; Laye, S. Polyunsaturated fatty acids and their metabolites in brain function and disease. Nat. Rev. Neurosci. 2014, 15, 771-785. (5) Hu, C.; van der Heijden, R.; Wang, M.; van der Greef, J.; Hankemeier, T.; Xu, G. Analytical strategies in lipidomics and applications in disease biomarker discovery. J. Chromatogr. B 2009, 877, 2836-2846. (6) Li, M.; Yang, L.; Bai, Y.; Liu, H. Analytical Methods in Lipidomics and Their Applications. Anal. Chem. 2014, 86, 161-175. (7) de Oliveira, M. A.; Porto, B. L.; Faria, I. D.; de Oliveira, P. L.; de Castro Barra, P. M.; de Jesus Coelho Castro, R.; Sato, R. T. 20 years of fatty acid analysis by capillary electrophoresis. Molecules 2014, 19, 14094-14113. (8) Christie, W. W. Gas Chromatography–Mass Spectrometry Methods for Structural Analysis of Fatty Acids. Lipids 1998, 33, 343-353. (9) Seppänen-Laakso, T.; Laakso, I.; Hiltunen, R. Analysis of fatty acids by gas chromatography, and its relevance to research on health and nutrition. Anal. Chim. Acta 2002, 465, 39-62. (10) Bicalho, B.; David, F.; Rumplel, K.; Kindt, E.; Sandra, P. Creating a fatty acid methyl ester database for lipid profiling in a single drop of human blood using high resolution capillary gas chromatography and mass spectrometry. J. Chromatogr. A 2008, 1211, 120-128. (11) Quehenberger, O.; Armando, A. M.; Dennis, E. A. High sensitivity quantitative lipidomics analysis of fatty acids in biological samples by gas chromatography-mass spectrometry. Biochim. Biophys. Acta 2011, 1811, 648-656. (12) Eder, K. Gas chromatographic analysis of fatty acid methyl esters. J. Chromatogr. B 1995, 671, 113-131.


Page 21 of 25

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


(13) Johnson, D. W. Contemporary clinical usage of LC/MS: analysis of biologically important carboxylic acids. Clin. Biochem. 2005, 38, 351-361. (14) Wang, M.; Han, R. H.; Han, X. Fatty Acidomics: Global Analysis of Lipid Species Containing a Carboxyl Group with a Charge-Remote Fragmentation-Assisted Approach. Anal. Chem. 2013, 85, 9312-9320. (15) Wei, G.-L.; Zeng, E. Y. Gas chromatography-mass spectrometry and high-performance liquid chromatography-tandem mass spectrometry in quantifying fatty acids. Trends Anal. Chem. 2011, 30, 1429-1436. (16) Trufelli, H.; Famiglini, G.; Termopoli, V.; Cappiello, A. Profiling of non-esterified fatty acids in human plasma using liquid chromatography-electron ionization mass spectrometry. Anal. Bioanal. Chem. 2011, 400, 2933-2941. (17) Zehethofer, N.; Pinto, D. M.; Volmer, D. A. Plasma free fatty acid profiling in a fish oil human intervention study using ultra-performance liquid chromatography/electrospray ionization tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2008, 22, 2125-2133. (18) Pettinella, C.; Lee, S. H.; Cipollone, F.; Blair, I. A. Targeted quantitative analysis of fatty acids in atherosclerotic plaques by high sensitivity liquid chromatography/tandem mass spectrometry. J. Chromatogr. B 2007, 850, 168-176. (19) Kamphorst, J. J.; Fan, J.; Lu, W.; White, E.; Rabinowitz, J. D. Liquid chromatographyhigh resolution mass spectrometry analysis of fatty acid metabolism. Anal. Chem. 2011, 83, 9114-9122. (20) Lacaze, J. P.; Stobo, L. A.; Turrell, E. A.; Quilliam, M. A. Solid-phase extraction and liquid chromatography--mass spectrometry for the determination of free fatty acids in shellfish. J. Chromatogr. A 2007, 1145, 51-57. (21) Hellmuth, C.; Weber, M.; Koletzko, B.; Peissner, W. Nonesterified fatty acid determination for functional lipidomics: comprehensive ultrahigh performance liquid



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 22 of 25

chromatography-tandem mass spectrometry quantitation, qualification, and parameter prediction. Anal. Chem. 2012, 84, 1483-90. (22) Li, X.; Franke, A. A. Improved LC-MS method for the determination of fatty acids in red blood cells by LC-orbitrap MS. Anal. Chem. 2011, 83, 3192-3198. (23) Sun, Y.; O'Riordan, M. X. Branched-chain fatty acids promote Listeria monocytogenes intracellular infection and virulence. Infect. Immun. 2010, 78, 4667-73. (24) Ran-Ressler, R. R.; Glahn, R. P.; Bae, S.; Brenna, J. T. Branched-chain fatty acids in the neonatal gut and estimated dietary intake in infancy and adulthood. Nestle Nutr. Inst. Workshop Ser. 2013, 77, 133-43. (25) Ran-Ressler, R. R.; Sim, D.; O'Donnell-Megaro, A. M.; Bauman, D. E.; Barbano, D. M.; Brenna, J. T. Branched chain fatty acid content of United States retail cow's milk and implications for dietary intake. Lipids 2011, 46, 569-576. (26) Onkenhout, W.; Venizelos, V.; van der Poel, P. F. H.; ven den Heuvel, M. P. M.; Poorthuis, B. J. H. M. Identification and Quantification of Intermediates of Unsaturated Fatty Acid Metabolism in Plasma of Patients with Fatty Acid Oxidation Disorders. Clin. Chem. 1995, 41, 1467-1474. (27) Psychogios, N.; Hau, D. D.; Peng, J.; Guo, A. C.; Mandal, R.; Bouatra, S.; Sinelnikov, I.; Krishnamurthy, R.; Eisner, R.; Gautam, B.; Young, N.; Xia, J.; Knox, C.; Dong, E.; Huang, P.; Hollander, Z.; Pedersen, T. L.; Smith, S. R.; Bamforth, F.; Greiner, R.; McManus, B.; Newman, J. W.; Goodfriend, T.; Wishart, D. S. The Human Serum Metabolome. PloS one 2011, 6, e16957. (28) Ecker, J.; Scherer, M.; Schmitz, G.; Liebisch, G. A rapid GC–MS method for quantification of positional and geometric isomers of fatty acid methyl esters. J. Chromatogr. B 2012, 897, 98-104.


Page 23 of 25

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


(29) Al-Dirbashi, O. Y.; Santa, T.; Rashed, M. S.; Al-Hassnan, Z.; Shimozawa, N.; Chedrawi, A.; Jacob, M.; Al-Mokhadab, M. Rapid UPLC-MS/MS method for routine analysis of plasma pristanic, phytanic, and very long chain fatty acid markers of peroxisomal disorders. J. Lipid Res. 2008, 49, 1855-1862. (30) ten Brink, H. J.; Stellaard, F.; van den Heuvel, C. M. M.; Kok, R. M.; Schor, D. S.; Wanders, R. J. A.; C., J. Pristanic acid and phytanic acid in plasma from patients with peroxisomal disorders: stable isotope dilution analysis with electron capture negative ion mass fragmentography. J. Lipid Res. 1992, 33, 41-47. (31) Ran-Ressler, R. R.; Devapatla, S.; Lawrence, P.; Brenna, J. T. Branched Chain Fatty Acids Are Constituents of the Normal Healthy Newborn Gastrointestinal Tract. Pediatr. Res. 2008, 64, 605-609. (32) Ran-Ressler, R. R.; Khailova, L.; Arganbright, K. M.; Adkins-Rieck, C. K.; Jouni, Z. E.; Koren, O.; Ley, R. E.; Brenna, J. T.; Dvorak, B. Branched Chain Fatty Acids Reduce the Incidence of Necrotizing Enterocolitis and Alter Gastrointestinal Microbial Ecology in a Neonatal Rat Model. PloS one 2011, 6, e29032. (33) Wongtangtintharn, S.; Oku, H.; Iwasaki, H.; Toda, T. Effect of Branched-Chain Fatty Acids on Fatty Acid Biosynthesis of Human Breast Cancer Cells. J. Nutr. Sci. Vitaminol. 2004, 50, 137-143. (34) Yang, Z.; Liu, S.; Chen, X.; Chen, H.; Huang, M.; Zheng, J. Induction of Apoptotic Cell Death and in Vivo Growth Inhibition of Human Cancer Cells by a Saturated Branched-Chain Fatty Acid, 13-Methyltetradecanoic Acid. Cancer Res. 2000, 60, 505-509. (35) Ran-Ressler, R. R.; Bae, S.; Lawrence, P.; Wang, D. H.; Thomas Brenna, J.: Branchedchain fatty acid content of foods and estimated intake in the USA. Br. J. Nutr. 2014, 112, 565572.



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 24 of 25

(36) Quehenberger, O.; Armando, A. M.; Brown, A. H.; Milne, S. B.; Myers, D. S.; Merrill, A. H.; Bandyopadhyay, S.; Jones, K. N.; Kelly, S.; Shaner, R. L.; Sullards, C. M.; Wang, E.; Murphy, R. C.; Barkley, R. M.; Leiker, T. J.; Raetz, C. R. H.; Guan, Z.; Laird, G. M.; Six, D. A.; Russell, D. W.; McDonald, J. G.; Subramaniam, S.; Fahy, E.; Dennis, E. A. Lipidomics reveals a remarkable diversity of lipids in human plasma. J. Lipid Res. 2010, 51, 3299-3305.


Page 25 of 25

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


For TOC only


High-Throughput Quantitative Lipidomics Analysis of Nonesterified

Recommend Documents